HN
ae ia 1 ho ath
a
iho%
)% \ 4
ERA Ries
nie vy")
aE why) PAG Basu ioe ade
ht Ant
’ ; DNA ek Ca hha
DAHER at
Utne ata
AE ay SH MUN
TORY! MDM wh 4 RW a ft ante wd vy)
te
Sy tn via lal
ani
9), ih DCN AYE Ae { "3 if My ay ‘i 4
Or
Me APS oh vw :
Oth
PMR rata bp u Ao Wars ta Beka Oe" Sia ay t]
see day ihe
Nin?
WR
anne
z
pay erika nae
pa
a
me aate iqite Hai ote yA pate yi a i}
eens z Sasa hese ts ase
“a ae ae aah
ve et ew i cg a ea
Se
2 ng aaa
Fat np res — = hapten
£
Ss - es pare
‘A
eee
aa “Ww = bed \iSAy ‘a ua hae ra Ay My >. if > ~ =) et 4 2 My Yi ¢ Z 2 CP, 7a : ALILSNI NVINOSHLINS ,
Saiuvual INSTITU
NVINOSHLINS S31uYVud
al
& 2 Z a : i= bie: < NG i he 3 a ie z NN Ly ¢ ay a 8 Ws 4 a > : a . = if K Ww 1 RAR I Be TE SONTAN INSTITUTION pNOLALILSNI NVINOSHLINS S3IYUVE eeu: ne, fo al loa <x} = < a < a 4 Oc = oc co = cn a co Be J fo) ie O . 2 on z se z pee aT QLILSNI_ NVINOSHLIWS SaIuvVYudIT LIBRARI ES oN x a a is fe D o zZ
Saiuvudi
INSTITUTION
sagiuvug
ARIES SMITHSONIAN INSTITUTION NOILALILSNI_ NVINOSHLINS S314VH
Linosm
RARITIES SMITHSONIAN
i
@ ; 7
= < z < 2 > 4 Zz a Urs = Z KOS = O 19 @, — iy NX ” ) Ww oN a) y, \ SS O aa ke: Oo g \ = as z 2. i vy = 7 Sy > 4 ap) we ierd Pa hi K NIIESN! _NVINOSHLINS Sa | YVYaIT LIBRAR IES SMITHSONIAN INSTITUT Zz 2 le — By i = 3 a, RAD
O fe)
Z z
,
4
INSTITUTION NOILALILSNI Sa1uVvy
INSTITUTION
INSTITUTION INSTITUTION NOILMLILSNI
NLILSNI Saluvual SMITHSONIAN INSTITL
IAN
NVINOSHLINS S3IYVYEIT LIBRARIES SMITH
NVINOSHLINS SJIMNVYYG@IT LIBRARIE
SMITHSONIAN
~
QO X aN SMITHSONI
SMITHSONIAN f
Saluvy
*
SARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS
LIBRARIES LIBRARIES NOILALILSNI
bE LY é fo Pes , ‘ te Ze LIBRARIES
NOILALILSNI
&. ~
G ee Minit anit RILSERECROCOL EE PIALO "Te rpus rte wm )dts at al > a ee ee er FA ADP RARIIA RY eee oe
ILSNI NVINOSHLINS S31uvua!t LIBRARIES
= Bae) = w = 5, HY * : > e obey “iy me - ie - = w = wn = ILSNI MWINGSHIINS aleve ail LIBRARI ES (SMITHSONIAN INSTITU ee = (a Ges Bae g af g : 2 NB - =) = Z ie ee = Bi ai ee a ARIES SMITHSONIAN _ INSTITUTION NOILNLILSNI MUIBOSHIWNS S34 1yYVvHX r dp) os aD | = % a a foctm ca Oc a o~ < pe < a < oc i oc c a © : 8 5 z aj a One a 2 i SJIYVYd!IT LIBRARIES SMITHSONIAN INSTITU C 2 : if = . ie _) = wo = ie = a So va OS > = Sea Ve Ee ee oe = a fe a Zz Hf R | ES SMITHSONIAN INSTITUTION NOILALILSNI SS a luv 3 : : : : 3 2 NN 2 g Z = EW & z : i= Zz > = VU >’ = > Z w Ni Zz vy = : _ NVINOSHLINS SAIYVYEaIT LIBRARIES _INSTITU =a af z oe 2 7) ul 7) = 7) = a <x a= a = = a = 5 bis O eae S) a Fad rae i Fe i Zz RARIES SMITHSONIAN INSTITUTION NVINOSHLINS S31¥YW? z2 - | ; ts : oye ee = a ‘a E es = = xD 2 — Wits 2 = 70 - 2 Yh m S m 8 = w = n = w = = ety WM y pg 2 = > 2
SMITHSONIAN SMITHSONIAN lip NVINOSHLIWS SMITHSONIAN
®
RIES SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S3IuYVe
LIBRARIES NOILNLILSNI LIBRARIES
LIBRARIES
NOILNLILSNI Phe aes
SS
er OG / 2
PROCEEDINGS
of the
Biological Society of Washington
VOLUME 87 1974
WASHINGTON PRINTED FOR THE SOCIETY
EDITOR
AvustTIN B. WILLIAMS
ASSOCIATE EDITORS
Classical Languages Invertebrates
GrorcE C. STEYSKAL THomas FE. BOWMAN
Insects Vertebrates
RosertT D. GORDON Victor G. SPRINGER and
WILLIAM R. TAayYLor
All correspondence should be addressed to the Biological Society of Washington, Smithsonian Institution Washington, D. C. 20560
ALLEN Press INC. LAWRENCE, Kansas 66044
OFFICERS AND COUNCIL of the BIOLOGICAL SOCIETY OF WASHINGTON FOR 1974-1975
OFFICERS
President THOMAS E. BOWMAN
Vice President CLYDE J. JONES
Secretary LOUIS S. KORNICKER
Treasurer OLIVER S. FLINT
Custodian of Publications JOHN H. MILES
COUNCIL
Elected Members RICHARD C. BANKS FENNER A. CHACE, JR. ISABEL C. CANET CLAYTON E. RAY PAUL J. SPANGLER
iu
TABLE OF CONTENTS
Volume 87
Baumann, Richard W. What is Alloperla imbecilla (Say)? Designation of a neotype, and a new Alloperla from eastern North America (Plecoptera: Chloroperlidae) —.____
Bogan, Michael A. Identification of Myotis californicus and M. leibi imy southwestern North Ament cays esas ee
Bowman, Thomas E. The “sea flea” Dolobrotus mardeni n. gen., n. sp., a deep-water American lobster bait scavenger (Amphipoda: Eusinridae)) 250 2
Bowman, Thomas E., and George A. Schultz. The isopod crustacean genus Munnogonium George and Stromberg, 1968 (Munnidae, “Asellota’) (2 ee Se
Child, C. Allan. Hedgpethius tridentatus, a new genus and new species, and other Pycnogonida from Key West, Florida, UGA co AE cL ia eR ee
Collette, Bruce B. Potamorrhaphis petersi, a new species of freshwater needlefish (Belonidae) from the upper Orinoco and. Rio, Negro. 22): we
Collette, Bruce B. A review of the coral toadfishes of the genus Sanopus with descriptions of two new species from Cozumel Island, México. 0000 on eh
Cowan, Richard S. A revision of the genus Bocoa (Caesal- pinioideae-Swartzieae)) 222 8 Ee Cressey, Roger F. A redescription of Hermilius pyriventris Heller (Copepoda: Caligoida) with the first description of the’ male eee ee
Dawson, C. E. Pseudophallus brasiliensis (Pisces: Syngnath- idae), a new freshwater pipefish from Brazil
Deevey, Georgiana B. Pelagic ostracods collected on Hudson 70 between the equator and 55°S in the Atlantic __.....___-_-__
Emerson, K. C., and Roger D. Price. A new species of Tricho- dectes (Mallophaga: Trichodectidae) from the yellow- throated marten (Martes flavigula) ~--- = ees
Flint, Oliver S., Jr. Studies of neotropical caddisflies, XIX: the genus Cailloma (Trichoptera: Rhyacophilidae) _____
Fraser, Thomas H. Redescription of the cardinal fish Apogon endekataenia Bleeker (Apogonidae), with comments on pre- vious usage of the name.) Ue
Gomon, Martin F. A new eastern Pacific labrid (Pisces), Decadon melasma, a geminate species of the western Atlantic Dy pellets: ise SEE)
257-264
49-56
129-138
265-272
493-500
31-40
185-204
235-244
405-410
351-380
77-80
473-484
3-10
205-216
Gopalakrishnan, T. C. A review of the copepod Scottocalanus securifrons (T. Scott) and a note on its synonym Scolecithrix cuneifrons Willey (Calanoida: Scolecithricidae) -----......
Gordon, Robert D. Additional notes on the genus Glaresis m@oleoptera;,. Scarabaeidae) Leet su ee Neal
Herald, E. S., and C. E. Dawson. Micrognathus erugatus, a new marine pipefish from Brazil (Pisces: Syngnathidae) __
Heyer, W. Ronald. Vanzolinius, a new genus proposed for Leptodactylus discodactylus (Amphibia, Leptodactylidae) __.
Heyer, W. Ronald, and M. Judith Diment. The karyotype of Vanzolinius discodactylus and comments on usefulness of karyotypes in determining relationships in the Leptodactylus- complex (Amphibia, Leptodactylidae) —___________
Hoffman, Richard L. A new polydesmid milliped from the southern Appalachians, with remarks on the status of Dixi- desmus and a proposed terminology for polydesmid gonopods
Holt, Perry C. An emendation of the genus Triannulata Good- night, 1940, with the assignment of Triannulata montana to Cambarincola Ellis 1912 (Clitellata: Branchiobdellida) __
Humes, Arthur G. Cyclopoid copepods (Lichomolgidae) from Sauconaceans in Madagascar 0 ae ee
imtonmations tor contributors 2 2
Jones, Meredith L. Brandtika asiatica new genus, new species, from southeastern Asia and a redescription of Monroika afri- cane (Monro) (Polychaeta: Sabellidae)
Kingsolver, John M., and Donald R. Whitehead. Biosystem- atics of Central American species of Ctenocolum, a new genus of seed beetles (Coleoptera: Bruchidae) _...
Light, William J. Occurrence of the Atlantic maldanid Asychis elongata (Annelida, Polychaeta) in San Francisco Bay, with BOUMnentSsONMItSeSyNODYIMNY) 22.2) Oe ee ee
Light, William J. A new species of Pseudopolydora (Poly- chaeta, Spionidae) from the Solomon Islands __.......______.
Lynch, John D. A new species of Eleutherodaciylus (Am- phibia: Leptodactylidae) from the Pacific lowlands of Ecuador
McLaughlin, Patsy A., and Julie H. Brock. A new species of hermit crab of the genus Nematopagurus (Crustacea: Decapoda: Paguridae)) from Hawai
Miller, Robert Rush. Cichlasoma regani, a new species of cich- lid fish from the Rio Coatzacoalcos basin, México _....._._
Novak, Allen, and Michael Salmon. Uca panacea, a new spe- cies of fiddler crab from the Gulf coast of the United States
(9)
273-282
91-94
27-30
81-90
327-336
345-350
57-72
411-438 |-2
217-230
283-312
175-184
389-394
381-388
245-256
465-472,
313-326
Olson, Storrs L. A new species of Nesotrochis from Hispaniola, with notes on other fossil rails from the West Indies (Aves: Beallichae,): <2 2 5 ocala lie el pce eet el oes ee
Orejas-Miranda, Braulio R., and George R. Zug. A new tricolor Leptotyphlops (Reptilia: Serpentes) from Peru ____--
Robbins, C. Brian. Comments on the taxonomy of the West African Taterillus (Rodentia: Cricetidae) with the descrip- tion, Of anew sSDECIES 22k eae Oke ee
Ross, Charles A., and Franklin D. Ross. Caudal scalation of Central American Crocodylus ____. eee me BE mererree Mere
Schwartz, Albert, and Orlando H. Garrido. A new Cuban species of Sphaerodactylus (Gekkonidae) of the nigropunc- tatus:; COMpIeRX, feces Ee ee
Serafy, D. Keith. A new heart-urchin, Araeolampas atlantica, new genus, new species, from the North Atlantic and notes on the closely related Homolampas
Steyskal, George C. Euxesta mazorca, new species, associated with ears of maize in South America (Diptera, Otitidae) __.
Thomas, Richard. A new species of Typhlops (Serpentes: Typhlopidae)) (from Hispaniolay 2
Vagvolgyi, Joseph. Eight new Ashmunellas from the south- western United States (Pulmonata: Polygyridae) —
Williams, Austin B. Allactaea lithostrota, a new genus and species of crab (Decapoda: Xanthidae) from North Carolina, USAGI ie
Williams, Austin B. Two new axiids (Crustacea: Decapoda: Thalassinidea: Calocaris) from North Carolina and the Straits of Florida 2.2. 2)
Williams, Austin B. A new species of Hypsophrys (Decapoda: Homolidae) from the Straits of Florida, with notes on related eras 30 DSN te
439-450
167-173
395-404
231-234
337-343
BIOLOGICAL SOCIETY OF WASHINGTON PROCEEDINGS
1046th Meeting—13 March 1974
NINETY-FIFTH ANNUAL MEETING
The meeting was called to order by Vice President Thomas E. Bowman in the absence of President Victor Springer at 2:10 p.m. in Room 43 of the National Museum of Natural History.
Reading of the minutes of the previous annual meeting was dis- pensed with, since they were printed in the Biological Society of Wash- ington Proceedings.
Treasurer Oliver Flint read the Financial Report. He thanked Mau- reen Downey for serving as Acting Treasurer for the past 6 months and for developing figures for the report. In answer to a question as to why an endowment fund repayment had not been made during the past year, Treasurer Flint said that it was an oversight and that $1700 is to be repaid in 1974.
Clayton Ray asked the Treasurer what percent of page-charges for papers published in the Proceedings came from the Smithsonian In- stitution. Treasurer Flint said that about one-half, but that some pay- ments of the previous year were not yet paid. In answer to a question from Stanwyn Shetler, Treasurer Flint said that page-charges were requested from all authors, but that payment was not a condition for publication.
Daniel Cohen, Chairman of the Financial Committee, stated that the Financial Report had been audited and found to be correct. A motion made to accept the report was seconded and passed. He then presented the budget for 1974.
Austin Williams, Chairman of the Editorial Committee, presented a report stating, Volume 86 of the Proceedings for 1973 consisted of 45 papers totalling 524 pages. The first issue of Volume 87, consisting of 14 papers, is now in galley proof. Ten of these are paid papers. Twelve papers for the second issue have been received and are under review. He also said that instructions entitled, “Information for Con- tributors,” will be included in Volume 87. George Watson suggested that the instructions not be republished every year, but that enough reprints should be printed to last about 5 years.
Secretary Louis Kornicker gave results of the election for Officers and Council members for 1974: President, Thomas E. Bowman; Vice President, Clyde J. Jones; Secretary, Louis S. Kornicker; Treasurer, Oliver S. Flint; Members of the Council, Richard C. Banks, Isabel C. Canet, Fenner A. Chace, Clayton E. Ray, and Paul J. Spangler.
vik
Secretary Louis Kornicker gave a report on membership (see Ap- pendix in Treasurer’s Report). Daniel Cohen asked how a new life member could have been elected since that category had been elimi- nated by the council some years back. He suggested that the Secretary check minutes of old meetings to ascertain whether the category of Life Member was still extant.
Vice President Bowman reported that the Council had 3 meetings during the past year and that a proposed Bicentennial Symposium had been cancelled because of being unable to obtain someone willing to organize it, and because of general lack of interest by possible partic- ipants approached. The subject of the symposium was to have been concerned with biological, geological and ecological changes that have taken place in the Washington area since 1776. He also announced that “The Panamic Biota” volume is being reprinted because of demand and should be available shortly. A policy statement concerning a pro- posed Sea-level Canal in Panama was distributed to other societies and was published in Science. A letter concerning this statement was re- ceived from the President of the Panama Canal Company and is avail- able to interested members.
Marian Pettibone made a motion thanking Officers and Council mem- bers for their efforts on behalf of the Society during the past year, and also Paul Spangler, who had served as Editor of the Proceedings until it was taken over by Austin Williams in the fall. The motion was sec- onded and passed.
Vice President Bowman then turned the meeting over to incoming President Bowman.
A motion to adjourn was seconded and passed.
Louis S. Kornicker Secretary
iii
~ 0673
Vol. 87, No. 0, pp. 1-2 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
INFORMATION FOR CONTRIBUTORS
1. Content. The Proceedings of the Biological Society of Wash- ington contains papers bearing on systematics in the biological sciences (both botany and zoology, including paleontology), and notices of business transacted at meetings of the Society. Except at the direction of the Council, only manuscripts by Society mem- bers will be accepted. Papers will be published in English, except for Latin diagnosis/description of plant taxa which should not be duplicated by an English translation.
2. Submission of manuscripts. Manuscripts should be sent to the Editor, Proceedings of the Biological Society of Washington, U. S. National Museum of Natural History, Washington, D. C. 20560.
3. Review. One of the Society’s aims is to give its members an opportunity for prompt publication of their shorter contributions. The Proceedings are issued in “occasional paper” style approxi- mately three times a year. Manuscripts will be reviewed in order of receipt by a board of associate editors and appropriate referees.
4. Presentation. Clarity of presentation and requirements of taxonomic and nomenclatural procedures necessitate reasonable consistency in the organization of papers. Authors are urged to follow recent issues of the Proceedings as models. Telegraphic style is recommended as the most economical of space for descrip- tions. Synonymy of abbreviated style (author, date, page) with full citations only in Literature Cited is also recommended.
The establishment of new taxa must conform with the require- ments of the appropriate international codes of nomenclature. Authors are expected to be familiar with these codes and to comply with them. New species-group accounts must designate a type- specimen deposited in an institutional collection.
(a) The sequence of material should be: Title, Text, Literature Cited, Appendix, Tables (each table numbered with an arabic numeral and heading provided), List of Figures (entire figure legends), Figures (each numbered and identified).
(b) Webster's Third New International Dictionary of the English Language Unabridged, the AIBS Council of Biology Editors Style Manual, United States Government Printing Office Style Manual, and the BIOSIS List of Serials with Title Abbrevia- tions are useful guides for style, spelling, abbreviations, and other technical matters.
0—Proc. Biot. Soc. WasuH., Vou. 87, 1974 (aL)
2 Proceedings of the Biological Society of Washington
(c) Type manuscripts double-spaced throughout (including tables, legends, and footnotes) on one side of paper measuring approximately 8% x 11 inches, leaving margins of at least one inch all around. Submit a facsimile with the original and retain an author’s copy. Number pages consecutively at the top. One manuscript page = approximately % of a printed page.
(d) Underline singly scientific names of genera and lower cate- gories; leave other indications to the editors. Use all capitals only in the title.
(e) Do not use periods after standard abbreviations such as measurements or compass points unless confusion might result from their omission: e.g.—mm, cm, m, C (Celsius), N, NW, ESE, but no. = number(s), Fig(s). = figure(s), mi. = mile(s).
(f) Figures and tables, with their legends and headings, should usually be self explanatory, not requiring reference to the text. Indicate their approximate placement by a pencil mark in the margin of the manuscript.
(g) Illustrations should be planned in proportions that will efficiently use space on the type bed of the Proceedings (9.7 X 16.8 cm). Legends require 4 mm of column length per line.
5. Charges. The high costs of publication make it necessary for the Society to solicit partial or complete payment of publication cost from authors who have funds available for this purpose through their grants or sponsoring institutions. Authors without access to publication funds should indicate this; cost of their publication will be borne by the Society's funds. Acceptance of a paper is in no way dependent on payment of publication costs, with the following exceptions: all authors are required to pay a handling charge for figures and to pay charges for all pages in excess of 16 that they publish in a single volume of the Proceedings.
6. Proofs. Galley proof will be submitted to authors for cor- rection and approval. Changes other than printer’s or editor's errors may be charged to authors. Reprint orders will be taken with returned proof.
mr C6 7/3
Vol. 87, No. 1, pp. 3-10 25 April 1974
PROCEEDINGS ORMInE
BIOLOGICAL SOCIETY OF WASHINGTON
REDESCRIPTION OF THE CARDINAL FISH APOGON ENDEKATAENIA BLEEKER (APOGONIDAE), WITH COMMENTS ON PREVIOUS USAGE OF THE NAME
By Tuomas H. FRASER! J. L. B. Smith Institute of Ichthyology, Rhodes University, Grahamstown, South Africa
In 1852 Pieter Bleeker described Apogon endekataenia from a single specimen (57 mm total length) taken at Banka in Indonesia. Subsequently, fifteen other specimens were added to the jar containing the type. According to the Auction Catalogue (Hubrecht, 1879) in which Bleeker’s fish collection was offered for sale, ten specimens of A. endekataenia were in the A series, two specimens each in the B, C and D series and none in the E series. Whitehead et al. (1966: 7-14) indicate some basic procedures involved in attempting to determine which specimens are type material in the Bleeker collections. In the case of A. endekataenia the criteria utilized were total length and the remains of faded color pattern. Fortunately, only one specimen (55 mm total length) was close to the 57 mm length given by Bleeker. The others were of various sizes, the nearest three 45, 63 and 64 mm total length. (I cannot account for the presence of two additional specimens in the jar unless a counting or typographic error was made by Hubrecht.) The 55 mm specimen is considered to represent Bleeker’s original specimen from Banka. This agrees with the opinion of M. Boeseman cited in Smith (1961).
The identification of the putative holotype fundamentally changes the concept of Apogon endekataenia; most past records are considered to be erroneous or doubtful identifications. The specimens subsequently added to the type jar were of
1 Present address: Division of Fishes, National Museum of Natural History, Wash- ington, D. C. 20560 U.S.A.
1—Proc. Biot. Soc. WaAsu., Vou. 87, 1974 (3)
4 Proceedings of the Biological Society of Washington
two species and may have been the source of some errors. Descriptions in all major faunal reviews examined indicate other species, or are composites of two or more species (Bleeker, 1874, 1873-76, Weber and de Beaufort, 1929; Fowler and Bean, 1930; Smith, 1961). Lachner (1953) did not reach a decision concerning the identity of Apogon endekataenia but he noted that none of the specimens available to him con- formed to the figure in Bleeker’s Atlas (1873-76). A re- description of the species is necessary and will contribute to the eventual resolution of the dark-striped complex of species in Apogon.
I thank M. Boeseman, Rijksmuseum van Natuurlijke His- torie (RMNH) for permission to examine Bleeker’s material and for his co-operation and valuable help during my stay at Leiden. A. C. Wheeler made possible the loan of speci- mens at the British Museum (Natural History) (BMNH). E. A. Lachner and M. M. Smith critically read a draft of the manuscript. Funds for the taxonomic study of fishes and attend- ing aspects were provided by the South African Council for In- dustrial and Scientific Research and by Rhodes University.
Counts and measurements were made according to Hubbs and Lagler (1958) except that the body depth was taken from the dorsal-fin origin to the pelvic-fin base. Measurements were made with dial calipers. The last dorsal ray and last anal ray, each split to its base, was counted as one element. The gillraker at the angle was included in the count for the lower arch.
Apogon endekataenia Bleeker, 1852 Figure 1
Description: For general body shape see Figure 1. Range of pro- portions (as percentages of standard lengths, putative holotype values in parentheses): body depth (33) 33-36; head length (36) 35-38; eye length (12) 11-12; snout length (8) 7-9; bony interorbital width (7) 6-7; upper jaw length (18) 17-18; caudal peduncle depth (14) 14-16; caudal peduncle length (25) 22-26; first dorsal spine length (4) 1-4; second dorsal spine length (12) 9-12; third dorsal spine length 19-23; fourth dorsal spine length 16-20; last dorsal spine length (17) 15-17; first anal spine length (2) 2; second anal spine length (13) 12-13; pectoral fin length (24) 23-25; pelvic fin length (21) 20-22.
Meristic counts and ranges for nine specimens examined: dorsal fin
Redescription of Cardinal Fish 5
Fic. 1. A. Apogon endekataenia, illustration from Bleeker 1873-— 1876. B. Apogon endekataenia, putative holotype, RMNH 5593 (43.9 mm SL). C. Apogon endekataenia, BMNH 1892.9.2.18—-20, one of three.
VII-I, 9; anal fin II, 8; pectoral fin 14; pelvic fin I, 5; well-developed gillrakers 13-15 (2-3 + 11-12), usually 13-14, including rudiments 18-21 (2-4 + 2-3; 11-12 + 2-4), putative holotype 3 + 3-12 +4 3; pored lateral-line scales (25) 24-25; longitudinal scale rows above lateral line (25) 24-25; transverse scale rows above lateral line 2; trans- verse scale rows below lateral line (6) 6-7; median predorsal scales (3) 38—4; cireumpeduncular scale rows 12 (5 + 2+ 5).
6 Proceedings of the Biological Society of Washington
Teeth in villiform bands on premaxilla and dentary; one or two rows on vomer and palatine; none on ectopterygoid, endopterygoid or basihyal.
Vertebrae 10 + 14. Five free hypurals, one pair of slender uroneurals, three epurals, a free parhypural. Three predorsals, two spines on first dorsal pterygiophore. Seven branchiostegal rays. Basisphenoid present. Supramaxilla absent.
Life colors unknown. In 70% ethanol six dark, longitudinal stripes present from dorsal-fin base to abdomen; large round basicaudal spot slightly off center (dorsally) of lateral line (Figure 1): 1. median stripe on head splitting at dorsal fin and extending along dorsum at base of fins, fading on caudal peduncle; 2. stripe extending from snout along occiput and nape paralleling above the lateral line and reaching nearly to basicaudal spot; 3. post-ocular stripe from upper part of eye and just below lateral line extending onto caudal peduncle; 4. median stripe ex- tending from eye above pectoral fin along side nearly to basicaudal spot; 5. stripe extending from lower part of eye and head through pectoral fin along side nearly to basicaudal spot; 6. stripe extending from lower jaw below pectoral fin along abdomen and onto last several anal-fin rays. Complete dark stripe present in basal portion of soft dorsal fin but no counterpart in soft anal fin. Peritoneum silvery, intestine black.
Distribution: Discounting the doubtful and erroneous reports, Apogon endekataenia is known from Borneo and Banka Island, near Sumatra, and recently I have seen material from Thailand at the California Academy of Sciences. Fowler (1937: 232-234, Fig. 23) identified two specimens from Bangkok as Apogon robustus (Smith and Radcliffe, 1911); how- ever, the illustration appears similar to A. endekataenia but with extra stripes. It has not been reliably reported in the western Indian or cen- tral Pacific Oceans.
Material examined: Putative holotype. RMNH 5593, 43.9 mm SL, 55 mm total length, Banka Island. Other material. BMNH 1892.9.2.18—20. (3 specimens, 52.1-62.1 mm SL), Borneo, X-ray. RMNH 23966 (5, 51.5-67.6 mm SL), East Indies.
Remarks: Apogon endekataenia differs from other dark-striped Apogon (Nectamia) in possessing the following combination of characters: a large, round, basicaudal spot, no dark stripe in the anterior part of the anal fin, and a long post-ocular stripe. Smith (1961: 400) believed that Bleeker’s illustration (1873-76: Pl. 32, fig. 2) was inaccurate. The illustration and putative holotype (Figs. 1A and B) agree better than figures by other authors of specimens purported to be A. endekataenia. The putative holotype also agrees with Bleeker’s original description.
Of the 18 specimens examined in the Bleeker collections, five speci- mens in RMNH 23966 are conspecific with the putative holotype and one other specimen in RMNH 23966 was unidentifiable; two specimens in RMNH 23968 may be conspecific with the type but are too poorly preserved to be certain; and all nine in RMNH 23967 are a different species, probably Apogon cookii Macleay, 1881.
Smith (1961) experienced difficulty in separating closely related spe-
Redescription of Cardinal Fish
€I-I1 61-91
Jourystpuy polfisuezur pur oseq [epneo ye APYsiys susp
uly jepreo 0}UO puecxy
yuosqy yussoIg
‘3 OV-GE ST
elles!
