Q iL U563 CRLSSI ' ’ Bult ; i ma ni MS i rt SYSTEMATICS AND ZOOGEOGRAPHY OF THE WORLDWIDE BATHYPELAGIC SQUID BATHYTEUTHIS (CEPHALOPODA: OEGOPSIDA) For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 — Price $1.50 (paper covers) ean ie “wy iar rf ; asian diel 4 4 bps + ti a a od , pa igeyll rates UNITED STATES NATIONAL MUSEUM BULLETIN 291 Systematics and Zoogeography of the Worldwide Bathypelagic Squid bathyteuthis (Cephalopoda: Oegopsida) Division of Mollusks Smithsonian Institution SMITHSONIAN INSTITUTION PRESS CITY OF WASHINGTON e 1969 Publications of the United States National Museum The scientific publications of the United States National Museum include two series, Proceedings of the United States National Museum and United States National Museum Bulletin. In these series are published original articles and monographs deal- ing with the collections and work of the Museum and setting forth newly acquired facts in the field of anthropology, biology, geology, history, and technology. Copies of each publication are distributed to libraries and scientific organizations and to specialists and others in- terested in the various subjects. The Proceedings, begun in 1878, are intended for the publication, in separate form, of shorter papers. These are gathered in volumes, oc- tavo in size, with the publication date of each paper recorded in the table of contents of the volume. In the Bulletin series, the first of which was issued in 1875, appear longer, separate publications consisting of monographs (occasionally in several parts) and volumes in which are collected works on related subjects. Bulletins are either octavo or quarto in size, depending on the needs of the presentation. Since 1902, papers relating to the bo- tanical collections of the Museum have been published in the Bulletin series under the heading Contributions from the United States Na- tional Herbarium. This work forms number 291 of the Bulletin series. Frank A. Taytor Director, United States National Museum IV Contents Introduction . Material and Methods Part I: SYSTEMATICS Historical Résumé ae . Bathyteuthis abyssicola Hoyle, “1885 : Bathyteuthis bacidifera Roper, 1968 Bathyteuthis berryi Roper, 1968 . Comparison of B. abyssicola and B. bacidifera Geographical Variation in Bathyteuthis Buccal Suckers . : Number of Arm Suckers . Gill Size . Relationship of Cae iete: to r Basngeoudntee Familial Relationships of the Bathyteuthidae Part II: ZOOGEOGRAPHY Review of Antarctic Ocean Oceanography . Soe Analysis of Environmental Parameters and Distribution . Geographic Distribution in Relation to Water Masses . Biological Factors Governing Distribution . Vertical and Regional Distribution Calculation of Maximum Depth of Caplin and Vertical eee . Vertical Distribution by Size . : : Depth of Capture in Relation to Dept of Oden 5 . Aspects of Regional Distribution and Relative Abundance . Summary and Conclusions . References . Appendix Page Oo Ol bo WwWwouo nr 68 68 75 78 85 89 95 96 102 140 157 165 165 ig! 175 176 197 199 209 a et - remus eat: ry ry | ll Systematics and Zoogeography of the Worldwide Bathypelagic Squid Bathyteuthis (Cephalopoda: Oegopsida)’ Introduction Study the deep-sea squid and see how he does only what he has to. Carl Sandburg Bathyteuthis abyssicola Hoyle has been recorded as an uncommon deep-sea species from widely scattered localities throughout the major oceanic regions of the world. In recent years the relatively large quan- tity of material that has accumulated allows a more thorough analysis of the systematics and distribution of Bathyteuthis. Newly described species and range extensions are represented in small collections from lower latitudes; however, the bulk of the material, useful in determin- ing details of horizontal and vertical distribution, comes from the Antarctic Ocean. The cephalopods of the open waters of the Southern or Antarctic Ocean and the shallower waters along the shores of the Antarctic Continent are very poorly known despite an impressive list of vessels and expeditions that have collected in these regions. Most of the species known in the Antarctic cephalopod fauna were described from collec- tions made during the great era of exploration prior to World War I. Berry (1917) reviewed the literature and listed the species that occur south of 60° south latitude; with the addition of Berry’s 5 new species the list at that time contained 17 species—11 species of the order Octo- poda and 6 species of the suborder Oegopsida. Thiele (1921) and Odhner (1923), reporting on the collections of the German and Swedish Expeditions (both 1901-1903), added to the knowledge of the Antarctic fauna. Robson (1930, 1932) struggled with the complexi- ties of the octopod fauna represented in the early collections of the Discovery. Dell (1959) reported on the cephalopod material col- lected during the British Australian New Zealand Antarctic Research Expedition (B.A.N.Z.A.R.E.) and listed the species that occur in the 1 Contribution no. 982 from the Institute of Marine Sciences, University of Miami and paper no. 65 from the Dana oceanographical collections. This paper was submitted in partial fulfillment of the requirements of the Degree of Doctor of Philosophy, University of Miami. 1 2 U.S. NATIONAL MUSEUM BULLETIN 291 zoogeographical regions defined by Powell (1951) for the Southern Ocean. Dell listed 20 species of cephalopods in the Antarctic Province that is bounded in the north by the mean location of the Antarctic Convergence. The list included 12 species of octopods and 8 species of oegopsids (4 of them cranchiids). Every species that Dell listed, endemic or cosmopolitan, presents a systematic problem, and the dis- tribution and biology of these forms have been virtually unknown. Some other groups of the Antarctic marine fauna, however, have been studied more thoroughly, primarily because of the extensive pro- gram carried on by the Déscovery office. More than 30 volumes of Discovery Reports have been prepared on the biology of Southern Ocean organisms. The systematics and distribution of planktonic and nektonic groups have been presented by Mackintosh (1934, 1937), Hardy and Gunther (1935), Fraser (1936), Baker (1954), Tebble (1960), Marr (1962), and many others. In 1962 the Office of Antarctic Programs of the National Science Foundation initiated its program in oceanography by the deployment of the USNS £itanin to the Southern Ocean. A biological collecting program of broad scope provided the opportunity to conduct detailed studies on the marine fauna of Antarctica. A grant to study the sys- tematics and distribution of Antarctic cephalopods was awarded to G. L. Voss of the Institute of Marine Sciences, University of Miami. As the large collections were sorted and identified, it became increas- ingly clear that the cephalopod fauna of Antarctic waters was con- siderably more extensive and more complex than had been indicated by all previous surveys. Preliminary sorting and identification of the collections by the writer in the winter of 1965 revealed that approxi- mately 30 species of the suborder Oegopsida occur in the Southern Ocean. Some of these species are relatively well known; some represent long extensions in range; several are undescribed. Nearly the same situation holds for the dozen or so nominal species of benthic octopods that are being studied by G. L. Voss. The finned octopods, a perpetual problem group, are represented by about a half-dozen species. Recent additions to the cephalopod fauna of the Antarctic include the Batoteuthidae, a new family of oegopsids (Young and Roper, 1968), the second and third specimens of the curious Promachoteuthis Hoyle (Roper and Young, 1968), the second known specimen of C7r- rothauma murrayi Chun, and new species in several oegopsid families (e.g., Cranchiidae, Histioteuthidae). While new or rare species were being added to the fauna with the increasing collections, one species, Bathyteuthis abyssicola Hoyle, 1885, emerged as the overwhelmingly dominant species of pelagic cepha- lopod. It seemed to be common everywhere in the Antarctic Ocean, BATHYPELAGIC SQUID BATHYTEUTHIS 3 where it was captured over a great range of depths; often it was taken in what is considered great abundance for a deep-sea squid, with as many as two-dozen specimens in a single midwater tow. Although B. abyssicola has been recorded in the literature from several different localities in the Atlantic, Pacific, Indian, and Antarctic Oceans, it has been regarded as a rather rare bathypelagic squid that exhibits a world- wide distribution. The overall objectives of the Antarctic Celphalopod Project are to delineate the Antarctic cephalopod fauna, to define the horizontal and vertical distributions of its components, and to determine its zoogeo- graphic relationships with the faunas of adjoining regions. The sys- tematics of the fauna is currently being worked out group by group, but it seemed that the ultimate objectives of the program could be achieved by first working out the distribution and zoogeographic rela- tionships of a single species as a model for comparison in future studies on other species. Bathyteuthis abyssicola was the obvious choice for the initial study because of its numbers and because it occurs throughout the area of operations of the E7tanin. In addition to the material from the Antarctic, a number of speci- mens of Bathyteuthis from the Atlantic, East Pacific, and Indian Oceans were available from the collections of the Dana, Pillsbury, Velero, Anton Bruun, Chain, and Bureau of Commercial Fisheries vessels. The work that 1s presented here is divided into two major parts. The part on systematics presents a review of literature on Bathyteuthis, detailed descriptions of B. abyssicola and of two newly named species of Bathyteuthis, bacidifera and berryi (Roper, 1968), a comparison of the species, a determination of the familial relationships of the Bathyteuthidae, an analysis of Ctenopterya, and a comparison of Ctenopteryx with Bathyteuthis to determine the validity of their in- clusion in the same family. The final part of the sections on systematics comprises an examination of the geographical variation in Atlantic, East Pacific, and Antarctic populations of B. abyssicola. The second part is a study of the distribution of Bathyteuthis. The bathymetric and geographic ranges of the species are established, and the physicochemical and biological factors that govern these distribu- tions are analyzed. The occurrence and abundance of the Antarctic population are examined in relation to area, environmental conditions, season, and other species of pelagic squid. The size-group composition of the population and the distribution of growth stages are presented. A detailed characterization of the Antarctic oceanic environment was a necessary prelude to the study of the distribution of B. abyssicola and of all other Southern Ocean species of cephalopods as well. Eltanin 4 U.S. NATIONAL MUSEUM BULLETIN 291 oceanographic data were available only in tabular form from Lamont Geological Observatory. Data in this form, however, are of little value to the biologist seeking an overall view of the physicochemical environment. Therefore, the oceanographic data were analyzed and compiled by hand as vertical sections of oceanographic parameters along meridians and latitudes. Seven groups of sections were con- structed along meridians between 25° west longitude and 160° west longitude; a pair of sections was made along 60° south latitude from 25° W to 160° W. A brief survey of the major features of Antarctic oceanography is supplemented by the detailed examination of the vertical sections, and. several previously suspected features of the oceanography of the Antarctic Ocean are verified. The analysis of oceanographic parameters has provided the infor- mation necessary for the determination of the many facets of the horizontal and vertical distribution of Bathyteuthis abyssicola. Furthermore, now that this environmental information has been com- piled, it can be used for determining and comparing the distributions of all pelagic species occurring in the Antarctic Ocean. Studies of this nature are moderately common in the literature of biological oceanography. In the field of systematics and biology of the Cephalopoda, however, this approach has not been taken because few large collections of oceanic squid, and particularly of bathypelagic species, have been available for study. Therefore, analyses of geo- graphic variation and ecological factors governing distribution are presented for a bathypelagic squid for the first time. A work of this scope is accomplished only with the aid and co- operation of other people and organizations. The Antarctic Cephalo- pod Project is supported by the Office of Antarctic Programs, Na- tional Science Foundation (Grants GA 103 and GA 253 to G. L. Voss, Institute of Marine Sciences [IMS] Miami). The Department of Biol- ogy of the University of Southern California is responsible for the macrobiological collecting program aboard the Z7tanin; I acknowl- edge the careful preservation and handling of specimens by the USC teams and their cooperation with Institute of Marine Sciences person- nel who have participated in Z/tanin cruises. 8. Jacobs of the Lamont Geological Observatory made suggestions and supplied listings of the oceanographic data collected aboard E7tanin., Additional material and assistance were available from several sources. J. Rosewater provided working space and facilities in the Di- vision of Mollusks, U.S. National Museum, Smithsonian Institution, where I examined the type of Benthoteuthis megalops Verrill, 1885, and miscellaneous collections; I am grateful to him for reviewing the manuscript. R. H. Backus loaned specimens that I sorted from the BATHYPELAGIC SQUID BATHYTEUTHIS 5 cephalopod collections at the Woods Hole Oceanographic Institute. W. Clench and G. Mead gave access to the cephalopod specimens in the Museum of Comparative Zoology at Harvard. N. Tebble, Curator of Molluscs at the British Museum (Natural History), provided fa- cilities for studying the type of Bathyteuthis abyssicola Hoyle, 1885. Cephalopod material collected during the Dana expeditions is avail- able at the Institute of Marine Sciences through the Carlsberg Founda- tion, E. Bertelsen, Director. Specimens collected by U.S. Bureau of Commercial Fisheries vessels are on deposit at IMS through H. Bullis, Pascagoula, Mississippi. R. E. Young, IMS, loaned material of a new species from the Velero collections in California waters. KE. McSweeny, W. Herrnkind, J. Walsh, and G. Hendrix of IMS participated in Eltanin cruises to collect material and make observations of Antarctic cephalopods. T. E. Bowman, U.S. National Museum, brought the quote by Sandburg to my attention. I am most grateful to all of these people and institutions for their aid. Several members of the faculty at IMS read all or portions of the manuscript and gave helpful suggestions. I am indebted to A. J. Provenzano, C. R. Robins, F. M. Bayer, J. Fell, L. P. Thomas, L. J. Greenfield, S. Broida, and E. S. Iverson. I am especially grateful to G. L. Voss for his suggestions and en- couragement throughout this work, for the stimulation he has provided by his continuous interest in and support of my studies, and for his review of the manuscript. Long discussions with R. E. Young have been particularly valuable ; I appreciate his suggestions and enthusiasm for this work. The illustrations of Bathyteuthis are the skillful work of Miss Con- stance Stolen; her patience and perceptiveness result in precise rendi- tions of the material. I am most grateful for her assistance. My wife, Ingrid, has compiled and plotted data, assisted in trans- lations, and prepared the graphs and charts for reproduction; for this and for her encouragement I am grateful. Material and Methods Most specimens used in this study were taken by the USARP vessel Eltanin, but material was available from other sources as well. Speci- mens are referred to by ship and station number in the text and figures. The following list gives the sources of material by ship and by geo- graphic area, and an example of the abbreviation of the reference number (ship and station). 