61-81 ou] [P1972] UO paie}UI9 “[[PWIS
jods epneo YA JUON{FUOD
uly [epneo 0UO pus}xXy
(q10yS ) UESe1g yuosolg
"YOIe }SITF FO SLOYVAT[IS 1aMoy pue ioddn sspnpuy +; "TS wu GQ—Fg ‘suouttosds (-G UO paseg
"TS wu Qg—CF ‘suouttoeds GT UO posed z "TS wu 9g—fF ‘suoutfoeds 6 UO poseg rt
L-9 QE-EE vI
eset
I¢-81
QUI] [¥19}e] JO 19}U99 jjo ApYSYS “os1e'T] yods |epneo
YA JUINTFUOI JON
ulj [Bpnvo UO JON (suo'T ) yUsSoLg yussqy
(TS Z) GO [eqqs010} UT
(US Go) GEE APMEL sABi [e10}09q ,SloyeATIIS pedojeacsp-[[PA SLoyep]Is
JO Joquinu [e}0 J,
yods [epney
ediys Apoqplyy sodiijs ureu IaMo] pue todd
od11}s Ie]Nd0-}sOg
uly jeue ul odi.y4s
Bae free ee 2 ee Se Se ee ee ee ee ee
eSNloydouan} “VY
al1y009 “VW
pDluapwyapua “VY
rapORVIVYyD
‘uosody pedtijs-y1ep FO sotoods 9a1y} IOF SioJOVIBYO Pezees FO uosuieduioyn “[ AIaV
§ Proceedings of the Biological Society of Washington
cies in the dark-striped complex of Apogon and believed that he was dealing with a polychromic species. He did not see the putative holo- type but did see two specimens (out of five) which are here con- sidered conspecific with the holotype. Nevertheless, he believed these specimens were identical with his relatively fresh material. None of the material so identified at the J. L. B. Smith Institute is Bleeker’s species but rather Apogon cookii and Apogon taeniophorus Regan, 1908, common species in the western Indian Ocean (see Table 1 for comparisons). The illustrations in Smith (1961) are of A. cookii. Both of these species also appear to be widespread in the eastern Indian and western Pacific Oceans, but have been misidentified or given junior synonyms, a prob- lem that will be discussed elsewhere.
Weber and de Beaufort (1929) probably based their description on the collection of mixed species at RMNH although it is not so stated. Their description is insufficient to determine whether more than one species was included.
Fowler and Bean (1930) examined only the type of A. wilsoni (Fowler, 1918), a different species.
Some of the dark-striped, seven-spined species of Apogon have re- ceived varied treatment: consideration as polychromic forms of the same species (Jordan and Seale, 1906; Smith, 1961), as subspecies (McCulloch, 1915), and as different species (Radcliffe, 1911; Lachner, 1953). McCulloch’s treatment would seem to be least acceptable on practical and theoretical grounds because several of these supposed subspecies are sympatric over a wide geographic range. The ideas of either a single variable species or several different species are more useful hypotheses to test at this stage. Radcliffe (1911) pointed out the changes in number of stripes in Apogon multilineatus (Bleeker, 1865) with growth, and it appears that stripes and the basicaudal spot may be altered with growth in other species. These changes are not described for other striped species and must be investigated.
LITERATURE CITED
BLEEKER, P. 1852. Bijdrage tot de kennis der ichthyologische fauna van het eiland Banka. Nat. Tijdschr. Ned. Ind. 3:443—460. 1873-76. Percoides I. Priacanthiformes, Serraniformes, Gram- misteiformes, Percaeformes, Datniaeformes. Atlas Ichthyol. Indes. Orient. Neerland. 7:1—126, pls. 249-320.
—. 1874. Revision des especes Indo-Archipelagiques du groupe des Apogonini. Verh. Holl. Maatsche. (3)2(1):1-82. Fowter, H. W. 1937. Zoological Results of the Third De Schauensee Siamese Expedition. Part VIII—Fishes obtained in 1936.
Proc. Acad. Nat. Sci. Phila. 89:125-264, 300 figs.
, AND B. A. BEAN. 1930. The fishes of the families Amiidae, Chandidae, Duleidae, and Serranidae, obtained by the United States Bureau of Fishes Steamer Albatross in 1907 to 1910,
Redescription of Cardinal Fish 9
chiefly in the Philippine Islands and adjacent seas. Bull. U. S. Nat. Mus. 100( 10) :i—ix, 1-334, 27 figs.
Husss, C. L., anp K. F. Lacter. 1958. Fishes of the Great Lakes Region. Cranbrook Inst. Sci. (revised ed.) Bull. 26:i—lxi, 1-213, 44 pls., 251 figs.
Huprecur, A. A. W. 1879. Catalogue des collections formees et laissees par M. P. Bleeker. Leiden iv + 71 pp.
JorDAN, D. S., AND A. SEALE. 1906. The fishes of Samoa. Description of the species found in the Archipelago with a provisional checklist of the fishes of Oceania. Bull. Bur. Fish. 25: 175-488, pls. 23-53, 111 figs.
LacHNER, E. A. 1953. Family Apogonidae: cardinal fishes. (In: Schultz, L. P., Fishes of the Marshall and Marianas Islands). Bull. U. S. Nat. Mus. 202(1):412—498, figs. 68-84, tabs. 38-44.
McCuttocu, A. R. 1915. Report on some fishes obtained by the F.1I.S. Endeavour on the coast of Queensland, New South Wales, Victoria, Tasmania, South and South Western Australia. Biol. Res. Endeavour 3(3):97—170, pls. 13-37.
RapcuiFFE, L. 1911. Notes on some fishes of the genus Amia, family of Cheilodipteridae, with descriptions of four new species from the Philippine Islands. Proc. U. S. Nat. Mus. 41(1853): 245- 261, pls. 20-25, 3 figs.
SmitH, J. L. B. 1961. Fishes of the family Apogonidae of the western Indian Ocean and the Red Sea. Ichthyol. Bull. Rhodes Univ. 29:373-418, pls. 46-52, 11 figs.
Weser, M. C. W., anp L. F. pE BeAurorr. 1929. The fishes of the Indo-Australian Archipelago. E. J. Brill, Leiden 5:i-xiv, 1-458, 98 figs.
WHITEHEAD, P. J. P., M. BOESEMAN, AND A. C. WHEELER. 1966. The types of Bleeker’ Indo-Pacific elopoid and clupeoid fishes. Zool. Verhand. 84:1—152, 19 pls., 3 tabs.
10 Proceedings of the Biological Society of Washington
ws O63
Vol. 87, No. 2, pp. 11-18 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
A NEW SPECIES OF TYPHLOPS (SERPENTES: TYPHLOPIDAE) FROM HISPANIOLA
By RicHARD THOMAS
Museum of Zoology, Louisiana State University, Baton Rouge, Louisiana 70808
Since the description (Thomas, 1965) of Typhlops syntherus, it has become evident to me that the Hispaniolan Typhlops, then regarded as T. lumbricalis Linnaeus, actually comprises two species. One of these, known from only a few localities in eastern Hispaniola but better represented from the Cul de Sac Plain of Haiti, differs in no major morphological feature, other than size, from the Cuban and Bahamian populations known as Typhlops lumbricalis. The other species, un- described, appears to be restricted to the southwestern quadrant of Hispaniola.
Aside from three old specimens, the hypodigm of this un- described species was obtained through the sponsorship and efforts of Albert Schwartz (ASFS designates the Albert Schwartz Field Series) and of Ernest E. Williams of the Museum of Comparative Zoology (MCZ). To Albert Schwartz I am greatly indebted for his support of my own efforts in the field and laboratory leading to the discovery of this new form. I also thank Lewis D. Ober, who generously provided some crucial specimens of T. lumbricalis (LDO designates the Lewis D. Ober private collection) and Douglas A. Ross- man for his careful critical reading of the manuscript. For the loan of specimens used in this study I wish to thank Richard G. Zweifel and George R. Foley of the American Museum of Natural History (AMNH), Edmond G. Malnate of the Acad- emy of Natural Sciences of Philadelphia (ANSP), Alice G. C. Grandison of the British Museum (Natural History) (BMNH), Neil D. Richmond and Clarence J. McCoy of the Carnegie
2—-Proc. Biot. Soc. WasuH., Vou. 87, 1974 (11)
12 Proceedings of the Biological Society of Washington
Museum (CM), Ernest E. Williams of the Museum of Com- parative Zoology (MCZ), and the late James A. Peters and George R. Zug of the National Museum of Natural History (USNM). Without the capable field work of Robert K. Bobilin, Patricia A. Adams, Ronald F. Klinikowski, Mark D. Lavrich, and David C. Leber so extensive a series of the new species would not have been collected. Material obtained from 1968 through 1971 was collected under National Science Foundation Grants GB 7977 and B-023603 to Albert Schwartz.
The undescribed species is the sixth known species of Typhlops from Hispaniola; in reference to this fact I propose the name:
Typhlops hectus new species
Holotype: MCZ 81149, taken at Martineau, ca. 9 km (airline) W Jérémie, Dépt. du Sud, Haiti, 12 March 1966 by Richard Thomas. Original number ASFS V9145.
Paratypes: HAITI. Derr. pu Sup: ASFS V9276-77, ca. 10 km WSW Moron, 475 m, 13 March 1966, R. Thomas; MCZ 74907-14, Marfranc near Jérémie, 27 December 1966, D. Hill and F. Vuilleumier; MCZ 64779, Carrefour Sanon near Jérémie, 14 December 1960, Luc and George Whiteman; ASFS V9147, Jérémie, 12 March 1966, native; ASFS V9495, Jérémie, 19 March 1966, native; ASFS V9601-12, Jérémie, 21 March 1966, natives; ASFS V25284, beach area within 1 km E Jérémie, 1 July 1971, Elie Joseph; ASFS V9389, 2 km SE Jérémie, 15 March 1966, native; MCZ 64773-78, *Place Négre near Jérémie, 14-15 December 1960, Luc and George Whiteman; MCZ 70044, *La Source near Jérémie, Decem- ber 1962, George Whiteman; MCZ 64780, *Laye near Jérémie, 14-15 De- cember 1960, Luc and George Whiteman; MCZ 70045, Tosia near (about 33 km SW) Jérémie, December 1962; ASFS V9348, ca. 5 km (airline) SE Marché Léon, 670 m, 15 March 1966, R. Thomas; ASFS V9514—16, ca. 7.5 km (airline) SSE Roseaux, est. 2 km W La Bastille, 20 March 1966, R. Thomas; ASFS V9519, ca. 3 km (airline) SW Corail, 20 March 1966, R. Thomas; MCZ 25552, Ile Grande Cayemite; ASFS V26459, Ile Grande Cayemite, vicinity of Anse 4 Macon, 6 August 1971, native; ASFS X3070, Camp Perrin, 26 July 1962, native; USNM 157928—29, ASFS X3141-52, Camp Perrin, 29 July 1962, native; MCZ R-123873-74, Marceline; ASFS V26251-53, Fond des Négres, 1 August 1971, R. Thomas; ASFS V26223, Paillant, 548 m, 31 July 1971, R. Thomas; MCZ 66323-24, CM 37953, *Butéte near Miragodne, 16 August 1961, Luc Whiteman; MCZ 66325, *Mingrette near Miragodne, August 1961, Luc Whiteman; MCZ 66326-28, CM 37921—22, *Pemel near Miragoane, 14 August 1961, Luc Whiteman. DEPT. DE L’Oursr: ASFS V982]1, ca. 2.4 km S Trouin, 243 m, 29 March 1966, R. Thomas; ASFS V8370-7]1, ca.
A New Typhlops from Hispaniola 13
Fic. 1. A, dorsal and lateral views of the head of ASFS V9604 (T. hectus); B, dorsal and lateral views of the head of ASFS V8395 (T. lumbricalis). Bar equals one mm.
7 km (airline) W Pétionville, N versant Morne lH6pital, ca. 609 m, 1 March 1966, R. Thomas; ASFS V8375, 9.6 km (road) W Pétionville, N versant More I|’H6pital, 822 m, 1 March 1966, native.
Non-paratypic specimens: HAITI. Dtpr. pE L’Ourst: ASFS V8295- 96, ca. 7 km SE Mirebalais; ASFS V26542—48, V26577, 3.55 km SW Lascahobas, 274 m, 15 August 1971, native. REPUBLICA DOMINI- CANA. Prov. SAn Juan: ASFS V512-15, Rio Arriba del Norte; ASFS V21567, 4 km N Sabaneta; ASFS V396—98, 7 km W Vallejuelo, 792 m. Prov. SAN RAFAEL: ASFS V21570-71, 1 km E Hondo Valle. Prov. INDEPENDENCIA: ASFS V20924, 15 km W Puerto Escondido; AMNH 41265-66, Duverge. Prov. BaraHona: ASFS V30493, El Iguito, 2.6 km NE Fondo Negro; ASFS V35641, El Iguito, 3.1 km NE Fondo Negro; AMNH 51496, above Delmonte’s Finca (near Barahona).
Localities indicated with an asterisk are not mapped; it is not certain whether the unmapped localities reckoned from Miragoane are in the Dépt. du Sud or the Dépt. de ]’Ouest.
Definition: A species of Typhlops having the preocular in contact with upper labial three (as opposed to contacting upper labials two and
14 Proceedings of the Biological Society of Washington
three), two postoculars, a primary scale row number of 20 with reduction to 18 rows absent or occurring far posteriorly, and high mid-dorsal scale counts (284-328). Further distinguished by a clavate rostral shape (narrow, posteriorly expanded and blunt); the posterior nasals being roughly parallel sided, not divergent in dorsal aspect; and the preocular being sharply pointed anteriorly (Fig. 1).
Distribution (Fig. 2): Southwestern Hispaniola south of the Cul de Sac-Valle de Neiba Plain (except for the Barahona Peninsula) and the southwestern part of the north island (that portion of Hispaniola north of the Cul de Sac-Valle de Neiba Plain). One certain record for the Cul de Sac-Valle de Neiba Plain (extreme eastern part).
Description of holotype: Total length 192 mm, tail length 4.6 mm; 307 mid-dorsal scales between rostral and caudal spine, 295 midventral scales posterior to mental and including last scale of anal flap; 20 scale rows anteriorly, reducing to 19 rows at ventral scale 277 and to 18 at ventral scale 279 (with some redividing and re-fusing of scale rows before final fusion); 2 postoculars and one parietal on each side. Colora- tion pale gray dorsally, fading out on lower sides to become unpigmented ventrally.
Variation: The paratypes vary in total length from 95 to 230 mm and in mid-dorsal scale counts from 284 to 324. The modal primary scale row number is 20 reducing to 18 at the level of midventral scale 194 to 288; 46 specimens (60 percent) have no reduction. A number of speci- mens show reduction abnormalities shortly anterior to the vent: the first reduction step (20 to 19 rows) may be made without the final (19 to 18 rows), or irregular fusion and dividing of the median and paramedian ventral rows may occur. Three specimens have 22 scale rows initially with reduction to 20 rows occurring from 35 to 131 ventrals posterior to the mental. A few have short regions of 22 scale rows anteriorly due to irregular division and fusion, and a few others have 21 scale rows for a short distance in the neck region. There are from 5 to 15 (mode 8-10) fewer ventrals than middorsals. Postoculars are 2/2 (66 specimens), 2/3 or 3/2 (2), or 3/3 (1); parietals are 1/1 (16), 1/2 or 2/1 (20), or 2/2 (25). Coloration is pale gray to tan but varies somewhat in intensity of pigmentation. The coloration may be bicolor with the transition from the pigmented dorsum to the unpigmented venter occurring in a narrow mid-lateral zone of one or two scale rows; or the pigmentation may extend well onto the venter, occasional individuals having almost entirely pig- mented venters with irregular non-pigmented areas occurring along the midventral region. In the more extensively pigmented individuals the intensity of pigmentation gradually decreases ventrally. Some bicolor specimens have an indistinct collar of dark pigment across the throat.
Twenty-four specimens from several scattered localities north of the Cul de Sac-Valle de Neiba Plain plus six from the eastern portion of the south island and the Valle de Neiba agree with Typhlops hectus in the configuration of the rostral-nasal complex, in having pointed preoculars, high longitudinal scale counts, and in having the reduction from 20 to 18
15
A New Typhlops from Hispaniola
‘UIR[ VQION Op 2I[BA-9VS 9p [ND oy} Jo spurl] oyeurrxoidde oy} soyeorpul SurTyoyeyY [v}I0A OY, “E ‘Sly UL poydeis snzoay ordAyeivd-uou Jo sojdures “e—-v sioqjey, ‘sngzoay ordAyered-uou 10} soiyeooy, ‘sojsueLy MOT[OY ‘snzoay oidAyered 10J sotppeooy, oye -IpUl sojsuel prog “snjoay *f pue ejoruedsipy Ul (SooIIO pl[os) synowquin] “J, JO UOTNGLys!Ip UMOUY SuIMOYS deyY “FZ ‘DIY
wy ST Ov Of OZ O1 O
16 Proceedings of the Biological Society of Washington
240 260 280 300 320
Fic. 3. Upper graph: Variation in reduction level (ventral scales posterior to mental). Range only is shown for samples of non-paratypic hectus (a-e); line histograms (smallest vertical unit represents one individual) are given for lumbricalis (L.) and paratypic hectus. The uppermost horizontal line of a pair indicates range in level of final reduc- tion step (19 to 18 scale rows); the lowermost line of a pair indicates the range in initial reduction step (18 to 19 rows). Crosses on lower range line for paratypic hectus indicate specimens that reduce to 19 but not to 18 scale rows; + 46 and + 1 indicate specimens that do not reduce from 20 scale rows. Lower graph: Line histograms showing variation in mid-dorsal scale counts of T. lumbricalis (L.) and T. hectus (H); the smallest vertical unit represents one individual. Histograms a—e show counts for geographic samples of non-paratypic hectus indi- cated in Fig. 2. The dashed line in both upper and lower parts separates graphs for lumbricalis and hectus.
A New Typhlops from Hispaniola 17
TasLE 1. Meristic and color variation among the Hispaniolan species of
Typhlops. Pre- Post- ocular ocular Scale Middorsal number number rows scales Coloration lumbricalis 1 1-2 20-18 231-294 bicolor 20-20 ss bicolor to hectus 1 2 90-18 284-328 Pee a pusillus 2 2; 20-18 952-315 bicolor to unicolor syntherus 1 2 WON) 304-353 bicolor sulcatus 1 1 20-20 400-452 bicolor
capitulatus 1 2; 20-20 358-418 unicolor
scale rows situated relatively far posteriorly. Despite an average far posterior reduction level, these specimens are somewhat heterogeneous in this respect (Fig. 3). The four from the eastern south island are most like paratypic hectus in reducing far posteriorly (one does not reduce). Of those from the north island, the specimens from the northern Valle de San Juan reduce to 18 rows farther anteriorly than those in any other sample and are therefore least like paratypic hectus in this respect. The sample from the vicinity of Lascahobas broadly overlaps the other samples of both paratypic and non-paratypic hectus in reduction level, and the three specimens from the Hondo Valle and Vallejuelo region are inter- mediate. One of the two specimens from E] Iguito in the eastern Valle de Neiba reduces far anteriorly and the other far posteriorly. Some of the samples of non-paratypic hectus are more heavily pigmented than paratypic hectus, and all have higher average mid-dorsal counts (Fig. 3). Two specimens of non-paratypic T. hectus have divided preoculars uni- laterally (AMNH 41266 from Duvergé and ASFS V398 from 7 km NW Vallejuelo); AMNH 41265 from Duvergé has an incompletely divided preocular on one side; ASFS V396 and V398 (Vallejuelo) and V513 (Rio Arriba) have indications of a groove on one preocular. Divided preoculars characterize Typhlops pusillus Barbour, a widespread Hispaniolan species. The significance of divided preoculars in specimens of T. hectus will be discussed in a comprehensive study of Antillean Typhlops.
Because of the isolation, heterogeneity in reduction level, and divergence of these eastern and northern samples from the balance of the specimens of T. hectus, I have not designated them as paratypes. It is hoped that the variation in these populations will be elucidated by more compre- hensive sampling.
Comparisons: Typhlops hectus is readily distinguishable from four of the other Hispaniolan species on rather gross morphological features
18 Proceedings of the Biological Society of Washington
(Table 1). From lumbricalis, however, hectus cannot be absolutely dis- tinguished on the basis of meristic features. There is a small amount of overlap between the two species in mid-dorsal counts (Fig. 3). In reduction level lumbricalis and paratypic hectus are almost completely separable; the non-paratypic samples are intermediate. The differences in head scale shape are subtle; some specimens are difficult to distinguish on head scale shape alone. In hectus the rostral is narrow in dorsal view, almost parallel sided but with a slight terminal expansion; in lumbricalis it may be narrow but is usually broader than in hectus, and the widest point is near the middle (Fig. 1). The preocular in hectus is more sharply pointed (the anterior edges form a smaller angle) than that of lumbricalis. A correlate of the low preocular angle in hectus is that the lateral edges of the postnasals are more nearly parallel with one another and with the lateral edges of the rostral, whereas they are divergent in lumbricalis (Fig. 1). In addition, the Cul de Sac lumbricalis typically have four parietals (32 of 37); only 25 of 61 paratypic hectus have four parietals. Ten of 11 of the eastern Hispaniolan lumbricalis have the first parietal (only two specimens have paired parietals and these only unilaterally ) greatly elongated and fused with the upper postocular, a condition not found either in hectus or other lumbricalis. T. lumbricalis also is a larger, stouter species than T. hectus.
Although lumbricalis and hectus have partially overlapping geographic ranges, their microsympatry is not established. At no time have the two been taken together, despite the collection of fair numbers of lumbricalis (and three other species) from the Cul de Sac. Microsympatry between lumbricalis and hectus will probably be most easily demonstrated in the area between the Cul de Sac and Pétionville.
Comparative material: Typhlops lumbricalis—HAITI, D&pr. DE L OvEst: MCZ 62637, 81150, CM 38886, ASFS V8185-86, Manneville; USNM 117270-72, 117275, ASFS V24346, Trou Caiman; BM 1948. 1.6.63-64 (2), Pont Beudet; USNM 75893, 123792, MCZ 51426, 62631- 33, Port au Prince; MCZ 65812, near Port au Prince; ASFS V8394—97, Chateau Blond, 6.4 km NE Pétionville, 160 m; ASFS V22440-44, LDO 7-6464-69, 4.8 km N Pétionville; ASFS V24346, 1.6 km NW Pétionville; MCZ 68571, Colombier near (22 km E) Saltrou. REPUBLICA DO- MINICANA, Prov. PEDERNALEs: ASFS V2604, 11 km SW Los Arroyos, 443 m; ASFS V2708, 21 km N Pedernales, 243 m. Prov. Azua: ASFS X8047, 2.9 km W, thence 16.4 km N, Azua. Prov. SAN PEDRO DE Macoris: AMNH 13630, San Pedro de Macoris. Prov. La Veca: ASFS V35897-900, V35904—05, 1 km W Jayaco, 274 m; ASFS V14143, 4 km S La Vega. Prov. SaAaMANA: USNM 55298, Sanchez. Prov. PuErro PLata: USNM 10276, Puerto Plata.
LITERATURE CITED
Tuomas, RicHarD. 1965. A new species of Typhlops from the Bara- hona Peninsula of Hispaniola. Copeia 1965(4):4386—-439.
mez) G 7 3
Vol. 87, No. 3, pp. 19-26 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
ALLACTAEA LITHOSTROTA, A NEW GENUS AND SPECIES OF CRAB (DECAPODA: XANTHIDAE) FROM NORTH CAROLINA, U.S.A.
By Austin B. WILLIAMS National Marine Fisheries Service Systematics Laboratory, Washington, D.C. 20560
In 1963 an unusual female crab of the family Xanthidae was brought to my attention in a collection of invertebrates made during a survey of benthic fauna off North Carolina, U.S.A., by M. J. Cerame-Vivas, then at Duke University Marine Laboratory. Description was delayed until males could be found and studied to allow more complete analysis of the species.
Genus Allactaea, new genus
Areoles of carapace prominent, indicated by large and small lobules often arranged in triads, tending to proliferate with increasing age. Mouth frame roughly quadrate, widest anteriorly. Third maxillipeds separated; merus wider than long, with anteroexternal corner produced, anterior margin sinuous, internal corner notched for insertion of palp. Lacinia of first maxillipeds moderate in size. Fused basal antennal seg- ment extending into inner orbital hiatus. First gonopods of male terminating in a more or less flat and moderately developed lobe pre- ceded by a tuft of long hairs in a nearly single file.
Type-species: Allactaea lithostrota.
Etymology: From the Greek “allos” other, and “actaea.” The gender is feminine.
Allactaea lithostrota, new species Figures 1-3 Holotypic male: Carapace xanthoid (Fig. 1B), wider than long, arched anteroposteriorly, regions prominently indicated by smooth raised lobules
arranged in a radiating pattern eminating laterally and anteriorly from urogastric region. Smooth tracts between lobules almost completely
3—Proc. Bion. Soc. WAsuH., Vou. 87, 1974 (19)
20 Proceedings of the Biological Society of Washington
Het
Fic. 1. A, right and left chelae of holotype ¢, frontal view; B, holo- type ¢, dorsal view; C, allotype ¢@, dorsal view; mm scales (upper, A-B; lower, C).
obscured by feltlike covering of clavate hairs anterior to urogastric region; hairy covering much thinner posteriorly.
Front fairly prominent and somewhat produced at each side of median, narrowly U-shaped notch; a dorsal, submarginal row of about 8
A New Xanthid Crab from North Carolina 21
coalesced lobules on each side extending from antennal notch to near midline giving front a transversely doubled appearance (Fig. 2). Lobules, exclusive of margin, arranged as follows (Fig. 1B): 2 F (see Rathbun, 1930, Fig. 3), 2 well separated laterally; 1 M, 3 completely coalesced posteriorly, less so anteriorly; 2 M, evidence of 8 or 9 clumped in two groups, a longitudinally oriented and fused group of 3 near midline having a deep lateral cleft, and a more loosely clumped group of 3 and rudimentary fourth along lateral margin of region with 2 associated lobules in intermediate space; 3 M, a row of 6 in midline, 4 in transverse basal row, each series more or less fused; 4 M, 3 separated in crescentic transverse row; 1 L, solitary and small near anterolateral border; 2 L, 2 closely approximated and among largest individual lobules; 3 L, 2 approximately size of 1 L and not coalesced on right side, but coalesced into single large lobule on left side; 4 L, solitary and small near fourth and fifth anterolateral teeth; 5 L, a complex of 2 groups of 3 nearly coalesced, among largest on carapace; finally, a complex composed of 1 P, 2 P, 1 R and 3 R arranged in irregular, transverse, lunate rows with fairly symmetrical fusion of lobules within rows. Orbital margin thickened, raised and smooth dorsally with fissures closed but evident as depressons; outer orbital hiatus a broadly open notch; obtuse suborbital tooth followed laterally by irregularly increasing suborbital rim. Eyestalks with a tuft of clavate hairs in corneal emargination preceded by 2 spiniform tubercles and a number of much smaller tubercles at base of cornea. Four prominent anterolateral teeth exclusive of outer orbital widely separated and reduced to anteriorly directed mammillary lobules; on margin a bit below these a row of smaller slender intercalary teeth; hepatic and sub- branchial region lightly tuberculate. Posterolateral border bearing a row of lobules.
Epistome with 2 tubercles on lower border of each antennular socket, mesial one largest. Mouth frame roughly quadrate (Fig. 2), widest anteriorly, anterior margin smoothly and sharply cristate and shaped into a low angular arch on each side. Third maxillipeds separated, di- vergent anteriorly; ischium with a shallow, smooth, longitudinal, ventral depression; merus broader than long with slightly sinuous anterior margin, notched for reception of strong palp and produced anterolaterally, tubercles clustered on ventral surface along mesial border and emargin- ation for palp but scattered elsewhere centrally and near anterior margin. Lacinia of first maxillipeds moderate in size covering roughly 40 percent of endostomial surface. Inner orbital hiatus filled by a club-shaped extension of fused basal antennal segment and first movable antennal segment.
Sternal plastron relatively smooth, bearing scattered clavate hairs. Ab- domen with telson wider than long, tip rounded, segments 3-5 fused.