6 U.S. NATIONAL MUSEUM BULLETIN 291 Eltanin Southern Ocean Elt. 354 Dana Atlantic, Bay of Panama D1208 VI Pillsbury Gulf of Guinea Pil. 300 Oregon Western Atlantic O 4296 Silver Bay Western Atlantic SB 2024 Pelican Western Atlantic Pel. 11 Chain Atlantic Ch. 17-18 Velero San Pedro Basin, California V 9661 The abbreviations used are standard, e.g., m for meters, mm for millimeters, ete. Abbreviations used in reference to morphometric dimensions are also standard in the teuthological literature; the most common is ML for mantle length. Others are explained in the legends or text. Definitions of these abbreviations are found in Voss (1963) and Roper (1966) so they need not be repeated. All raw data, measurements and indices, and complete station data are on file in the Division of Mollusks, U.S. National Museum, Smith- sonian Institution. PART L. SYSTEMATICS Historical Résumé The initial account by Hoyle of the cephalopods collected by HMS Challenger appeared in the narrative of the cruise that was published in May 1885. Hoyle’s brief comments dealt primarily with the more unusual cephalopods captured, and a number of new genera and species were erected. One very curious decapod, according to Hoyle, was the small specimen taken in the Southern Ocean between Prince Edward and Crozet Islands by a dredge (trawl) haul that fished at 1600 fathoms (2900 m). Appropriately, he named this species Bathyteuthis abyssicola. The original description was brief, but, in conjunction with its accompanying illustration, it adequately established the species. Hoyle listed these important features (p. 272): body tapering to a blunt point; fins small, round; head broad, with prominent eyes; oral membrane with suckers; arms short with minute, biserially arranged suckers; tentacles with unexpanded clubs but gradually tapering to a point armed with numerous, very small suckers; pen resembling that of Ommastrephes. Hoyle stated that the structure of this form seemed to adapt it for life in the great depths, justifying the belief that it actually came from the 1600-fathom depth reached by the trawl. He further stated (pp. 272-273) that the “... small fins are in marked contrast to those of pelagic species, while the small suckers and delicate tentacles are equally little fitted for raptorial purposes; but, on the other hand, the large circumoral lip would seem well suited for collecting nutritive matters from an oozy bottom.” Shortly after the narrative was published Hoyle brought out his diagnosis (1885a) of new species of cephalopods collected during the cruise of HMS Challenger. In this he not only described new species which had not been described in the narrative (primarily species of Loligo and Sepia), but also included the previously described Promachoteuthis and Histiopsis. However, Bathyteuthis, the only other new decapod in the Challenger collections, curiously was omitted. The third catalog of Mollusca published by Verrill (1885) contained a number of new species of cephalopods that were captured by the U.S. Fish Commission Steamer A/batross in 1884 off southeastern New England. Among the new species were two small specimens that repre- sented a new genus as well, which Verrill named Benthoteuthis megalops. As was his practice, he did not assign the new genus to a family. 8 BATHYPELAGIC SQUID BATHYTEUTHIS 9 Verrill gave detailed descriptions of the new genus and species, ac- companied by a single illustration “of one of the types.” Apparently this is the female, mantle length 57 mm, to which he referred several times in the text. I have examined the remains of the type in the U.S. National Museum; it bears the catalog number 39968, and it is a mature female. No additional mention is made in Verrill’s text of the other specimen, number 39967. Verrill mentioned (p. 402) that Benthoteuthis displays “marked em- bryonic or primitive characters,” typified in the young stages of Loligo and Ommastrephes, that “are seen especially in the small size, posterior position and form of the fins; in the form of the body, head and mantle; in the small short arms, with the dorsal pair shortest; in the small simple suckers; in the want of differentiation of the tentacular club and the uniformity of its minute suckers.” Verrill thought that the affinities of the genus probably lay with the Ommastrephes group because of the “distinct eye lids and sinus, and by the character of the connective cartilages of the mantle.” He also noted that the pen was like that of Zo/igo, but that the pen is appar- ently of little value in determining relationships of squids. It is curious that he should choose for showing affinities such a general and inde- cisive character in the oegopsids as the eyelid and such a singularly decisive familial character as the mantle-connective cartilages of Om- mastrephes. Nowhere in Verrill’s description is there an indication that the locking apparatus in Benthoteuthis resembles the strong, inverted T-structure of ommastrephids. Shortly after the diagnosis on the Challenger cephalopods appeared, Hoyle published his preliminary report (1885b), which was identical to the earlier diagnosis (1885a) except for the inclusion of an intro- duction, a generic diagnosis, and a description of B. abysstcola. In addition, he gave a footnote (p. 282) : “This [Bathyteuthzs| seems to be at all events congeneric with a form which Professor Verrill has recently dredged in the North Atlantic, and named Benthoteuthis megalops.” In the generic synonymy Hoyle (p. 308) listed Bathyteuthis Hoyle, May 1885 and, as a synonym, Benthoteuthis Verrill, July 1885. This is the first mention of these two genera after their original descrip- tions. In synonymizing Benthoteuthis, Hoyle had to determine the month of publication; apparently he felt that no future conflict would exist over the dates, for he did not discuss the point in the text. The generic characterization emphasized the blunt body shape, the small, rounded, subterminal fins, the simple, elongate mantle-connec- tive, the large, very broad head with prominent eyes, the short arms with minute suckers in two rows, the large 7-pointed buccal membrane 10 U.S. NATIONAL MUSEUM BULLETIN 291 bearing 1 to 2 suckers, the long, slender tentacles with unexpanded clubs bearing numerous minute suckers, and the “Ommastrephes-like” anterior section of the gladius. The specific description is an amplification of the generic diagnosis. Hoyle gave the following details: arm formula 4.3.2.1; arms only one- fourth as long as the body ; delicate protective membranes on the arms; suckers spheroidal, nearly embedded, sucker rings smooth with 2-3 rows of conical papillae; hectocotylus lacking; club one-eighth as long as tentacle; club sucker rings smooth with two rows of papillae. Hoyle’s accompanying illustration (fig. 2, p. 309) indicates that the specimen had undergone distortion during capture and preservation; this may account largely for the discrepancies in shape and proportion of fins, head, mantle, and arms noted by later authors, particularly Chun (1910) and Pfeffer (1912). Hoyle’s comprehensive report on the Challenger cephalopods ap- peared in 1886. The first section of this work is a synoptic listing of the recent cephalopods. The genus Bathyteuthis Hoyle, 1885, is listed under the family Ommastrephini Steenstrup, 1861, subfamily Ommastre- phidae Gill, 1871, with Benthoteuthis Verrill, 1885, as its synonym. Hoyle listed two species in this genus: his own Bathyteuthis abyssicola and Verrill’s B. megalops. Hoyle mentioned in the discussion the pos- sibility that the two species might ultimately prove to be conspecific (p. 169) ; in the zoogeographic and bathymetric sections of this mono- graph, however, he continued to list them separately. Hoyle, in an attempt to delimit the distributions of cephalopods, defined four major zoogeographic regions of the oceans and seven bathymetric zones between the surface and 3000 fathoms (5500 m). B. megalops occurred in the Atlantic Ocean region, 600-1073 fathoms (1100-1950 m); B. abyssicola inhabited the Indian and Southern Ocean regions at 1600 fathoms (2925 m). In his discussion of bathymetric distribution, Hoyle stated that there is reason to regard Bathyteuthis and Benthoteuthis (which “ob- viously belongs to the same genus” [p. 232]) as deep-sea inhabitants. Hoyle drew attention to Verrill’s observation of embryonic characters in this genus, but he considered these as characters that indicate a deep-sea habitat. In addition, he believed that the small fins were ill- adapted for pelagic existence, that the minute suckers were not fitted for raptorial purposes, and that the large buccal membrane was well suited for collecting food from an oozy bottom. The generic diagnosis and the specific description are nearly “verba- tim” duplications of those given in his preliminary report (Hoyle, 1885b) with the exception of two phrases. In the earlier work he stated it was impossible to determine if the gladius formed a terminal conus, BATHYPELAGIC SQUID BATHYTEUTHIS ial while in the Challenger Report he said that the gladius was damaged in dissection but that it was still possible to determine that had no terminal conus as in Ommastrephes or Taonius. In the earlier report he had described the arm suckers as being arranged in two rows, but in the later work the statement had been altered to “two or four rows.” In stating that B. abyssicola and B. megalops might prove to be the same species, Hoyle noted the following differences (p. 169): B. abyssicola lacks an angular sinus on the eyelid, has a large head, has suckers more nearly in two very irregular rows (“if slightly more [irregular they] might be regarded as four”). The type specimen is well illustrated on his plate 29, figures 1-7. Hoyle stated his uncertainty to which family Bathyteuthis rightly belongs and suggested that the establishment of a new family might be necessary. Hoyle (1886b) again listed Bathyteuthis (with Benthoteuthis as a synonym) and included both B. abyssicola Hoyle and B. megalops (Verrill) in the genus as he had done in the Challenger Report. Hoyle reported upon the existence of cephalopods in the deep sea at the meeting of the British Association for the Advancement of Sci- ence held in September 1885. He considered Cirroteuthis, Mastigoteu- this, and Bathyteuthis as true representatives of the deep sea because they had not been captured prior to the days of deep-sea investiga- tions, and Hoyle indicated that Bathyteuthis possessed structural pe- culiarities that fitted it for abyssal existence, some of which it shared with Mastigoteuthis. He did not enumerate these characters. Goodrich (1892) published one of the first keys for the identifica- tion of oegopsid genera. He characterized Bathyteuthis by the features that he considered most diagnostic (p. 318) : “siphon without bridles, arms with four rows of smooth suckers, club of tentacle with many rows of minute suckers, small [eye] sinus, pen feather-shaped, suckers on the seven lobes of buccal membrane, fins subterminal, rounded.” Pfeffer (1900) erected and characterized the new family Bathy- teuthidae. The familial name first appeared in the key to the families where it was separated out on the basis of its peculiar characters: free funnel, simple locking apparatus, Loligo-like gladius with free rhachis and leaflike vane, and in particular, the size, number and arrangement of suckers on the arms and clubs. After a short diagnosis of the family, Pfeffer stated that it could not be said with certainty whether Bathyteuthis and Ctenopteryx, both of which he included within the family, truly constitute a natural group. He described the two genera in some detail and gave partial synon- ymies; Bethoteuthis Verrill was listed as a synonym of Bathyteuthis 321-534 O—69 9 ~ 12 U.S. NATIONAL MUSEUM BULLETIN 291 Hoyle. Pfeffer utilized the descriptions of both Hoyle and Verrill, and his description of Bathyteuthis is a composite of the two orignal de- scriptions. Species were listed with their synonyms and distributions. Here, for the first time, Benthoteuthis megalops Verrill, 1885, is synon- ymized with Bathyteuthis abyssicola Hoyle, 1885, although Hoyle had anticipated it earlier (1886). Pfeffer stated that it was difficult to say if Hoyle’s and Verrill’s specimens belong “exactly” to the same species (p. 173). Chun’s classic work on the eyes and light organs in deep-sea cepha- lopods (1903) was based on material gathered during the German Deep-Sea Expedition. He made a detailed study of the eye of Bathy- teuthis. Chun placed Benthoteuthis in parentheses after Bathyteuthis, indicating that it was a synonym. Hoyle (1904) reported on the cephalopods captured by the U.S. Fish Commission steamer Albatross during two cruises to the tropical east- ern Pacific (1891 and 1899-1900). Two specimens of Bathyteuthis abyssicola were recorded from off Cape Mala, Gulf of Panama. No de- scription of the specimens was given; however, a colored illustration was presented, made when the larger specimen was still fresh, and Hoyle commented about slight inaccuracies in the drawing (head and arms too large, nuchal mantle point too pointed). In addition, it should be noted that the eyes are too prominent, having been forced out through the eye openings, and that the fins are displaced posteriorly. All of these discrepancies are attributable to stresses of capture. Hoyle’s (1904a) key to the genera of recent dibranchiate cephalo- pods included Bathyteuthis and Ctenopteryx in the family Bathyteu- thidae. Hoyle mentioned that the association of these two genera in the same family is rather artificial and that further information is necessary to determine their true relationships. In the second supplement (1909) to this catalog, Hoyle listed Bathyteuthidae Pfeffer, 1900, and Ctenopteryx, and he gave the synonymy of Ctenopteryx sicula (Ruppell). Chun (1910) covered the Bathyteuthidae in his monograph based on the cephalopods collected during the German Deep-Sea Expedition aboard the Valdivia; Bathyteuthis, in particular, was considered in detail. Chun attempted to determine whether Benthoteuthis megalops Ver- rill or Bathyteuthis abyssicola Hoyle had priority; he decided that the question was answered by the notation “April, 1885” on page 399 of Verrill’s paper which was published in the Transactions of the Connecticut Academy of Arts and Sciences. Since Hoyle’s description was published in May in the narrative of the Challenger expedition, Chun considered that Verrill’s Benthoteuthis had priority by one BATHYPELAGIC SQUID BATHYTEUTHIS 13 month. Therefore, Chun, in attempting to establish priority, intro- duced further uncertainty that, in spite of Hoyle’s attempts to rectify the mistake, has persisted to the present, primarily because of the stature of Chun’s monograph. Unfortunately, Chun used the printer’s signature date, which indicates only the date that sheet 50 of Trans- actions of the Connecticut Academy of Arts and Sciences was run off the press. Signature dates are not necessarily valid dates of publication. Chun quoted Pfeffer’s original familial diagnosis and further stated that he retained the Bathyteuthidae not only on the basis of the known characters, but in addition on the basis of a number of unusual features of the internal structure of Benthoteuthis. Chun emphasized that until that time the only anatomical work had been on Ctenopteryx from the Mediterranean and that this knowledge should not be applied to Benthoteuthis. He differentiated these genera on the differences in size and structure of the fins and the absence of light organs in Ctenopteryex. Chun’s specimens were captured by the Valdivia in the Benguela Current south of the Cape of Good Hope, off northwest Sumatra, southwest of Ceylon, and north of the Chagos Archipelago. All of the specimens were small, ranging from 9 to 18 mm in mantle length. (The largest specimen probably belongs to the new species described from the Bay of Panama.) Since two species are represented in Chun’s material, it is impossible to know to which species the description refers except in sections where he mentioned individual specimens. Presumably most of the description, and especially that of the eye and internal anatomy, is based on the largest specimen (only 18 mm in mantle length). This probably is the only specimen of Chun’s that belongs to the new species described from Panama. The description of the species is extremely detailed and compre- hensive, covering both external and internal features. Chun gave de- tails concerning the consistency and gross structure of the skin and muscles. The funnel organ and olfactory papilla were described for the first time. Chun’s observations of 1903 on the structure of the eye were included. He also commented on the structure and position of the light organs. From these observations he deduced that Benthoteu- this isa true deep-sea cephalopod. In attempting to determine the relationships of the Bathyteuthidae within the Oegopsida, Chun decided that the family held a truly iso- lated position. He stated that the nature of the buccal membrane connectives to the arms corresponds to the condition in the Enoploteu- thidae, Histioteuthidae, and Ommastrephidae, but that the form and position of the liver and pancreas preclude any close relationship. 