Chelipeds (Fig. 1A) dissimilar, right larger than left. Palm and carpus of each ornamented on outer surfaces with well separated distally di- rected lobules varying from sharpest along dorsal crest of carpus and palm to flared at top with narrower bases on external surface of carpus and
22 Proceedings of the Biological Society of Washington
Fic. 2. Portion of frontal, suborbital and mouth field regions of allotype @ showing edge of front, antennules, left antenna and eye, epistome, left third maxilliped, lacinia of right first maxilliped in situ with overlying mouthparts turned aside, and outline of mandibles in situ.
dorsal portion of palm, becoming smaller, sharper and tending to align- ment in longitudinal rows on outer surface of palm. Carpus with a spiniform lobule at inner angle and another below it; anterior border with a few small teeth hidden in hairs. Dorsal and outer surfaces of chela, carpus and merus with spaces between lobules and tubercles covered with dense coating of clavate hairs, inner surfaces smooth. Fingers dark colored, a crest of tubercles at base of each dactyl; teeth distinct but forming a continuous row on occlusal surface of each finger, proximal teeth of major chela somewhat molariform. Merus of each cheliped with a dorsal crest of sharp forward-curved spines flanked laterally by a few tubercles.
Second and fifth pereopods with external surfaces covered by a dense ragged coat of clavate hairs (third and fourth pairs of pereopods missing). Dorsal crest of merus, carpus and propodus of each leg armed with erect, well separated, distally trending spines; a secondary lateral row of spiniform tubercles nearly hidden in hairs on carpus and propodus.
A New Xanthid Crab from North Carolina U8}
SS SS SSS S) &
Fic. 3. First right gonopod of paratype ¢ (USNM 143774), terminal portion; A, mesial view, B, lateral view; scale = 0.1 mm.
Dactyl of second legs densely hairy on all surfaces, of fifth legs nearly naked on inner surface, both with darkened corneous tip.
First gonopod of male (Fig. 3) rather stout, approximated along middle part but tips diverging ventrolaterally, terminating in a membranous rounded lobe. An irregular row of articulated retrogressive spines along sternal edge near tip grading from scattered and smallest proximally to
24 Proceedings of the Biological Society of Washington
denser, longer and more slender distally, and merging finally into an irregular subterminal file of elongate hairs; a broader tract of similar scattered spines on mesioventral surface and a lateral cluster at base of terminal lobe. In addition, a row of slender short spinules on mesial surface following course of sperm duct.
Measurements in mm: Holotypic male: greatest width of carapace at level of fourth anterolateral tooth, 16.2, length in midline, 11.6; telson width, 1.8, length, 1.6.
Allotypic female: Lobular pattern (Fig. 1C) basically same as in holotype but specimen larger and more omate suggesting greater age. Coalescence of lobules varying in minor details from that in holotype with introduction of additional small solitary lobules at edge of and between areoles, especially behind orbits, on branchial regions and antero- lateral borders. Lunate arrangement or rows of lobules behind urogastric region in 1 P, 2 P, 1 R, 2 R and 3 R not so evident as in holotype.
Epistome with tendency to formation of a row of tubercles on lower border (Fig. 2) of antennular socket, 3 on right side, 2 on left. Lobulation at corners of buccal frame and subhepatic region much bolder and merus of third maxillipeds with ventral tubercles stronger than in holotype. Sternal plastron lightly punctate and hairy on exposed parts becoming ornamented with obsolescent granules along anterior and anterolateral borders at base of third maxillipeds and chelipeds.
Abdomen ovate, densely edged with long hairs; segments lightly covered externally with clavate hairs except on worn raised portions. Telson broadly rounded, broader than long.
Chelipeds with disparity in size and shape more pronounced than in holotype. Lobules not pointed but many with tips rounded, flattened and flared forming knobs. Anterior edge of carpus with a row of spines, a strong spine at inner angle followed by a second almost equal below it. Proximal teeth of major chela more molariform than in holotype.
Remaining pereopods essentially as in holotype, each with crest of distally trending spiniform lobules, many of them club shaped, along upper border of merus, carpus and propodus; crest most pronounced in coalesced ornate lobules on superodistal angle of merus. Each leg with an external flanking row of lobules on carpus and propodus, and a scattering of lobules on merus of fifth leg. Lower edges of each merus with ornamentation grading from proximal granules to attenuated distal lobules. Legs clothed in dense growth of clavate hairs, longest and most ragged along crests.
Measurements in mm: Allotypic female: greatest width of carapace at level of fourth anterolateral tooth, 28.1, length in midline, 18.1; telson width, 4.9, length, 4.0. Ovigerous female paratype: greatest width of carapace at fourth and fifth anterolateral teeth, 17.2, length in midline, le
Type locality: Approximately SE Cape Lookout, N. C., 33°43’ N, 76°40’ W, 90 m-33°42.7’ N, 76°40.2’ W, 110 m, Eastward Sta. 1087.
A New Xanthid Crab from North Carolina 25
Material studied: Specimens studied are confined to the type series deposited in collections of the National Museum of Natural History (USNM), Washington, D. C., and University of North Carolina Institute of Marine Sciences (UNC-IMS), Morehead City, N. C.
USNM 143770 6 holotype; USNM 143772 ¢ juv. paratype, Eastward Sta. 1087, 27 Apr. 1965, L. R. McCloskey, col. USNM 143771 off Cape Lookout, N. C., 91 m, Oct. 1963, 2 allotype, M. Cerame-Vivas, col. USNM 143773 32°39’ N, 78°32’ W, 91-95 m, 28 July 1960, 2 (ov) paratype, R/V Silver Bay Sta. 2266, clam dredge. USNM 143774 33°25.7’ N, 77°01’ W, 60-70 m, 17 Mar. 1969, ¢ paratype (frags.), Eastward Sta. 11552, dredge, N. Chamberlain, col. UNC-IMS 2299, 34°11’ N, 76°10’ W, 60 m, 4 May 1969, @ paratype, Eastward Sta. 11943, F. J. Vernberg, col. UNC-IMS 2366, about 34°08’ N, 76°10’ W, 50 m, 27-30 Apr. 1966, 2 juv. (dry) paratype, Eastward, C. E. Jenner and class, col.
Name: The name is from the Greek “lithostrotos” meaning inlaid with stones, a term descriptive of the symmetrical lobate ornamentation.
Remarks: Allactaea lithostrota is unusual among numerous species of the actaeid group in that lobules of the carapace are large, smooth, and either clustered or solitary in each area. Tending to be rather sharply pointed in juveniles, the lobules become progressively blunter as they increase in size, many of them finally becoming flattened knoblike ex- pansions on somewhat constricted bases. In addition, new small lobules proliferate with increasing size. Clusters on frontal and medial areas tend to be tripartite with an opening in the cluster directed forward and outward. All lobulations of the carapace tend to be peripherally directed away from the urogastric area.
Surface ornamentation indicates close relationship to none of the American actaeas, but characters of the mouth field and orbito-antennal region are similar to Actaea de Haan 1833 (emend.) and those of the first male gonopod resemble Paractaea Guinot, 1969 (Guinot, 1967; 1969; 1971). On the basis of these comparisons, the mouth field of Allactaea lithostrota has a remote resemblance to that of Indo-Pacific Actaea calculosa (H. Milne Edwards, 1834) and A. glandifera Rathbun, 1914 (Guinot, 1969), and the pleopods to those of western Atlantic Paractaea rufopunctata nodsa (Stimpson, 1860) (Guinot, 1969).
Allactaea lithostrota may be confined to the Lithothamnion reeflike structure described by Menzies, Pilkey, Blackwelder, Dexter, Huling and McCloskey (1966), and Menzies, George and Rowe (1973) in warm water near the edge of the continental shelf off North Carolina.
I am indebted to M. J. Cerame-Vivas, N. A. Chamberlain, C. E. Jenner, L. R. McCloskey and F. J. Vernberg for gifts of specimens, H. B. Roberts for counsel, F. A. Chace, Jr., B. B. Collette and I. Pérez Farfante for criticism of the manuscript, and M. Diéguez for drawings and assembly of figures. The type, allotype and two paratypes have been transferred to USNM from UNC-IMS lots 2365, 1899 and 2374.
26 . Proceedings of the Biological Society of Washington
LITERATURE CITED
Gurnor, D. 1967. Recherches préliminaires sur les groupements naturels chez les crustacés décapodes brachyoures. III. A propos des affinités des genres Dairoides Stebbing et Daira de Haan. Bull. Mus. Natl. Hist. Nat. Paris Ser. 2, 39(3):540-563.
—. 1969. Sur divers Xanthidae notament sur Actaea de Haan et
Paractaea gen. noy. (Crustacea Decapoda Brachyura). Cah.
Pac. No. 3:223—267.
1971. Recherches préliminaires sur les groupements naturels
chez les crustacés décapodes brachyoures. VIII. Synthése et
bibliographie. Bull. Mus. Natl. Hist. Nat. Paris Ser. 2, 42(5):
(for 1970) : 1063-1090.
Menzies, R. J., R. Y. GEorcE AND G. T. Rowe. 1973. Abyssal environ- ment and ecology of the world oceans. John Wiley and Sons, New York. xxiii + 488 pp.
, O. H. Pmxey, B. W. BLACKWELDER, D. DEXTER, P. HULING AND L. McCtoskey. 1966. A submerged reef off North Caro- lina. Int. Rev. Gesamten Hydrobiol. 51(3):393—431.
Ratusun, M. J. 1930. The cancroid crabs of America of the families
Euryalidae, Portunidae, Atelecyclidae, Cancridae and Xanthidae. U.S. Nat. Mus. Bull. 152: 1-609.
MX G&S (ENC er
Vol. 87, No. 4, pp. 27-30 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
MICROGNATHUS ERUGATUS, A NEW MARINE PIPEFISH FROM BRAZIL (PISCES: SYNGNATHIDAE)
By E. S. Heratp' anp C. E. Dawson Steinhart Aquarium, California Academy of Sciences, San Francisco, California 94118, and Gulf Coast Research Laboratory Museum Ocean Springs, Mississippi 39564
In the course of recent fieldwork by the junior author in Central and South America, we have been fortunate to obtain a single specimen of an unusual marine pipefish. Characters of this new species are so distinctive that we do not believe it necessary to delay its description until more specimens are available.
The holotype has been deposited in the Museum de Zoologia da Universidade de Sao Paulo (MZUSP), Brazil. Measure- ments are in millimeters (mm).
Appreciation is expressed to Dr. Naercio A. Menezes (MZUSP) for making this specimen available for study. The entire J. L. Almeida family of Salvador, Brazil is thanked for their hospitality and active participation in the field. This work was supported in part by National Science Foundation Grant GB 31053X.
Micrognathus erugatus, new species
Figure 1
Holotype: MZUSP 9408 (64 mm SL, female); Brazil, Bahia, rock tidepool about 1 km S of Arembepe, approx. 12°45’ S, 38°10’ W, 0-1.3 m, chemical ichthyocide; 26 August 1972, C. E. Dawson and party.
Diagnosis: Dorsal-fin rays 19, located on 0.5 trunk rings and 4.5 tail rings; trunk rings 20; tail rings 36; pectoral-fin rays 13 (2); anal-fin rays 2; caudal-fin rays 10; head length 5.9; snout length 1.9; length of
1 Deceased.
4—Proc. Biot. Soc. WasH., Vou. 87, 1974 (27)
28 Proceedings of the Biological Society of Washington
Fic. 1. Head and anterior body of Micrognathus erugatus MZUSP 9408; holotype; 64 mm SL.
dorsal-fin base 4.9; caudal-fin length 2.0; body ridges smooth, poorly developed.
Description: Head ridges indistinct, only snout, supraorbital and opercular ridges visible. Opercle with 20 or more rows of minute pock- like depressions radiating from the indistinct median ridge. Excurrent gill openings extend to top of head. Body ridge pattern typical of Micrognathus, i.e., lateral trunk ridge deflected ventrad at anal ring thence continuous with inferior tail ridge; superior tail ridge deflected at dorsal fin to become lateral tail ridge, ending free near middle of last trunk ring. Body ridges rounded, indistinct; intermedial scutellar plates on trunk and anterior portion of tail equal about half width of individual rings. Pectoral cover plate without ridges but with a few pocklike de- pressions; pectoral-fin length equals about 1.75 trunk rings. Head-in-SL 10.8; snout-in-head 3.1; dorsal-fin base-in-head 1.2.
Color in alcohol mainly dark brown with indications of 5 darker bands on upper sides and dorsum of trunk and 10 on tail; tip of upper jaw white; narrow white cross bar below angle of gape; white surrounds nares and continues as narrow, irregular bar across snout; streaks of white radiate from eye and one streak extends posteriad as narrow blotch on opercle; pectoral-fin base and top of head mottled with white; pale blotches bisect dark trunk and tail rings. Dorsal-fin rays streaked with brown near bases, fin otherwise pale; caudal fin mainly brown, but uppermost ray and distal margin white.
Etymology: erugatus, from the Latin e, not, -- rugatus, wrinkled, in allusion to the almost smooth head and body.
Discussion: The genus Micrognathus now includes the subgenera Micrognathus, Anarchopterus and Minyichthys (Herald and Randall, 1972). Counts of 13-17 trunk rings occur in all but the two known spe- cies of the Pacific Minyichthys, which is characterized by 19-21 trunk rings and well-developed head and body ridges. The 20 trunk rings of M. erugatus suggest a close relationship with Minyichthys, but its smooth head and poorly developed ridges are atypical of Minyichthys. In the latter characters, Micrognathus erugatus is similar to the smooth-bodied Atlantic American Anarchopterus, but the anal fin is absent in this subgenus.
A New Pipefish from Brazil 29
The unique holotype of Micrognathus erugatus was collected with eight specimens of the wide-ranging (Florida—Brazil) M. vittatus (Kaup). This is the only Atlantic representative of the subgenus Micrognathus, although seven of its nine Pacific congeners fall in this characteristically sharp-ridged subgenus. The western Atlantic species Micrognathus dawsoni (Herald, 1969) was originally placed in the subgenus Micro- gnathus, but its body ridges are not sharp edged and it is the only long- snouted species in a genus otherwise noted for short-snouted forms.
The Atlantic Micrognathus group presents somewhat of an enigma in that there are now five species, representing four intrageneric trends, and two species (M. dawsoni and M. erugatus) do not fall within described subgeneric limits. These two species are known only from unique female holotypes and additional study material will hopefully help clarify relationships within the genus.
LITERATURE CITED
Heratp, E. S. 1969. A new pipefish from the Virgin Islands, Micro- gnathus dawsoni. Occ. Pap. Calif. Acad. Sci. 73:1-3.
, AND J. E. RANDALL. 1972. Five new Indo-Pacific pipefishes. Proc. Calif. Acad. Sci., 4th Ser. 39(11):121—140.
30 Proceedings of the Biological Society of Washington
Mn fa. CJD F 7
Vol. 87, No. 5, pp. 31-40 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
POTAMORRHAPHIS PETERSI, A NEW SPECIES OF FRESHWATER NEEDLEFISH (BELONIDAE) FROM THE UPPER ORINOCO AND RIO NEGRO
By Bruce B. COLLETTE
National Marine Fisheries Service Systematics Laboratory National Museum of Natural History, Washington, D.C. 20560
While reviewing the South American freshwater needlefish genus Potamorrhaphis, an undescribed species was found from the upper tributaries of the Orinoco in Venezuela and Co- lombia and the upper Rio Negro in Brazil. Because completion of the generic review has been delayed, I have decided to describe the new species separately. It is herein named in honor of the late Dr. James A. Peters, who was best known as an authority on Neotropical amphibians and reptiles. He directly aided my studies of the South American freshwater needlefishes Potamorrhaphis and Pseudotylosurus, and I take great pleasure in naming this species after him.
Study materials are in the collections of the Academy of Natural Sciences, Philadelphia (ANSP); the American Mu- seum of Natural History (AMNH); California Academy of Sciences, San Francisco (CAS and SU); Muséum National d'Histoire Naturelle, Paris (MNHN); INDERENA, Bogota; Museo de Biologia, Universidad Central de Venezuela, Caracas (MBUCV); and the National Museum of Natural History (USNM). This paper is one in a series on the systematics of the Synentognathi; see Collette (1966) on Belonion and Collette (in press) on Pseudotylosurus for studies on the other two genera of South American freshwater needlefishes.
Discovery of a second species of Potamorrhaphis and sepa-
5—Proc. Brot. Soc. WAsH., Vou. 87, 1974 (31)
32 Proceedings of the Biological Society of Washington
OCI 9 vv VL th OF chy [230.1 ce OVE 2 6) 8) OE Ge I HO ZEUVeee al €9 9'EE © p= LO, OGL Ss seueine) GG OES I S01 9 O0OULIQ 12MO'T
sisugupins *g
6 GILV Gene sie. VL [F79,L I Se I . Owlo,, ON G OV ee | erlvInbisey-0130N “yo Joddq 9 STV Ce SS ooouLIQ, 1eddq
isuajad *g
Seeks ee ee ee ee ee N x Vise Sie cre i OF 6G. 8c LS 9S ese Vou ce. 6G ee OF SvIGOVIOA [epNney
OGL CL ce ve oF Vesle to Gyec [F791 ct OGL le 9 SS eee ee a WMOPARLEN chee NO co GGL 6 i” 4 BO Ee SBUBITIS) GG VGL [ee he OMT) RNA |
sisuaupina ‘J
6 © 18 lege ee Oe Oe [810,L T LL T OWOT, ONY G C08 cE <I d1evInbise7-O1s9N “Ye toddq 9 o68 Weenie = a 1G ooounlg 1cddq
istajad “gq
a nn nn LLL N s cg #8 €8 68 18 08 6GL 8L LL 9L GL FL EL GL TL OL 69 89
dVIGOWIZA [VIO J, eS eS
‘sisuaupin? “g Jo suonepndod sary} pure rsiajad srydpyliowmzog JO syuNOd AvI-UlZ pUe [eIGeHoA “T AIAV],
0C6 186 € 6.L 8€ TOL9F IT @ [e101 oe SHG ih 2 6-16 6le Se uoZeULY 18MO'] SIL 68% tT If 46 9¢ G TI seueime) GG 8 LG r= C Vi s= 6 ODOULIQ, IdA\O'T sisuaupins *q Ss 6 ge Fes O=e se oe LV Sor AONE, = I GE I OWLOT, OLY = G OVE G sreinbisen-o1s0eN “YyY ioddq <x 9 COE ee eS os0ulI9Q 1eddq 3 18.1ajad * o N x 6€ 8& LE 9F st ve €€ CF Té O& 66 8G LE 9G GZS Ss svi UlJ-[eUY = = 0cG GCE eI T€ OS 88 Ge GI T [eIO.L = ages CGE J (GAL OG 8 =, uozeuly IaMO0'T S 6 Ow b & I€ SF SIT seurm ce ole 6 Wl pv ODOULIQ) IOAMO'T > SisuaUuDING * J Soest OCs Sl al 1s te. Ca [210.1 z I LE T OWOy, Or < G 08 6 ovinbdisvy-O1seN ‘yY 1oddq 9 © OV en Be oe ce ooouLIQ, reddy wiajad “J N x cy Cy ly OY GE (8 Ze 9s So 176 eS ce e 0S) Gc
SABI ULF-[eSIOG
‘penunuoy ‘[ a1aVvy,
34 Proceedings of the Biological Society of Washington
TaBLE 2. Measurements (as percent of body length) of Potamorrhaphis guianensis from the lower Orinoco and P. petersi.
P. guianensis P. petersi Character Range x N Range x N Body length (mm) 87.8 119 1OL9 23. (92770 162) MeO Head length 55.2 59.2 56.91 11. 47.8 (Sb. O lesa Snout length 37.0. 42.10 "40152 1 S39) Valor oma Postorbit length NOS pA Be 28} 94 11.1 AGO 9 Preopercle length Sto GO Olin 3 5.2) | 15.9 aaa) Orbit length AVON Olly Age) PAB 3.9 | AiG 424 Interorbit width AW 6) EN IIA BAS} AD AQY Ai529 Head depth Bee, Ga) OM WB A9 60 554 9 Head width Sele > Oo aS AT (6.30 or A0men Pectoral-fin length 10337 13S .27 11 s6r as 9.7 118 1061 9 Pelvic-fin length dh TR BAD), 8} 6.2)" 73°) Gione9 Pectoral fin to pelvin fin 47.4 52.0 4944 23 435 468 44.94 9 Pelvic fin to
caudal-fin base 46.0 51.4 48.97 23 516 542 53.18 9
ration (Collette, in press) of Pseudotylosurus microps (Ginther) from Ps. angusticeps (Gtinther) raises the total number of species of freshwater synentognaths known from South America east of the Andes to seven. As previously noted (Collette, 1966), the other three include two neotenic needle- fishes Belonion dibranchodon Collette (described from the same Rio Atabapo locality as three paratypes of P. petersi) and B. apodion Collette plus the halfbeak Hyporhamphus brederi (Fernandez-Yépez).
Potamorrhaphis petersi, new species
Figures 1-3
Diagnosis: A species of the genus Potamorrhaphis that differs from the other nominal species of the genus, P. guianensis (Schomburgk) and P. eigenmanni Miranda Ribeiro, primarily in having the posterior portion of the body greatly elongated. This is best reflected in the higher counts (Table 1): greater total number of vertebrae (77-85 vs. 64-77); caudal vertebrae (38-44 vs. 28-37); dorsal-fin rays (37-43 vs. 27-36); and anal-fin rays (32-39 vs. 24-31). Predorsal scales are more nu- merous than in the Orinoco population of P. guianensis (110-141, x 128 vs. 97-112, x 102).
Morphometrically (Table 2), this difference is shown by the distance
30
‘[izeig “uozeury
TOMOT “TQNI— ONY “WU BOT “LEG6LT WNSN ‘sisuaupins -g ‘0 “eIquIC[OD “ee OLY “WU ETT “ESSOTT dSNV ‘Stsuaunins gq -q ‘ejonzoueA ‘Oo0UTIO Joddn ‘eping ‘ij Jeou “YyySue] Apog wu FET ‘6196 HNIWY ‘edAyered ‘isuajad srydoysowpjog ‘e ‘[ ‘ol
eo)
alee aA Bal le pte a .—
Sn SAO
A New Needlefish from South America
WY NN CAN WMS
WW KK SX S MSV
36 Proceedings of the Biological Society of Washington
Fic. 2. Diagrammatic representation of pigment patterns of five specimens of Potamorrhaphis petersi. From top to bottom: USNM 210546, holotype, 162 mm BL, junction Rio Guaviare and R. Inirida; USNM 210547, paratype, 137 mm, same data; AMNH 9619, paratype, 133 mm, upper Orinoco; MBUCV-V-6132, paratype, 124 mm, Rio Casiquiare; and INDERENA G.A-P-0066, 111 mm BL, Rio Tomo.
from the pelvic-fin origin to the caudal-fin base (P:-C) being much greater than the distance from the pectoral-fin origin to the pelvic-fin origin (P:—P.) in P. petersi (PC 51.6-54.2% vs. Pi-P2 43.5-46.8% body length). These distances are about equal in P. guianensis from the lower Orinoco (P:—-C 46.0-51.4% vs. P:—P2 47.4--52.0% body length). In- creasing the P:-C distance in P. petersi also increases the body length
37
A New Needlefish from South America
‘(S}Op) sisuaupina
‘d
jo suonejndod ureyj10u pue (sieys) ssvajzad srydpy.lownjog JO UOTNGIISIq ‘¢ ‘oI
sisusunin6 @ Issojed 4 siydpyssowpjog
aN ce S3OWNOS Va Ay i » =
4 i SS SA XQ Rivinobvs Vf AEE SS \) . VOIUINI SS auyiAvN 2p) .
38 Proceedings of the Biological Society of Washington
thereby decreasing other body proportions. The type-series of P. petersi is completely separated from the Orinoco population of P. guianensis by having a smaller Pi-P. distance and is nearly completely separated in having a narrower interorbital distance and a shorter head (Table 2). Other proportions are also smaller but overlap with the Orinoco popu- lation of P. guianensis.
Potamorrhaphis petersi is also distinct in its pigment pattern. The deeper lying brown pigment that forms bars (Fig. lc) or a continuous lateral band (Fig. 1b) in P. guianensis is absent and replaced by scattered large black melanophores that appear to be more superficial. These melanophores vary in number, size and position from specimen to specimen (Fig. 2) and from side to side in a single specimen.
Types: Holotype—USNM 210546, 162 mm body length, Colombia, Laguna Coco northeast of Puerto Inirida, pool near junction of Rio Guaviare and Rio Inirida, 17 January 1972, P. Cala. Paratypes: USNM 210547 (1: 137), same data as holotype. AMNH 9619 (1: 133), Vene- zuela, Cano Pescado, about 8 km north of Esmeralda near Mt. Duida, 9 March 1929, G. H. Tate. MNHN 87-655-6 (2: 97.5-123), and USNM 210861 (1: 108), Venezuela, Amazonas, Rio Atabapo at San Fernando de Atabapo, October 1886; J. Chaffanjon. MBUCV-V-6132 (1: 124), Venezuela, Amazonas, Cafo Beripamoni, tributary of Rio Casiquiare, 29 January 1969, F. Mago L., J. Moscé, A. Machado. CAS 27587 (1: 92.7), Brazil, Rio Negro, Sao Gabriel rapids, above Camanaos, 1 February 1925, C. Ternetz.
Other specimen examined: INDERENA G.A-P-0066 (1: 111), Co- lombia, Rio Tomo near entrance into Orinoco.
Comparative material: Only the data on Orinoco Potamorrhaphis guianensis is presented here; extensive material from the other parts of the range will be reported on in a subsequent generic review. Seven series containing a total of 24 specimens (62.7-113 mm body length) were taken in the lower Orinoco in Cafto Quiribana at Caicara by C. Ternetz in April and May, 1925: SU 52684, SU 52683, CAS 28322, SU 52686, CAS 28323, SU 58818, USNM 209303. One specimen was recently collected in a western tributary of the Orinoco: ANSP 116533; 119 mm; Rio Meta, Cafo Emma at Finca El Viente south of Matuzal; 4°08’ N, 72°39’ W, 18 March 1973, J. E. Bohlke, W. Saul, and W. Smith-Vaniz.
Discussion: Discovery of P. petersi adds another species to the list of fishes common to the upper Orinoco, upper Rio Negro, and the connecting Rio Casiquiare (Mago Leccia, 1971). Potamorrhaphis is probably derived from an inshore marine needlefish similar to Strongylura and has secondarily developed a larger number of caudal vertebrae and dorsal- and anal-fin rays (Collette, 1966: 21). In its even higher counts, P. petersi seems to be a specialized derivative of the wide-spread P. guianensis. Additional collecting is needed in the upper Orinoco, Rio Tomo to the Rio Meta and in the upper Rio Negro to ascertain the
A New Needlefish from South America 39
distributional limits of P. petersi and to determine if it is sympatric with P. guianensis in the Orinoco or Amazon.
The INDERENA specimen from the Rio Tomo has the lowest counts of vertebrae and dorsal- and anal-fin rays of the small series of P. petersi. It also has traces of the usual P. guianensis pigment pattern underlying the large scattered melanophores characteristic of P. petersi. This speci- men is, therefore, not designated a paratype because of these hints of possible intergradation with downstream populations of P. guianensis. However, the recently collected ANSP specimen of P. guianensis from the Rio Meta agrees well with the 24 specimens from Caicara on the lower Orinoco.
Additional information about four collections of P. petersi is available from the following references: Chaffanjon (1889), for the Rio Atabapo specimens; La Monte (1929), for the AMNH specimen from near Mt. Duida; Bohlke (1953), for Ternetz material from the upper Orinoco and Rio Negro; and Mago Leccia (1971), for the Casiquiare specimen.
Acknowledgments: Museum material and information on_ localities was kindly provided by: Drs. Donn E. Rosen and Gareth J. Nelson and Miss Avis Rumney (AMNH); Dr. James E. Bohlke and Messrs. William Saul and William Smith-Vaniz (ANSP); Drs. William N. Eschmeyer, Warren C. Freihofer, and Tomio Iwamoto and Miss Pearl Sonoda (CAS); and Sr. Francisco Mago Leccia (MBUCV). Dr. Neal R. Foster (ANSP) put me in contact with Dr. Plutarco Cala (Universidad Nacional de Colombia, Bogota) who lent me the INDERENA specimen and graciously donated two important specimens of P. petersi to the USNM collections. Mrs. Keiko Hiratsuka Moore drew the figures except for Fig. lc which was drawn by Mrs. Mildred Carrington; Mr. George Clipper made most of the radiographs and counted vertebrae; and Drs. Daniel M. Cohen, Victor G. Springer, and Stanley H. Weitzman commented on the manuscript.
LITERATURE CITED
BOHLKE, JAMES E. 1953. Studies on fishes of the family Characidae. No. 5. A review of the genus Microschemobrycon with descrip- tions of two new species. Ann. Mag. Nat. Hist., ser. 12, 6: 841-849.
CHAFFANJON, J. 1889. L’Orénoque et la Caura. Relation de voyages exécutés en 1886 et 1887. Librairie Hachette, Paris, 351 pp.
CoLLETTE, BrucE B. 1966. Belonion, a new genus of fresh-water
needlefishes from South America. Amer. Mus. Nat. Hist.
Novitates No. 2274, 22 pp.
in press. South American needlefishes (Belonidae) of the
genus Pseudotylosurus. Zool. Meded., Leiden.
La Monte, FraNcEsca. 1929. Two new fishes from Mt. Duida, Vene- zuela. Amer. Mus. Nat. Hist. Novitates No. 373, 4 pp.
Maco Leccia, Francisco. 1971. La ictiofauna del Casiquiare. Rev. Defensa Naturaleza (Caracas) 1(4):5—10.
40 Proceedings of the Biological Society of Washington
ae
me Oe
Vol. 87, No. 6, pp. 41-48 25 ‘April 1974
PROCEEDINGS OP Une
BIOLOGICAL SOCIETY OF WASHINGTON
A NEW HEART-URCHIN, ARAEOLAMPAS ATLANTICA, NEW GENUS, NEW SPECIES, FROM THE NORTH ATLANTIC AND NOTES ON THE CLOSELY RELATED HOMOLAMPAS
By D. Kerry Serary!
New York Ocean Science Laboratory Drawer EE, Montauk, New York 11954
INTRODUCTION
In the summer of 1971 two lots of an undescribed species of Homolampas were discovered in the collections of the National Museum of Natural History, Smithsonian Institution (USNM). Unfortunately, both specimens were badly broken. An attempt was made to reconstruct the specimens, but in neither case was a complete specimen obtained. A similarly incomplete lot was found in the Museum of Comparative Zoology at Harvard University (MCZ). During 1972 three more lots of fragments were discovered in the unidentified echinoid collec- tions of the University of Miami, School of Marine and Atmospheric Sciences (UMML). A description of this new species has been withheld for three years in the unfulfilled hope of finding a complete specimen. A complete specimen may be very difficult to obtain since the species occurs in deep water and has a very fragile test.
A review of Homolampas revealed two distinct species groups. One group, including H. fragilis and H. lovenioides, has a small test (less than 35 mm TL), no peripetalous fasciole, deeply sunken areoles that form internal pouches, globiferous pedicellariae, non-crenulate primary tubercles
1 The first part of this work was done in the Department of Zoology, University of Maine at Orono.
6—Proc. Brot. Soc. WaAsH., Vou. 87, 1974 (41)
42 Proceedings of the Biological Society of Washington
which do not occur on interambulacrum 5 and three (rarely four) genital pores. The other group, including H. fulva, H. glauca, H. hastata, H. rostrata and the new species, has a large test (up to 110 mm TL, except for H. hastata which is less than 35 mm TL), peripetalous fasciole, slightly sunken areoles that form slight internal bulges, no globiferous pedi- cellariae, crenulate primary tubercles extending on to inter- ambulacrum 5 and four genital pores. On the basis of the peripetalous fasciole alone, Mortensen (1950: 265) believed that Homolampas may include two distinct genera, but as the species were poorly known, he retained the single genus. The presence of the peripetalous fasciole and the other char- acters mentioned above are believed to merit generic dis- tinction. Since the type-species is Homolampas fragilis (A. Agassiz, 1869) the group including H. fragilis and H. lovenioides is included in Homolampas, strict sense. The other group including the five remaining species comprise the new genus described herein. A key to the species of Homo- lampas s.s. and Araeolampas n. gen. is given below.
Kry TO THE SPECIES OF Homolampas s.s. AND Araeolampas N. GEN.
1. No peripetalous fasciole; no primary tubercles on interambulacrum 5; deeply sunken areoles forming internal pouches; 3 (rarely 4) genital pores; globiferous pedicellariae present; primary tubercles non-crenulate; small forms generally less than 35 mm TL 2s ABTA va Pie eo ip ee lee eee ee Homolampas s.s. 2
1’ Peripetalous fasciole present; primary tubercles on interambulacrum 5; areoles not forming internal pouches; 4 genital pores; no glo- biferous pedicellariae; primary tubercles crenulate; generally large forms up to 110 mm TL (except A. hastata, which is less than: 35mm (1s) 3 ae eee Araeolampas n. gen. 3
2. Posterior prolongation of labrum not reaching beyond second pair of adjoining ambulacral plates (tropical western Atlantic) MEN rs Paes ser Ek volte! Homolampas fragilis (A. Agassiz, 1869)
2’ Posterior prolongation of labrum reaching middle of third pair of adjoining ambulacral plates (Malay Region) CIMT AIM of Bile Meee RR OA a Homolampas lovenioides Mortensen, 1948
3. Posterior end of test forming a rostrum surrounded by subanal fasciole; periproct visible from above (Indonesia) —.------__-- Sine esl sal Wedica hd hp OR RC Araeolampas rostrata (de Meijere, 1902)
3’ Posterior end of test not forming a rostrum; periproct not visible from above .2.2.2 3s ee ee ee 4
A New Heart-Urchin from North Atlantic 43
4, Only a single primary tubercle in each series of interambulacra 1, 4 and 5; small forms less than 35 mm TL (tropical eastern
ACTIC)) pees ee Araeolampas hastata (A. Agassiz, 1898 ) 4’ Several primary tubercles in each series of interambulacra 1, 4 aval Be lava iwormane wre wo, JMC) more Ib; 5
5. Frontal notch shallow (about 2 percent TL) (Indian Ocean) _ erro Araeolampas glauca (Wood-Mason & Alcocks, 1891) HaErontalsnoteh deep (about 5 percent PI) 220. 6 6. Tridentate pedicellariae with very narrow blade (Figure 1d and e); periproct small (width plus height about 15 percent TL); pores of ambulacra double adapically (central and eastern IRACHIC) preter ti Loe Araeolampas fulva (A. Agassiz, 1879) 6’ Tridentate pedicellariae with wide blade (Figure lc); periproct large (width plus height 19 to 24 percent TL); pores of ambulacra single adapically (North Atlantic) _.. Araeolampas atlantica n. sp.
Homolampas, s.s.
Diagnosis: Small forms up to 35 mm TL; test fragile with a distinct frontal notch; ambulacral plates flush with interambulacral plates, not sunken to form petals; ambulacral pores double, placed obliquely near inner edge of plates; primary tubercles perforate, non-crenulate and present on interambulacra 1—4; primary tubercles with deeply sunken areoles that form internal pouches, as in Lovenia; primary spines long, curved, uniformly dentate; subanal fasciole present; apical system with 3 (rarely 4) genital pores; peristome anterior; labrum with long posterior prolongation reaching to second or third pair of adjoining ambulacral plates; anterior portion of sternum naked of primary spines; globiferous, tridentate, ophicephalous and_ triphyllous pedicellariae present; sphaeridia not placed in pits.
Type-species: Lissonotus fragilis A. Agassiz, 1869.
Araeolampas, new genus
Diagnosis: Small to large forms up to 110 mm TL; test fragile with distinct frontal notch; ambulacral plates flush with interambulacral plates, not sunken to form petals; ambulacral pores single or double, placed obliquely near inner edge of plates; primary tubercles perforate, crenulate, present on interambulacra 1-5; primary tubercles with slightly sunken areoles forming slight internal bulges; primary spines long, curved, unilaterally dentate; peripetalous and subanal fasciole present; apical system with 4 well-developed genital pores; peristome anterior; labrum with long posterior prolongation; anterior portion of sternum naked of primary spines; tridentate, rostrate and triphyllous pedicellariae present; sphaeridia not placed in pits.
Type-species: Homolampas fulva A. Agassiz, 1879.
Etymology: The generic name is derived from the Greek araeo, thin, and the Greek lampas, f., lamp. Gender, feminine.
44 Proceedings of the Biological Society of Washington
oosD
0.5mm
Q 5 6, ) ot 04 0 6 0 b 0 0 Wh
Fic. 1. Pedicellariae. a—c, Araeolampas atlantica n. sp., 92 mm TL, USNM 24598, holotype; d-f, A. fulvua, 108 mm TL, USNM E889. a, rostrate pedicellaria; b, small tridentate pedicellaria; c, large tridentate pedicellaria; d, large tridentate pedicellaria; e, large tridentate pedicellaria (side view of 1d); f, rostrate pedicellaria.
Araeolampas atlantica, new species
Figures la—c, 2a and b
Material Examined: 1 specimen (Holotype), 92 mm TL; off Virginia, Albatross station 2105, 37°50’ N, 73°03.8’ W; 2575 m; 6 November 1883; USNM 24598. — fragments of 1 specimen; off the Azores, Atlantis sta- tion 20, 37°50.5’ N, 26°00’ W; 2585 m; 18 August 1948; USNM E7948. — fragments of 1 specimen; west of Dry Tortugas, Gulf of Mexico; 3545 m; MCZ 2910. — 1 specimen, 82 mm TL and fragments of 2 specimens; west of Georgia, Pillsbury station P-120, 31°48’ N, 76°38’ W; 1920 m; USNM E13095. — fragments of 1 specimen; west of Gabon, Gulf of Guinea, Pillsbury station P-292, 0°12’ N, 5°11’ E; 3595 m; 23 May 1965; USNM E13096. — fragments of 1 specimen; north of Haiti, Pillsbury station P-1429, 21°19.2’ N, 73°45.5’ W; 2532 m; USNM E13097.
Description: Test large (up to 92 mm TL), fragile; aboral aspect heart- shaped (test width about 86 percent TL and test height about 37 percent
A New Heart-Urchin from North Atlantic 45
Frc. 2. a, aboral aspect of Araeolampas atlantica n. sp., 92 mm TL, USNM 24598, holotype; b, lateral aspect of A. atlantica n. sp. (holo- type); c, aboral aspect of A. fulva, 108 mm TL, USNM E889; d, lateral
aspect of A. fulva.
TL); periproct large (width plus height 19 to 24 percent TL), not visible from above; slight indentation above periproct at ambitus; pos- terior part of test truncate vertically; frontal notch deep (5 percent TL):
apical system anterior (32 percent TL from anterior ambitus) with 4
46 Proceedings of the Biological Society of Washington
TasLE 1. Morphometric characters of Araeolampas atlantica and A. fulva.*
Distance Test apex to Periproct Depth —_———__——- anterior ———_———__ frontal Species length width height ambitus width height notch A. atlantica 91.9 86.0 37.0 31.9 Wore MILE 4.9 Sik5 33.0 FB MILL 4A
A. fulva 107.9 80.2 39.9 30.7 8.4 6.9 4.8
* Test length in mm; all other measurements expressed as percent TL.
well developed genital pores; peripetalous fasciole present; primary spines restricted to interambulacral areas and to within the peripetalous fasciole except for interambulacrum 5 (Loven’s system) where they occur outside the peripetalous fasciole; areoles of primary tubercles about 2.8 percent TL: posterior part of areoles of primary tubercles about 2.8 percent TL; posterior part of areole deeply sunken, anterior part slightly sunken; areoles produce slight internal bulges, but not deep internal pouches as in Homolampas; primary tubercles perforate and crenulate; subanal fasciole present; tridentate pedicellariae with wide valves (Figure Ib and c); rostrate and triphyllous pedicellariae the same as in A. fulva; pores of ambulacra near apical system single; oral surface unknown.
Etymology: The specific name refers to occurrence in the Atlantic Ocean.
Distribution: In the western North Atlantic A. atlantica occurs from off Virginia to north of Haiti, and in the eastern North Atlantic from the Azores to off Gabon. It is known from 1920 to 3595 m.
Taxonomic Affinities: The closest relative of A. atlantica (Figure 2a and b) is A. fulva (Figure 2c and d) from the Pacific. A. fulva differs from the new species by having tridentate pedicellariae with narrow blades (Figure 1d and e), a smaller periproct and double ambulacral pores adapically. The morphometric characters of these two species are shown in Table 1. The Indian Ocean species A. glauca can be distin- guished from A. atlantica by its shallow frontal notch.
DISCUSSION
Agassiz (1881) considered many of the generic differences discussed above for Homolampas s.s. and Araeolampas n. gen. to be due to age. He assumed that large specimens of H. fragilis would acquire a peripetalous fasciole and four genital pores. Although some specimens of H. fragilis do have four genital pores, these specimens are not necessarily the largest specimens. No remnants of a peripetalous fasciole have ever been ob- served on H. fragilis while small specimens of A. fulva and A. atlantica n. sp. have both a distinct peripetalous fasciole and four well-developed
A New Heart-Urchin from North Atlantic AT
genital pores. The peripetalous fasciole is one of the more conservative fascioles. No instance has ever been cited where this fasciole was lost during ontogeny, as has been the case for some other fascioles (e.g. subanal and lateroanal). All other generic characters mentioned above are also present on small A. fulva and A. atlantica n. sp.
The large specimen identified by H. L. Clark (1941) as H. fragilis (MCZ 2910) is really A. atlantica.
ACKNOWLEDGMENTS
I thank R. H. Chesher of the Marine Research Foundation for dis- covering the first specimen of the new species and encouraging me to describe it. H. B. Fell (MCZ), D. L. Pawson (USNM) and G. L. Voss (UMML) provided me with specimens. F. J. Fell and D. L. Pawson critically read the manuscript. Partial support was provided by NSF grant GB-3532 to H. B. Fell and a Sigma Xi Grant-in-Aid of Research to D. K. Serafy.
LITERATURE CITED
Acassiz, A. 1881. Report on the scientific results of the voyage of the H.M.S. Challenger during the years 1873-76; Report on the Echinoidea. Challenger Rep. 3(1):1-321, 45 pls.
Ciark, H. L. 1941. Reports on the scientific results of the Atlantis Expeditions to the West Indies, under the joint auspices of the University of Havana and Harvard University: The echino- derms (other than holothurians). Mem. Soc. Cubana Hist. Nat. 15(1):1-154, 10 pls.
MortTENSEN, T. 1950. A Monograph of the Echinoidea. 5(1). Spat- angoida I. C. A. Reitzel, Copenhagen, 432 pp., 25 pls.
48 Proceedings of the Biological Society of Washington
Vol. 87, No. 7, pp. 49-56 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
IDENTIFICATION OF MYOTIS CALIFORNICUS AND M. LEIBII IN SOUTHWESTERN NORTH AMERICA
By MicHaEL A. BOGAN Bird and Mammal Laboratories National Museum of Natural History Washington, D.C. 20560
Within temperate North America there are, perhaps, no two bats more similar than Myotis californicus (Audubon and Bachman) and Myotis leibii (Audubon and Bachman). Over much of their respective ranges they are allopatric, and speci- mens may be identified on strictly geographic grounds. How- ever in the southwestern United States and Mexico where their ranges overlap broadly and where californicus is known to be highly variable, problems of identification arise. Most students of southwestern bats agree that the two kinds are closely related but nonetheless distinct. For example, Findley (1972) includes the two species in the “Leibii group” of the subgenus Selysius within Myotis.
Most problems with the allocation of specimens appear to center in the southwestern portion of New Mexico and ad- jacent Arizona, Texas and Mexico (Barbour and Davis, 1969; Anderson, 1972; Findley et al., in manuscript), probably due to the increased variability of californicus in this area (Bogan, 1973). The collection of good series of both kinds of bats from Hidalgo County, New Mexico, by J. D. Druecker and others from the University of New Mexico offered the chance to ascertain the status of the two bats in this area and, hopefully, to establish reliable quantitative procedures for identifying these bats, both in the field and in the museum.
METHODS AND MATERIALS
Thirty-four M. leibii and 50 M. californicus, all adults, from various localities in Hidalgo County, New Mexico, were examined. Sexes were
7—Proc. Biot. Soc. WasH., Vou. 87, 1974 (49)
50 Proceedings of the Biological Society of Washington
TaBLE 1. Traditional characters used to separate Myotis californicus and M. leibii.
Character californicus leibii Facial coloration Brown Black Ear color Brown Black Dorsal pelage Dull-tipped, not Shiny-tipped, burnished or glossy glossy or burnished Third metacarpal As long as forearm Shorter than (over 30.5 mm) forearm (less than 30.3 mm) Braincase Rounded Flattened Forehead profile Rises abruptly Rises gradually Skull size Smaller Larger
combined for this analysis, as I have found that sexual dimorphism is not marked in these two species. The fourteen characters utilized, measured as described in Bogan (1973), are: total length, tail length, hindfoot length, ear length, forearm length, tibia length, condyle-premaxillary length, condylocanine length, maxillary toothrow length, cranial breadth, cranial depth, least interorbital breadth, rostral breadth, rostral length.
The specimens were initially identified using only published characters (Table 1). The crania were examined closely to determine additional characters for identification. Bats were grouped into samples for sta- tistical analysis in two ways: first, using just two samples, Hidalgo County californicus and Hidalgo County leibii; and second, by sub- dividing the two species into samples from eastern, central and western Hidalgo County. These samples correspond to the three major mountain ranges in Hidalgo County and likewise to most of the productive collecting localities in the county (Findley and Traut, 1970).
Frequency distributions for each variable were prepared to examine normality of distributions and degree of overlap between californicus and leibii. Means and variances were then calculated for the two species. Additionally, bivariate scattergrams were prepared for most pairs of variables.
The data matrix was subjected to a two-group discriminant analysis (BMD04M ) which gives a discriminant function as well as the dis- criminant coefficients for each variable enabling the identification of new specimens. Next, a stepwise discriminant analysis (BMD07M) yielding five canonical variates and a classification matrix based on the posterior probabilities and generalized distance values was performed, allowing an estimate of the degree of phenetic overlap. Additionally, the data were subjected to a principal component analysis (BMD0O1M) and to a numerical taxonomic analysis (NTSYS, developed by F. James
Identification of Two Myotis Species 51
TABLE 2. Basic statistics for 14 variables of Myotis californicus and
M. leibii from Hidalgo County, New Mexico. The mean plus or minus
one standard deviation and the range (in parentheses) are shown for each sample.
M. californicus M. leibii Character (N = 50) (N = 84) Total length 84.52 + 4.69 89.06 + 4.16 (73.0-94.0) (80.0-99.0) Tail length 39.00 + 3.03 41.50 + 2.71 (32.0-45.0) (37.0-49.0) Hindfoot length 6.82 + 0.92 7.71 = 0.62 (5.0-8.5) (6.0-9.0) Ear length 137/22 1-03 14.70 + 0.97 (11.0-15.0) (12.0-16.0) Forearm length COAG) a= 10 33.49 + 0.93 (30.45-35.22) (31.28-35.77) Tibia length 13.98 = 0.83 13.88 = 0.60 (12.30-15.75) (13.11-15.34) Condyle-premaxillary length 12.69 + 0.27 IBS se OM (12.10-13.33) (12.73-13.88) Condylocanine length 11.98 + 0.26 12.60 + 0.26 (11.50-12.60) (11.99-13.07) Maxillary toothrow length 4.99 = 0.13 5.26 + 0.13 (4.64-5.25) (5.02-5.57) Cranial breadth GP az Oz 6.49 + 0.13 (5.95-6.65) (6.24-6.79) Cranial depth 4.58 + 0.17 4,42 + 0.13 (4.19-4.95) (4.13-4.80) Interorbital breadth 3.09 + 0.09 3.25 == 0.14 (2.90-3.28) (2.91-3.52) Rostral breadth 4.81 + 0.13 Sroot== 038 (4.51-5.10) (5.00-5.63) Rostral length Foo = 028 5.56 + 0.26 (4.32-6.00) (5.02-6.30)
Rohlf and associates). These last two analyses supported the discriminant analyses and therefore are not reported herein.
RESULTS AND DISCUSSION
An impressive array of investigators (Hall and Kelson, 1959; Cockrum, 1960; Barbour and Davis, 1969; Anderson, 1972; Armstrong, 1972; Findley et al., in manuscript) have used the same basic set of characters [first presented by Miller and Allen (1928)] to separate Myotis cali- fornicus and M. leibii (Table 1). In the Southwest, however, no single
CRANIAL DEPTH
TABLE 3. Discriminant coefficients for original in computing discriminant scores for unknowns.
Proceedings of the Biological Society of Washington
9.5
5.0
4.5
4.0
Fic. l.
M. californicus
45 4f¢ 90
9.9 ROSTRAL BREADTH
60
Relationships of two cranial measurements in Myotis cali- fornicus and M. leibii. Measurements are in millimeters.
Variable
Total length
Tail length
Hindfoot length
Ear length
Forearm length
Tibia length Condyle-premaxillary length Condylocanine length Maxillary toothrow length Cranial breadth
Cranial depth
Interorbital breadth Rostral breadth
Rostral length
variables to be used See text for details.
Coefficient
0.00508 0.00019 —0.02895 —0.01790 0.01046 0.02176 —0.18182 0.02349 —0.06011 —0.00339 0.30601 0.14769 —0.43839 0.07154
Identification of Two Myotis Species 53
californicus leibii
FREQUENCY
33 BO) TBE ae | qn fee DISCRIMINANT FUNCTION
Fic. 2. Frequency distribution of Hidalgo County samples of Myotis californicus and M. leibii on the discriminant function computed by the two-group discriminant analysis.
qualitative character provides certain identification. Rather, all such characters must be considered simultaneously in a typically time- consuming operation. The quantitative character most often cited, that of third metacarpal length, was so variable within californicus that it was not further considered as an effective means to distinguish cali- fornicus from leibii. The basic statistics and amount of character over- lap between the two species in Hidalgo County are shown in Table 2. The character showing the least amount of overlap in my analysis is rostral breadth as measured at the junction of M:i—M2; 96% of cali- fornicus not exceeding 5.0 mm, and 92% of leibii equalling or exceeding 5.2 mm.
Most of the bivariate scattergrams were useless for separating the two species. The exceptions were rostral breadth plotted against other skull parameters; the most discriminating being rostral breadth against cranial depth (Figure 1). This useful combination is a reflection of Myotis leibii having a flattened skull, and M. californicus possessing a more globose skull. I have subsequently plotted rostral breadth against cranial depth for numerous specimens of both species from elsewhere in the southwestern U. S. and Mexico with equally good separation. Only rarely do specimens overlap the 52° line separating the clusters of leibii and californicus.
The results of the bivariate scattergram of Figure 1 were extended by performing a discriminant analysis to maximize the separation of the two species of bats. The results of this analysis are presented in Figure 2, and show a clear separation between the two groups. This analysis yielded the discriminant coefficients listed in Table 3 which can be used in determining identification of unknowns. The procedure, described in
54 Proceedings of the Biological Society of Washington
: / L East -3.300 i LWest | / Least Central
ib LWest Central leibii
californicus
6) -|.800 1.200 4.200 7.200 -3.300 -.300 2.700 5.700
JL
Fic. 3. Distribution of Myotis californicus and M. leibii on the first two canonical variates computed by the stepwise discriminant analysis. Each species is represented by three samples from western, central and eastern Hidalgo County. Lines enclose all bats within a sample.
- 4.80 -6.300
Choate (1973) and elsewhere, is to multiply the value for each character by the discriminant coefficient, sum these values, and then plot the value on the discriminant function as shown in Figure 2. As Choate (1973) points out, the importance of extreme care in taking the measure- ments cannot be overstressed.
The multiple stepwise discriminant analysis using six groups (three californicus and three leibii samples) is of interest for at least two reasons. First, since the variables are entered in a stepwise fashion, the analysis enables the investigator to determine which variables possess the greatest discriminating power; and second, since the samples are sub- divided it is possible to define a five-dimensional space within which the bats are dispersed. As seen in Table 4 the variable possessing the greatest discriminatory power among the 14 original variables is rostral breadth. Figure 3 is a plot of the six samples on the first two canonical variates and clearly shows that the two species do not overlap. The first two canonical variates account for 84% and 8% of the variance, respectively, and the first four variates account for 98.7% of the total variance.
These analyses demonstrate the distinct nature of the two taxa, and thus support the opinions of other investigators, and they enable the
Identification of Two Myotis Species 5D
TaBLE 4, Rank order of variables as determined by the stepwise dis- criminant analysis. Those variables at the top of the column possess the greatest discriminatory power.
Rank Variable i Rostral breadth D Hindfoot length 3 Cranial depth 4 Ear length 3S Maxillary toothrow length 6 Total length 7 Cranial breadth 8 Tibia length 9 Rostral length 10 Condyle-premaxillary length ILL Condylocanine length 12 Interorbital breadth 13 Forearm length 14 Tail length
rapid allocation of specimens of unknown identity. Proper identification, however, requires the presence of a clean, intact skull. It is disappointing that this study did not reveal any completely reliable character facilitating positive identification of these species in the field. Field identification still requires the utilization of traditional characters coupled with con- siderable experience. The most important feature is that leibii is usually darker in color than californicus. The exceptions are found in populations of dark-colored californicus occurring at higher elevations in the southwestern U. S. and Mexico. Furthermore, californicus usually appears to be distinctly smaller and more aggressive than leibii.
Having demonstrated that the two taxa are distinctive the intriguing question is, how are they partitioning the environment in areas of sympatry? Black (1972), based on his study of bat food habits, sug- gests that one member of the pair might be a “beetle strategist” while the other may prey more heavily on moths. Husar (1973) has recently described such a situation in Myotis evotis and M. auriculus, two very similar bats with rather narrow regions of overlap in the southwest. Geluso (1972) has shown that californicus and leibii are physiologically different in urine concentrating abilities and in kidney morphology. These investigations suggest that californicus and leibii are at least as different ecologically and physiologically as they are morphologically and they may be minimizing competition through differences in diet, roosting or foraging sites, or foraging times. Such partitioning should be demon- strable through observable differences in dietary, behavioral, or physio- logical parameters.
56 Proceedings of the Biological Society of Washington
ACKNOWLEDGMENTS
Several people and institutions were instrumental in bringing this report to completion. The University of New Mexico and the Smithsonian Institution provided computer time. James S. Findley generously assisted in initial stages of identification of bats. Alfred L. Gardner, Charles O. Handley, Jr., Clyde Jones, Richard W. Thorington, Jr., and Don E. Wilson all read and commented on the manuscript. To all the above I am grateful.
LITERATURE CITED
ANDERSON, S. 1972. Mammals of Chihuahua. Bull. Amer. Mus. Nat. Hist. 148:149—410.
ARMSTRONG, D. M. 1972. Distribution of mammals of Colorado. Mono- graph, Mus. Nat. Hist., Univ. Kansas 3:1-415.
Barsour, R. W. ann W. H. Davis. 1969. Bats of North America. Univ. Press Kentucky, Lexington. pp. 285.
Buack, H. L. 1972. A north temperate bat community: structure, food habits and prey community. Unpubl. Ph.D. Dissertation, Univ. New Mexico, Albuquerque.
Bocan, M. A. 1973. Geographic variation in Myotis californicus in the southwestern United States and Mexico. Unpubl. Ph.D. Dissertation, Univ. New Mexico, Albuquerque.
CuHoateE, J. R. 1973. Identification and recent distribution of white- footed mice (Peromyscus) in New England. J. Mammal. 54: 41-49.
Cocxrum, E. L. 1960. The recent mammals of Arizona: their tax- onomy and distribution. Univ. Arizona Press, Tucson. viii + 276 pp.
FINDLEY, J. S. 1972. Phenetic relationships among bats of the genus Myotis. Systematic Zool. 21:31-52.
FINDLEY, J. S., A. H. Harris, C. J. Jones anD D. E. Witson. Mammals of New Mexico. Manuscript.
FINDLEY, J. S. and G. L. Traut. 1970. Geographic variation in Pipistrellus hesperus. J. Mammal. 51:741-765.
Getuso, K. N. 1972. Urine concentrating abilities, kidney structure and drinking water dependence of insectivorous bats. Unpubl. Ph.D. Dissertation, Univ. New Mexico, Albuquerque.