14 U.S. NATIONAL MUSEUM BULLETIN 291 (The value of these characters for determining relationships between the higher taxa is discussed below.) Finally Chun agreed that Pfeffer’s placement of Ctenopteryx and Benthoteuthis in the same family should be upheld. Hoyle’s (1910) list of generic names included Bathyteuthis with Benthoteuthis as its synonym. In a footnote Hoyle gave the sources and dates of publication for the two monotypic genera. Hoyle (1912) reported on the cephalopods captured by the Scottish National Antarctic Expedition’s vessel Scotia. A single specimen of Bathyteuthis was taken off Coats Land at 71°22’S, 18°15’W. Hoyle listed the locality information and previous records but did not de- scribe the specimen, Instead, he attempted to settle the nomenclatural disharmony that had persisted since Bathyteuthis abyssicola Hoyle and Benthoteuthis megalops Verrill had been erected in 1885. He chided Chun (1910) for using the sheet (signature) date, April 1885, to determine priority for Verrill’s name, but he emphasized that if sheet dates are to be used, then Bathyteuthis would still have priority be- cause the date on sheet 34 in which Bathyteuthis abyssicola was de- scribed is “1884.” Hoyle apparently made an extensive investigation to determine that Verrill’s paper was not released as separate signa- tures in April, May, and June but that it came out as a single unit in June or, more probably, in July 1885. Pfeffer’s extensive monograph (1912), based primarily on the Plank- ton-Expedition cephalopod material, gave a key to the families of the Oegopsida that included the family Benthoteuthidae. This key is more detailed than the one given in his synopsis (1900), although the basic arrangement and the characters used are the same. Pfeffer gave a detailed diagnosis of the family and discussed the similarities and differences of Benthoteuthis and Ctenopteryx,; he concluded that the strong differences that exist between the adults are small or non- existent in the young stages, and that these genera form a unit distinct from all other oegopsids. At no point in the text did Pfeffer mention that he was altering the family name to Benthoteuthidae, nor did he give Bathyteuthidae as a synonym. In fact, even in his section on history and synonymy, he failed to mention the erection of the family Bathyteuthidae 12 years earlier in his “Synopsis der Oegopsiden Cephalopden.” Pfeffer, in a short sentence following his diagnosis of the genus Benthoteuthis (p. 324), casually stated that it need only be mentioned that, as Chun had determined, the generic name Benthoteuthis Verrill is one month older than Bathyteuthis Hoyle. Pfeffer apparently did not make his own investigation into the matter of priority but simply accepted the authority of Chun. (Perhaps this is understandable judg- BATHYPELAGIC SQUID BATHYTEUTHIS 15 ing from Chun’s contemporary prominence and stature as a scientist.) Pfeffer did include his use in 1900 of Bathyteuthis and B. abyssicola in the synonymy of Benthoteuthis and B. megalops. Pfeffer’s unexplained substitution of Benthoteuthidae for Bathy- teuthidae is considered improper and not in the best interests of sta- bility in nomenclature. It is generally regarded that a familial name, once properly erected, as the Bathyteuthidae was by Pfeffer himself in 1900, should not be changed because of subsequent altera- tions to the generic name (International Code, Art. 40). Of course, the problem in this case is resolved with the establishment of priority for Bathyteuthis. Pfeffer presented a comparison of previously described and illus- strated specimens, analyzed and explained the inconsistencies, and concluded that all specimens represent the same species which has nearly a worldwide distribution. Pfeffer gave a very detailed descrip- tion of his specimens, which came from the western North Atlantic at 40.4°N, 57°W, the South Equatorial Current at 05.1°S, 14.1°W, and the Mediterranean. Apparently no details of the Mediterranean speci- men are available except that it was deposited in the Hamburg Museum and apparently came from Messina. The validity of the record has been questioned by some subsequent compilers (e.g., Grimpe, 1922) of the Mediterranean cephalopod fauna. Pfeffer’s plate 27, figures 12-15, presents illustrations of the specimens, but Pfeffer’s caption to the figures is “Benthoteuthis abyssicola Verrill,” an unfortunate com- bination of names. The specimen from Messina in the Hamburg Mu- seum (figs. 14 and 15) has a mantle length of only 3 mm and does not resemble Bathyteuthis; it looks very much like the larva of Cteno- pterye illustrated by Naef (1923, figs. 116, 117). Since this specimen is the first of two doubtful records of Bathyteuthés in the Mediterran- ean and since Ctenopteryx is so common there, it is very probable that this larva has been misidentified. This matter is considered in detail in the discussion of Joubin’s 1920 work. Figures 12 and 18 are the Atlantic specimens and look like Bathyteuthis larvae with mantle lengths of about 3 mm and 4.5 mm. Berry (1912) presented some nomenclatural changes in the Ceph- alopoda and discussed briefly the problem of priority for Bathyteuthis or Benthoteuthis. He accepted Chun’s reasoning that Benthoteuthidae should be the familial designation. Apparently Berry had not seen Hoyle’s discussion (1912) that appeared in May, seven months prior to Berry’s paper. Joubin (1912) noted some of the more interesting cephalopods that were captured during the 1911 cruise of the Princesse-Alice. He re- corded one specimen of Benthoteuthis megalops caught between 0 and 16 U.S. NATIONAL MUSEUM BULLETIN 291 4500 m and stated that only one other sample, the original material from America, was known. Apparently Joubin was unaware of Chun’s (1910) Valdivia specimens and of the nomenclatural controversy that persisted concerning Benthoteuthis and Bathyteuthis. Massy (1916) listed some cephalopods from the Irish Atlantic Slope; the small collection contained a young specimen of B. abyssicola that was taken at 55°22’N, 11°40’W in 700-750 fathoms. This locality constitutes the northernmost record of B. abyssicola. Massy (1916a) recorded B. abyssicola in her report on the cephalo- pod collections of the Indian Museum. Two small specimens (about 5 and 12 mm mantle length) were captured in the Bay of Bengal by the /nvestigator. The brief description noted the following features: 2 rows of suckers on arm IV, 3 to 4 irregular rows of suckers on arms I-III (2 rows on all of the arms of the larva) ; 4 blunt, widely sepa- rated teeth on the club suckers, 5 to 6 teeth on the arm suckers (4 to 5 in the larva); distinct photophores and pigment. Measurements are given for the larger specimen. In his work on the paleobiology of dibranchiate cephalopods, nearly half of which concerns living forms as well, Abel (1916) mentioned Bathyteuthis and Benthoteuthis in his discussion of retention of larval characters in adult forms. In a number of highly specialized species of oegopsids that inhabit the deep sea, the terminal fin is divided even in the adult, whereas in almost all known young stages of Oegopsida the terminal fins are divided only during the larval life. The comments by Abel, a paleontologist, are undoubtedly based on the suggestions and observations of Verrill, Hoyle, Chun, and Pfeffer. Naef (1916) listed the family Benthoteuthidae Pfeffer, 1900, and Ctenopteryx sicula. In his incorrect assignment of the family name, Naef erred in one of two ways. He intended to use either Bathy- teuthidae Pfeffer, 1900, or Benthoteuthidae Pfeffer, 1912. It was not until 1912 that Pfeffer, without explanation but apparently following Chun’s decision that Benthoteuthis had priority, made the unwar- ranted change in the family name. In his 1921 survey of Mediter- ranean dibranchiate cephalopods, Naef correctly listed the family Bathyteuthidae but erroneously referred it to Pfeffer’s 1912 work, which included only Benthoteuthidae. This work again lists Cteno- pteryx Appellof and adds Bathyteuthis Hoyle, probably on the basis of Pfeffer’s (1912) record from the Mediterranean. Berry (1917) listed Hoyle’s Scotia specimen of B. abyssicola as a representative of the Antarctic cephalopod fauna in the historical sec- tion of his report of the Cephalopoda captured during the Australian Antarctic Expedition, 1911-1914. Joubin (1920) reported on the cephalopods captured during the BATHYPELAGIC SQUID BATHYTEUTHIS ie cruises of the Princesse-Alice from 1898-1910. One small specimen of Benthoteuthis megalops, only 7.5 mm from “mouth” to posterior end of body, was taken between the Azores and northern Portugal. Joubin’s description was brief and dealt primarily with color obser- vations. Another specimen, only 3.5 mm in length, was taken in the western Mediterranean and, according to Joubin, very closely resem- bled the specimen illustrated by Pfeffer (1912) on plate 27, figures 14 and 15. Joubin’s specimen was smaller than Pfeffer’s and Joubin felt that it belonged in the same series but listed it only as Bentho- teuthis sp. in the event of a future reconsideration. Judging from the illustration it appears that Joubin’s specimen, like Pfeffer’s, is a larva of Ctenopteryx. Therefore, the only two records of supposed Bathyteuthis (reported as Benthoteuthis) in the Mediterranean ac- tually represent Ctenopteryx larvae. Further verification of this is found in Naef’s (1923) discussion and illustrations of the juvenile stages of Ctenopteryax siculus. Joubin’s Benthoteuthis sp. (pl. 12, fig. 9) and Pfeffer’s Benthoteuthis megalops (mislabeled as “Benthoteuthis abyssicola Verrill,” pl. 27, figs. 14 and 15) very closely resemble Naef’s illustrations of the larvae of Cteno- pteryx siculus (p. 253, figs. 116 and 117). In addition, all of these specimens come from the western Mediterranean, Messina, and Naples. The larvae illustrated by the three authors range in size from less than 2 mm to about 3.5 mm in mantle length and, taken together, rep- resent a good growth series of larval Ctenopteryz. Robson (1921) described Chunoteuthis minima, a new genus and spe- cies, which he placed in the family Benthoteuthidae. The description of the new genus and species is based upon a single larval specimen only 83 mm in length from the apex of the mantle to the base of the arms. Nothing in the description would lead one to believe that this form was a bathyteuthid instead of the larva of any number of other Oegopsida. Robson admitted that the specimen was the source of considerable trouble because it was very shriveled and many of the external features were obliterated. He could not align it with any known genus, and furthermore, “even its family relationships are very doubtful.” Nevertheless, he saw fit (p. 482) to “assign it to the Benthoteuthidae on the strength of its general superficial appear- ance”! Robson’s entire presentation is the absolute epitome of the fallacious belief held by some earlier workers that every specimen in a collection, regardless of size or condition, must have a name applied to it. Nothing can be ascertained from the description concerning the true identity of this larval form. The type in the British Museum 18 U.S. NATIONAL MUSEUM BULLETIN 291 (NH) has been examined by G. L. Voss; it is hard and unmanage- able and its identity cannot be determined. The illustration reveals nothing that would qualify Chunoteuthis as a bathyteuthid; the arms and tentacles are too long and slender, no “web” connects the basal parts of the arms, the clubs are too ex- panded, and the arm suckers are too prominent. Chunoteuthis almost certainly is not a bathyteuthid. Unfortunately, Robson’s name has to be reckoned with since he considered the specimen a bathyteuthid. The only recourse is to pro- claim Chunoteuthis minima Robson, 1921, a nomen dubium. Thiele (1921) reported on the Cephalopoda obtained during the German South-Polar Expedition, 1901-1903. Specimens of Cteno- pteryx sicula and Benthoteuthis megalops were included, which he placed in the Benthoteuthidae following Chun and Pfeffer. Two speci- mens 11 mm in mantle length were referred without doubt to B. mega- /ops; one was captured in the South Atlantic at 35°10’S, 02°33’K, while the other came from northwest of Prince Edward Island at 43°04’S, 36°22’. A third specimen was questionably referred to this species; it was taken south of the Cape Verde Islands at 05°27’N, 21°41’°W. Naef’s work on fossil cephalopods (1922) included a classification of both fossil and Recent cephalopods. The family Bathyteuthidae was listed as the first family under the suborder Metateuthoidea Oegopsida Naef, 1916. The genera Bathyteuthis and Ctenopteryx were listed. Grimpe (1922) erected several new taxa in association with the Bathyteuthidae. He listed as the first group in the oegopsids Bathy- teuthina, a new “family group” that included the Bathyteuthidae Pfeffer, 1900. In addition, he erected two new subfamilies, the Ctenopteryginae and the Bathyteuthinae. The former contained only Ctenopteryx, while the latter included Bathyteuthis and Indoteuthis Grimpe, 1922, p. 45 (an improper replacement name for Chunoteuthis Robson, 1921; see discussion under Grimpe, 1925). Grimpe considered that it might be a juvenile Bathyteuthis. No diagnoses are given for the new taxa. Grimpe gave the distribution of species in seas around Europe. B. abyssicola is recorded from the northwest Atlantic and from the Mediterranean with question, apparently querying Pfeffer’s (1912) record. Naef’s monograph (1923) has devoted a chapter (p. 251) to the Bathyteuthidae that includes a detailed diagnosis and discussion of the family, a generic diagnosis of Ctenopteryx, and a detailed descrip- tion of C. siculus. Apparently Naef had no specimens of Bathyteuthis abyssicola, for he referred readers to Chun’s (1910) coverage, upon which he must have relied heavily for his information about the genus. BATHYPELAGIC SQUID BATHYTEUTHIS 19 Naef again listed the Bathyteuthidae as being erected by Pfeffer in 1912 instead of 1900. In the first volume of the monograph, Naef (1921a, p. 48) listed the families and genera of cephalopods and gave Benthoteuthidae Pfeffer, 1900, and Benthoteuthis Verrill, 1885. It is curious that Naef always confused the dates of the familial synonyms. Naef emphasized that although Ctenopteryx and Bathyteuthis are not close relatives, they show special relationships to each other that are best expressed by retaining them in one family; together they represent a common contrast to all other oegopsids. According to Naef, the characters of the family are partly juvenile characters, and in part they may be regarded as primitive for the entire Oegopsida. In particular, Naef gave the shape of the gladius and the subterminal fins as general juvenile characters, and the occurrence of suckers on the buccal membrane and the suckers on the arms in four rows and on the clubs in many rows as primitive characters in all oegopsids. He gave a number of other characters which, though less clearly understood, indicated the primitive nature of the Bathyteuthidae. Naef presented his ideas on the phylogeny of the Cephalopoda in a phylogenetic bush (p. 795). He showed the Bathyteuthidae aris- ing obscurely with the rest of the Oegopsida and coming off as the oegopsid family closest to the Myopsida. Finally, Naef gave a sys- tematic review of the dibranchiates (p. 809) in which he included the Bathyteuthidae as the first family in the Oegopsida, with two genera Bathyteuthis and Ctenopteryzx. Odhner (1923) reported on the small collection of cephalopods taken by the Swedish Antarctic Expedition, 1901-1903. He recorded one damaged specimen of B. abyssicola 50 mm long (total length?) which “agreed in all details with the illustration of Chun.” The specimen was captured at 48°27’S, 42°36’W. He gave the range of the species in the subantarctic area based on the Challenger, Valdivia, and Scotia specimens. Although he gave no discussion, Odhner apparently based his decision to use the name Bathyteuthis abyssicola on the arguments presented in Hoyle’s rejection (1912) of Chun’s statement of priority for Benthoteuthis megalops. A major report on the cephalopods collected by the Prince of Monaco was published by Joubin in 1924. Three specimens of Benthoteuthis megalops were recorded from the Azores- Portugal region of the eastern Atlantic. Again, Joubin added little to the description of the species except the statement that the head is more darkly colored than illus- trated by Chun. In addition, Joubin stated that the little white pearls (photophores) at the bases of the arms are replaced by masses of dark chromatophores. This character is quite variable, however, and de- 20 U.S. NATIONAL MUSEUM BULLETIN 291 pends upon the age and the state of preservation of the specimens. Throughout his work Joubin utilized only the name Benthoteuthis megalops, apparently relying on Chun’s and Pfeffer’s usage. The final section of Grimpe’s major work on the cephalopod fauna of the North Sea (1925) consists of an annotated systematic review. The scheme of classification is similar to that given in his work of 1922 including the use of the “provisional family-group” Bathyteu- thina, Bathyteuthidae Pfeffer, 1900, and the subfamilies Ctenoptery- ginae and Bathyteuthinae. Grimpe maintained that the differences between Ctenopteryx and Bathyteuthis were so considerable as to war- rant their subfamilial separation. In the footnote to Bathyteuthis, Grimpe fully concurred with Hoyle’s arguments for priority of that genus. Referring to the myopsid-reminiscent characters of Bathyteuthis and Ctenopteryx mentioned by Chun (1910) and Naef (1921a), Grimpe placed the family at the beginning of the Oegopsida in the special family-group Bathyteuthina to indicate that a certain relation- ship exists between it and the Myopsida. The footnote for Zndoteuthis emphasizes the brevity and inconclu- siveness of Robson’s description of Chunoteuthis (1921) and Grimpe questioned the placement of the form and, in fact, the reality of the genus. Grimpe (1922) had changed the name to J/ndoteuthis because it too closely resembled Chunioteuthes Grimpe, 1916, a cirrotuthid octopod. A change of this nature, however, is illegal under the Code of Nomenclature; /ndoteuthis is a junior objective synonym of Chuno- teuthis, which is a nomen dubium. Massy (1928) briefly described B. abyssecola in her report of the cephalopods of the Irish coast. Massy first listed the specimen in 1916; it was taken at 50°22’N, 11°40’W by the Helga in a midwater otter trawl that fished in 700-750 fathoms. The larva, only 3.5 mm in mantle length, had distinct light organs and pale reddish coloration. Massy gave the distribution and vertical range from the literature. The first description of the male reproductive system of Bathy- teuthis was presented by Robson (1932a) who used the name Bentho- teuthis. Robson found two male specimens in the cephalopod material obtained by the RRS Discovery; only the larger one (40 mm mantle length) was mature. The “most striking” external feature, Robson noted, was the total absence of hectocotylization, and in this feature Benthoteuthis agrees with “essentially abyssal forms.” He gave a very brief description of the genitalia but deferred a more detailed discus- sion until a later date. Robson concurred with Chun that the ink-sac ismuch reduced and must be regarded as atrophied. Robson’s report on the vast collections of decapod cephalopods taken BATHYPELAGIC SQUID BATHYTEUTHIS | by the Discovery never appeared, so we have no additional details on the structure of the male genitalia. Furthermore, we have no clarifica- tion of Robson’s statement (p. 375), “. . . it will be seen that the speci- mens obtained by the Discovery are probably not referable to B. megalops, the single species hitherto described.” Since Robson’s Discovery specimens came from the same Antarctic localities (ca 55°S, 52°W) that have been thoroughly collected by the Z7tanin as well, it is doubtful if they represent anything other than Bathyteuthis abyssicola. Thiele (1935) listed the family Bathyteuthidae and gave a short diagnosis for the family and for Bathyteuthis (with Benthoteuthis as as synonym) and Ctenopterya. He mentioned Robson’s (1921) Chuno- teuthis minima as a bathyteuthid and Grimpe’s (1922) subsequent changing of the generic name to /ndoteuthis. Allen (1945), reporting on the planktonic cephalopod larvae of eastern Australia, split off the family Ctenopterygidae from the Bathyteuthidae based on the literature and on one larval specimen of Ctenopteryx sicula. Apparently, she had no specimens of Bathy- teuthis and relied entirely on the description of this form in Hoyle’s Challenger report. Although Allen’s decision was based on totally irrelevant taxonomic characters, current information indicates that she was correct in her elevation of Grimpe’s subfamilial designation. This is discussed in more detail in the section in which Ctenopteryx and Bathyteuthis are compared. In a posthumous publication Robson (1948) reported on the cepha- lopods caught during the Arcturus Expedition of 1925. Contrary to his previous publications, Robson used the family name Bathyteuthidae and listed ten specimens of B. abyssicola taken around the Galapagos and Cocos Islands of the eastern Pacific., He stated that he was in agreement with Naef and Grimpe in believing that Hoyle’s B. abys- sicola had priority over Verrill’s Benthoteuthis megalops. Robson described one small specimen (mantle length 7.0 mm) as Bathyteuthis sp. which he believed differed from B. abyssicola in “head and body-shape and in general proportions...” (p. 117). Then he stated, “On the whole the features agree fairly well with abys- sicola.” Fortunately, on this occasion he was unwilling to base the type of a new species on such a young and immature specimen. The Bathyteuthis sp. probably falls within the normal range of variation for abyssicola, particularly if it differs only in shape and general pro- portions, features that are readily altered by rigors of capture and preservation. Voss (1956) reviewed the extensive collections made in the Gulf of Mexico by the U.S. Bureau of Commercial Fisheries R/V Oregon. U.S. NATIONAL MUSEUM BULLETIN 291 22 606I ‘297 a8SaIUlld Lo70b1jsaauy Lojobysaauy ba ‘dxq uo} UBT g ‘dxq uo} xuUR[g Dy0sg DUIPIDA DLALPIV A DURIDIDA DMUDIDA $80.709Q]¥ $80.4.091P 880.709] ¥ $80.4.09]¥ Jabuaqyoy 9 uouipad xa 10 128894 you paeyory [Med 19440 J0}@Mprul you uo} yueld you uo} yueld Jou [BOI}IOA 4ou [BOT}IOA Jou [BOT}IOA you [BOT}IOA [Mzi} WIvaq V3.1] pei} Wveq 8318] [4e1} Wveq VBI] [Mv} WIv9q IBIVT [sey u103}0q L094) 00T—-0 028-0 0S9E-0 GLET —08<6T 00F O1ZP 0002 0002 008 00SG OFT] OTOT OOTT 0961 0062 suajau $4006 fig payovas yidaq BAIB] ol G BO G's ct 81 TI-6 &% LG z OF 1 $4979UL1)] TUL yjbua) a7jUD Ay = SN = I suaut -1ads fo LaQuUn ny UBdIO JIVUBT}Y YON [esuog jo Avg useyyNog jedsueg jo Avg useyynog ses] YSItIg Yo orpueyyyW YWON queting [euoyenby g ‘uB—I0Q0 JI}ULTTY wmBelyg JIny ‘ouepy WON puey soivon JO osvyodiyo1y sosvyO jo N ‘uvadQ uUvBIpUy uopsaQ Jo MY ‘UID UBIPUy eiyeuNng MN Yo ‘uvs09 uviIpuy svy[n3sy edeyg yo Yuolnyg vjensueg eulvueg jo j[ny ‘eleyy oden yO sureuvg jo sny ‘eye edeg YO pivAoutA 8,eyWVyy JO pivAoulA §,eyyIVY JO SpuUvIs] 19Z019 puv pIeMpy e0UlIg UesMjJoq UBIDQ Way NOG Daly sIqyneyAyyeg fo spuovas snoraaid fo sisdoufig—] @14aV J, M,ZF.6T N, POE H/SZ0.L8 N.80.60 H, 20098 Ns 160 M,OFOTT NsGG.0¢ MoI'FT Sot'S0 MoLG Nov OF MSTo8T S,GooTL G/F2.EL S,S0.60 H/ST.8L N,9So.v0 H/8F.86 N,€6.90 A SELT S,€S.9€ M.8%.6L N.900L0 M.¥¥T8 N,OE.90 M,96.0L N,6¥.6€ M.S8ToIL NiGF6E AH LE.8F S,IT9PF u01Dnr0T OZ6T ‘uIqnor BOT6I ‘Asse 8261 ‘OT6I ‘Asse ZI6L ‘9PM ZI61 ‘TAO OI6T ‘unyO FOGT ‘oso G88T ‘TIAA egst ‘asoH Loymny 23 BATHYPELAGIC SQUID BATHYTEUTHIS uo0baiQ snunpoLy SNUNJILY Ssninjosy SnUNnpoLy SnunjoLy SninjoLpy SninpoLy SNINpILy J fisaaoosig J fisaaoasiq [ fisaaoosig ZIGI ‘9927 a8 S9IUL I TI6T 992V ISSIIUL ‘dxq o1yoiejUuy YsIpomg ‘dxq ivpod ynog uBvULIex) [Mey 8B OF [avy uas1o}0g you MO} [aviy uasi9e}0g you M04 you MO4 you MOY [Mey uasie}0g [Mey uasi0}0g you veinog you voinog you uo} xuryd 4ou [BOIJIOA you [BOI}IAA 4ou [BOI}IOA OOFT O8cT OOTT O8ZI OOTT 000T 0€L og¢ O8ZI 00€T—-00GT 000€-0 00SE—-0 00S2 00ST 00SZ 0008 bP Go-G (cPINPCUIUIL, , €$ OF BAIR] 0€ OS + TT Il re ms ON rl eS ost oe ‘gsoy} ATUO pouoTyUoUr ynq sueTUpeds Azacods1q JO 1equINU B peYy AT}JUeIVdde UOSqoY ¢ *So]0BJUE} BUIPN]OX9 YIZUT [BIOL 5 “UTeJIIOUN UOTJVOYWUEPT ¢ “sdojnbaw siyjnajoyjuag Jo odA J, z “ppooisshiqn siyynazhyjog Jo odAL, 1 ODIXdTT JO JINH puryst sosedyley puvys] sosvdeyer puvrysy sosedeery PUvIS[ S090 jo YyNOYG pus] SOd0D JO YYNOG puvys] sosedy[er) jo qs puvysy sosedyyey Jo WN puvys[ sosevdv[ery jo FN uva0Q ULeyNOG UBID_) UJIYINOG uBvIDg) ULIYNOG BILOPBIA, Sol0ZY uva0Q dTJURY YINOYg puvys] 9pio, adeg jog uva0Q URIpUy YING uvadQ IURTW YING M8ToL8 N/8G08% M.¥E.16 S/LT.00 M,LPo1T6 N,ZF.00 M.¥E.16 S,L1.00 M,00.L48 N,0G.%0 M,€F.98 N,ZGoE0 M.,0£.68 $,00.60 M.20.06 N,LT.00 MTGo88 N/OF.00 M.GE.0E S,S°8hoSS M,LT.06 NGPolE MTGoTG N,6009E M.9€06P S,LE.8F Mi T¥olG Ns LZ0¢0 H,ZZ.9E S,POEP H,€E.60 S,OT.S€ 9G6T ‘SSO A SFEI ‘UOSqoy ZE6T ‘UOsqoy F261 ‘UIqnor E261 ‘19UYPO TZ61 ‘OTPIYL 24 U.S. NATIONAL MUSEUM BULLETIN 291 B. abyssicola was represented by one specimen 44 mm in mantle length. Voss gave a detailed description of the specimen and in his remarks mentioned the possibility of the presence of two species of Bathyteuthis based on the presence or absence of light organs, a feature that gen- erally merits specific distinction. Dell (1959) worked up the B.A.N.Z.A.R.E. cephalopod collections. Although no specimens of B. abyssicola were captured during the ex- pedition, Dell included the species in his comprehensive list of the cephalopod fauna of the Antarctic Province. Powell’s (1960) compilation of Antarctic and Subantarctic Mol- lusca lists the family Bathyteuthidae and Bathyteuthis Hoyle with the type locality in the Southern Ocean. Powell listed only Odhner’s (1923) record for the range of the species in these waters and omitted the Antarctic and Subantarctic specimens recorded by Hoyle (1912), Thiele (1921), and Robson (1932). A. H. Clarke (1962) gave the distribution of B. abyssicola as cosmo- politan in temperate and tropical seas at depths of 0-2000 fathoms (0-8660 m); apparently he overlooked all the records from boreal waters. M. R. Clarke (1966), in a work published since the present research was completed, reviewed the previous records of B. abyssicola and gave additional records from the collections of the Discovery II. Clarke estimated the vertical range and ecological conditions occupied by the species. Roper (1968) gave preliminary descriptions of two new species of Bathyteuthis; B. bacidifera and B. berry are described in greater de- tail in the present work, and information about their distribution is included. Family BATHYTEUTHIDAE Pfeffer, 1900 Bathyteuthidae Pfeffer, 1900, p. 152, 171.—Chun, 1903, p. 68.—Hoyle, 1904, p. 33; 1904a, p. 3, 14; 1909, p. 271.—Chun, 1910, p. 185.—Berry, 1912, p. 645.— Hoyle, 1912, p. 282.—Naef, 1921, p. 5385; 1922, p. 298.—Grimpe, 1922, p. 45.— Naef, 1923, p. 251, 809.—Grimpe, 1925, p. 95.—Thiele, 1935, p. 970.—Robson, 1948, p. 116.— Voss, 1956, p. 142.—Powell, 1960, p. 185. Benthoteuthidae Pfeffer, 1912, p. xx, 323.—Berry, 1912, p. 645.—Naef, 1916, p. 14.—Joubin, 1920, p. 57.—Thiele, 1921, p. 453.—Naef, 1921a, p. 48.—Joubin, 1924, p. 74. Not Benthoteuthidae—Robson, 1921, p. 482 (for Chunoteuthis minima). Typr-Genus.—Bathyteuthis Hoyle, 1885. Dracnosis.—Mantle-funnel locking apparatus simple with straight ridge and groove; buccal connectives attach dorsally to arms IV; club short, unexpanded, with more than four rows (8-10) of numerous, small suckers; arms with two, increasing to more than two rows (4) of irregularly arranged suckers; lappets of buccal membrane with BATHYPELAGIC SQUID BATHYTEUTHIS 25 small suckers; fins subterminal; gladius simple with long, narrow rhachis, broad, thin vane and open, thin terminus (no conus). Genus Bathyteuthis Hoyle, 1885 Bathyteuthis Hoyle, May 1885, p. 272; 1885b, p. 282, 308; 1886, p. 36, 167, 232, 236; 1886a, p. 247; 1886b, p. 1064.—Goodrich, 1892, p. 318.—Pfeffer, 1900, p. 172.—Chun, 1903, p. 69, 71, 77, 85-90.—Hoyle, 1904, p. 33; 1904a, p. 3, 14; 1909, p. 271.—Chun, 1910, p. 186.—Hoyle, 1910, p. 408; 1912, p. 282.— Pfeffer, 1912, p. 324.—Berry, 1912, p. 645.—Abel, 1916, p. 98.—Massy, 1916, p. 114; 1916a, p. 241.—Naef, 1921, p. 535; 1922, p. 298.—Grimpe, 1922, p. 45.—Naef, 1923, p. 795, 809.—Odhner, 1923, p. 1—Grimpe, 1925, p. 95, 101.— Massy, 1928, p. 30.—Thiele, 1935, p. 970.—Allen, 1945, p. 328.—Robson, 1948, p. 116.—Voss, 1956, p. 142. Benthoteuthis Verrill, July 1885, p. 401.—Hoyle, 1885b, p. 282, 308; 1886, p. 36, 167, 169, 232.—Chun, 1910, p. 185, et. seqq.—Hoyle, 1910, p. 408.—Joubin, 1912, p. 396.—Pfeffer, 1912, p. 324, et. seqq. (pars).—Berry, 1912, p. 645.— Abel, 1916, p. 93.—Joubin, 1920, p. 57 (pars).—Thiele, 1921, p. 453.—Naef, 1921a, p. 48.—Joubin, 1924, p. 74.—Robson, 1932, p. 375. ?Chunoteuthis Robson, 1921, p. 482 (nomen dubium).—Grimpe, 1922, p. 45; 1925, p. 95, 102.—Thiele, 1935, p. 970. ?2Indoteuthis Grimpe, 1922, p. 45 (replacement for Chunoteuthis Robson, 1921; not Chunioteuthis Grimpe, 1916) ; 1925, p. 95, 102.—Thiele, 1935, p. 970. Typr-species.—Bathyteuthis abyssicola Hoyle, 1885. By monotypy. Draenosis.—Tentacular clubs with 8-10 rows of numerous, minute suckers; arms I-III with 2-4 rows of irregularly placed suckers; arms IV with 2 rows of suckers; fins short, round, subterminal; eyes directed anteriorly; color a deep maroon; simple photophore at aboral base of arms I-III; arms I-III connected by a broad “web”; integu- mentary layers thick, honey-combed, semigelatinous; liver with large oil chambers; hectocotylus absent. Bathyteuthis abyssicola Hoyle, 1885 PLATES 1-5 Bathyteuthis abyssicola Hoyle, May 1885, p. 272-8, fig. 108; 1885b, p. 282-8, 309-10, fig. 2; 1886, p. 36, 168-9, 203, 213, 229, pl. 29, figs. 1-7; 1886a, p. 247.— Pfeffer, 1900, p. 173.—Hoyle, 1904, p. 33, pl. 1, fig. 2; 1909, p. 271.—Chun, 1910, p. 186.—Hoyle, 1910, p. 408; 1912, p. 273, 282.—Pfeffer, 1912, p. 327, et seqq.—Massy, 1916, p. 114; 1916a, p. 241.—Berry, 1917, p. 7—Grimpe, 1922, p. 45.—Odhner, 1923, p. 1—Grimpe, 1925, p. 95, 101.