Hau, E. R. anp K. R. KEtson. 1959. The mammals of North America. Vol. I. Ronald Press, New York. xxx + 546 + 79 pp.
Husar, S. L. 1973. Behavioral character displacement: evidence of food limitation in bats. Unpubl. M.S. Thesis. Univ. New Mexico, Albuquerque.
Miter, G. S., Jr. anD G. M. ALLEN. 1928. The American bats of the genera Myotis and Pizonyx. U. S. Nat. Mus. Bull. 144:viii ++ 218 pp.
Vol. 87, No. 8, pp. 57-72 25 April 1974
PROCEEDINGS OF TEE
BIOLOGICAL SOCIETY OF WASHINGTON
AN EMENDATION OF THE GENUS TRIANNULATA GOODNIGHT, 1940, WITH THE ASSIGNMENT OF TRIANNULATA MONTANA TO CAMBARINCOLA
ELLIS 1912 (CLITELLATA: BRANCHIOBDELLIDA )
By Perry C. Hout Department of Biology and Center for Systematic Collections Virginia Polytechnic Institute and State University Blacksburg, Virginia 24061
The nominal genus Triannulata Goodnight, 1940, is com- posed of two species, T. magna Goodnight, 1940, the type- species, and T. montana Goodnight, 1940. Superficially, these large worms resemble members of the genus Cambarincola Ellis, 1912, and I thought they were species of the latter when I began this study. But magna is a representative of a dis- tinctive, and as of now, monotypic, genus, while montana is indeed a member of the genus Cambarincola. The following redefinition of the genus Triannulata, a redescription of T. magna and the reassignment and emended description of montana are offered as a part of my ongoing effort to describe and classify the North American branchiobdellid fauna.
Other than the original treatment of these species (Good- night, 1940:56-58), their possible inclusion in such compendia as Pennak (1953:300) and brief statements in Hoffman (1963:281, 295), Liang (1963:570) and Holt (1969:195), nothing else, to my knowledge, has been written about them. In former works (Holt, 1965; 1968a; 1969), I ignored Triannulata, believing its species to belong to Cambarincola.
The illustrations herein are so oriented that the anterior of all animals, or parts thereof, is to the reader’s right. Measure- ments given are approximations, roughly correct to the nearest
8—Proc. Biot. Soc. WaAsH., VoL. 87, 1974 (57)
58 Proceedings of the Biological Society of Washington
0.1 mm. The animals are large and the details shown in the illustrations are essentially free-hand sketches, based upon pro- portions established with the camera lucida. Further, the drawings are conventionalized (as in all my works on the branchiobdellids): stippling indicates glandular cells or struc- tures; line hatching, muscular structures or investments. The initials “PCH” indicate collection numbers of branchiobdellids in the collections of the VPI&SU Center for Systematic Collections.
I am grateful to Mrs. Virgie F. Holt, my wife and constant field companion, for helping in the collecting of topotypical material and other specimens treated herein; to Dr. Marian H. Pettibone, Curator, Division of Worms, Department of In- vertebrate Zoology, National Museum of Natural History, Smithsonian Institution, for allowing me to examine the holo- types of Triannulata magna and Cambarincola montanus; to Dr. Horton H. Hobbs, Jr., for supplying the identifications of the host animals and for a critical reading of the manuscript; to the National Science Foundation (grant GB-372) for financial support of my field work.
Goodnight (1940:56) presented the following diagnosis of the genus Triannulata: “With the characteristics of the sub- family; spermatheca not bifid; no accessory sperm tube; body cylindrical, not flattened; head roughly triangular in shape with protruding lips; major annuli of most segments redivided to give the appearance of three annuli per segment; this is especially evident in the median segments and moderately contracted specimens; anterior nephridia opening to the out- side through separate pores in the dorsal half of segment III.”
When this diagnosis is examined in the light of our knowl- edge of the branchiobdellids as it now exists, it is found to be inadequate.
“With the characteristics of the subfamily” means that the worms produce spermatozoa in both segments V and VI (Goodnight, 1940:27), instead of possessing testes and male funnels only in segment V. The latter is true of species of the Eurasian genus Branchiobdella Odier, 1823, but is not known for any North American genus (Holt, 1967:8). In including the statement, “no accessory sperm tube” in his diagnosis,
An Emendation of Triannulata 59
Goodnight was referring to what is now known as the prostate (Holt, 1960:63). Triannulata magna clearly does not possess a prostate; montana is characterized by a very large and dis- tinctive one. The only known genus with a bifid spermatheca is the monotypic, North American Bdellodrilus Moore, 1895. Members of the American genera Xironodrilus Ellis, 1918, and Xironogiton Ellis, 1919, are flattened, rather than terete, and are not at all closely related; the remaining 15 genera of the order, including Triannulata, are composed of animals with cylindrical bodies. The “triangular” shape of the head (“cone- shaped” would be more descriptive) with protruding lips are minor features, worthy of mention only as specific characters, and cannot be taken seriously as distinctive of a genus, since many intrageneric variations of this type are known among the branchiobdellids (cf. Hoffman’s 1963 treatment of Cambarincola). The same objection applies to the use of “redivided” major annuli as a generic character. Not all of the segments of members of the two species that Goodnight placed in Triannulata present a tripartite appearance; in those that do, the redivision takes the form, mainly ventrally, of a rather minute diminution of the diameter of the anterior annulus immediately anterior to the usual distinct sulcus that sets off the minor (posterior and shorter) annulus of a seg- ment. This tripartite appearance of some segments may simply be a consequence of the large size of the worms and is also characteristic of the anterior segments of the likewise large Cambarincola ingens Hoffman, 1963, and Stephanodrilus (= Cirrodrilus) truncatus Liang, 1963. [See Holt (1967:2-3) for a discussion of the synonymy of Cirrodrilus Pierantoni, 1905, and Stephanodrilus Pierantoni, 1906].
Goodnight’s belief that Triannulata is characterized by separate nephridiopores on the dorsum of segment III must be considered carefully; the manner of opening of the anterior nephridia is a recognized generic character. T. magna is ge- nerically distinct on the basis of features of the reproductive systems, but Goodnight appears to be mistaken in his statement that the dorsum of segment III bears two nephridiopores. The holotype is mounted with the dorsal side uppermost on the slide and I cannot find two nephridiopores; rather, middorsally,
60 Proceedings of the Biological Society of Washington
there appears to be a single pore, but the nephridia are very difficult to see in this specimen and impossible to trace un- mistakably to their opening. Immature specimens from the type locality, that I am confident on the basis of body shape and the structure of the jaws are conspecific with the type, have only one nephridiopore. In all branchiobdellids, the nephridiopores are frequently difficult to locate unambiguously in specimens mounted entire. Often, not always, their position can only be determined in mature specimens by means of serial sections. Though Goodnight (1940:8-9) mentioned the use of serial sections, nowhere does he refer to their use in the discussion of any species in his monograph or present any drawings or photographs of sections. This is noticeably true of his treatment of T. magna. Unfortunately, I was able to take only a few specimens of T. magna in the limited time I could devote to collecting in Washington and Oregon. Of these, I chose to dissect for a study of the reproductive systems all but one which is mounted entire. Yet for the reasons cited, I am convinced that Triannulata is partly characterized by a common opening of the anterior nephridia. The point is im- portant for any consideration of the phylogenetic relations of the genus. |
Part of these conclusions were anticipated by Liang (1963: 570) on the basis of an anlysis of Goodnight’s descriptions. That is, Liang was unable to distinguish between Goodnight’s description of Triannulata and Stephanodrilus (= Cirrodrilus) and placed both Triannulata and Stephanodrilus (= Magmato- drilus Holt, 1967) obscurus (Goodnight, 1940) together in the invalid genus Stephanodrilus. Liang adequately described and illustrated the reproductive systems of his Chinese worms, accepting my earlier contentions that only detailed study of the reproductive systems furnish an adequate basis for the classification of the branchiobdellids (Holt, 1949, et seq.; Hoffman, 1963.
Triannulata Goodnight, 1940
Triannulata Goodnight, 1940:56-58—Pennak, 1953:300.—Hoffman, 1963:281, 295.—Holt, 1969:195. Stephanodrilus.—Liang, 1963:570 [in part].
An Emendation of Triannulata 61
Type-species: Triannulata magna Goodnight, 1940:56-57, by original designation.
Diagnosis: Moderately large branchiobdellids (about 4.5 mm in aver- age length); 2 pairs of testes; unpaired nephridiopore on dorsum of segment III; body terete, without peristomal tentacles or dorsal pro- jections, head large, lips (peristomium) prominent; some, mostly III-V, segments superficially triannulate; spermiducal gland large, subspherical to subcubical, vasa deferentia entering entally, with small deferent lobes, without prostate or prostatic protuberance; no ejaculatory duct; bursa with large ectal spherical atrium, long muscular eversible penial sheath; spermatheca with thick outer muscular wall, internally essentially filled with tall columnar epithelial cells, lumen reduced.
Distribution and affinities: With the removal of montana from the genus, Triannulata becomes monotypic. Goodnight (1940:57) records T. magna from two localities in Washington and three in Oregon. I have material from the type-locality and one other locality in Washington. The species is probably widespread in the Cascade and Coastal Ranges of the Pacific Northwest.
The affinities of the genus must be sought among a group of pre- sumably primitive branchiobdellids among the Sathodrilus—lineage whose relationships have been discussed (Holt, 1969:195-198; 1973:35): Sathodrilus Holt, 1968b; Ceratodrilus Hall, 1914; Magmatodrilus. In addition, the Asiatic genus Caridinophila Liang, 1963, must be included as a possible relative of these genera.
On the assumption that an eversible penis and the ental entry of the vasa deferentia into the spermiducal gland are primitive features, Satho- drilus has been placed near the beginning of a lineage that culminates in such genera as Pterodrilus and Cambarincola with single anterior nephridiopores, well defined ejaculatory ducts, prostates always present, ental points of entry of the vasa deferentia into the spermiducal gland and protrusible, cone-shaped muscular penes.
The species of Sathodrilus have single anterior nephidiopores, ejacu- latory ducts that usually are short and relatively thick, prostates or prostatic protuberances may or may not be present, the vasa deferentia always enter the ental end of the spermiducal gland and the penes are eversible, though their finer structure differs among the species that are included in the genus. Sathodrilus ranges from northwestern South Carolina to southern Mexico (Holt, 1973a; 1973b).
Ceratodrilus is composed of two allopatric species from the Great Salt Lake and Snake River drainages. The anterior nephridiopore is single. A well defined, but relatively short though prominent ejaculatory duct, a prostatic protuberance, ental entry into the spermiducal gland of the vasa deferentia and an eversible penis are shared with other members of the Sathodrilus lineage. The genus is distinguished by long peristomal and dorsal tentacles and appendages (Holt, 1960).
Magmatodrilus, a monotypic genus from northern California, is similar to Ceratodrilus in the above respects, except that the places of
62 Proceedings of the Biological Society of Washington
a aa eecr
a ee
Fic. 1. Triannulata magna. a, holotype, ventral view; b, same, outline of ventral jaw; c, same, ventral view of reproductive systems.
entry of the vasa deferentia into the spermiducal gland are widely separated, the latter is long and slender without any indications of a prostatic protuberance, the penial sheath enclosing the eversible [not protrusible, contra Holt, 1967] penis is much shorter, tentacles and dorsal projections are absent (Fig. 4).
The Chinese genus Cardinophila has no spermatheca, the spermiducal gland is small and there are no vasa deferentia (the vasa efferentia enter the spermiducal gland at four separate places) and an ejaculatory duct is present (Liang, 1963: 569). Nothing is known of the penis of C. unidens.
The condensed account just given of the major features of these genera constitute convincing evidence of their close phylogenetic relationships. Their generic distinctiveness is attested by the differences mentioned in addition to those in the overall facies of the jaws, the presence (in Ceratodrilus) or absence of peristomial and dorsal body appendages and striking, but difficult to describe succinctly (see Liang, 1963, and Holt, 1960; 1967; 1968a; 1969) variations in the minor features of the reproductive systems.
An Emendation of Triannulata 63
is Ee v = ue Zz f ; \ \S Uf | i Ne
Fic. 2. Triannulata magna. Latero-dorsal view of reproductive sys- tems from a dissection: ba, bursal atrium; ps, penial sheath; sb, sperma- thecal bulb; sd, spermathecal duct; sq, spermiducal gland; vd, vas deferens.
It is obvious that Triannulata is related to this group of genera. The absence of an ejaculatory duct in Triannulata immediately separates it from the others. Or, if one wishes to consider what I have called the penial sheath of Triannulata an ejaculatory duct, the eversibility of the ejaculatory duct does so. But until more study is devoted to these struc- tures of the male reproductive system of the branchiobdellids, I cannot more precisely place the genus in the Sathodrilus lineage. I can only say that these four genera are closely related, that they appear to be phylogenetic relics and that guesses as to which is nearer in structure to the postulated ancestor of branchiobdellids with a single anterior nephridiopore are futile.
Triannulata magna Goodnight, 1940 Figures 1-3
Type-specimens: Holotype, USNM 20567 from Naches, Washington, on Pacifastacus sp.
Diagnosis: As for the genus.
Description: The worms are large, the holotype, the only mature speci- men I have seen from the type-locality, has the following dimensions: total length, 3.8 mm; greatest diameter, segment VII, 1.2 mm; head length, 0.9 mm; head diameter, 0.9 mm; diameter, segment I, 0.6 mm;
64 Proceedings of the Biological Society of Washington
Fics. 3-4. Longitudinal section of male copulatory apparatus. Fig. 3. Triannulata magna. Fig. 4. Magmatodrilus obscurus: ba, bursal atrium; ejd, ejaculatory duct; p, penis; pl, lumen of penis; ps, penial
sheath.
diameter, sucker, 0.8 mm. A specimen from Cowlitz County, Wash- ington, is somewhat larger with the following dimensions: total length, 4.9 mm; greatest diameter, 1.2 mm; head length, 1.1 mm; head diameter, 1.0 mm; diameter, segment I, 0.7 mm; diameter, sucker, 0.9 mm. The other two mature specimens from Cowlitz County that I have examined were dissected for a study of the jaws and reproductive systems, but were of comparable size.
The peristomium is divided by lateral indentations into upper and lower lips that lack emarginations or lobes, but that noticeably protrude and are less in diameter than the head. There are no detectable oral papillae. The head, execpt for the region of lessening diameter of the peristomium, has no external sulci and internally there is oa one, though prominent, pharyngeal sulcus.
The anterior annuli of the trunk segments are only very slightly
An Emendation of Triannulata 65
greater in diameter than the posterior ones, hence there are no noticeable dorsal ridges. The anterior annuli of segments I-V are subdivided, a subdivision that appears rather superficial and does not involve the longi- tudinal muscles. The anterior nephridiopore has been mentioned above.
The jaws are subequal in size, triangular in en face view, without lateral teeth, therefore, a 1/1 dental formula. They are dark brown.
The following description of the reproductive system is based upon dissected material and it is difficult or impossible to present the relative size of its components in comparison to the diameter of the animals.
The spermiducal gland is large, subspherical, almost subcubical in shape. Its ventral border is at the level of the dorsal border of the bursa and the vasa deferentia enter it at opposite angles of its ventral (ental) side. There are small, indistinct deferent lobes that are not at all prominent. A prostrate is totally absent.
The penis of the branchiobdellids is the ectal end of the ejaculatory duct. For these traditionally separately described organs in Triannulata magna, two interpretations are immediately suggested by the bursal- penial complex: (1) the ejaculatory duct is absent and the penial sheath is long and in its totality eversible as the penis; or (2) the ejaculatory duct is itself a heavily muscular, eversible organ with only its ectal portion representing the penis. In actuality, neither interpretation is satisfying. I have chosen to interpret the eversible, presumably intro- mittent portion of the male copulatory apparatus of T. magna as the penis and deny the animal an ejaculatory duct. At a deeper level of interpretation the problem is probably semantic. It is most likely that in the ancestors of the branchiobdellids a muscular tube, undiffer- entiated into penial sheath, penial and ejaculatory duct portions, was everted as the intromittent organ through a somewhat heavier muscular, but small, bursal atrium. T. magna, with an increase in the muscularity of this intromittent organ, has, then retained this postulated ancestral arrangement.
In any case, adopting the first of the interpretations above for the sake of simplicity of description, the eversible penis, comprised of what in other members of the order would be known as the ejaculatory duct plus the penis, is composed of, other than the investing peritoneum, a prominent outer layer of encircling muscles and a much thicker layer of longitudinal (in reference to the organ itself) ones. The lumen is distinct throughout and convoluted, particularly ectally.
The bursa is spherical, with a rather short outlet duct, and is composed almost entirely of the atrial portion: the penis projects outward only a short distance into what in other branchiobdellids is the penial sheath. The organ is, nonetheless, relatively large.
The spermatheca has a comparatively short ectal duct; most of the organ consists of a clavate bulb with a thick muscular investment and very tall columnar glandular cells which almost obliterate the lumen, leaving only a minute space near the median portion for the storage of spermatozoa.
66 Proceedings of the Biological Society of Washington
OV
Fic. 5. Cambarincola montanus. a, lateral view of animal from Douglas County, Oregon; b, same, lateral view of jaws; c, same, lateral view of reproductive systems.
Variation: In the limited material at my disposal, no significant varia- tions are detectable. Segment VI may be sometimes triannulate, but on the other hand it may always be so and the extra “annulus” is obscured in the holotype and other specimens I have seen.
Affinities: Discussed above as those of the genus.
Distribution: As discussed above for the genus.
Hosts: Pacifastacus |. leniusculus (Dana, 1852) and P. I. klamathensis (Stimpson, 1857).
Material examined: The holotype, several immature topotypical speci- mens (PCH 1811) taken on Pacifastacus leniusculus klamathensis from the Naches River just above its confluence with the Tieton River, Yakima County, Washington, 13 August 1964, by Perry C. and Virgie F. Holt.—4 specimens (PCH 1814) taken from the Kalami River on P. I. leniusculus about 8 miles south of Kelso, Cowlitz County, Washington, 15 August 1964, by Perry C. and Virgie F. Holt.
Cambarincola Ellis, 1912
Astacobdella Leidy, 1851:206.
Branchiobdella——Moore, 1894:427 [in part].
Bdellodrilus.—Pierantoni, 1912:24 [in part].
Cambarincola.—Ellis, 1912:481; 1919:263.—Hall, 1914:190.—Stephen- son, 1930:801.—Yamaguchi, 1932:454; 1934:189.—Goodnight, 1940: 30.—Holt and Hoffman, 1959:97.—Hoffman, 1963:271.—Hobbs, Holt and Walton, 1967:52.—Holt, 1969:197; 1973a:84; 1973b:9.
Diagnosis (modified from Hobbs, Holt and Walton, 1967:52): Body terete without specialized projections other than peristomial tentacles in some species; anterior nephridia opening through common pore on
An Emendation of Triannulata 67
dorsum of segment III; deferent ducts entering ental end of spermiducal gland; prostate and ejaculatory duct both present; penis non-eversible; bursa subpyriform to obcordate; spermatheca present, never bifid.
Type-species: Cambarincola macrodonta Ellis, 1912, by original designation.
Cambarincola montanus (Goodnight, 1940), new comb. Figure 5
Triannulata montana Goodnight, 1940:57.—Pennak, 1953:300.—Hoff- man, 1963:281, 295.—Liang, 1963:570.—Holt, 1969:195.
Type-specimens: Holotype, USNM 2056, from the Kalami River, Wash- ington, on Pacifastacus sp.
Description: Goodnight’s description of Triannulata magna is con- fined almost entirely to features of the body and jaws and omits any diagnostic reference to the reproductive systems. In the following emended description of Cambaricola montanus, I shall quote all of Goodnight’s relevant statements while adhering to my previously de- veloped format for species descriptions.
Specimens of Cambarincola montanus are large worms. Averages of 5 mature individuals from Polk County, Oregon, selected at random, have the following dimensions (ranges in parentheses): total length, 5.8 mm (4.8-6.3 mm); greatest diameter, 0.8 mm (0.6—-1.0 mm); head length, 1.0 mm (0.9-1.4 mm); head diameter, 0.7 mm (0.5-0.8 mm); diameter segment I, 0.6 mm (0.4—0.7 mm); diameter, sucker, 0.6 mm (0.5-0.8 mm). These measurements are concordant with Goodnight’s description (1940:57) of a worm 5.0 mm long.
The peristomium (lips) are provided with 4 dorsal tentacles, 2 lateral lobes on each side and 4 ventral lobes, which accords with Good- night’s (1940:57) statement “peristomium divided into twelve lobes . . . which may be extended into tentacular appendages, dorsal longer than ventral or lateral.” Experience with a variety of species, e.g., those of Ceratodrilus (Holt, 1960) and Cambarincola fallax Hoffman, 1963, enables one to distinguish between lobes and tentacles of the lips. No oral papillae are detectable. The peristomium, lateral indentations of which form the lips of the branchiobdellids, is set off from the remainder of the head by a marked narrowing in diameter, often with about 3 annular indentations. There are no other external sulci of the head and only one prominent internal (pharyngeal) sulcus. The marked narrow- ing of the peristomium and the relatively large diameter of the head in contrast to the lesser diameter of trunk segment I confer a distinct cone-shaped appearance to the head.
The anterior (major) annuli of the trunk segments are not noticeably greater in diameter than the posterior ones, there are, therefore, no dorsal ridges, but in segments III to V the major annuli are subdivided ventrally to give a triannulate appearance to these segments. The anterior nephridiopore is not prominent, but contra Goodnight (1940:57) it is
68 Proceedings of the Biological Society of Washington
a single pore located in the usual position on the dorsum of the major annulus of segment III.
The jaws are massive in appearance, but not disproportionately large, and dark brown. They are triangular in shape in en face view and sub- retangular to rounded triangular in lateral view. The usual dental formula is 1/1 with prominent blunt teeth. Younger, though large, speci- mens from the Kalami River, the type-locality, have a dental formula of 5/5, but the lateral teeth are obscure and probably wear away with age, so I cannot dispute Goodnight’s statement that the dental formula is 7/5. There may well be this much variation with age in the number of lateral teeth.
Goodnight (1940:57) has nothing further to say of any diagnostic value, remarking merely that the reproductive organs are in their normal position in segments V and VI. The male reproductive system of Cambarincola montanus furnishes, however, one of the most distinctive features of the species. In its totality it is of normal proportional size for the genus, but the worms are large and its components are often compressed underneath the gut.
The spermiducal gland, though, is relatively small, in length, about 1/3 the diameter of segment VI, usually flexed and partially obscured in whole mounts by the prostate. The vasa deferentia enter the gland at widely separated regions, that is, there is a prominent anterior deferent lobe. Otherwise, it is not unusual.
The prostate is the truly remarkable feature of the male system. Proportionately huge, it exceeds somewhat the spermiducal gland in length and is at least 144 times the diameter of the latter. Yet it is composed of only the usual single layer of columnar glandular cells which are highly vacuolated, that is, differentiated. There is no ental bulb of the prostate.
The ejaculatory duct is relatively short and thick. The bursa is somewhat more than 14 the diameter of its segment, a little more than Y% its length in breadth and subpyriform in shape. Internally, its structure is typical of that of other species of the genus, a short penial sheath enclosing a cone-shaped protrusible penis, a rather short atrial portion and relatively long bursal “outlet canal.”
The spermatheca is composed of a long ectal duct and a globose bulb. Because of the bending of the ectal duct, the total length of the spermatheca is difficult to estimate, but it is about 14 the diameter of segment V.
Variation: There is considerable variation in the size of mature animals, but all are larger than those of most members of the genus. The reproductive systems may appear to differ from one worm to another, because of the differing positions in which they lie with reference to the gut. The peristomial tentacles may vary in degree of extension, which is of no consequence; they are always distinctly tentaculate and borne on the dorsal lobes of the upper lip. More significantly, the jaws of most speci- mens appear to bear only one tooth each, with undulations along the
An Emendation of Triannulata 69
normally tooth-bearing margins, but as remarked, this is probably a function of wear and the dental formula in young animals is either 7/5 or 5/5, possibly varying between these two.
Affinities; At this stage of our knowledge of the genus Cambarincola, it is futile to speculate as to which of its species is closest to C. montanus. The differentiated prostate, the peristomal tentacles, and almost surely the large total size mark it as an advanced member of the genus. The lack of a prostatic bulb removes it from the philadelphicus section (Hoffman, 1963), yet in overall facies C. montanus is closest to mem- bers of this no longer valid section (Holt, 1973a, b) of the genus. Of the much better known eastern species of the genus, C. ingens Hoffman, 1963, rivals or exceeds C. montanus in size, but the prostate is much longer than the spermiducal gland, less in diameter, and possesses an ental bulb in the former. C. fallax Hoffman, 1963, has four dorsal peristomial tentacles and a 5/5 dental formula, but otherwise is a much smaller worm without the discordance in size of the prostate (which also has an ental bulb) and the spermiducal gland. Further comments must await a detailed study of the genus in western North America. For the present, the size of the total animal, the prominence of the tentacles of the upper lip and the size and nature of the prostate readily separate C. montanus from all of its known congeners.
Hosts: Pacifastacus leniusculus leniusculus (Dana, 1852), P. l. klamathensis (Stimpson, 1857), P. I. trowbridgii (Stimpson, 1857).
Distribution: Streams of the Coastal and Cascade Ranges of the Pacific drainage in western North America from Santa Barbara County, California, to northern Washington.
Material examined (all collected by Perry C. and Virgie F. Holt): 5 specimens (PCH 1110) taken on Pacifastacus leniusculus klamathensis from Myrtle Creek, 8.3 miles south of Tiller, on state highway 42, Douglas County, Oregon, 11 July 1960.—7 specimens (PCH 1113) taken on P. l. klamathensis from stream tributary to the Umpqua River, 12.6 miles south of junction of state highway 23 and U. S. highway 99 on U. S. 99, Douglas County, Oregon, 11 July 1960.—2 specimens (PCH 1116) taken on P. |. leniusculus from Mary’s River at Philomath, Benton County, Oregon, 12 July 1960.—2 specimens (PCH 1119) taken on P. l. klamathensis from a small tributary to the Yaquina River, 14.4 miles east of Toledo, Lincoln County, Oregon, 12 July 1960.—5 specimens (PCH 1124) taken on P. 1. klamathensis from South Yamhill River, 1.6 miles west of Valley Junction, Polk County, Oregon, 13 July 1960.—5 speci- mens (PCH 1127) taken on P. I. klamathensis from Butte Creek, on state highway 213 at Marion-Clackamas County line, Oregon, 13 July 1960.—5 specimens (PCH 1130) taken on P. 1. klamathensis from Gray’s River on U. S. Highway 830, Wahkiakum County, Washington, 14 July 1960.—One specimen (PCH 1133) taken on P. I. klamathensis from Humptulips River at Humptulips, Gray’s Harbor County, Washington, 16 July 1960.—2 specimens (PCH 1137) taken on P. I. trowbridgii from Mill Creek, 0.9 miles south of Forks on U. S. Highway 101, Clallam
70 Proceedings of the Biological Society of Washington
County, Washington, 16 July 1960.—3 specimens (PCH 1813) taken on P. |. trowbridgii from the Chehalis River at Adna, Lewis County, Wash- ington, 15 August 1964.—4 specimens (PCH 1814) taken on P. l. leniusculus from the Kalami River about 8 miles south of Kelso, Cowlitz County, Washington, 15 August 1964.
LITERATURE CITED
Evuis, Max M. 1912. A new discodrilid from Colorado. Proc. U. S.
Nat. Mus. 42(1912):481—486.
1918. Branchiobdellid worms from Michigan crayfish. Trans.
Amer. Microsc. Soc. 37:49-51.
1919. The branchiobdellid worms in the collections of the
United States National Museum, with descriptions of new
genera and new species. Proc. U.S. Nat. Mus. 55:241—265.
GooDNIGHT, CLARENCE J. 1940. The Branchiobdellidae of North American crayfishes. Illinois Biol. Monogr. 17(3):1-75.
Hatt, Maurice C. 1914. Descriptions of a new genus and a new species of the discodrilid worms. Proc. U. S. Nat. Mus. 48(2071): 187-193.