—Massy, 1928, p. 30.— Thiele, 1935, p. 970.—Allen, 1945, p. 328.—Robson, 1948, p. 117.—Carcelles, 1953, p. 226.—Voss, 1956, p. 142, fig. 11d.—Dell, 1959, p. 104.—Powell, 1960, p. 185.—Clarke, A. H., 1962, p. 75.—Clarke, M. R., 1966, p. 166. Benthoteuthis megalops Verrill, July 1885, p. 402-3, pl. 44, fig. 1—Hoyle, 1885b, p. 282; 1886, p. 36, 169.—Chun, 1910, p. 185-199, pl. 24-27.—Hoyle, 1910, p. 408; 1912, p. 282.—Joubin, 1912, p. 396.—Pfeffer, 1912, p. 325-331, pl. 27, figs. 12-138 (as “Benthoteuthis abyssicola Verrill’’) (pars; figs. 14, 15= Ctenopteryz).—Joubin, 1920, p. 57, pl. 18, fig. 4—Thiele, 1921, p. 4538.— Joubin, 1924, p. 74.—Robson, 1932, p. 375, figs. 1-2—Johnson, 1934, p. 161. 26 U.S. NATIONAL MUSEUM BULLETIN 291 Bathyteuthis megalops—Hoyle, 1886, p. 36, 213, 229; 1886a, p. 247. Not Benthoteuthis sp.—Joubin, 1920, p. 57, pl. 12, fig. 9 (=Ctenopteryz). ?Chunoteuthis minima Robson, 1921, p. 432, pl. 65, fig. 2 (nomen dubium) .— Thiele, 1935, p. 970. ?*Indoteuthis minima—Grimpe, 1922, p. 45 (replacement for Chunoteuthis). ?Bathyteuthis sp. Robson, 1948, p. 117, fig. 1. Dracenosis.—Protective membranes low, fleshy, no enlarged trabecu- lae; tentacles and clubs relatively short; arm suckers relatively few; sucker rings with 8-18 protuberances; gills short, narrow. List of Material Antarctic Ocean The majority of specimens that were captured by the Hltanin in the Southern Ocean were taken by 10’ Isaacs-Kidd midwater trawls; they are listed below followed by the few that were taken by other types of gear. Complete station data are recorded in Savage and Cald- well (1965) and “University of Southern California” (1966). Depth of Eltanin Number Sizerange ML, Number Size range ML, capture, Sta. No. of males millimeters of females millimeters meters 97 - - 3 31-40 1830 99 3 10-21 2 11-19 1210 5 larvae, size range 8.5-11 mm 110 1 43 1 48 2891 123 2 38-44 1 48 2439 125 2 36-38 2 19-27 1830 132 2 41-43 - - 1219 133 ~ ~ 2 32-44 2196 137 5 28-41 5 22-53 1556 142 1 19 1 54 1830 148 - - 1 20 1226 149 1 32 1 37 2105 154 13 33-46 i 31-56 2105 175 - - 3 15-38 2893 190 - _ 3 21-36 2891 213 2 14-21 - - 550 1 larva, 6 mm 215 1 13 1 17, 1220 235 3 19-41 1 32 1830 247 1 49 1 41 1830 248 6 28-40 6 42-45 1370 252 1 42 3 13-46 1570 262 1 35 - - 2400 274 1 a5 4 31-55 1875 275 - ~ 2 36-40 1885 296 1 37 - _ 1880 BATHYPELAGIC SQUID BATHYTEUTHIS Eltanin Number Sizerange ML, Number Sta. No. — of males 310 - 313 ~ 325 = 354 355 3 © 360 361 woe 364 368 381 382 383 388 392 449 571 580 1 592 = —_ Ke or Onl | 597 2 605 626 668 683 687 691 692 714 718 737 = 741 742 —_ Ce | wore 743 1 767 771 778 woe 779 782 789 792 793 796 | wnnnd wv 321-534 O—69——3 Size range ML, millimeters of females millimeters - 2 13-43 2 larvae, 6 mm = 1 11 25. 5-47 5 42-54 41-49 1 49 1 larva, 9 mm 37 2 12-55 15-38 2 25-50 1 larva, 7 mm - 1 47 - 2 19-20 28-48 - = 20-42 5 18-56 22-44 4 15-43 22-37 2 25-37 44 3 45-52 - 1 35 = 1 49 29 - - 1 larva, 9 mm 24-40 - = 35 - = - 1 55 - 1 33 33 - = 35 2 22-48 - i 47 — 2 26-32 - 1 22 23 2 41-43 - 1 51 17 - - 13-21 1 16 2 larvae, size range 8-12 mm 17 4 13-29 1 larva, 9 mm 37 - = 32-42 1 45 = 2 12-16 1 larva, 10 mm 34-41 3 31-48 14-36 4 34-52 27-31 1 50 12-45 2 27-33 19-40 - - - 1 52 1 larva, 7 mm 27 Depth of capture, meters 1425 810 933 2150 2440 1680 2100 713 914 1870 1280 1695 1420 2330 1610 1491 2743 2562 1922 1510 1647 1280 1690 2214 1510 1034 1097 1065 2575 2325 864 1830 1373 2400 1251 1910 2860 2196 1481 2260 2475 28 U.S. NATIONAL MUSEUM BULLETIN 291 Depth of Eltanin Number Sizerange ML, Number Size range ML, capture, Sta. No. — of males millimeters of females millimeters meters 802 4 28-40 3 42-49 1603 812 1 34 5 30-53 823 832 - ~ 2 39-46 2200 836 5 28-48 2 35-45 2055 846 4 20-42 2 35-63 1866 847 6 34-47 3 15-41 1991 850 2 15-23 3 43-46 1785 852 - - 2 11. 5-48 2917 854 - ~ 1 26 1285 855 2 22-53 1 25 2320 858 - 30-52 5 31-56 2099 864 - - 3 24-39 1285 866 1 13 1 14 1127 867 2 36-43 5 31-55 2642 874 2 12-16 1 24 1391 877 1 25 ~ - 1718 891 1 19 - - 1347 949 - - 1 33 1028 1038 - - 1 47 1281 1057 ~ - 1 25 812 1099 1 22 - ~ 956 1106 1 37 - - 769 1107 1 16 - - 714 1132 - - 2 45-49 1603 1133 1 38.5 6 45-61 685 1137 1 36 - - 626 1162 - ~ 1 20 803 1163 ~ - 1 28 626 1170 4 11-14 4 11-13 1080 1185 1 46 1 36 2200 1187 2 31-46 2 36-56 1550 2 unmeasured males 1195 1 41 - _ 1650 1201 10 14. 3-49 10 18-46 1120 1214 1 49 3 17-21 2150 1216 2 17-34 2 15-16 1000 1218 2 15-38 3 15-42 1850 1224 - - 1 34 1600 1236 2 23-24 - - 900 1262 2 12-21 3 12-21 1230 1269 10 14-45 t 15-27 1248 1270 2 16-38 - - 933 1286 1 26 4 34-38 2500 1287 4 15-40 4 28-37 2269 1288 1 15 2 13-39 2620 1 larva, 10 mm 1299 1 24 2 17-21 1190 1302 1 15 - - 685 1303 3 16-39 4 15-45 1281 BATHYPELAGIC SQUID BATHYTEUTHIS 29 Depth of Eltanin Number Sizerange ML, Number Size range ML, capture, Sta. No. — of males millimeters of females millimeters meters 1304 1 12 1 16 864 1307 6 18-50 - a 1373 1319 - - 6 37-48 1867 1320 1 47 4 33-58 2288 1323 2 20-48 2 36-41 2560 1324 a 26-46 8 21-49 1958 1327 7 33-46 2 37-45 2060 1328 6 18-46 5 35-48 1775 1358 2 41 4 35-44 1702 1359 1 17 3 30-52 2416 1361 1 46 1 22 1903 1364 8 22-35 3 22-37 1848 1365 3 21-43 2 34-56 2434 1376 - - 1 13 705 1 larva, 10 mm 1383 9 15-49 4 18-54 914 1384 1 21 i 32 1262 1388 2 22-27 3 19-40 942 1389 1 46 - - 1710 1392 - - 3 16-35 897 1393 4 15-34 2 22-43 1537 1396 - - 1 46 2525 1448 6 22-42 i 17 2275 1454 1 20 - - 825 1456 2 40-43 2 49-54 1000 1462 4 23-29 1 21 1010 1470 1 32 - - 1800 1471 1 32 = = 311 1473 1 21 - - 415 The following specimens were collected in Blake trawls, beam trawls, and dredges. Depth of Eltanin Number Sizerange ML, Number Size range ML, capture, Sta. No. — of males millimeters of females millimeters meters 230 1 44 - ~ 1150 353 1 45 1 37 3642 450 - - 1 54 1110 462 ~ ~ 1 48 3404 870 1 18 = = 5014 1117 = oa 1 46 4813 1148 2 34-35 - - 4850 1199 1 41 - ~ 4374 1292 ~ ~ 1 34 4941 1363 1 36 2 16-32 2763 30 Ser ML M 56mm F 49 mm F 47 mm F 45 mm M 44mm F 39 mm F 38 mm F 37 mm F 36mm M 33mm F 30mm M 25mm M 24mm M 24mm F 23 mm F 22 mm ie 20 mm M 19mm- F 19 mm M 17mm M 17mm L! 16mm F 16 mm F 15 mm M 15mm F 14 mm L 14 mm L 138 mm L 13 mm L 12 mm L 12 mm i) Aldeamm L 10mm L 9mm L 9mm L 8mm L 8mm U.S. NATIONAL MUSEUM BULLETIN 291 Atlantic and Eastern Pacific Oceans Ship Sta. Pel. 11 Ch. 17-18 D3981 I SB 2024 O 4296 Pil. 20 Ch. 17-13 Ch.(RHB) 978 D3996 I D4000 I D1208 XIV Pil. 300 D1208 XV D1208 XV D4003 I D1208 XV Pil. 37 D3998 VIII D1209 I Pil. 39 Pil. 21 D4000 VI D1162 II D1162 II D1022 D1165 VIII D3998 XI D3998 II D1209 III D4003 IV D4003 IV Pil. 300 D4003 V D3998 X D4201 XIX D4005 III D3996 II Location 27°52’N 79°45’W Sargasso Sea 19°16’S 01°48’W 28°26.5'N 80°11.5’W 07°55'N 53°55’W 04°56’N 00°13’E 05°N 15°W 01°44’S 27°44’W 15°41’S 05°50’ W 00°31’S 11°02’W 06°48’N 80°33’W 02°05’N 04°50’E 06°48’N 80°33’W 06°48’N 80°33’W 08°26’N 15°11’W 06°48’N 80°33’W 04°00’N 02°46’W 07°34’S 08°48’W 07°15’N 78°54’ W 04°24’N 03°00’ W 05°07’N 00°05’E 00°51’S 11°02’ W 13°35'N 30°11/W 13°35’N 30°11/W 39°51/N 48°20’W 12°11’N 35°49’W 07°34’S 08°48’W 07°34’S 08°48’ W 07°15’N 78°54’W 08°26’N 15°11’/W 08°26’N 15°11’W 02°05’N 04°50’E 08°26/N 15°11’W 07°34’S 08°48’ W 47°02’N 31°45’W 13°31’N 18°03’W 15°41’S 05°50’ W Date 11 III 56 61 19 II 30 27 IV 60 22 III 63 27 «VV «64 IV 61 27 II 63 25 II 30 4 III 30 16. 122 24 V 65 16 122 16 I 22 9 III 30 16 122 29 V 64 1 III 30 te 1 22 30 ~V 64 27 ~V 64 4 III 30 6 XI 21 6 XI 21 16 IX 13 9 XI 21 1 III 30 1 IIL 30 17 122 9 III 30 9 III 30 24 V 65 9 III 30 1 III 30 20° VES 12 IIT 30 25 I1I30 'L refers to larval or juvenile specimens of undetermined sex. Depth, meters 430 500 | 36 914 1650- 2125 750 1000-— 3500 2000 500 1550 2000— 3000 1300 1300 3000 1300 490 2000 1750 300 1150- 1450 3000 200 200 1500 1500 500 200 1250 1500 1500 2000— 3000 1000 1000 1500 1500 1500 Gear 40’ Otter trawl 10’ IKMT E 300 10’ seallop dredge 65’ shrimp trawl 10’ IKMT 10’ IKMT 10’ IMKT EK 300 E 300 S 150 10’ IKMT S 150 S 150 E 300 S 150 10’ IKMT S 150 8S 150 10’ IKMT 10’ IKMT E 300 5 200 5S 200 P 200 S 150 S 150 S 150 S 150 S 150 5S 150 10’ IKMT S 150 S 150 S 200 8S 150 S$ 150 BATHYPELAGIC SQUID BATHYTEUTHIS 31 Descriprion.—The mantle is short and broad; it is widest just posterior to the mantle opening. The mantle tapers gradually to a bluntly rounded tip, giving it a bullet-shaped outline (pl. 1). In adults the greatest mantle width is about 48% of the mantle length, and the average width is 44%. The margin of the mantle opening has three small, anteriorly projecting triangular lobes that mark the articula- tion points of the two ventral mantle-funnel locking cartilages and the single dorsal mantle-nuchal cartilage. The wall of the mantle is rela- tively thick and muscular, although in individuals that have recently spawned there may be some degradation of the muscle tissue and a thinning of the mantle wall. The integumentary layers on the mantle form a thick sheath (pl. 2B). The outermost integument consists of small, closely aligned patches that give the skin a velvet-like appearance when observed in natural size. Under magnification, however, the outer integument looks like a very fine-meshed knotless net. Though somewhat variable, the meshes are primarily pentagonal, occasionally hexagonal. The margins of the web-units are very thin and membranous and stand perpendicu- lar to the plane of the mantle wall. These margins of the web are deep- ly colored by maroon-brown pigment. One or two chromatophores of the same color lie in the thin membrane that underlies and intercon- nects the walls of the web. A thin sheet containing chromatophores lies immediately beneath the outer reticular layer. The sheet of chromatophores overlies a rela- tively thick, watery-gelatinous, transparent, unpigmented layer. This gelatinous layer is supported by a network of ridges that are of greater density than the nearly fluid contents of the pockets made by the in- tersections of the supporting reticulation. These pockets are also pen- tagonal but are larger than those of the outer layer of integument. Two additional layers of maroon chromatophores are located be- neath the gelatinous layer. A thick elastic layer of connective tissue and fibers binds the integumentary layers to the muscles of the mantle wall. A few small widely spaced chromatophores le even in the con- nective tissue very close to the muscular wall of the mantle. The fins are small and nearly circular in outline (pl. 1). The bases of the fins are separated posteriorly by the blunt termination of the mantle; they are broadly separated anteriorly by the dorsal surface of the mantle. The bases of the fins are robust and muscular but radial- ly the fin muscle becomes thin and weak, so that the borders of the fins are thin, nearly membranous. The margins of the fins are damaged even in the best-preserved specimens. The rounded anterior fin lobes protrude well forward of the bases of the fins; the rounded posterior lobes extend beyond the end of the mantle. 32 U.S. NATIONAL MUSEUM BULLETIN 291 The funnel is long and narrow; it extends nearly to the level of the posterior edge of the eye opening (pls. 1p; 2a). Between the components of the locking apparatus the posteroventral margin of the funnel is broadly U-shaped and is weakly muscled. The tube of the funnel is long and narrow; the opening is small. The collar of the funnel is thin-walled and weakly developed; it does not form a strong anterior fold. The funnel retractor muscles are strong and robust. They extend posteriorly from the dorsal wall of the funnel and the bases of the locking-apparatus components to the posterior end of the mantle where they flare and insert along the shell-sac and the mantle wall. The bridles of the funnel are broad, thin, weak bands deeply embedded in the posterior depression of the funnel groove. Integument and subcutaneous gelatinous connective tissue blend the funnel smoothly with the surface of the head around the collar and along the lateral borders of the funnel tube. The anterior end of the funnel, however, lies free in the funnel groove. A small pore lies in the midline of the funnel groove; a tube leads dorsally from the pore, through the gelatinous tissue, and between the bridles. I am conducting a separate study of this previously unreported struc- ture which occurs in several oegopsid families. The cartilaginous components of the funnel-making locking ap- paratus belong to the simple type. The funnel component is a long, narrow structure with a smooth median sulcus that is deep and narrow anteriorly but shallow and flaring posteriorly (pl. 24, £). The funnel component is bordered anteriorly and laterally by a thin, mem- branous lip that fuses posteriorly with the thin, posteroventral edge of the funnel. The mantle component of the locking apparatus is a simple straight ridge that is most pronounced and distinct in its anterior half; it diminishes and broadens posteriorly until it is flush with the inner wall of the mantle (pl. 2a, p). The dorsal member of the funnel organ is a broad, inverted, Y-shaped glandular structure (pl. 2r). The diagonal limbs are long, slender, unsculptured, and rounded posteriorly. The median (anterior) limb is short and blunt; it terminates with an erect, spatulate spire that protrudes from the dorsal wall into the chamber of the exhalant tube. The ventral members of the funnel organ are elongate, rounded pads (pl. 2r). The median border of each is nearly straight and the lateral border is broadly curved. The anterior and posterior ends are rounded. The funnel valve is a large semicircular flap that extends across the dorsal surface of the funnel tube just posterior to the opening of the funnel (pl. 2r). A deep pocket is formed by the fusion of the BATHYPELAGIC SQUID BATHYTEUTHIS 33 lateral and posterior borders of the funnel valve with the dorsal wall of the funnel tube. The head is broad and short (pl. 1). The huge, anteriorly directed eyes give the head a distinctly swollen appearance. The head narrows markedly at the base of the brachial crown just anterior to the eye openings. The dorsal and ventral surfaces of the head are nearly flat and the head tapers anteriorly into the brachial crown. No nuchal crests or folds occur in the region of the neck. The nuchal component of the mantle-nuchal locking apparatus is a strong, elongate, cartilaginous structure with a straight median ridge bordered on each side by a smooth sulcus (pl. 1c). The edge of the nuchal component is bordered by a broad membranous skirt. The cartilaginous mantle component of the nuchal lock is com- plementary in structure to the nuchal component. A deep median sulcus is bordered on each side by a long, straight ridge. The mantle component lies directly ventral to the anterior terminus of the gladius, and its form is imposed by the chitinous rhachis that has a deep median groove ventrally and is supported by strong, rodlike edges. The olfactory papillae lie on the posterior ventrolateral surface of the head just anterior to the collar. They are short, small, slightly swollen, lobate, or flaplike structures that le beneath the anterior border of the mantle. The integumentary layers on the head are similar to those on the mantle. The gelatinous layer is particularly thick on the dorsal and posterior regions of the head. In general, the pigmentation on the head and arms is more concentrated than it is on the mantle, so that the color isa dark maroon. A small, simple photophore is located at the base of each of the first, second, and third arms (pl. 5a, £). No trace of a photophore is present at the bases of the ventral arms. The photophore is readily visible on smaller specimens as a dark, elongate, rounded patch, often with a pale central spot. In larger specimens, however, the light organs become less distinct because they are partially overgrown with integument and gelatinous tissue. The organs arise in one of the deeper chromatophore layers where they are surrounded by an area of maroon pigment which