Hosss, Horton H., Jr., Penny C. HOLT AND MARGARET WALTON. 1967. The crayfishes and their epizoGtic associates of the Mountain Lake, Virginia, region. Proc. U. S. Nat. Mus. 123(3602):1-84.
HoFFMAN, RicHarp L. 1963. A revision of the North American annelid worms of the genus Cambarincola (Oligochaeta: Branchiobdel- lidae). Proc. U. S. Nat. Mus. 114(3470) :271-371.
Hott, Perry C. 1949. A comparative study of the reproductive sys- tems of Xironogiton instabilius instabilius (Moore) and Cam- barincola philadelphica (Leidy). (Annelida, Oligochaeta, Branchiobdellidae). J. Morph. 84(3) :535-572.
——. 1960. The genus Ceratodrilus Hall ( Branchiobdellidae, Oligo- chaeta ), with the description of a new species. Virginia J. Sci., ON:S: )) 11(2)2538=77,
—. 1965. The systematic position of the Branchiobdellidae (An- nelida: Clitellata). Syst. Zool. 14(1):25-32.
———. 1967. The status of the genera Branchiobdella and Steph- anodrilus in North America with description of a new genus. Proc. U.S. Nat. Mus. 124( 3631 ):1-—10.
—. 1968a. The Branchiobdellida: epizodtic annelids. The Biol- ogist 50( 2-3 ):79-94.
———. 1968b. New genera and species of branchiobdellid worms (Annelida: Clitellata). Proc. Biol. Soc. Wash. 81:291-318.
—. 1969. The relationships of the branchiobdellid fauna of the southern Appalachians. in Holt, Perry C. (ed.). The distri- butional history of the biota of the southern Appalachians. Part I: Invertebrates. Res. Div. Monogr. 1, Virginia Poly- technic Institute: 191-219.
An Emendation of Triannulata 71
—. 1973a. Epigean branchiobdellids (Annelida: Clitellata) from Florida. Proc. Biol. Soc. Wash. 86(7):79-104.
—. 1973b. A summary of the branchiobdellid (Annelida: Clitel- lata) fauna of Mesoamerica. Smithsonian Contr. Zool., 142: i-iii, 1-40.
—, anpD Ricwarp L. Horrman. 1959. An emended description of Cambarincola macrodonta Ellis, with remarks on the diag- nostic characters of the genus (Oligochaeta: Branchiobdellidae ). J. Tennessee Acad. Sci. 34(2):97—104.
Leiwy, JosEpH. 1851. Contributions to helminthology. Proc. Acad. Nat. Sci. Phila. 5:205—209.
Lianc YAN-LIN. 1963. Studies on the aquatic Oligochaeta of China. I. Descriptions of new naids and branchiobdellids. Acta Zoologica Sinica 15( 4) :560-570.
Moore, J. Percy. 1894. On some leech-like parasites of American
crayfishes. Proc. Acad. Nat. Sci. Phila., (for 1893) :419-428.
1895. The anatomy of Bdellodrilus illuminatus, an American
discodrilid. J. Morph. 10(2):497-540.
Opirer, AucusTE. 1823. Memoire sur le branchiobdelle, nouveau genre d’Annelides de la famille de Hirudinees. Mém. Soc. Hist. Nat. Paris 1:71-78.
PENNAK, RoBERT W. 1953. Fresh-water Invertebrates of the United States. The Ronald Press, New York, 769 pp.
PIERANTONI, UMBERTO. 1905. Cirrodrilus cirratus, n.g. n.s.p. Parasita
dell’ Astacus japonicus. Ann. Mus. Zool. R. Univ. Napoli,
(N.S.) 1(31):1-3.
1906. 2 nuovi discodrilidi del Giappone e della California.
Ann. Mus. R. Zool. Univ. Napoli. (N.S.) 2:11.
1912. Monografia dei Discodrillidae. Ann. Mus. Zool. R.
Univ. Napoli, (N.S.) 3(24):1-28.
STEPHENSON, J. 1930. The Oligochaeta. Oxford Univ. Press, Oxford, 978 pp.
YamacucHi, Hipeyi. 1932. A new species of Cambarincola, with re- marks on spermatic vesicles of some branchiobdellid worms. Proc. Imp. Acad. [Japan] 8(9):454—456.
—. 1934. Studies on Japanese Branchiobdellidae with some re- visions on the classification. Journ. Fac. Sci., Hokkaido Imp. Univ. ser. VI. Zool. 3(3):177-219, pls. 12, 13.
72 Proceedings of the Biological Society of Washington
Vol. 87, No. 9, pp. 73-76 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
EUXESTA MAZORCA, NEW SPECIES, ASSOCIATED WITH EARS OF MAIZE IN SOUTH AMERICA (DIPTERA, OTITIDAE)
By Georce C. STEYSKAL
Systematic Entomology Laboratory, Agr. Res. Serv., U.S. Department of Agriculture, Washington, D. C1
Specimens of a species of Euxesta have been received on a few occasions from Colombia, Ecuador, and Peru accompanied with the notation that they were in some manner associated with ears of corn (maize, Zea mays Linnaeus), although exact data concerning their relations with the plant are not avail- able. A few specimens of the species have been in the United States National Museum collection for many years determined as Euxesta obliquestriata Hendel, which was de- scribed from Brazil, Paraguay, and Bolivia. The species herein described, however, differs in several ways from Hendel’s description and figure.
Euxesta mazorea, new species Figures 1-2
Male and female. Length of wing 4-O—4.83 mm. Agrees with E. obliquestriata Hendel in basic pattern of wing, interantennal space of same color as lower parts of face (not dull black), wholly dark metallic blue-green abdomen, and largely black legs.
Head as in Fig. 2; largely yellowish, metallic greenish on frontal triangle, parafrontal plates, vertex, and occiput; middle of face and sides of clypeus infuscated to variable extent; antenna usually wholly orange, but 3rd segment sometimes a little infuscated above; front largely almost polished, narrow orbits silvery tomentose.
Thorax wholly metallic greenish, mesoscutum lightly pale grayish tomentose with indistinct more blackish arcuate postsutural transverse band, especially evident above wing bases; scutellum blackish. Angle
1 Mail address: c/o U. S. National Museum, Washington, D. C. 20560. 9—Proc. Biot. Soc. WaAsH., VoL. 87, 1974 (73)
74 Proceedings of the Biological Society of Washington
Fics. 1-2, Euxesta mazorca, new species: 1. wing; 2. lateral view, head of holotype, with anterodorsal view of front.
between line connecting posterior dorsocentrals bristles and each acro- stichal bristle 45°.
Legs largely metallic greenish; forecoxa (except basilateral oval dark spot), most of inner face of forefemur, fore- and midbasitarsi, hind- basitarsus (except more or less of tip), and distal end of forefemur tawny.
Wing as in Fig. 1; medial hyaline band attaining costa in nearly full width, impinging on only small apical part of Ist vein (R:); preapical hyaline band somewhat narrower than apical brown band; section of Ath vein crossing preapical hyaline band about equal in length to apical section in brown area; length of apical brown area in discal cell less
A New Dipteran on Maize in South America 75
than half as great on 5th vein as on 4th vein; brown area in base of 2nd costal cell squarish, not extended along costa; extension of tip of anal cell about as long as greatest width of cell.
Abdomen wholly metallic bluish green, very lightly pale tomentose.
Holotype @, allotype, and 1 @ and 3 ¢@ paratypes, ECUADOR: Quito, Estacién Experimental Santa Cataline, August, 1966, feeding on corn ear (V. Vazquez), No. 72830 in United States National Museum. Also examined: ECUADOR: 1 9, 1 sex P (head and abdomen missing ), Azogues; 3 2, 1 6, Paute [Canar Province, near Azogues], 18 June 1959, in corn buds (G. Merino); COLOMBIA: 12, 3 4, Obonuco, 16 January 1973, mazorca choclo [= green ear of corn] (B. Yanguatin); 1 ¢, Sonson, Antioquia, 16 November 1955, 2500 meters, Solanum andigenum (L. Posada); PERU: 1 sex ? (abdomen missing), Ajipampa, Cajamarca, EF] Chota mountain, 1800 meters, 29 January 1968, trap with hydrolyzed protein (B. G. Casanova C.); 1 9, Huadquifia, Cuzco [at base of Cerro Macchupicchu], 5000 meters, 30 July 1911. All specimens are in the United States National Museum collection.
The species-name is an American Spanish word for “ear of maize,” treated as Latin and used as a noun in apposition.
76 Proceedings of the Biological Society of Washington
Vol. 87, No. 10, pp. 77-80 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
A NEW SPECIES OF TRICHODECTES (MALLOPHAGA: TRICHODECTIDAE) FROM THE YELLOW-THROATED MARTEN (MARTES FLAVIGULA)
By K. C. Emerson AND RocER D. PRICE
2704 North Kensington Street, Arlington, Virginia 22207, and Department of Entomology, Fisheries, and Wildlife, University of Minnesota, St. Paul, Minnesota 55101
Dr. Howard B. Emery and his colleagues of the Arun Valley Wildlife Expedition recently collected a series of Mallophaga in Nepal on the Yellow-throated Marten. We believe that these
lice represent an undescribed species and herewith describe and illustrate it.
Trichodectes emeryi, new species
Figures 1-5
Holotype male: Total length, 1.28 mm. External morphology and chaetotaxy as shown in Figure 1. Anal region as shown in Figure 4. Genitalia as in Figure 3; genital sac large, with prominent spines; parameres unmatched in size and form, each curved outwardly.
Allotype: Total length, 1.64 mm. Female paratypes varying in length from 1.41 to 1.64 mm, reflecting differences in mounting. External morphology and chaetotaxy as shown in Figure 2, with ventral terminalia as in Figure 5.
Discussion: Trichodectes emeryi does not cloesly resemble any known species, although it does have a number of characters in common with certain other species. The head, in both sexes, resembles in shape that of T. octomaculatus Paine, 1912, found on the Raccoon, and T. pinguis Burmeister, 1838, found on the European Brown Bear. The chaetotaxy of T. emeryi consists of smaller setae on the dorsum and lateral margins of the head than T. pinguis, and the forehead is somewhat more rounded than that of T. octomaculatus. Setae on the posterior margin of the pterothorax are more prominent and numerous than T. pinguis and T. octomaculatus.
Dorsal abdominal chaetotaxy is of the type found on T. octomaculatus and T. fallax Werneck, 1948, taken from the Crab-eating Raccoon; how-
10—Proc. Brot. Soc. WasH., Vou. 87, 1974 (CT)
78 Proceedings of the Biological Society of Washington
Fics. 1-5. Trichodectes emeryi, new species. 1, dorsal-ventral view of male. 2, dorsal-ventral view of female. 3, male genitalia. 4, dorsal terminalia of male. 5, ventral terminalia of female.
ever, both sexes of T. emeryi usually have three prominent setae (less often two or four) on the posterolateral margins of segments II-VI, instead of two, and the female lacks any medial tergal setae on III- VIII, instead of having the single seta on each side as the other two species or the male of T. emeryi. Ventral abdominal chaetotaxy of
New Louse from Yellow-throated Marten fig
segments II-VI is heavier and more numerous than that of T. octo- maculatus and T. fallax.
The female genital region is distinctive, as shown in Figure 5. Male genitalia have uniquely structured parameres and sac. T. zorillae Stobbe, 1913, found on the Libyan Striped Weasel, is the only previously known species with dissimilar parameres and an armed genital sac. However, the parameres of T. emeryi are short, stout, and curved outwardly, whereas the parameres of T. zorillae are long, slender, and curved inwardly at the distal tip.
Type-host: Martes flavigula (Boddaert, 1785).
Type-material: Holotype male, allotype, and 22 paratypes collected off the type-host at Sankhuwa Sabha, Nepal, on 8 February 1973. The holotype and allotype will be deposited in the National Museum of Natural History, Smithsonian Institution.
80 Proceedings of the Biological Society of Washington
Vol. 87, No. 11, pp. 81-90 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
VANZOLINIUS, A NEW GENUS PROPOSED FOR LEPTODACTYLUS DISCODACTYLUS (AMPHIBIA, LEPTODACTYLIDAE )
By W. RonaLp HEYER Division of Amphibians and Reptiles, National Museum
of Natural History, Smithsonian Institution, Washington, D.C. 20560
The relationships of Leptodactylus discodactylus Boulenger have been obscure. In the first analysis of the relationships of discodactylus since Boulenger’s description, I (1970) placed this species in the Melanonotus group of the genus Lepto- dactylus, based primarily on external morphology. This re- lationship was questioned when certain life history data were gathered (Heyer and Bellin, 1973). In a recent analysis of the relationships of the genus Adenomera within the subfamily Leptodactylinae (Heyer, 1974), I concluded that disco- dactylus was most closely related to Lithodytes lineatus, and with a certain amount of hesitation redefined Lithodytes to include discodactylus. Since then, the karyotype has been determined for discodactylus (Heyer and Diment, in prep.) and I have been able to examine more material of Lithodytes lineatus, which has resulted in a change of one character state. With this additional information, I believe that there is now sufficient evidence to firmly establish the generic identity of discodactylus.
Previously (Heyer, 1974), primitive and derived states were defined for 50 taxonomic characters, and polarities of character states were inferred for members of the subfamily Leptodactylinae. Alternate phylogenies were constructed based on shared derived character states. One phylogeny was chosen that was thought to best represent the actual relationships
11—Proc. Brot. Soc. WasH., Vou. 87, 1974 (81)
82 Proceedings of the Biological Society of Washington
within the subfamily. The analysis showed that Adenomera, Leptodactylus, Lithodytes lineatus, and discodactylus con- stitute a tight taxonomic cluster. It is the relationships among these taxa that need to be reevaluated; this study, therefore, is limited to examining the relationships among these four taxa.
MertTHOops AND MATERIALS
The same methods and data are used as in Heyer (1974) with the following modifications: 1) If all four taxa share the same character state, that character is not used; 2) If a derived state is unique to one taxon, it is not used as it gives no information on common ancestry; 3) The units of com- parison are different, necessitating a redescription of the character states (see below). In the previous study, the unit used in the analysis was the species. In this study, as generic relationships are the focus, the unit of comparison is the genus; 4) Four characters (pterygoid-parasphenoid overlap; iliacus externus muscle; adductor longus muscle; gluteus muscle), were shown to be suspect in the original determination of direction of change of states (Heyer, 1974), and as no addi- tional evidence has been accumulated to resolve these char- acters, they are omitted in the present study.
Character state descriptions: External vocal sacs—Primitive state: No external vocal sac. State 1: No external vocal sac or indications of lateral vocal folds. State 2: No external vocal sac, indications of vocal folds, or well-developed lateral vocal sacs. The direction of change of character states is: P>1—2.
Male thumb—Primitive state: Nuptial adspersities present in form of pad. State 3: No nuptial adspersities. State 4: Either spines on thumb, or in case of Fuscus group, no adspersities (see Heyer, 1974). The directions of change of character states are: 3<-P—4.
Body glands—Primitive state: No well-defined glands. State 5: No glands and/or dorsolateral folds.
Toe disks—The dorsal toe disks of Lithodytes lineatus have dorsal scutes, similar to those of Hylodes and distinct from the longitudinal grooves in the disks of discodactylus. Primitive state: No disks. State 6: No disks and/or disks lacking dorsal grooves or scutes. State 7: Toe disks with
A New Genus of Leptodactylid Frog 83
scutes on dorsal surface. State 8: Toe disks with longi- tudinal grooves on dorsal surface. The directions of change of character states are: P—6-—>7.
\
8 Toe web—Primitive state: Toes fringed. State 9: Toes
fringed or free. State 10: Toes free. The direction of change of character states is: P—9—10.
Egg pigment—Primitive state: Melanocytes present or absent. State 11: No melanocytes.
Geniohyoideus lateralis muscle—Primitive state: No lateral flare or slip. State 12: Lateral flare or slip present.
Sternohyoideus muscle origin—Primitive state: Single or double slip from sternum. State 13: Double slip from sternum.
Sternohyoideus muscle insertion—Primitive state: Lateral edge of hyoid plate. State 14: Some fibers near midline of hyoid plate.
Gracilis minor—Primitive state: Broad. State 15: Rudi- mentary.
Frontoparietal fontanelle—Primitive state: None. State 16: Small fontanelle present or absent.
Anterior articulation of vomer—Primitive state: Vomer articulating with maxilla or premaxilla or neither. State 17: Vomer articulating with maxilla or premaxilla.
Sphenethmoid-optic foramen relationship—Primitive state: Posterior extent of sphenethmoid far from optic foramen. State 18: Posterior extent of sphenethmoid far from to bordering optic foramen.
Anterior extent of sphenethmoid—Primitive state: To mid- vomer. State 19: To mid-vomer or more anteriad.
Terminal phalanges—Primitive state: Knobbed. State 20: T-shaped.
Karyotypes—Heyer and Diment (in prep.) argued that two aspects of karyotypes yield phylogenetic information on the genera Adenomera and Leptodactylus: diploid number and presence or absence of acrocentric chromosomes. The prim- itive karyotype was argued to have a diploid number of 26 with acrocentric chromosomes.
Diploid number—Primitive state: 2N = 26 or 24. State 21:
84 Proceedings of the Biological Society of Washington
2N = 22. State 22: 2N = 18. The direction of change of states is: P2122.
Acrocentric chromosomes—Primitive state: Present. State 93: Present or absent. State 24: Absent. The direction of change of character states is: P—~23—24.
RELATIONSHIPS
The four taxa have the following advanced states: Adeno- mera—l, 3, 6, 10, 11, 12, 14, 15, 16, 18, 19, 20; Leptodactylus— 2, 4, 5, 9, 12, 16, 18, 19, 21, 23; lineatus—3, 5, 7, 10, 11, 13, 14, 17, 20, 22, 24; discodactylus—3, 8, 11, 13, 15, 17, 20, 21.
Two taxa pairs share the most derived character states (7): Adenomera-lineatus and lineatus—discodactylus. As Lepto- dactylus and discodactylus only share one derived state, the best phylogeny using the Adenomera-lineatus pair cluster is that shown in Fig. 1, A. Two alternate phylogenies are possible using the lineatus—discodactylus cluster (Fig. 1, B, C). Adeno- mera shares 6 states with Leptodactylus (Fig. 1, C) and 5 states with discodactylus (Fig. 1, C, D). The phylogenies figured (A-E) are all of the reasonable possibilities.
In evaluating which phylogenies are likely to give a truer reflection of the actual relationships among these taxa, three criteria may be applied to the phylogenies. The first criterion was used in constructing the phylogenies—maximizing the number of shared character states, or in Hennig’s terms, seek- ing sister-groups. Using this criterion, the phylogenies of Fig. 1, A, B, C are preferred over those of Fig. 1, C and D. A second criterion is the number of convergences of character states required by each phylogeny. The fewer the number, the more likely that the phylogeny is correct. The number of convergences in each of the phylogenies pictured are: Fig. 1, A and B, 13 convergences; Fig. 1, C, 11 convergences; Fig. 1, D and E, 15 convergences. A third criterion is to evaluate the ancestral clusters in terms of non-convergent character states. That is, if a character state is already convergent in
>
Fic. 1. Five possible phylogenies for Adenomera, Leptodactylus, Lithodytes lineatus, and discodactylus.
A New Genus of Leptodactylid Frog
A Adenomera Lineatus LIZ AS MGSO Seo seiliaceacoe4, 9,10,14 discodactylus 8,13,15,17;21 3,6,11,20 Leptodactylus
132,4,539,12,16,18,19,21 ,23
B lineatus discodactylus 5,7,.9,10,14,. 8,15 2252324 Adenomera WSU 62 1,9,10,12,14, ; 15,16,18,19 3,6,11,20 Leptodactylus ee Se errr nae C lineatus discodactylus Adenomera Leptodactylus 5,7,9,10,14 8,15 3,6,10,11 25455321,23 22523524 14,15,20 1,9,12,16,18,19 SLASUU SSS l/pAU all D Adenomera discodactylus 1,9,10,12,14,16,18,19 8,13,17,21 Ss 15 Lineatus 5,7;,9,10,13,14,17521 522,23,24 BAG5 U1 520) Leptodactylus 1,2,4,5,9,12,16,18,195,21 ,23 E Adenomera discodactylus Leptodactylus lineatus 1,9,10,12,14, 8,13,17,321 152,4,12,16,18,19 SRON OSs sili. 16,18,19 17,20,22,24
3,6,11,155,20 559521523
86 Proceedings of the Biological Society of Washington
a given phylogeny, its phylogenetic information content is not as great as non-convergent character states. A set of four shared derived character states that are not convergent in the phylogeny is more robust than a set of four derived states that are convergent within the total phylogeny. The phylog- enies of Fig. 1, A and B are strong, with six non-convergent states ancestral to either Adenomera and lineatus (Fig. 1, A) or lineatus and discodactylus (Fig. 1, B) and four non- convergent states ancestral to Adenomera, lineatus, and discodactylus. The phylogeny of Fig. 1, C is almost as strong as Adenomera and Leptodactylus share five non-convergent states; however, lineatus and discodactylus share but two non- convergent states. The phylogeny of Fig. 1, D has the same non-convergent states ancestral to Adenomera, discodactylus, and lineatus that are found in the phylogenies of Fig. 1, A and B. Within the cluster of these three taxa, the phylogeny of Fig. 1, D is weakest, as Adenomera and discodactylus only share one non-convergent state while each of these taxa shares two non-convergent states with lineatus (Fig. 1, A, B). The phylogeny of Fig. 1, E is weak, Adenomera and discodactylus sharing only one non-convergent state and lineatus and Lepto- dactylus sharing two non-convergent states.
The phylogenies of Fig. 1, A, B, and C are the most robust. All three have the maximum number of shared derived states for a species pair. The phylogenies of Fig. 1, A and B have the most shared non-convergent ancestral states, while that of Fig. 1, C has the least number of convergent states. The phylogenies of Fig. 1, D and E are much weaker in comparison and are not further considered.
It is necessary at this point to reiterate my concept of a genus: it should 1) be monophyletic, 2) be reasonable in size (number of species), and 3) represent a distinct adaptive com- plex if possible. Criterion 2 does not apply here, as Adeno- mera, Leptodactylus and Lithodytes combined have about 40 species. As developed elsewhere (Heyer, 1973, 1974), Adeno- mera is a wet forest genus, Leptodactylus is a savanna genus and each has evolved in response to the very different selective pressures associated with the two environments. The resolu- tion of how many genera are represented among Adenomera,
A New Genus of Leptodactylid Frog 87
Leptodactylus, discodactylus, and lineatus thus hinges on adaptive complexes, which information is not complete for discodactylus and lineatus. The available evidence is sufficient to determine the broad adaptive relationships of lineatus and discodactylus, however. No egg clutches have been found for either form, so it is not known whether either has a foam nest. In both, the eggs are non-pigmented (determined from ad- vanced ovarian eggs), which strongly suggests that the eggs are hidden in some fashion. In the case of lineatus, the eggs are large and of the same size found in Eleutherodactylus, which has direct development. The eggs are small in disco- dactylus and are probably laid at the male calling sites, which are at the edge of seepage areas in naturally occurring en- closures formed by tree roots and leaves (Heyer and Bellin, 1973). The calling sites of lineatus are not known, but collec- tions indicate that they are terrestrial and/or standing water breeders. With what is known, discodactylus can be cate- gorized as having adaptive complexes distinct from both Adenomera and Leptodactylus. Briefly, discodactylus is adapted to the slow moving stream way of life, Adenomera is adapted to the terrestrial, wet-forest way of life, and Lepto- dactylus is adapted to xeric environments.
Combining this information with the phylogenies of Fig. 1, A, B, C, the following nomenclatural decisions are possible. If Adenomera, Leptodactylus, and discodactylus are ge- nerically distinct, then lineatus would have to be a distinct genus according to the phylogeny of Fig. 1, A, or it could be combined with discodactylus in a common genus in the phylogenies of Fig. 1, B and C. Lithodytes lineatus has a number of derived states in addition to those shared with discodactylus, however, including dorsolateral folds, free toes, and a derived karyotype. These unshared derived states, together with the probability that lineatus and discodactylus differ in mode and habitat of egg and larval development convince me that lineatus and discodactylus are generically distinct.
A comment might be helpful in explaining why I think this decision is the proper one to make. In the previous study (Heyer, 1974), the evidence then available dictated that
88 Proceedings of the Biological Society of Washington
discodactylus was certainly not a Leptodactylus. To remove discodactylus from the genus Leptodactylus, a decision had to be made to include the species in the genus Lithodytes or to name a new genus for it. I chose the former course because it was nomenclaturally conservative and I hoped that more in- formation on life history and karyotypes would resolve the matter. Combining lineatus and discodactylus in a common genus was believed to be an unsatisfactory solution, however, as the two species appear very dissimilar. The karyotypic evidence further supports the generic differentiation of these two species. The relationships of discodactylus have been of concern because externally discodactylus looks like members of the Melanonotus group of the genus Leptodactylus, or if the toe fringe were removed, like members of the genus Adenomera. It is apparent that the similarities of discodactylus to certain species of Leptodactylus and Adenomera are based on shared primitive character states, however. The analysis of derived shared character states clearly demonstrates the divergent evolutionary pathways of these three taxonomic units.
TAXONOMIC CONCLUSIONS
Four genera are recognized in the Leptodactylus-complex: Adenomera, which contains five species, Leptodactylus, which contains about 35 species, Lithodytes, which contains one species, and a genus for discodactylus. Definitions of Adeno- mera and Leptodactylus may be found in Heyer (1974); a definition of Lithodytes may be found in Lynch (1971). As no generic name has been proposed for discodactylus, a new genus is described as follows:
Vanzolinius, new genus
Type species: Leptodactylus discodactylus Boulenger, 1883.
Diagnosis: Vanzolinius is unique among leptodactylid frogs in possess- ing a bony mesosternum and expanded toe disks with longitudinal grooves on the dorsal surfaces. All other genera with a bony mesosternum either do not have toe disks or, if disks are present, either do not have any dorsal modifications or have dermal scutes.
Definition: Pupil horizontal; tympanum distinct; vocal sac internal; male thumb without nuptial adspersities; body without well defined glands; toes disked with dorsal surfaces with 3-5 longitudinal grooves;
A New Genus of Leptodactylid Frog 89
tarsal fold present; metatarsal tubercles neither pronounced nor cornified; toes with lateral fringes; eggs lacking melanophores; large clutch size (> 1000 eggs); depressor mandibulae condition DFsq (large slip originating from dorsal fascia, small slip originating from squamosal area); geniohyoideus medialis continuous medially; geniohyoideus lateralis without lateral flare or slip; anterior petrohyoideus insertion on edge of hyoid; sternohyoideus origin with distinct slips from anterior meso- sternum and another from posterior meso, and/or xiphisternum; sterno- hyoideus insertion near lateral edge of hyoid; omohyoideus insertion on hyoid plate and fascia between posterolateral and posteromedial processes of hyoid; tendon of semitendinosus confluent with posterior portion of sartorius insertion on knee and tendons of gracilis minor and major passing dorsad to tendon of semitendinosus; iliacus externus extending from % to full length of iliac bone; tensor fasciae latae insertion pos- terior to iliacus externus on iliac bone; gracilis minor narrow; interior and exterior portions of the semitendinosus uniting in common tendon distally, exterior portion larger or equal to interior (smaller) portion; sartorius moderately developed; accessory head of adductor magnus without distinct tendon; adductor longus well developed; gluteus insertion on cruralis and knee; quadratojugal well developed, con- tacting maxilla; maxillary teeth present; nasals widely separated; no frontoparietal fontanelle; zygomatic ramus of squamosal just longer than, just shorter than, or equal to otic ramus; vomerine teeth present; vomers not in medial contact; vomer articulation with premaxilla and/or maxilla; posterior extent of sphenethmoid widely separated from optic foramen; sphenethmoid extending anteriorly to middle of vomers; occipital condyles widely separated; no anterior processes of hyale; alary process of hyoid somewhat narrow and stalked; posterolateral process of hyoid present; ilium with well developed dorsal crest; terminal phalanges T-shaped, expanded; mesosternum a bony style; xiphisternum entire, single; diploid chromosome number 22, one pair of acrocentric chromosomes.
Etymology: The genus, masculine in gender, is named for Dr. Paulo E. Vanzolini, in recognition of his work on the South American herpeto- fauna.