gradually diminishes away from the organ. In long-preserved speci- mens that have become bleached, the photophores usually are indis- tinguishable. The arms and tentacles extend from the base of the narrow brachial crown. The arms are short, subequal in length. The arm formula occur- ring with the greatest frequency is 4.3.2.1, followed by 4=3=2.1 and 4.3=2.1. The length of the arms is measured from the basal portion of the sucker-bearing area to the tip of the arm. The measurement on the 34 U.S. NATIONAL MUSEUM BULLETIN 291 ventral arms may be misleading in some specimens, because the sucker- bearing portion may originate more distally than on the remaining three pairs of arms, giving the impression that the fourth arms are shorter. Generally, however, the fourth pair of arms is the longest or equals in length the third and/or second arms. The first pair of arms is always the shortest. The four pairs of arms are connected by a fleshy web (pl. 3a, B) that is an extension of the integumentary layers of the head and arms. The web that connects the first arms is the deepest, and web depth decreases gradually between the first and second, the second and third, and the third and fourth arms. No web exists between the fourth arms. The web between the third and fourth arms is the continua- tion of the lateral membrane (or “tentacular sheath”) along the dor- sal aboral angle of the fourth arms. This thick, fleshy web makes the arms look shorter and the head (anterior to the eye opening) longer than they actually are. The arms are elongate cones, thick at the bases and evenly tapered to the tips (1.e., not sharply attenuate). Swimming keels and mem- branes exist on the aboral surfaces of the arms, but they occur only as fleshy ridges rather than as distinct, muscular (or membranous) keels. The keel on the-dorsal arms is a raised fleshy ridge along the distal half of the arm. The keel on the second arms is noticeable proxi- mally as a thickened ridge that becomes raised and thinner distally. The greatest development of a true swimming keel occurs on the third arms where the keellike ridge is raised and relatively thin (though still fleshy) along the distal three-fourths of the arm. On the ventral arms the lateral membranes, which partially envelop the tentacular stalks during swimming, are moderately developed, although they too are fleshy, particularly proximally, and not membranous as in most other oegopsids. All arms in B. abyssicola are supplied with protective membranes (pl. 3a) that are extremely variable; they are thick and fleshy. Typi- cal thin, muscular, strutlike trabeculae are lacking. The membranes appear to be the result of the fusion of the edges of flattened trabec- ulae. The membranes arise abruptly on the oral surface of the first three pairs of arms, about in line with the base of the proximalmost sucker. Along the proximal two-thirds to three-fourths of the first three pairs of arms the protective membranes are well developed and the edges stand well above the level of the suckers. Distally, how- ever, the membranes become merely low ridges that form the oral angles of the arms primarily; suckers stand well above the borders of the membranes. Borders of the protective membrane may be straight, gently undulate, or scalloped to varying degrees. The protective mem- BATHYPELAGIC SQUID BATHYTEUTHIS 35 branes of the fourth arms are very weakly developed along their entire length. The oral surfaces of the arms are covered with numerous, small suckers that are irregularly arranged in up to four longitudinal rows particularly on the dorsal three pairs of arms (pl. 3a). One or two suckers arise proximally close to the midline of the arm; these may be in asingle row or slightly biserial. The next several pairs of suckers are generally clearly distributed in two biserial, often well-separated rows. It is difficult to determine exactly how many longitudinal sucker rows exist, because about a third to halfway out the arms the distribution of suckers becomes irregular. Generally, however, the rows can be separated down the midline of the arm where no suckers occur so that on each side a band of biserially arranged, closely packed suckers occurs that continues toward the tip. When the sucker stalks on each side of the midline emerge from the oral surface of the arms they are in biserial arrangement, but beneath the surface their bases arise nearly in line. The sucker arrangement on the ventral arms usually remains uncomplicated with a normal biserial distribution. Close to the arm tips the suckers diminish in size and number. The sucker-bearing por- tion of the arm terminates proximal to the extreme tip of the arm, and the last traces of the protective membranes fuse in the middle; a sudden decrease in the diameter of the arm occurs distal to this point. The extreme tip of the arm may be entirely devoid of suckers, or one or two minute suckers may be present (pl. 3c). In either case the tip is short, almost blunt. The texture of the arms is the same as that of the head and mantle; the arms are covered with the reticulate integument and the subcu- taneous gelatinous layer. The tentacles are extremely long, thin and muscular (pl. 1). They range in length from about 100% to 130% of the mantle length. (These values may vary considerably depending on the amount of contraction at the time of fixation.) The tentacular stalks are nearly round in cross section except along the flat to slightly concave oral surface. The longitudinal concavity or depression extends nearly the entire length of the tentacular stalk; it is less pronounced toward the base of the stalk. Occasionally a slightly raised ridge appears in the depression, but this is a factor of contraction of muscles during fixation. The tentacular club is relatively short, narrow, and unexpanded (pl. 4a) ; it belongs to the simple type and is not divided into distinct regions of carpus, manus, and dactylus. The club occupies from about 20% of the mantle length in younger specimens up to about 30% in adults. The club length is about 20% of the tentacle length. The tentacular stalk maintains about the same diameter along its entire 36 U.S. NATIONAL MUSEUM BULLETIN 291 length to the base of the club, but the club tapers gradually to a blunt termination. The sucker-bearing portion of the club is covered with numerous, minute, closely packed suckers. The suckers originate with a single sucker proximally; the next few suckers are shghtly scattered and set apart, but the suckers increase in number rapidly and become very closely packed. Suckers are evenly distributed over the club so that no clearly defined pattern of rows exists; however, eight to ten longitudinal rows of suckers occur on the sucker-bearing surface for nearly the entire length of the club. Suckers decrease in numbers only at the very proximal and distal ends of the club. The sucker-bearing portion of the club terminates in a bluntly rounded area of suckers. An extremely small, papilla-like tip protrudes distal to the sucker-bearing area and is continuous with the aboral sur- face of the club (pl. 48). A few minute knobs are located on the prox- imal portion of the terminal papilla; these knobs probably are precursors to suckers. Possibly the terminal papilla represents the growing portion of the tentacular club. No protective membranes exist along the borders of the club; only a slight, pigmented line outlines the sucker-bearing portion of the club, especially in its proximal half. A thin, low keel originates on the dorsal aboral surface of the tentacular stalk just proximal to the first (proxi- malmost) sucker of the club. The keel extends distally and broadens gradually in the distal third of the club; it terminates near the tip of the club just proximal to the end of the sucker-bearing area. No addi- tional keels or membranes are present. The suckers on the club are small and numerous; about 525 suckers, the largest 0.14 mm in diameter at the aperture, occur on the club of a specimen 56 mm in mantle length. The diameter of the sucker aper- tures in adults varies between 0.08 and 0.14 mm, depending upon the size of the specimen. Some variation in dentition occurs, but in general about 8-12 teeth and knobs are set around the aperture. The teeth that occur on the distal half of the border generally are truncate, longer than broad, and widely separated, although occasionally the teeth will be shorter, almost square, or rounded, or very occasionally, triangular. The larger suckers of a club from a specimen 28 mm in mantle length (Elt. 1201; pl. 47) have about 8-12 points around the aperture; those with 8 or 9 points have 4 to 5 long, truncate, widely spaced teeth and 4 to 5 small, rounded bumps or knobs; those with 11 or 12 points have 6 teeth and 5 to 6 knobs. Sometimes two teeth will be closely set and very narrow as though earlier in ontogeny they had been a single tooth that split in two. Larger specimens have club suckers with a few more enlarged teeth so that rings with 9-13 points may have 5-8 long, trun- cate teeth and 4 to 5 knobs (pl. 4K). Some large teeth are notched on BATHYPELAGIC SQUID BATHYTEUTHIS 37 the ends, and sometimes the notch is so deep that the tooth will have two small, pointed cusps. The largest specimen in the collection, 75 mm in mantle length, has as many as 17 protuberances around the aperture; of these, 7 are proper teeth, truncate to rounded and a little more closely packed than usual, while the remainder are merely low knobs that give a scalloped appearance to the posterior half of the border of the ring. Occasionally large suckers will have nearly smooth or scalloped apertures. The sucker rings from the arms of a specimen 56 mm in mantle length are described first. The outer rings on the suckers of the arms bear 3 to 4 concentric rows of small chitinous papillae (pl. 41). The papillae are minute knobs on the outer margin and gradually increase in length on the rows toward the aperture. The innermost row that borders the aperture has numerous, small, elongate papillae that project into the aperture and give the appearance of being small teeth on the inner sucker rings. Together the teeth of the inner sucker ring and the papillae of the outer ring make a double-rowed armature. The inner chitinous sucker rings are subglobular; the walls are broadest distally, and they taper to their narrowest dimension proxi- mally. The sucker rings on the first arms bear 9-15 small, truncate to rounded teeth. The maximum diameter of the apertures is about 0.20 mm. The largest sucker ring aperture on the second arms is also 0.20 mm. The dentition of the rings is variable. The larger rings have about 12-18, usually 14-16 small, low, truncate, subtruncate, or slightly pointed teeth. The inconsistency in dentition holds for smaller suckers also, but they generally have about 9-12 teeth. The sucker rings may be nearly smooth or scalloped, or they may have truncate or pointed teeth. The suckers on the third arms have maximum ring apertures of about 0.20 mm. The rings bear from 9-18 very small teeth, which are generally truncate, but sometimes rounded in shape. The sucker rings on the fourth arms are about 0.16 mm in maximum diameter, and they bear between 8 and 18 teeth, more frequently 10-14. Generally the smaller rings have fewer teeth but this does not always hold; one me- dium-sized ring had 18 closely spaced teeth. Normally the larger suckers have 10-14 teeth. The teeth usually are truncate to slightly rounded with their bases set apart, but occasionally the teeth will be very closely packed. A few rings are nearly smooth to slightly scalloped. Since dentition varies slightly with the size of the animal, the fol- lowing description of suckers from a specimen 38 mm in mantle length is given. The diameter of the outer ring from the largest sucker on arm I is 0.22 mm and of the inner ring aperture 0.16 mm (pl. 4c). The largest 38 U.S. NATIONAL MUSEUM BULLETIN 291 rings have 12 teeth and the medium-sized rings have 9-10. The teeth are always small; they are more elongate and truncate in the distal portion of the ring and shorter, more rounded, and knoblike proxi- mally. The teeth on the larger rings often are separated at their bases by the widths of several teeth; teeth are widely spaced on the smaller rings. The rings from the first arms of a specimen of 28 mm mantle length bear eight teeth: four long, truncate, widely spaced teeth distally and four small, round, widely set knobs proximally (pl. 4p). These rings measure 0.12 to 0.14 mm in diameter. The outer ring has 3 to 4 concentric rows of chitinous knobs that become elongate and toothlike on the inner row. The outer ring fits closely around the aperture of the inner ring; in this position the inner row of toothlike knobs serves as an additional row of teeth that are very closely associated with the teeth of the inner ring. On the second arms the outer rings measure 0.24 mm and the inner rings 0.14 mm (pl. 4n). There are 9 to 10 teeth around the aperture. The teeth are generally short, truncate to rounded in shape; some are worn so that the tips are concave. The largest suckers on arm III have outer rings 0.24 mm in diameter and inner rings 0.16 mm in diameter. The larger inner rings bear about 12 evenly spaced teeth (pl. 4), while the smaller rings have 8 teeth that are more elongate and widely spaced than those of the larger rings. The diameter of the outer ring of the largest sucker on the fourth arm is 0.28 mm and of the inner ring 0.20 mm. On the largest inner rings appear 12 to 14 small, widely spaced truncate to rounded teeth (pl. 4a) ; 9 teeth appear on the medium-sized rings where the 3 to 4 distalmost teeth tend to be more elongate than the others. The suckers on the buccal lappets possess outer rings that measure 0.18 mm in diameter and inner ring apertures that measure 0.12 mm across. The 9 to 10 teeth on the inner ring are widely spaced and elongate, particularly along the distal border (pl. 41). The buccal membrane is large, fleshy, and rugose (pl. 3a). The buccal connectives are attached to the dorsal, dorsal, ventral, dorsal oral edges of arms I-IV respectively (pl. 38). The buccal lappets bear from 0 to 3: minute suckers that have 8-12 teeth on the chitinous rings (pl. 41). The significant geographical variation that occurs in the number of buccal suckers is discussed in detail later. The buccal mass, comprised of the muscular bulb that encases the mandibles, radula, and other mouth parts, is relatively smaller in B. abyssicola than in other oegopsids (e.g., ommastrephids, enoploteu- thids, gonatids). The beaks (pls. 3p, £) are small; they are darkly pigmented where the rostra are exposed, but the pigmentation light- BATHYPELAGIC SQUID BATHYTEUTHIS 39 ens on the lamellae. The rostrum of the upper beak is curved and sharply pointed; the curved jaw angle has a slight protuberance. The insertion plate of the palatine lamella is nearly two times longer than high and the posterior end is angled. The rostrum of the lower beak is relatively blunt and forms nearly a right angle with the long, nar- row rostral lamella. The gular lamella is broad; it is bluntly rounded posteriorly. The radula consists of seven transverse rows of pointed teeth and two rows of elongate lateral plates (pl. 3r). The rhachidian has a broad, low base; the median cusp is moderately long and bluntly pointed. No lateral cusps exist, but the basal, concave, lateral borders of the rhachidian are often set with irregular protuberances. Some of these may approach the size of lateral cusps, but still they remain as protuberances and are not cusps. The first lateral teeth are broadly crescent-shaped with convex medial and concave lateral borders; the blunt cusp is about as long as that of the rhachidian. The second lat- erals are long, slender and crescent-shaped; the cusps are slightly longer and more pointed than the first laterals or the rhachidian. The basal lateral borders of the first and second lateral teeth are also marked with slight irregularities or small protuberances, but these are not incipient cusps. The third lateral teeth are very long, slender, scythe-shaped ; they terminate in moderately sharp points. The lateral plaques are elongate plates set on the radula ribbon at about 45° to the long axis of the radula. The long sides are concave and the ends are bluntly rounded. In other specimens the plates become very knobby or irregular, or they may even be divided into two or more small masses. The gladius is embedded in the inner surface of the mantle wall. The circular muscles of the mantle are continuous over the gladius so that it is not visible in the dorsal midline. The anterior end of the rhachis les on the inner surface of the mantle covered only by a thin sheath. At its anterior tip the rhachis is covered by a thin cartilaginous band. This band and the exposed rhachis posterior to it make up the mantle component of the mantle-nuchal locking apparatus. Immedi- ately posterior to the nuchal locking apparatus the gladius is bound to the inner surface of the mantle wall by a strong, thick muscle that is an extension of the inner portion of the collar muscle. This sheath spreads posteriorly and ventrally and envelops the liver in a strong muscular sheath. The strong dorsoanterior band is continuous ventral to the mantle, but it diminishes posteriorly, so that the fibers of the liver sheath attach to the shell-sac along the lateral edges of the rhachis. The remainder of the gladius is visible in the shell-sac pos- teriorly extending from the heavy muscle band to the termination. 40 U.S. NATIONAL MUSEUM BULLETIN 291 The powerful funnel retractor muscle passes posteriorly over the ventrolateral surface of the liver, and in the region dorsal to the branchial heart it flattens out into a broad sheet. Dorsally it merges with the fibers of the liver sheath and attaches to the shell-sac along the edge of the rhachis and vane of the gladius. The funnel retractor sheath attaches to the lateral wall of the mantle. The vane of the gladius arises from the rhachis at a point in line with the branchial hearts. The vane broadens gradually posteriorly and curves ventrolaterally with the mantle wall. The visceral dome and the visceral-pericardial coelom lie directly ventral to the vane and are attached to the shell-sac with weak connective tissue. The relation- ship of the posterior end of the gladius with the tip of the mantle is difficult to ascertain, because the mantle tip always seems to be vio- lently contracted. The contractions push the vane of the gladius an- teriorly and ventrally in a series of irregular folds. The rhachis of the gladius is very long and narrow (pl. 2a). The free portion of the rhachis accounts for over half the length of the gladius. The lateral edges of the rhachis are strong, straight, cylindri- cal rods that are joined dorsomedially by a concave strip. A deep groove lies between the lateral rods; anteriorly the groove of the rhachis holds portions of the muscles of the collar and liver-sheath that are attached to the shell-sac. In the posterior half of the gladius the rods of the rhachis converge and taper gradually until they disap- pear in the posterior end of the vane. The vane is a thin, blade-shaped structure that arises at about the midpoint of the gladius. It broadens gradually to its widest point posteriorly then quickly terminates in a bluntly rounded tip. No conus is present. The anterior portion of the vane is nearly flat, but the broader posterior section is deeply concave, conforming with the shape of the tapering mantle. Considerable variation exists in the shape of the vane. The spermatophores (pl. 2H) are 4.5-7 mm in total length, depend- ing upon the size of the male. The proportions of the components, however, remain relatively constant. The sperm mass occupies about 68-72% of the total length; the cement body makes up 6-8% of the length, and the ejaculatory apparatus occupies the remaining 20-25%. The base of the ejaculatory apparatus is goblet-shaped; the basal sec- tion is a thickened collar that fuses with the cigar-shaped cement body. No hectocotylus exists. Hotoryrr.—British Museum (Natural History) ; BM 1890, 1.24.15. Typr Locatrry.—Between Prince Edward Island and the Crozets Islands at 46°16’S 48°27’E. HMS Challenger, 30 December 1873. Disrrisution.—Bathypelagic in the Southern Ocean and in the pro- BATHYPELAGIC SQUID BATHYTEUTHIS 41 ductive waters of the East Pacific, Atlantic, and Indian Oceans. Ranges between 300 and 3000 m, mostly 1000-2500 m (see Part IT). MorpHoMetry.—Measurements of various body parts have been made to determine allometric relationships and variability of propor- tions that occur in Bathyteuthis abyssicola during growth. All meas- urements are compared with the standard of size, the mantle length, which is plotted as the abscissa. Measurements were made on 121 speci- mens of B. abyssicola from the Southern Ocean. These specimens rep- resent about one-fourth of the total sample population that was avail- able at this stage of the study; the sample consists of 53 males and 68 females of all available sizes taken throughout the study area. The following measurements were made: mantle length, mantle width, head length, head width, basal fin length, total fin length, fin width, tentacle length, club length, arm lengths (I-IV). The values of these measurements are plotted as the ordinate against mantle length on scatter diagrams (figs. 1 to 13). The plots of all of the characters, with the exception of tentacle length and club length, show little spread in the values, and the rela- tions of mantle length to the other variates appear to be linear. The broader spread of points on the scattergrams for tentacle length espe- cially and less so for club length (figs. 7, 8) reflects a condition of preservation rather than of irregular or unordered growth. The elastic tentacles are subject to contraction or stretching, depending on the condition of the specimen at fixation. Even so, a more or less linear growth is indicated for the tentacles (fig. 7). Clubs make up a small proportion of tentacle length so they do not readily show the effects of preservation. Club length vs. tentacle length is plotted (fig. 9) ; a linear relation- ship is indicated here, too, although in the extreme upper and lower range of sizes club lengths appear to be plateaus. These may be a result of the conditions of preservation, or they may reflect a natural sig- moidal pattern of growth. Allometric growth appears to occur as a general feature in the squids that have beén studied in this regard. Chiroteuthis exhibits an extreme growth pattern in which drastic changes in proportions occur between the “Doratopsis” stage and adulthood. Also, Adam (1952, p. 63) has demonstrated that the fins and ventrolateral arms of males of Alloteuthis africana cease to grow allometrically above a certain size (age). In B. abyssicola the allometric relationships of mantle width, head width, and fin width, for example, may be slightly altered at the maximum size of the species, although too few very large specimens are available currently to substantiate this. Therefore, within the ex- tent of these data, B. abyssicola exhibits well-defined allometric growth curves. 42 U.S. NATIONAL MUSEUM BULLETIN 291 Discussion.—I have examined the holotypes of both Bathyteuthis abyssicola Hoyle, 1885, and Benthoteuthis megalops Verrill, 1885. Hoyle’s type in the British Museum (NH) is in fair condition only, but no particular points are in question about this specimen from the Southern Ocean. Light organs cannot now be detected at the bases of the arms, but, as mentioned in the preceding description in reference to recently preserved material, this is a feature of preservation and is not due to the absence of the photophores. The type of Benthoteuthis megalops, deposited in the U.S. National Museum, is in poor condition; it has dried up in the past and now is hard and unmanageable. That it is conspecific with the Atlantic form of Bathyteuthis abyssicola, however, is still verifiable. The nomenclatural problem that exists between Bathyteuthis abys- sicola and Benthoteuthis megalops seems not to have arisen until Chun (1910) announced that Verrill’s Benthoteuthis megalops had priority. Actually, Benthoteuthis was synonymized with Bathyteuthis in the same year that the two genera were introduced; Hoyle (1885c, p. 282) considered Verrill’s specimens at least congeneric with Bathy- teuthis and, further, determined that Benthoteuthis megalops was not published until July 1885, while Bathyteuthis abyssicola appeared in May 1885. In his Challenger Report, Hoyle (1886, p. 169) sug- gested that B. abyssicola and B. megalops were conspecific, but he re- tained them as separate species (p. 36). Again in 1886a (p. 274) Hoyle listed the two species of Bathyteuthis. Pfeffer (1900, p. 173) synon- ymized B. megalops with B. abyssicola, although not without some uncertainty. In 1903 Chun referred to “Bathyteuthis (Benthoteuthis)” (p. 85) and from then on to Bathyteuthis, a clear indication that he considered Benthoteuthis a synonym. But Chun (1910, p. 186) changed his mind and claimed that Benthoteuthis megalops Verrill had priority over Bathyteuthis abyssicola Hoyle by one month on the basis of the April 1885 signature date on the sheet that contained Verrill’s de- scription. Shortly thereafter, however, Hoyle (1912, p. 282) emphati- cally opposed Chun’s decision claiming that Verrill’s work was not published in separate sheets. “Therefore, under the most. favourable construction, it cannot possibly have appeared before June, and careful inquiries which I made at the time led me to the conclusion that. it did not make its appearance till July” (p. 283). Since that time the usage of the names seems to have been more a matter of preference than of priority. I have made an exhaustive search through the records of the Smith- sonian Institution Library, the Yale University Library, and the Con- necticut Academy of Arts and Sciences in an effort to determine the exact date of publication of Verrill’s Benthoteuthis megalops. I have 43 BATHYPELAGIC SQUID BATHYTEUTHIS found no evidence to prove that the description was published in the nomenclatural sense prior to June 1885. Bathyteuthis abyssicola Hoyle, 1885 (May), therefore, must take priority. Bathyteuthis bacidifera Roper, 1968 PLATES 6-10; 12 G,H ?Benthoteuthis megalops, Chun, 1910, p. 185-199, pls. 24-27 (pars; station 221, 18 mm specimen only).—Pfeffer, 1912, p. 325-331 (pars; using Chun’s description ). Bathyteuthis bacidifera Roper, 1968, p. 163, pls. 1-4. Dracnosis.—Protective membranes on arms lacking; long, free, fin- ger-like trabeculae present; tentacles and clubs relatively long; suckers on arms numerous; sucker rings with 18-34 protuberances; gills long, broad. List of Material ML, milli- Depth, Sex meters Ship Sta. Location Date meters! Gear HouLoryPe: F 37.—s Et. 34 07°47’S 81°23’W 7 VI 62 683 10’ IKMT PARATYPES: F oi Elt. 34 07°47’S 81°23’W 7 NI62 683 10’ IKMT F 34 ~—sE It. 54 18°23’S 72°39’W 16 VI 62 1373. 10’ MW Beam M 28 Bilao LO 06°54’N 79°57’W 3 V 67 23182 40’ otter F 26 D1208 XIV 06°48’N 80°33’W 16 I 22 1550 38 150 OTHER MATERIAL: F 19 D1203 XIII 07°30’N 79°19’W 11 ~=I[ 22 1000 $S 100 M Ti D1208 VIII 06°48’N 80°33’W 16 I 22 750 S100 M 12 D1209 III 07°15'’N 78°54’ W lit, el 22 1250 S 150 M 11.5 D1208 VI 06°48’N 80°33’W 16 I 22 1250 S$ 150 F 11.5 D1203 XVI 07°30’N 79°19’W 11 I 22 750 8 150 L4 10 —s Ellt. 34 07°47’S 81°23’W 7 VI 62 683 10’ IKMT F 9 D1203 XII 07°30’N 79°19’W ss 11-—so: 22 1250 S 150 L 6 D1208 XVI 06°48’N 80°33’W 16 I 22 1050 §$ 150 1 Estimated depth of capture. 2 Bottom depth fished by otter trawl. 3 Stramin tow-net of 150 cm diameter. 4L refers to larvae. Description.—The mantle is short, broad, and bullet-shaped with ‘gently curving sides; it has approximately the shape of a truncated ellipse, terminating in a bluntly rounded tip (pl. 6). The mantle width 321-534 O—69——4 44 U.S. NATIONAL MUSEUM BULLETIN 291 is about 50% of the mantle length. The location of the mantle-funnel locking apparatus is indicated ventrally on each side of the mantle opening by a small triangular projection. The anterior projection on the dorsum of the mantle is a blunt lobe. The muscle that makes up the wall of the mantle is moderately thick and well developed. The in- tegumentary layers are relatively thick, similar to those of B. abys- sicola; they are bound to the mantle wall by a semigelatinous, fibrous matrix of connective tissue. The fins are short, small, and paddle-like; they are nearly circular in outline (pl. 6). The fins are subterminal; the bases are separated posteriorly by the blunt end of the mantle and anteriorly by the broad dorsal surface of the mantle. The bases of the fins are thick and muscu- lar, but the margins are thin and fragile, almost membranous; they are easily torn so that the actual outline of the fins is difficult to deter- mine. The anterior and posterior fin lobes project well beyond the bases of the fins and the posterior lobes extend beyond the end of the mantle. The funnel is broad at the base and tapers anteriorly (pls. 6B; 74). The exhalant opening extends nearly to the level between the posterior margins of the eye openings. The posterior border along the base of the funnel is deeply concave between the funnel components of the locking apparatus. The funnel retractor muscles are strong and robust. The bridles are weak, thin bands of muscle. The collar is a simple, thin-walled band of muscle that passes dor- sally from the base of the funnel to the nuchal lock. The funnel and the collar are bound to the head by the integument and the subcutaneous, semigelatinous connective tissue so that only the anterior portion of the tube is free. The funnel groove is short and relatively shallow. A small median orifice in the integumentary tissue occurs near the posterior end of the funnel groove, dorsal to the exhalant tube (pl. TF). The orifice marks the opening of a narrow tube that passes dor- sally through the gelatinous tissue and between the median edges of the bridles. This unusual structure, which occurs in some other groups, is currently being investigated. The funnel component of the locking apparatus is a long, narrow, cartilaginous structure of the simple type (pl. 7a, 8). The rounded posterior end is broader than the rounded anterior end. The sulcus of the component is smooth and relatively shallow; it is slightly deeper and narrower anteriorly. A narrow, membranous lip outlines the struc- ture. The mantle component of the locking apparatus is a simple, low ridge; it is highest and most pronounced anteriorly and diminishes posteriorly (pl. 7a, c). BATHYPELAGIC SQUID BATHYTEUTHIS 45 The dorsal member of the funnel organ is an inverted Y-shaped structure (pl. 7k). The posterolateral limbs are long, narrow, and rounded; the anterior limb is short, broad, and blunt. A long, flattened papilla protrudes from the apex of the anterior limb. The lateral borders of the organ are slightly concave, while the medial borders form a curved, open V. The ventral members are elongate, oval patches. The funnel valve consists of a free, semicircular flap anteriorly and a muscular base posteriorly that forms a flat, deep pocket against the dorsal wall of the funnel. The head is short and broad; it is rounded and swollen laterally by the huge, bulbous eyes that are directed anterolaterally. The eyes con- stitute a very large proportion of the head. Anterior to the eye open- ings the head narrows abruptly where the base of the brachial crown originates. The head appears longer than it actually is because the web that interconnects the arms is continuous with the integument of the head. The eye openings are circular; an optic sinus is lacking. The olfactory papillae are minute projections located on the posterolateral curvature of the head. The dorsal and ventral