Content: Monotypic. For further details of morphology and dis- tribution, see Heyer (1970).
ACKNOWLEDGMENT
George R. Zug, National Museum of Natural History, Smithsonian Institution, critically read the manuscript.
LITERATURE CITED
Boutencer, G. A. 1883. On a collection of frogs from Yurimaguas,
Huallaga River, northern Peru. Proc. Zool. Soc. London 1883: 635-638.
90 Proceedings of the Biological Society of Washington
Heyer, W. R. 1970. Studies on frogs of the genus Leptodactylus
LYNCH,
(Amphibia, Leptodactylidae). VI. Biosystematics of the Melanonotus group. Contrib. Sci., Los Angeles County Mus. Natur. Hist. 191:1—48.
1973. Systematics of the Marmoratus group of the frog genus Leptodactylus (Amphibia, Leptodactylidae). Contrib. Sci., Los Angeles County Mus. Natur. Hist. 251:1-50.
1974. Relationships of the Marmoratus species group (Am- phibia, Leptodactylidae) within the Subfamily Leptodactylinae. Contrib. Sci., Los Angeles County Mus. Natur. Hist. 253:1—46. AND M. S. Betuin. 1973. Ecological notes on five sympatric Leptodactylus (Amphibia, Leptodactylidae) from Ecuador. Herpetologica 29:66-72.
J. D. 1971. Evolutionary relationships, osteology, and zoo- geography of leptodactyloid frogs. Univ. Kansas Mus. Natur. Hist. Misc. Publ. 53:1-238.
wna
Vol. 87, No. 12, pp. 91-94 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
ADDITIONAL NOTES ON THE GENUS GLARESIS (COLEOPTERA: SCARABAEIDAE)
By Rosert D. GorDON
Systematic Entomology Laboratory, Agricultural Research Service, USDA‘
Since the genus Glaresis was revised for the U. S. and Canada (Gordon, 1970), specimens have been examined that extend the known ranges of some species. The most interesting record is of a specimen of G. ecostata Fall collected by H. Howden at Angostura Dam, Hot Springs, South Dakota in July 1968, along with specimens of G. canadensis Brown and G. inducta Horn. The latter two species would be expected at that locality, but ecostata had previously been recorded only from California, Arizona and New Mexico. I collected another specimen of ecostata at Angostura Dam in August 1973. Speci- mens of G. phoenicis Fall have been examined from Mesquite, Nevada, and White Sands, New Mexico. G. confusa Brown was previously known only from the unique type (Gordon, 1970), but specimens from two additional localities have since been observed. The data on these specimens is as follows: Bull Creek, 5000’, 15 mi. S Hanksville, Utah, VII-6-1968, A. T. Howden; Cochise Co., Arizona, 5 mi. N Benson, VII-26-1969, A. R. Hardy, F. G. Andrews, J. W. Smith. Of 15 Glaresis specimens from Benton, Mono Co., California, sent to me by Alan Hardy, 13 proved to be G. medialis Gordon, a species previously known only from Utah (Logan and Richfield) and Washington (Toppenish). The other two specimens are G. clypeata Van Dyke and G. mendica Horn. Hardy also sent four Glaresis specimens from another locality, these being the type material of the new species described below.
1 Mail address: c/o U. S. National Museum, Washington, D. C. 20560. 12—Proc. Brot. Soc. WaAsH., Vou. 87, 1974 (91)
92 Proceedings of the Biological Society of Washington
Notes on Glaresis 93
(f\
6
Fic. 6. Glaresis arenata, new species (holotype), ventral view of male genitalia.
The antenna of species of Glaresis has been recorded in the literature as having nine segments. In my 1970 paper I accepted this without examining the antenna myself. I am indebted to Paul Ritcher, Oregon State University, for calling my attention to the fact that Glaresis has ten-segmented antennae. I have personally confirmed this by examining the antenna of five species of Glaresis.
In addition to Dr. Ritcher, I wish to thank Alan Hardy, California Department of Agriculture, for the loan of speci- mens; Henry Howden, Carleton University, Ottawa, and E. C. Becker, CNC, Ottawa, for data obtained from specimens in their care.
The Scanning Electron Microscope time for this paper was supported in part by the University of Maryland Center of Material Research, Department of Mechanical Engineering and Electron Microscope Central Facility, College Park, Maryland.
Glaresis arenata, new species
Male: Length 3.80 mm, greatest width 2.05 mm. Form elongate, widest posterior to middle of elytra. Color pale brownish yellow, head and pronotum slightly darker than elytron, reddish brown. Head smooth, extremely feebly tuberculate; anterior clypeal margin straight, distinctly
< Fics. 1-5, Glaresis arenata, new species (paratype): 1, head; 2, apex of mesotibia; 3, teeth on posterosuperior margins of hind trochanter and
femora; 4, enlarged view of teeth on posterosuperior margin of hind femora; 5, hind tibia.
94 Proceedings of the Biological Society of Washington
margined with traces of tubercles, anterior angle angulate, not toothed (Fig. 1); mandible slender, sinuate. Pronotum slightly more than half as long as wide; short, longitudinal carinae present except in impressions; surface nearly smooth, impressions feeble; lateral and posterior margins with fringe of coarse setae; a distinct fovea medially near lateral margin and a feeble, longitudinal, median impression present. Elytron smooth; interval nearly without costa, flat with tiny, widely interrupted remains of a costa present; sparsely pubescent. Mesotibia sinuate on outer margin, five spines present on outer margin from middle nearly to apex, projecting at apex (Fig. 2). Posterosuperior margin of hind trochanter with a single, large tooth near femur (Fig. 3); posterosuperior margin of hind femur with two small teeth (Figs. 3, 4). Hind tibia broadly tri- angular, a strong projection on outer margin medially, elongate tubercles irregularly scattered over surface, inner margin strongly pubescent (Fig. 5). Genitalia with apex of aedeagus rounded in ventral view (Fig. 6), curved upward in lateral view.
Female: Length 3.56 mm, greatest width 2.00 mm. Similar to male in external appearance.
Type-material: Holotype, male, sand dunes, 7 mi. SW Kelso, San Bernardino Co., California, IV-20-21-1969, fluorescent black light, M. S. Wasbauer and R. R. Pinger (USNM). Paratypes, three, same data as for holotype except one collected IV-16-17-1969 and one collected beating Larrea divaricata. Paratypes in collections of USNM, California Acad- emy of Sciences and the collection of Alan Hardy, Sacramento, Calif.
Variation: Length 3.25 to 3.80 mm, greatest width 1.78 to 2.05 mm. The number of spines on the outer margin of the mesotibia varies from four to five, and one specimen of the type series had a feeble second tooth present on the posterosuperior margin of the trochanter.
Remarks: This species goes to couplet ten in the key to species (Gordon, 1970) and would key to cartwrighti Gordon. G. arenata differs from cartwrighti in having the apex of the aedeagus broadly rounded, the elytral intervals not distinctly costate, the teeth on postero- superior margin of hind femur widely separated, the outer margin of hind tibia strongly projecting medially and the outer margin of middle tibia with spines present, distinctly projecting apically. The lack of dis- tinct elytral costae is unusual, and in this respect arenata resembles medialis Gordon and, to a lesser degree, ecostata Fall, but these two species do not resemble arenata in any other characteristics.
LITERATURE CITED
Gorpon, R. D. 1970. <A review of the genus Glaresis in the United States and Canada (Coleoptera: Scarabaeidae). Trans. Amer. Entomol. Soc. 96:499-517.
Vol. 87, No. 13, pp. 95-128 25 April 1974
PROCEEDINGS OF THE
BIOLOGICAL SOCIETY OF WASHINGTON
A REVISION OF THE GENUS BOCOA (CAESALPINIOIDEAE-SWARTZIEAE )
By RicHarp S. CowANn Smithsonian Institution, Washington, D.C. 20560
INTRODUCTION
Aublet (1775) described the genus Bocoa, with a single species, B. proudcensis, in a supplemental section of his classic work on the plants of French Guiana. When Schreber (1789- 1791) and Willdenow (1800) considered the species of what now is Swartzia, no mention was made of Bocoa, and, except for the confusion with Inocarpus and Etaballia discussed below, the genus was ignored thereafter. In fact, taxa con- generic with B. prouacensis were described subsequently as species of Swartzia by Bentham (1840) and others. Tulasne (1843) even redescribed the genus under a new name, Trischidium, based on a species published by Bentham three years earlier. It remained for Amshoff (1939) to recognize the affinities of Bocoa, although she considered it a synonym of Swartzia. My earlier study of this subtribe convinced me of the distinctness of the two genera and in my monograph of the species of Swartzia (Cowan, 1968), I excluded several as taxa belonging to Bocoa; the new combinations were not made, pending further study of specific characteristics.
If one consults Index Kewensis, Bocoa will be found listed as a synonym of Inocarpus Forster, an Old World genus of which one species is widely cultivated on islands of the Pacific Basin. Another synonym that appears under this genus is Etaballia Bentham. Although the three genera were main- tained by Bentham as separate taxa in the beginning, he was less than convinced in his 1862 paper on the subject and later in the Genera Plantarum he synonymized his genus under
13—Proc. Bron. Soc. WAsH., Vou. 87, 1974 (95)
96 Proceedings of the Biological Society of Washington
Inocarpus. Comparison of herbarium material of the species of Etaballia and Inocarpus does raise considerable doubt con- cerning their separateness and, in fact, they have been con- sidered to be congeneric by some later workers. Most re- cently, Rudd (1970) concluded they should be treated as separate genera, pending further detailed studies.
Whether or not two genera are involved, the species of Etaballia and Inocarpus are clearly papilionate legumes and have little similarity to the caesalpinioid genus Bocoa, except for the unifoliolate leaves of many of the taxa in question. The flowers of B. prouacensis, the only species Bentham accepted in the genus, are quite distinctive in being apetalous, other species of the genus having a single obovate petal; in con- trast, flowers of Inocarpus and Etaballia have five, strap-like petals and a very different androecium. The latter genera have indehiscent fruits while those of Bocoa are dehiscent in two valves along both sutures.
The characteristics separating Bocoa and Swartzia can be set out clearly in a key:
Anthers oval or oblate to oblong, dorsifixed; stamens usually many and dimorphic, the filaments several times the length of the anthers; seeds arillate; herbage not foetid; leaflets rarely alternate 200 oy Swartzia
Anthers linear to narrowly oblong, basifixed; stamens 7-30, the filaments and anthers of similar length; seeds usually exarillate; herbage often foetid in fresh state; leaflets often alternate to subopposite =... 20. DS eee Bocoa
There is greater similarity morphologically between these genera than there is between Swartzia and other genera of this complx. One is tempted to think of Bocoa as a phylo- genetic derivative of Swartzia, derived from such species as S. arborescens (Aublet) Pittier and S. auriculata Poeppig, both of which have monomorphic stamens. However, I doubt that the situation is anything so simple and I have no “hard” evidence on which even to speculate profitably with respect to the true relationship of these genera. Similarly, the order of the species in the following treatment is largely arbitrary, but generally, evolutionary loss of flower parts is viewed as derived and connation of parts more advanced than totally free mem- bers of a series.
A Revision of the Genus Bocoa O77
POLLEN MORPHOLOGY
In my monograph of Swartzia, I described pollen shapes (Cowan, 1968:6) as “globose .. . to oval, elliptic, deltoid- globose, or deltoid.” These forms were from observations of pollen in lacto-phenol aniline-blue on slides prepared by an in- experienced technician (the author). I am now convinced that deltoid-globose or deltoid grains do not occur in the genus or its near-relatives. This conclusion is based on study of the pollen of fourteen species of Swartzia, representing all the major subgeneric groupings, four species of Bocoa, two of Aldina, and one each of Zollernia, Lecointea, Candolleo- dendron, and Holocalyx. I am greatly indebted to Joan Nowicke, Julianne Piraino, and Dieter Wasshausen for the preparation of the pollen samples for study by both scanning electron and light microscopy, as well as for their consultative assistance in my analysis of the data.
I undertook this preliminary review of the pollen of the Swartzieae because of my long held suspicion that this collec- tion of genera is an unnatural assemblage which has two characters in common but whose members are not otherwise closely related—an entire calyx and an indefinite number of stamens. It seemed possible that palynology might be help- ful in sorting out the several taxa. However, this has not proven to be the case and this fact, in addition to greater familiarity with the constituent genera, leads me to believe now that there may indeed be considerable evolutionary relationship among them. Certainly pollen morphology does not refute this viewpoint, for the similarities palynologically are great.
To document this conclusion and to provide a brief summary of the diversity observed among the pollens of the limited sampling of species enumerated below, the following account and illustrations are offered. It is not intended to be a com- prehensive, analytical description of pollen morphlogy in Bocoa or of any of its relatives. Rather, it is hoped that this presentation, however incomplete, will stimulate further, more detailed palynological investigations of the caesalpinioid legumes.
98 Proceedings of the Biological Society of Washington
All species examined had pollen grains which were: monads with the longest axis ca. 18-30 wm, spheroidal to subprolate or prolate, 3-colporate, the colpi short, elongate, or syncolpate, the os mostly circular and frequently prominent, sometimes bridged by the ektexine, ektexine more or less psilate, punctate, finely or coarsely rugose, or striato-reticulate.
Based on a combination of colpus length and ektexine pattern, five pollen types can be recognized. Unfortunately, these groupings, appear to correlate with nothing else and certainly provide no new insights into the phylogeny of the Swartzieae. Also, it should be emphasized that these pollen types are not well separated and that palynological investi- gation of additional species of either Bocoa or Swartzia may reduce the already limited distinctions among these groupings:
Group I: Colpi short or elongate but not syncolpate. Type I-A: Ektexine rugose. Type I-B: Ektexine striato-reticulate, sometimes very finely so. Type I-C: Ektexine punctate. Group II: Colpi syncolpate. Type III-A: Ektexine rugose. Type II-B: Ektexine punctate.
Following is a listing of the species examined in each pollen- type, the number of the illustration in parentheses just after each name, any special notes about the morphology, citation of the voucher collection, and the source herbarium for the sheet sampled:
Comments Voucher
_Species of Type I-A: Bocoa alterna (Fig. 1) Ducke 24203 (US) B. mollis (Fig. 2) Blanchet 2774 (G) B. prouacensis Stahel 69 (G) B. racemulosa (Fig. 3) Snethlage 10058 (G) Swartzia apetala Thick-walled grains Santos et al. 24166 (US) S. aptera Many grains abnormal Maguire et al. 47056 (US) S. brachyrachis Some grains syncolpate, Ducke 17038 (US)
others not S. flaemingii (Fig. 4) Finely rugose Froes 11753 (US) S. guianensis (Fig. 5) Very coarsely rugose Boyan 67 (US)
and thick-walled
A Revision of the Genus Bocoa 99
Comments Voucher Species of Type I-B: S. amplifolia Very finely striato- Klug 3122 (US) reticulate S. panacoco (Fig. 6) Finely reticulate Cardona 1142 (US) S. cuspidata (Fig. 7) Os prominent Maguire et al. 60409 (US) S. jorori (Fig. 8) Cardenas 5211 (US) S. arborescens (Fig. 9) Very finely reticulate, Lizots.n. (US) appearing psilate by light microscopy S. leptopetala Finely striato- Wurdack et al. 41367 reticulate (US) Aldina latifolia (Fig. 10) Maguire/Wurdack 35595 (US) A. macrophylla Wurdack/ Adderley 43431 (US) Zollernia vogelii (Fig. 11) Handro 45344 (US) Cyathostegia matthewsti (Fig. 12) Ferreyra 89lla (US)
Species of Type II-A:
Swartzia benthamiana (Fig. 13) Thick-walled with Irwin 57641 (US) os prominent
Species of Type II-B:
Swartzia panamensis (Fig. 14) Punctate Standley 55317 (US) S. grandifolia Coarsely punctate Holt/Blake 673 (US)
It is interesting to note that all the species of Bocoa have the same pollen-type, in common with several Swartzia species representing disparate phylogenetic groupings into which the genus is divided. Also the species of Aldina, Cyathostegia, and Zollernia have similar pollen morphology which appar- ently is slightly different from that of all other species ex- amined. (Cyathostegia is now considered a separate genus but was earlier viewed as a part of Swartzia.)
The pollens of the three other genera in this complex do not fit into the “classification” above. The pollens of Lecointea (Ducke 1702-US) and Holocalyx (Pedersen 1422a-US) are more similar to each other than to those of any of the other taxa studied: under light microscopy the grains appear thin walled, more or less psilate, 3-colporate, the colpi not con- tinuous at the poles. On the other hand, the pollen of Can- dolleodendron (Froes 32364-US) is intermediate in several respects, just as is the gross morphology of other parts of the plant: the grains are prolate and rugose but some are syn- colpate while others (fewer) are not.
100 Proceedings of the Biological Society of Washington
The following figures are arranged by pollen-type to facilitate review of the observations made. The arrangement represents, then, an attempt to organize the data for report- ing purposes, rather than to suggest phylogenetic relationships of the taxa in each grouping or between groupings.
SYSTEMATIC TREATMENT
Bocoa Aublet
Bocoa Aublet, Pl. Guiane Fr. Suppl. 38, t. 391. Jun. 1775. Trischidium Tulasne, Ann. Sci. Nat. (Ser. 2) 20:141, t. 4. Sep. 1843.
Shrub or small to large tree up to 30 m tall and 25 cm in diameter, the wood and leaves foetid, the branchlets glabrous or strigulose to tomentulose and glabrescent, the stipules minute and more or less tri- angular or lanceolate and up to 5 mm long; leaves 1- to 9-foliolate, imparipinnate, the petioles and rachis terete or occasionally flat or canaliculate on the upper surface, leaflets alternate to opposite, the blades glabrous to tomentulose, oval, elliptic, rounded, sometimes ovate- elliptic or lanceolate-elliptic, the base rounded to acute, the apex acute to acuminate or obtuse and retuse to emarginate; inflorescences racemose, infrequently compound racemes, ramuligerous just below the leaves of the current season, or occasionally axillary, sometimes very short but up to 10 cm long, the axis glabrous, puberulous, strigulose, or tomentulose, the pedicel scars strongly raised, the bracts caducous to persistent, tri- angular to semicircular, about 1 mm long and wide; bracteoles absent; pedicels 0.5-8 mm long at anthesis, becoming nearly twice as long with mature fruits, pubescent like the inflorescence axis; buds ellipsoid to oblong-ellipsoid, more or less pubescent externally, glabrous within, opening into 3 or 4 segments, free or remaining united in basal half, reflexed to strongly revolute after anthesis; petal present or lacking, white and usually glabrous, caducous to persistent, the claw 2-6 mm long, the blade round to oblate, 4-8 mm long and 3-7 mm wide; stamens 7-30, uniform, the filaments glabrous, sometimes shortly united basally, 2-6 mm long, the anthers oblong, basifixed, 1.5-4 mm long, 0.5-0.9 mm wide, glabrous; gynoecium glabrous to densely strigulose, the stigma truncate to capitellate, sometimes obliquely so, the style about as long as the elliptic ovary, the ovules 10-14, in two rows but only the one nearest the style maturing, the gynophore basicentric, to 2 mm long; fruit oval, inflated, one-seeded, the surface usually reticulate-veined, glabrous to densely strigulose, 1-3 cm long, the seed yellow brown, tan, or black, oval, arillate in two species but usually exarillate, the funicle elongate (to 3 m) in the type species.
Type-Species: Bocoa prouacensis Aublet, Pl. Guiane Fr. (Suppl.) Soh ts Goll, U1.
A Revision of the Genus Bocoa
Be
Fic. 1. Bocoa alterna—Type I-A. a. Equatorial view. x 3500. b. Portion of surface. >< 10,000.
102 Proceedings of the Biological Society of Washington
Fic. 2. Bocoa mollis—Type I-A. a. Equatorial view, showing prom- inent os. x 3900. b. Polar view. x 3900.
A Revision of the Genus Bocoa 103
Fic. 3. Bocoa racemulosa—Type I-A. a. Polar view, showing short colpi. xX 3700. b. Equatorial view. x 3700.
104 Proceedings of the Biological Society of Washington
Fic. 4. Swartzia flaemingii var. psilonema—Type J-A. a. Polar view. X< 3000. b. Equatorial view. x 3000.
Fic. 5. Swartzia guianensis—-Type I-A. Polar view. x 3000. b. Slightly oblique equatorial view, showing very coarsely rugose ektexine. < 3000.
106 Proceedings of the Biological Society of Washington
Fic. 6. Swartzia panacoco var. cardonae—Type I-B. a/b. Equa torial views. x 3000.
A Revision of the Genus Bocoa
Fic. 7. Swartzia cuspidata—Type I-B. a. Oblique polar view, show- ing striato-reticulate ektexine. > 3000. b. Equatorial view. x 3000.
108 Proceedings of the Biological Society of Washington
Fic. 8. Swartzia jorori—Type I-B. a. Oblique equatorial view. < 3000. b. Portion of surface of ektexine, showing relatively narrow, interwoven muri. xX 10,000.
A Revision of the Genus Bocoa 109
Fic. 9. Swartzia arborescens—Type I-B. a. Equatorial view, showing prominent os and discrete orbicules. > 3600. b. Enlargement of mid- section of colpus. x ca. 10,000.
110 Proceedings of the Biological Society of Washington
Fic. 10. Aldina latifolia—Type I-C. a. Equatorial view. x 3000. b. Oblique polar view. x 3000.
A Revision of the Genus Bocoa feet:
Fic. 11. Zollernia vogelii—Type I-C. a. Polar view. Xx 3000. b. Slightly oblique equatorial view. x 3000.
112 Proceedings of the Biological Society of Washington
Fic. 12. Cyathostegia matthewsii—Type I-C. a. Slightly oblique equatorial view. >< 3000. b. Equatorial view. x 3000.
A Revision of the Genus Boco
Fic. 13. Swartzia benthamiana—Type I-A. a. Polar view, showing syncolpate condition. x 3000. b. Equatorial view. > 3000.
114 Proceedings of the Biological Society of Washington
Fic. 14. Swartzia panamensis—Type II-B. a. Polar view, showing syncolpate condition. x 3000. b. Oblique equatorial view. >< 3000.
A Revision of the Genus Bocoa 115
KEY TO THE SPECIES
im leeaveseunitoliolate or titoholate) =. 2 2} eavesiwithi more than 13 leaflets 22208 ee 4 2. Flowers lacking a petal; leaf rachis and petioles about equal in length or leaves unifoliolate; seeds obviously arillate; large trees
WARIO RINOe Call emcee ee ee ee 5 Flowers with a petal; leaf rachis several times as long as the petioles; seeds exarillate; shrub or small tree 3. B. racemulosa
3. Leaves trifoliolate, the stipules deciduous, lanceolate, 3-4.5 mm long; inflorescence tomentulose; stamens 25-30; funicle not GOING AUC Clem Mem eerie Sek se eet ln 4. B. viridiflora
Leaves unifoliolate, the stipules persistent, broadly triangular, 1.5 mm or less in length; inflorescence glabrous; stamens 7-10; HUMIC oWIb Oo TN MONG fee Ne 5. B. prouacensis
4. Leaflets alternate, the apices acute to acuminate, about 7-13 cm
JMS? ei Ni a ec ere oe 5 Leaflets opposite to subopposite, the apices obtuse to rounded, usually at least some emarginate or retuse, 0.5-5 cm long 6
5. Petal deciduous or caducous; stamens 13-15; stigma obliquely capitellate; inflorescence 30-50 mm long; calyx segments free throughoutcdiem length 28s 1. B. alterna
Petal persistent even with old flowers; stamens 6-11; stigma trun- cate; inflorescences 3-10 mm long; calyx segments not separating irom oaSaleemotsleme tines. 200 Neel ee Bee 2. B. limae
6. Branchlets, lower surface of leaflets, and inflorescence axes puber- ulous to tomentulose; leaflets 35-55 mm long, the lower ones oval to elliptic; inflorescence axes 9-20 mm long; calyx segments be- coming strongly revolute; ovary and fruit glabrous _ 6. B. mollis
Branchlets, lower surface of leaflets, and inflorescence axes minutely strigulose, glabrous or glabrescent; leaflets 3-22 mm long, the lower ones round; inflorescence axes 2—2.5 mm long; calyx seg- ments reflexed; ovary and fruit strigulose __________ 7. B. decipiens
1. Bocoa alterna (Bentham) Cowan, comb. nov.
Swartzia alterna Bentham, Hook. Jour. Bot. 2:89. 1840.
Description: Shrub or small tree 3-8 m tall, the trunk slender, to 10 cm in diameter, the bark when bruised with a foetid odor, the branchlets puberulous to strigulose, sometimes glabrescent, the stipules usually per- sistent, more or less triangular 0.3-0.4 mm long, strigulose externally; petioles terete to subterete, (5—)12-18(-25) mm long, more or less puberulous, sometimes glabrescent, the rachis (3—)6—8(—11.5) cm long, terete to subterete, more or less puberulous; leaflets (3—)5—7, subopposite to alternate, always with one terminal on the rachis, the petiolules (1—)2- 4(-6) mm long, usually sparingly puberulous, the blades mostly elliptic to ovate-elliptic, sometimes narrowly elliptic, lanceolate-elliptic, or ovate (4.5—)7—12(-20.5) cm long, (2.8—)3.5-6(-7.5) cm wide, the base
116 Proceedings of the Biological Society of Washington
cuneate and acute or rounded but finally acute, the apex acute to acuminate with the tip acute to obtuse, usually glabrous except pu- berulous on the costa above, occasionally soft-puberulous on the undersurfaces generally, often somewhat lucid on one or both surfaces, the venation prominulous on both surfaces or subobscure above, the costa impressed on the upper surface, salient beneath; inflorescences (1—)3-5 (—8) em long, racemose or sparsely branched panicles of racemes, usually several arising together at the same node below the leaves of the current season but sometimes also axillary, the axis strigulose or puberulous, the flower scars prominently raised, the bracts about triangular and 1 mm long and wide, persistent or tardily deciduous, the pedicels 4-7 mm long in flower, 6-13 mm long in fruit, sparingly strigulose or puberulous, the buds 5-6 mm long, 3 mm in diameter, ellipsoid, sparsely strigulose except for the tufted apex; calyx segments 3 or rarely 4, free to the base, revolute tightly at anthesis, glabrous on the inner surfaces; petal one, white, glabrous, obovate to oblanceolate, the claw (1—)3-6 mm long, the blade 5-8 mm long, 3-5 mm wide; stamens (12—)13-15(-—18), the filaments filiform, 3.5-5 mm long, usually free but sometimes shortly joined basally, the anthers narrowly oblong, 3-4 mm long, 0.7—0.8 mm wide; gynoecium glabrous or occasionally with a few hairs on the ovary and rarely puberulous, the stigma obliquely capitellate, the style 3-5.5 mm long, the ovary oblong, 1.5-4.5 mm long, 1-2.5 mm wide, the gynophore 0.8-2 mm long, ovules 14-16 in each ovary but only one (usually nearest the style) maturing; fruits dehiscent, usually glabrous, rarely puberulous, more or less reticulate-veined on the surface, oval to round in outline, 1-2 cm long, 1-1.5 cm wide, the seed black, oval to oblong-oval, 8-10 mm long, 5—5.5 mm in diameter, exarillate.
Type-Collection: 1. Riedel 1 (holotype K), “Barra do Rio Negro in Brasilia.” The specimens cited below as “Riedel s.n.” are from the type locality and may be isotypes.
Distribution: Guyana, Amazonian Brazil, and western Peru in dense primary or open, dry secondary forest on sandy or clay soils above annual flooding by the rivers in the lowlands and up to 800 m elevation.
Discussion: Since there are small local morphological differences among the collections representing the species sensu lato, the following are cited under each of several unnamed “phases” I have distinguished. They are not worthy of formal recognition but should not be passed over without notice.
Key to PHASES OF B. ALTERNA
1. Leaflets soft-pubescent on the lower surfaces eR Gn sence | nc A Ne “Pubescent Leaflet Phase”
Leaflets glabrous or minutely puberulous on costa 2 2 Eiriitss pubescent mamas. aie nna enanne nna “Pubescent Fruit Phase” Eruits glabrous. 2... ee eee 3
3. Fruits about 10 mm long on pedicels 4-5 mm long —_------------------ . “Small Fruit Phase”
A Revision of the Genus Bocoa M7
MNES OS ..F NY EE ee ut “Acuminate Leaflet Phase” Leaflets acute to acuminate, elliptic, ovate-elliptic, or lanceolate- GIONS = 2 a Se SE eek eee a et “Glabrous Phase”
(1) The “Glabrous Phase” is glabrous in most parts, particularly the leaflets and the gynoecium. Collections representing this phase are: GUYANA: basin of Essequibo R., near mouth of Blackwater Creek, December 1937, A. C. Smith 2828 (A, K, NY); Wawitau, Kanuku Mtns., September 1948, Wilson-Browne 176 (Record No. 5669 (K, NY). PERU: Depto. Junin, Puerto Bermudez, July 1929, Killip & Smith 26502 (US). BRAZIL-AMAZONAS-vicinity of Manaus: Igarapé do Binda, April 1956, Coelho (INPA No.) 3740 (US); Barcelos, September 1962, Duarte 6978 (US); Flores, January 1941, Ducke 371 (MADw); Cachoeira Grande, April 1937, Ducke 454 (A,F,LAN,MG,MO,NY,US); Cachoeira below Taruma, 21 September 1956, Francisco & Luiz 4188 (IAN); Estrada do Mundu, February 1945, Froes 20465 (IAN,K,NY,US); Pai Raimundo, Rio Demeni, Froes & Addison 29016 (IAN); Rio Taruma, Aug. 1949, Froes 24946 (IAN); Flores, March 1924, Kuhlmann 96 (P, U) and 1607 (RB,US); Cachoeira baixa de Taruma, September 1956, Mello & Coelho (INPA No.) 4188 (US); Kms 65-70 da Rodovia, Manaus-Itacoatiara, 22 Oct. 1963, Oliviera 2742 (IAN); Estrada do Aleixo, March 1947,
ires 213 (US); Road Igarapé Leaé 25 km north of Manaus, November
1966, Prance et al. 3137 (U,US); estrada Manaus-Itacoatiara km 70, October 1960, Rodrigues & Coelho 1877 (US); Estrada do Taruma, January 1962, Rodrigues & Chagas 4112 (US); Igarapé do Buiao, No- vember 1962, Rodrigues & Chagas 4723 (US); Km 22 da estrada Manaus-Caracaroi, November 1962, Rodrigues et al. 4811 (US); Manaus, Schwacke 3942 (RB); between Santarem and Barra do Rio Negro along the Amazon River, October 1850, Spruce s.n. (BM,G,K,NY,RB); Paruna-miri dos Ramos, Oct. 1850, Spruce 1106 (K,P). PERNAMBUCO: Recife, Dois Irmaos, Feb. 1948, Ducke 2116 (IAN).
(2) The “Pubescent Leaflet Phase” has leaflets that are softly pubes- cent on the undersurfaces, but like the typical phase in other respects. All the collections are from BRAZIL-AmMazonas-Manaus: B.A.M., BR 17, Km 9, December 1955, Dionisio 3076 (MG); Flores, October 1936, Ducke 296 (A,F,IAN,K,MG,MO,NY,US); between the city and Flores, July 1929, Ducke (HJBR No.) 23342 (G,K,RB,US); near the city, August 1931, Ducke (HJBR No.) 24203 (G,K,P,RB,U,US); road to Aleixo, August 1936, Krukoff 7944 (A,BM,G,K,MO,P,U,US); Margem do Igarapé do Buiao, July 1956, Luis 13979 (IAN); Sept. 1945, Pires & Black 930 (IAN); Barra do Rio Negro, Riedel s.n. (A,G,K,P); Igarapé do Santa Maria, June 1962, Rodriguez and Chagas 4473 (US), cultivated at Rio de Janeiro, Dec. 1878, Glaziou 9767 (K,P,U).
(3) The “Acuminate Leaflet Phase” differs in having relatively narrower leaflets that have long-acuminate apices and it occurs in the Kanuku Mountains of Southwestern Guyana: March—April 1938, A. C. Smith 3147 (F,IAN,K,NY,P,U,US) and 3575 (G,F,IAN,K,MADw,MO,
118 Proceedings of the Biological Society of Washington
NY,U,US). One additional Guyana collection without locality should be assigned here: Robt. Schomburgk 293 (Rich. Schomburgk 585) (K,P).
(4) The “Pubescent Fruit Phase” is separable only by the puberulous fruits and is represented by a single collection from Guyana: basin of Essequibo River near mouth of Onoro Creek, December 1937, A. C. Smith 2772 (A,F,G,K,MADw,MO,NY,U,US).
(5) The “Small Fruit Phase” has noticeably smaller fruits, somewhat shorter pedicels, and longer inflorescences: Serra do Navio, Terr. do Amapa, Rio Amapari, Brazil, November 1954, Cowan 38157 (F,NY).
(6) The “Unifoliolate Phase” is identical with the “typical phase” except that the leaves are uniformly unifoliolate. The only two collections are from the Brazil-Guyana boundary in the Akarai Mountains, at 600— 800 m elevation between the drainages of the Rio Mapuera and Shodikar Creek: A. C. Smith 2991 (A,F,C.K,MADw,MO,NY,U,US) and 2996 (A,F,G,K,MO,NY,P,U,US).
The closest relationship of B. alterna is with B. limae, which shows once again a familiar distributional pattern (in the legumes at least )— a wide-ranging species with a disjunct, very similar, related species in the coastal forest of southeastern Brazil. The two taxa differ in the length of the inflorescence, persistence of the petal, and the post-anthesis union of the calyx-segments.
2. Bocoa limae Cowan, sp. nov.
Description: Frutex 1.5-2 m altus, trunco ca. 3 cm diametro, ramulis strigulosis, stipulis caducis, non visis; petioli 3-4(-6) mm longi, sparse strigulosi, subteretes, rhachibus 4.5-8 cm longis, sparse strigulosis, sub- teretibus; foliola 4-7, alterna, imparipinnata, petiolulis 1-2 mm _ longis, laminis superioribus amplioribus, 6—20 cm longis, 2.5—7.5 cm latis, laminis inferioribus minoribus, 3-5.5 cm longis, 1.5-3 cm latis, laminis omnibus ellipticis, ad basim angustatis acutis vel tandem obtusis, ad apicem argute vel obtuse acutis, sparse strigulosis infra sed mox glabrescentibus, lucidis, manifeste venulosis in ambobus paginis; inflorescentiae ramuligerae vel axillares, axe 3-5(-10) mm longo, striguloso, bracteis deciduis, tri- angularibus, strigulosis extus; pedicelli 4—5.5 mm longi, strigulosi, ala- bastris oblongis, sparse strigulosis; calycis segmenta 3 vel 4, revoluta, ad basim incomplete fissa; petalum unum, album, persistens, glabrum, unguiculo 3.5-4 mm longo, lamina rotundata sed attenuata ad basim, 5-7 mm diametro; stamina 6-11, filamentis 3-5 mm longis, antheris anguste oblongis, ca. 4 mm longis et 1 mm latis; gynoecium glabrum, stigmate truncato, stylo ca. 5 mm longo, ovario elliptico, ca. 3.5 mm longis et 1.8 mm latis, gynophoro ca. 1 mm longo; fructus ovalis, glaber, reticulatus, ca. 18 mm longus et 12 mm latus, rostellatus, semine uno, nigro, exarillato, ovali ca. 10 mm longo et 7 mm lato.
Type-Collection: A. Lima 65-4270 (holotype US, sheet no. 2639742, isotype IPA), “Bahia. Salvador. Areias em torno da Lagéa do Abaeté,” Brazil, 26 January 1965.
Distribution: Coastal brush-forest from Pernambuco to Bahia, Brazil,
A Revision of the Genus Bocoa IL,
on sandy soils. BRAZIL-PERNAMBUCO: Recife, woods near Dois Irmaos, 29 October 1949, Lima 49-369 (IPA); along road to Aldeia, 17 March 1952, Lima 52-1007 (IPA, US); mata de Dois Irmaos, 2 June 1966, Soares 2185 (US).
Vernacular Name: “Feijao brabo” (Soares 2185).
Discussion and Etymology: There is a great resemblance of B. limae (named for the collector of most of the known material of the new taxon) to B. alterna but close examination reveals a number of differences in addition to the obvious geographic disjunction. Unlike its nearest relative, the new species has a very short inflorescence, at times the flowers almost seeming to arise from the branchlets directly; the calyx segments do not open completely, remaining partially united basally; the petal is persistent, even with quite old flowers; the stamens are fewer, and the stigma is truncate; vegetatively, the leaflets are obviously venulose and the costa is salient on both surfaces.
3. Bocoa racemulosa (Huber) Cowan, comb. nov.
Swartzia racemulosa Huber, Bol. Mus. Goeldi 5:395. 1909.
Description: Shrub or small tree, the branchlets sparingly strigulose minutely, glabrescent, the stipules not seen, the petioles 2-10 mm long, sparingly strigulose but glabrescent soon, rachis 2.5-6 cm long, glabrous, subterete; leaflets 1-3, the lateral ones smaller (6-11 cm long, 3-7 cm wide) ovate, with petiolules 1-3 mm long, the terminal leaflet 12—23.5 cm long, 5-12 cm wide, the petiolules 3-4 mm long, base of leaflets rounded-obtuse (some lateral ones) to rounded and acute, or simply acute, the apex sharply to bluntly acute, sometimes emarginate, glabrous, venation prominulous; inflorescences 2.5-3 cm long, ramuligerous, racemose, minutely strigulose, the bracts persistent, ovate-triangular, about 1 mm long, minutely strigulose externally; pedicels about 2.5 mm long in flower and 4 mm long with fruit, minutely strigulose sparingly, the calyx minutely strigulose externally, opening only about half of length in 2-4 segments, ca. 3.5-4 mm long, segments erect or somewhat reflexed, the calyx usually falling in one piece; petal white, glabrous, obovate, 7 mm long (fide Huber), cuneate-unguiculate; stamens 12—14, the filaments 2-3 mm long, the anthers oblong, 2.5-2.8 mm long, 0.5 mm wide; gynoecium glabrous, the stigma capitellate, the style 3.5 mm long, the ovary elliptic, ca. 2 mm long and 0.8 mm wide, the gynophore ca. 1 mm long; fruit ca. oval in outline and 1.5 cm long, 1 cm wide, reticulate- veined on surface, glabrous, the seed exarillate, black, oval-oblong.
Type-Collection: A. Ducke 7870 (holotype presumably in Brazil but not at Museu Goeldi in Belém (fide the curator), isotypes F,G), “in silvis flumenis Trombetas ad vicum Oriximina, Para,” Brazil, 8 December 1906.
Distribution: Lowland forest of Para in east-central Brazil, near the junction of the Amazon and the Rio Trombetas: Sao Jorge, Municipio de Faro, November 1950, Black & Ledoux 50-10669 (IAN,US); Orixi-
120 Proceedings of the Biological Society of Washington
mina, 23 November 1907, Ducke 8865 (BM,G,MG); Rio Tapajoz, Mangabal, February 1917, Ducke 11184 (HAMP No. 16755) (BM,G, MG,P,RB,U,US); S. Luiz do Rio Tapajoz, December 1915, Ducke 11185 (HAMP No. 15856) (BM,MG,RB); Obidos, November 1919 (fl.), June 1926 (frt.), Ducke 11186 (G,K,P,RB,U,US); Rio Tapajoz, Villa Braga, November 1908, Snethlage 10058 (BM,G,MG).
Discussion: While clearly related to B. alterna, this species is easily distinguished by its dimorphic leaflets. The pair of lateral ones is ovate and about half the length of the elliptic terminal one. In addition, the calyx segments in B. racemulosa only separate about halfway and they are more or less reflexed, in contrast to the totally free, retrorsely revolute segments in B. alterna.
4. Bocoa viridiflora (Ducke) Cowan, comb. nov.
Swartzia viridiflora Ducke, Arch. Inst. Biol. Veg. 2:44. 1935.
Description: Tree to 30 m tall and 43 cm diameter, with glabrescent, lucid branchlets, strigulose at very early stages, the stipules deciduous, lanceolate, 3-4.5 mm long, 1-1.5 mm wide, strigulose externally at first but soon glabrescent; leaves glabrous, trifoliolate (unijugate and im- paripinnate ), the petioles 1-4 cm long, terete, the rachis 2—4.3 cm long, terete; lateral leaflets opposite, occasionally subopposite, petiolules 3-6 mm long, transversely rugose, the blades subcoriaceous, lucid, 7-15 cm long, 3-6.5 cm wide (the terminal leaflet larger), elliptic, the base acute, the apex acuminate to bluntly acute, costa and 3-5 pairs of primary veins more or less salient on both surfaces; inflorescences 5-9 cm long, to 20 cm long in fruit, ramuligerous, axes albo-tomentulose, the peduncle 3-4 mm long, the bracts persistent, darker colored, semicircular, cucullate, about 1 mm long and broad, puberulous externally; flowers greenish, fragrant, the pedicels tomentulose, about 1 mm long; calyx segments deciduous, sparingly tomentulose externally, glabrous on inner surfaces; petal absent; stamens equal, 25-30, the filaments 6 mm long, the anthers basifixed, oblong ca. 1.5 mm long and 0.5 mm wide, gynoecium glabrous, the stigma capitellate, the style 3-4 mm long, the ovary oblong, 3 mm long 1.5 mm wide, the gynophore 1.5-2 mm long; fruits oval in outline, ca. 3 cm long and 1.5 cm in diameter, the surface tessellate-scaly, the “scales” darker, the seed oval, only slightly smaller than the fruits, the aril small.
Lectotype-Collection: A. Ducke (H.J.B.R. No.) 24219 (flowering por- tion). (Lectotype RB, isolectotypes F,G,K,MADw,NY,P,U,US), “circa Manaos (civ. Amazonas) in silvis loco alto ultra Flores,” Brazil, 29 No- vember 1932.
Distribution: Vicinity of Manaus, Brazil, southern Guyana and Suri- name, in riverine lowland forest. BRAZIL-Amazonas: Vicinity of Manaus above Flores, 18 April 1933, Ducke (H.J.B.R. No.) 24219 (fruiting portion) (F,G,MADw,NY,P,US) and 1 March 1946, Ducke 1923 (A,F,IAN,K,MG,NY,US). BRAZIL-ParaA: Tinguelim km 21, 30
A Revision of the Genus Bocoa 121
March 1970, Silva 3024 (IAN). GUYANA: “Plot 3, hill top 1000 feet high, 7 miles E. of Onoro mouth, Upper Essequibo,” 30 September 1952, Guppy 304 (Record No. 7280) (NY). SURINAME: Fallawatra, 8 Nov. 1971, Jimenéz-Saa 1568 (K).
Choice of a lectotype is required in this instance because Ducke in- cluded two collections under one number, probably from the same tree but this is not at all certain; both parts of the type collection are cited here.
The collection from Guyana is sterile but it almost certainly represents this species.
Vernacular Names: “Gombeira amarela” (Silva 3024).
Discussion: This is one of the two species of the genus which are char- acterized by apetalous flowers and arillate seeds. It is distinct from its near relative, B. prouacensis, by the number of leaflets, stipule size, and number of stamens. Like B. racemulosa, although less pronounced, B. viridiflora has trifoliolate leaves with the terminal one the largest. This species has the largest stipules, the longest inflorescences, and the largest number of stamens.
5. Bocoa prouacensis Aublet, Pl. Guiane Fr. Suppl. 38, t. 391. 1775.
Swartzia prouacensis (Aublet) Amshoff, Meded. Bot. Mus. & Herb. Rijks. Univ. Utrecht 52:40. 1939. Swartzia minutiflora Kleinhoonte, Rec. Trav. Bot. Neer]. 22:408. 1925.
Description: Tree 14-30 m tall, the trunk 19-25 cm diameter, straight and cylindrical, the bark brownish-grey with small rectanglar scales, the branchlets glabrous, the stipules persistent, rigid, 0.7-1.3 mm long and wide, broadly triangular-ovate, acute, glabrous; leaves glabrous, uni- foliolate, the petiole subterete, non-alate, 6-16 mm long or occasionally totally suppressed, petiolules 4-7 mm long, transversely corrugate, the blades coriaceous, 9—-18(-—21) cm long, 4.5-8(-10) cm wide, mostly elliptic but sometimes elliptic-ovate or broadly ovate, sometimes asym- metric, the base rounded and obtuse to subobtuse, infrequently acute, the apex bluntly acute, infrequently acuminate, the costa and its primary branches more or less salient on both sides of the leaflet-blade; in- florescences 2-5 cm long, ramuligerous, axes glabrous, the bracts per- sistent, cucullate, broadly ovate, about 0.6-0.8 mm long and 1 mm wide, glabrous except for ciliolate margin, bracteoles lacking; pedicels 0.3-0.6 mm long, rarely to 2 mm long, glabrous, the calyx opening nearly to the base in three, more or less equal segments which often fall in one piece; petal absent; stamens 7-10, uniform, the filaments 4.5 mm long, the anthers oblong, 1.3-1.5 mm long, ca. 0.5 mm wide; gynoecium glabrous, the stigma capitellate, the style filiform, 1.7-2 mm long, the ovary 1.8—2 mm long, 0.8—1 mm broad, oblong, the gynophore 1—-1.5 mm long; fruit oval in outline, 2.5-3 cm long, 1.5-2.5 cm wide, dehiscent, the seed yellow or light brown (fide Amshoff) 1-2 cm long, the aril white, laciniate, the funicle filiform, 1.5-3 m long.
122 Proceedings of the Biological Society of Washington
Type-Collection: J. B. C. F. Aublet s.n. (BM), “Habitat in sylvis Caux,” French Guiana.
Distribution: Suriname and French Guiana in lowland forest. SURI- NAME: Marechalskreek, 9 July 1921, BW 5441 (U) and 21 Jan. 1922, BW 5769 (U); Brokopondo District, 8 km ESE of Brownsweg Village, 26 Mar. 1965, van Donselaar 2273 (U); Tapanahoni River, 11 Nov. 1918, Gonggrypp 4175 (IAN,U); Nassau Mts., 9 Mar. 1949, Lanjouw & Lindeman 2545 (U); Boven Coesewijne, ca. 20 km SW of Poika, Schulz 7926 (U); Sectie “O,” 23 Apr. 1915, Suriname Forestry Department 342 (K,MO,U ),—23 Oct. 1915, 1191 (U) (type of Swartzia minutiflora Kleinhoonte),—27 Oct. 1915, 1198 (K,MO, U),—Nov. 1915, 1434 (U),—22 Nov. 1915, 1560 (U),—26 Feb. 1916, 1660 (U),—22 Dee. 19116, 2559 (U) 5 Mar. 1917, ou qunurs) ae 29 Aug. 1918, 3974 (U),—2 Sept. 1918 3984 (NY,U,US)._4 Sep 1918, 3986 (IAN,U),—9 Dec. 1918, 4185 (U),—30 May 1919, 4326 (U),—28 Nov. 1919, 4455 (MO,U),—28 Nov. 1919, 4477 (U),— 21 Jan. 1920, 4521 (U),—6 Mar. 1920, 4569 (IAN,K,U,US),—3 Nov. 1920, 4781 (U),—12 Nov. 1921, 5556 (U),—17 May 1922, 5839 (K,NY,U),—16 Mar. 1923, 6084 (IAN,U); Brownsberg Forest Re- serve, Suriname Forestry Department 6130 (NY,U); Sectie “O,”’ Dec. 1942, Suriname Woodherbarium (Stahel) 69 (A,IAN,K,MADw,NY,U ); Brownsberg Forest Reserve, 10 Oct. 1969, Tawjoeran 12581 (U). FRENCH GUIANA: St. Laurent, Feb. 1956, BAFOG 339M (U), 7236 (U), and 7249 (P,U), 7 Feb. 1956, Bena 1117 (U); Placeau no. 2—Carreau no. 56—Route de Mana,” 3 Jan. 1956, French Guiana Forestry Department 7126 (NY,P,U), 26 Mar. 1956, 7415 (U); Sinna- mary River, Crique Grégoire, 28 Apr. 1968, Oldeman B-1614 (P); Camopi River, 1 km downstream from Saut Ouasseye, 11 Dec. 1967, Oldeman 2625 (P); Karouany, 1859, Sagot 1210 (BM,K,P); with- out locality or date, Wachenheim s.n. (BM,K,P,US).
Vernacular Names: “Bois bobo,” “caux bois boco” (Aublet s.n.); “boko” (Oldeman B-1614); “aie oudou” (F. G. For. Dept. 7126); “itikiboroballi hohorodikoro,” “ijzerhart” (Sur. Woodherb. 69); “yzerhart” (Sur. For. Dept. 4326).
Discussion: Bocoa prouacensis is one of the two most frequently collected species in the genus and certainly one of the most distinctive. One characteristic alone separates it from all its relatives; the single seed in each fruit at the time it dehisces is suspended on a thread up to three meters long. (An interesting description of this phenomenon is given by Stahel in the Journal of the New York Botanical Garden, vol. 45:265-268. 1944). This attribute plus an obvious, fleshy aril is surely of some adaptive significance, probably for dispersal by animals. The aril in this species, and in the preceding, links the genus to Swartzia, most of whose species have arillate seeds. The unifoliolate leaves of B. prouacensis serve to separate it from all its near-relatives, the closest of which is probably B. alterna.
A Revision of the Genus Bocoa 123
6. Bocoa mollis (Bentham), Cowan, comb. nov.
Swartzia mollis Bentham, Hook. Joum. Bot. 2:89. 1840.
Trischidium vestitum Tulasne, Ann. Sci. Nat. (Ser. 2) 20:14], t. 4. Sept. 1843.
Swartzia cearensis Ducke, Anais da Academia Brasileira de Ciencias 31: 295. 30 June 1959.
Description: Shrub or small tree 2—4 m tall, foetid in fresh state, the young branchlets tomentulose densely; stipules early caducous, minute, ca. 0.5 m long, densely strigulose; petioles 6-11 mm long, tomentulose, terete, the rachis 25-50 mm long, tomentulose, terete; leaves impari- pinnate, the leaflets 5-9, 3.5-5.5 cm long, 1.5-3 cm wide, subopposite or rarely opposite, the petiolules 0.7-1.5 mm long, tomentulose, the blades oval to elliptic or ovate to lanceolate, the base rounded, obtuse or slightly cordate, the apex obtuse, retuse, to sub-emarginate, puberu- lous to tomentulose beneath, rarely only on costa, puberulous to strigulose above, the venation subobscure to barely prominulous, plane except the costa salient on lower surface; inflorescences ramuligerous, racemose, the axes 9-20 mm long, densely tomentulose-pilosulose, the bracts early deciduous, broadly triangular-ovate or semicircular in outline, about 1 mm long and wide, densely strigulose externally, glabrous within, the pedicels 4-12 mm long at anthesis, 7-15 mm long with mature fruits, pilosulose; buds elliptic in outline, subappressed- tomentulose, ca. 5 mm long and 3.5 mm diameter, the calyx opening only about half-way into 3 or 4 segments; petal one, deciduous to caducous, white, the claw 2-6 mm long, sparingly villosulose externally or glabrous, the blade glabrous, oblate, 4-7 mm long and 5-7 mm wide; stamens (12—)20-22, the filaments 4—7 mm long, slightly joined basally, the anthers narrowly oblong, 2-3 mm long, 0.7-0.9 mm wide; gynoecium glabrous except sometimes a few long hairs on ovary, the stigma capitellate, sometimes obliquely so, the style 2-3.5 mm long, the ovary elliptic, ca. 3.5 mm long and 2 mm wide, the gynophore ca. 1 mm long; fruit obliquely ellipsoid, beaked with the 1-2 mm long remnant of style, 10-15 mm long, 8-10 mm diameter, reticulate, the carpophore about 1-4 mm long; the seed tan to black, exarillate, 5.0.-7 mm long, 4.5-5 mm diameter, shiny, ellipsoid to subrotund.
Type-Collection: J. S. Blanchet 2774 (holotype K, isotypes F (frag.), G,K,P) (also the type collection of Trischidiwm vestitum Tulasne). “In Brasilia prope Utinga, Fazenda in Certao de San Francisco,’ Bahia, 1839.
Distribution: Near-interior of Ceara and Rio Grande do Norte to Bahia, Brazil, on sandy soils in scrub forest. BRAZIL-Crara: Baturité, 11 Apr. 1909, Ducke 1980 (BM,G,US); Itaitinga, Pé do Serrote da Pedreira, 11 August and 23 December, 1955, Ducke 2467 (IAN,INPA,K,MG,NY,US) (type of Swartzia cearensis Ducke) and Ducke 2582 (IAN,K,NY,RB,US); along the road from Morada Nova to Bixoba, 14 February 1960, Lima 60-3443 (IPA,US); no specific
124 Proceedings of the Biological Society of Washington
locality, Dias da Rocha (HJBR No.) 11183 (RB); without locality or date, Allemao & Cysneiros 497 (P). Rio GRANDE po Norte: Acu, along the margin of the left bank of the Rio Acu, Lima 60-3505 (IPA). PERNAMBUCO: Margem da estrada Ibimirim-Joazeiro dos Candidos, Lima 50-700 (IPA); Manicobal, Fazenda Contenda, 20 June 1952, Lima & Magalhaes 52-1050 (IPA). Banta: Antonio B. de Oliviera (HJBR No.) 68404 (RB); Raxo da Catarina, Labouriou 912 (F, RB,US). Mrnas Gerais; 4 km N of Vasante de San Francisco, Municip. Januaria, 29 Sept. 1953, Magalhaes 6082 (IAN,RB).
Vernacular Names: “Brinquinho” (Oliveira 68404).
Discussion: This species and B. decipiens form a small subgeneric grouping that may represent a separate phylogenetic offshoot of the genus. Both have several pairs of emarginate leaflets and small fruits, and both apparently are shrubs to small scrubby trees in the arid parts of southeastern Brazil. The smaller, glabrous or glabrescent leaflets, longer inflorescence, strongly revolute calyx segments, and _ larger number of stamens amply separate B. mollis from its closest relative.
7. Bocoa decipiens Cowan, sp. nov.
Swartzia decipiens Holmes, Pharm. Journ., ser. 4, 3:2. 4 July 1896. (Provisional Name)
Diagnosis: Frutex foliorum foliolis basalibus rotundatis 3-9 mm longis, 5-5.5 mm latis; gynoecium sicut fructus strigulosum.
Description: Shrub with the branchlets minutely strigulose, the stipules minute; petioles 1-2 mm long, lightly canaliculate on upper surface, minutely strigulose, the rachis 9-28 mm long, minutely strigulose, obviously canaliculate-marginate on the upper surface; leaf- lets 5-7, imparipinnate, the laterals subopposite, the petiolules 0.2—0.5 mm, glabrous, the blades glabrous, smaller and rounded near the base of the leaves but progressively larger and more elliptic toward the leaf apex, the basal ones 3-9 mm long, 5-5.5 mm wide, the upper ones and the terminal one 14-22 mm long, 6-10 mm wide, the base acute except the basal ones rounded-obtuse, the apex rounded, emarginate, glabrous or sparingly puberulous minutely on the costa, the venation subobscure, the costa salient on both surfaces; inflorescences axillary, racemose, few-flowered, the axis 2-5 mm long, strigulose, the bracts persistent, ca. 0.5 mm long, the fruiting pedicels 8-10 mm _ long, glabrous; calyx segments 3 or 4, glabrous, separating to the base, about 4.5 mm long, reflexed; stamens 8-10, glabrous, the filaments 2.5-3 mm long; gynoecium densely strigulose; fruit more or less strigulose, the basal 1-2 mm of the style persistent as a beak, the body of the fruit 8-9 mm long and wide, broadly oval in outline, the surface more or less reticulate, gynophore and carpophore densely strigulose, 0.5-1 mm long; seed black, exarillate, oval in outline, 5 mm _ long, 3.5-4 mm wide.
Type-Collection: A. Lima & M. Magalhaes 52-1075 (holotype US,
A Revision of the Genus Bocoa 125
sheet no. 2639739, isotype IPA), “Nos campos da Serra Araripe, prox. Est. Exp. Araripina,” Pernambuco, Brazil, 24 June 1952.
Distribution: Known certainly only from the type collection but circumstantial evidence in the following discussion indicates a broader range.
Discussion: In the latter part of the 19th Century, Europe was importing for medicinal purposes