surfaces of the head are flattened across the median areas, but they curve latérally. Nuchal folds or crests are lacking. The nuchal component of the nuchal-mantle locking apparatus is long and narrow with rounded ends (pl. 7p). It is bordered by a thin, narrow, membranous skirt. A distinct cartilaginous ridge lies in the midline; a narrow deep sulcus lies along each side. These features con- form to the cartilaginous mantle component that has a deep median sulcus with a high ridge on each side; this is formed by the anterior end of the rhachis of the gladius. A small, flat, ovoid photophore is located near the base of each of the dorsal six arms (pls. 64; 104). The light organs are embedded in the integument and, as in B. abyssicola, are more difficult to detect in large, darkly pigmented specimens. The simple organs consist of a white or cream-colored central area encircled by a darkly pigmented ring that is broadest posteriorly. In young specimens the photophores are readily visible as raised, light or pearly organs (pl. 10a) ; the pig- mented ring is much less pronounced than in larger forms. The arms are relatively short and conical; the tips are pointed but not attenuate. The ventral three pairs of arms are subequal in length; the dorsal arms are always the shortest. The most frequent arm formula is 4.3=2.1 followed by 4=3=2.1. A deep web joins the four pairs of arms; it extends out about one- third the length of the arms (pls. 8a; 10B, p), except between the ventral arms where no web occurs. The web is a continuation of the integument of the head and arms, so it is relatively thick and fleshy, 46 U.S. NATIONAL MUSEUM BULLETIN 291 not membranous. Because of this the web makes the head appear longer and the arms shorter than they actually are. The section of web between the third and fourth arms continues out the fourth arm as the broad lateral keel or “tentacular sheath.” All arms have aboral swimming keels or membranes. The keel on the dorsal arms occurs as a low, fleshy ridge along the distal third of the arms. The dorso- and ventrolateral arms have relatively well- developed keels along their distal two-thirds. The lateral keel of the fourth arms has been described above. The most distinctive feature of B. bacidifera is the structure of the protective membranes that occur on all arms. Protective membranes in the usual sense are lacking along the proximal half of the arms and are replaced by long, fleshy, finger-like or rodlike projections that roughly resemble thick, blunt cirri (pls. 8 a, B; 10 B, p). A trace of a low membrane occurs between the bases of the projections, or trabec- ulae. The trabeculae are not strong muscular rods, but are soft and fleshy, nearly semigelatinous, i.e., the same consistency as the normal protective membranes of Bathyteuthis. The trabeculae attain their maximum dimensions immediately distal to their origin at the bases of the arms. The ventral trabeculae of the dorsal three pairs of arms are longer and more robust than those along the dorsal edges. Distally the trabeculae become shorter, broader, and more lobate which gives a scalloped effect to the protective membrane that is fully developed on the distal quarter of the arms. The protective membranes are normal toward the arm tips. The ventral arms have fewer, smaller, less strongly developed trabeculae that give way almost immediately to scalloped protective membranes. The comblike rows of trabeculae appear early in ontogeny and are present even in the smallest larva (6 mm ML) in the present sample. In fact, this feature alone permits definite identification of the larvae of this species. The oral surfaces of the arms are covered with numerous small to minute suckers arranged in irregular rows. The arrangement of suckers is biserial at the bases of the arms but quickly becomes irregularly 3- to 4-rowed along the middle portion of the arm (pls. 84,8; 10p,D). The suckers become extremely numerous and closely packed in the distal quarter of the arms, and toward the tips they become minute. Suckers do not occur on the extreme distal tips of the arms. The arrangement of suckers on the ventral arms is nearly the normal biserial distribution. The tentacles are very long, thin, and muscular; they are about 125- 150% of the mantle length. The tentacles are nearly round in cross section except along the oral surface which is flattened and has a BATHYPELAGIC SQUID BATHYTEUTHIS 47 shallow, narrow depression running the entire length of the tentacular stalk, The tentacular club is relatively long, narrow, tapering and simple with no differentiation into carpus, manus, or dactylus (pl. 94). The club is about 25-30% of the mantle length in smaller specimens and 30-35% in the large specimens; the club makes up about 23-27% of the tentacle length. The sucker-bearing portion of the club is cov- ered with numerous, minute, closely packed suckers that originate proximally as one or two scattered suckers. The suckers increase very rapidly in number distally and are so numerous and closely packed that a definite linear pattern is not obvious, but there are about 8-10 suckers across the club. Suckers decrease in numbers only at the extreme end of the club. The left club of the holotype has approxi- mately 615 suckers. Although the club tapers gradually posteriorly it terminates abruptly in a blunt tip. In smaller specimens the tip is slightly drawn out. This papilla-like tip bears minute granules that appear to be precursors of suckers. The tips of the clubs of the two larger specimens (37 mm ML) are blunt nubs with only a very few nonchitinous suckers; no papilla-like growing tips exist, so these clubs may be approaching the maximum size for these specimens. No protective membranes occur along the lateral borders of the club. A narrow, weak swimming membrane arises on the dorsolateral sur- face of the tentacular stalk just proximal to the sucker-bearing region ; it runs the length of the club and terminates just short of the tip. The apertures of the largest inner sucker rings from the first arms of the holotype measure 0.12-0.14 mm in diameter and possess 20-34 (average 24) minute teeth (pl. 9B, c). Medium-sized suckers have around 16 teeth. In general the teeth are small, short, and closely packed ; they are long and truncate distally and grade to stubby knobs proximally. The teeth on some rings are so closely packed that their lateral edges are in contact; these teeth tend to be rounded to knobby and not truncate. The maximum diameter of the sucker apertures on the second arms is 0.14-0.16 mm. These rings bear 20-23 (average 22) teeth (pl. 9p). The small, short teeth are truncate to rounded in shape distally and grade to knobs proximally. Smaller rings tend to have only rounded or knobby teeth. Occasionally a tooth will be broad (about two times wider than its neighbors) or long (about two times longer than broad). The third arms have sucker apertures with maximum diameters of 0.16 mm. Around the aperture are 21-25 (average 23) small, closely packed truncate (separate) or rounded (borders touching) teeth (pl. 9F). Smaller rings have about 16 teeth. The apertures of the 48 U.S. NATIONAL MUSEUM BULLETIN 291 suckers from the third arms of a 26 mm specimen are 0.10 mm in dia- meter and bear about 20-22 closely packed protuberances. The most distal 4 or 5 teeth are relatively long and robust and are bluntly rounded. The four lateral teeth on each side are reduced in size and these grade into the eight blunt knobs that occur on the proximal part of the ring. The apertures of the largest suckers from the fourth arms are 0.14 mm in diameter and bear 18-26 (average 21) short, closely packed, truncate teeth that are longer distally, blunter and stubby proximally (pl. 9a). The largest sucker rings from the tentacular club are 0.08-0.10 mm in diameter and bear 8-10 extremely small, widely spaced, truncate teeth that are elongate on the distal half of the aperture and knoblike on the proximal half (pl. 91). The aperture of the tentacular club sucker from the specimen 27 mm in ML is 0.06 mm in diameter and bears about eight small protuber- ances: four relatively long, bluntly rounded, evenly spaced teeth on the distal border; two slender, blunter teeth on the lateral border; and two short, blunt, widely spaced knobs on the proximal border. The suckers of the buccal lappets of the holotype have aperture diameters of 0.08-0.10 mm, and they possess 8 to 12 widely spaced, minute, truncate teeth (pl. 91). The apertures from the 26 mm speci- men measure 0.06 mm in diameter. During ontogeny there is an increase in the number of teeth on the sucker rings; plate 10¥F, G, H, shows rings with 7, 13, and 18 teeth from specimens of 11, 16, and 29 mm ML. The large, fleshy, rugose buccal membrane has seven points; the buccal connectives have an attachment formula of DDVD (pl. 8F). From one to five minute suckers occur on each of the buccal lappets (pl. 8a). The suckers are about 0.06-0.10 mm in diameter across the aperture and bear 8-12 truncate to rounded teeth (pl. 9H). The buccal mass and the beaks are very small for the size of the animal. The rostra of the beaks are black but the pigmentation de- creases markedly on the lamellae. The rostra are strong but the lamellae are very weak, thin, and fragile. The rostrum of the upper beak is long, curved, and sharply pointed (pl. 8c). The jaw angle at the junction of the rostrum and the rostral lamella is nearly a right angle, and the anterior border of the rostral lamella is slightly convex. The dorsal outline of the palatine lamella is nearly flat, not curved; the ventral outline is nearly semicircular. The rostrum of the lower jaw is short and blunt; the jaw angle is a smooth curve (pl. 8p). The rostral lamella is long; the insertion plate of the gular lamella is broad and, posteriorly, it is subangular. BATHYPELAGIC SQUID BATHYTEUTHIS 49 The radula has seven transverse rows of pointed teeth and two rows of marginal plates (pl. 8k). The rhachidian has a broad, low base and the median cusp is relatively short, triangular, and bluntly pointed. No secondary (lateral) cusps occur on the rhachidian, but the concave borders of the tooth are occasionally interrupted with small, ragged protuberances. The first lateral has a long, low base and a moderately long, pointed cusp. The second lateral has a long, sharply pointed, crescent-shaped cusp. The third lateral has a very long, slen- der, scythe-shaped cusp. The marginal plates vary considerably in shape but are generally irregularly rectangular. The rhachis of the gladius is long and slender (pl. 7a). The anterior tip is thin and weak, but almost immediately the lateral edges are rolled under to form strong, heavy rods that extend posteriorly nearly the entire length of the gladius. The rods taper gradually along the vane and terminate just anterior to the end of the gladius. The rhachis is free for more than half the length of the gladius. No median ridge or rod occurs. The vane is very thin, weak, and fragile, almost membranous. From its origin at the rhachis the vane broadens and becomes wide posteriorly, terminating suddenly in a very blunt, rounded, weak tip. No conus exists. The entire animal is maroon colored. The pigmentation is darker on larger specimens, particularly in the region of the head and arms; there the color may mask the photophores. On younger, paler speci- mens, however, the photophores are readily seen. The tentacles alone lack pigmentation. Spermatophores from the only mature male available (28 mm ML) are 3.8-4.4 mm in length (pl. 12 cj). The sperm mass is 63-66% of the total length of the spermatophore, the cement body is 17-20% and the ejaculatory apparatus is 15-20%. The cement body is cigar-shaped and elongate with a slightly flaring hp anteriorly. The base of the ejaculatory apparatus is barrel-shaped. No hectocotylus is present. Hotoryrr.—United States National Museum, number 576148. Tyre Locatiry.—Off northern Peru at 07°47’S 81°23’W. USNS Eltanin Sta. 34, 683 m, 7 June 1962. DisrrinuTion.—Bathypelagic in the productive waters of Eastern Pacific Equatorial Water Mass and possibly in the Indian Ocean Equa- torial Water Mass (based on Chun’s [1910] single specimen). Erymotocy.—The specific name bacidifera is a neo-Latin word meaning “bearing little rods”; this is derived from the old Latin bacu- lum, a staff, stick or rod, the diminutive -¢diwm, and -fer, a suffix mean- ing bear, carry. The name alludes to the outstanding characteristic of the species, the rodlike trabeculae. 50 U.S. NATIONAL MUSEUM BULLETIN 291 MorpHometry.—Although relatively few specimens of B. bacidifera are available, they represent a broad range of sizes from larvae of 6 mm to fully ripe adults of 37 mm mantle length. Standard measure- ments have been made to determine the gross features of growth of B. bacidifera and to compare these features with B. abyssicola. The morphometric values are plotted as pluses (+) on the scattergrams for B. abyssicola (figs. 1 to 13) ; this allows a direct comparison of the morphometric characteristics of the two species. In B. bacidifera, as in B. abyssicola, the data indicate that the growth of the body parts that were analyzed is allometric with respect to mantle length. Discussion.—The Dana and Eltanin material used in this work show that B. abyssicola and B. bacidifera are sympatric in the tropical eastern Pacific Ocean. Certain statements in the literature indicate that the two species may also be sympatric in the equatorial waters of the Indian Ocean. Reports of specimens from these two areas may refer to either species or both. In most instances, however, it is difficult to determine the identity of the material. Chun (1910) had five specimens of Bathyteuthis taken by the Vail- divia; four were captured in the Indian Ocean Equatorial Water Mass. The specimens ranged from 9-18 mm mantle length. In a separate sec- tion on the details of arm structure, Chun noted that the largest speci- men lacked the protective membrane between the strong muscle sup- ports (trabeculae) and that the trabeculae looked like cirri. All smaller specimens had well-developed protective membranes. Therefore, Chun’s largest specimen probably represents B. bacidifera, while the remainder of the specimens probably refers to B. abyssicola. Unfortu- nately, it is impossible to be sure to which species Chun’s description refers except in sections where he mentions particular specimens. Since the largest specimen is only 18 mm in mantle length, probably most of the description applies to it, especially the descriptions of the eye and internal structures. Pfeffer (1912) had three larval specimens that he considered to be Bathyteuthis (Benthoteuthis). The specimen from the Mediterranean was a larva of Ctenopteryx (see Historical Résumé). The two remain- ing specimens were from widely separated localities in the Atlantic. Since the two larvae were only 3 and 4.5 mm mantle length, it seems unlikely that Pfeffer could have derived his detailed description from them. For the greater part, it appears that Pfeffer’s information is based on Chun’s description. Pfeffer’s specimens from the Plankton Expedition were undoubtedly B. abyssicola, but since most of his de- scription is based on Chun’s work, it necessarily includes a mixture of both species. Pfeffer stated that the protective membranes of larval specimens disappear in older specimens and that only the cross-struts BATHYPELAGIC SQUID BATHYTEUTHIS ip (trabeculae) remain as a comblike structure reminiscent of the am- bulacral stalks of some asteroids. Therefore, Pfeffer’s description, like Chun’s must be considered a mixture of the characters of two species. Both of these works, nevertheless, are still usable in a general way be- cause they do not deal with the characters (other than the trabecula- comb) that distinguish B. abyssicola from B. bacidifera. Naef’s conclusions about Bathyteuthis, which he stated are based primarily on Chun’s and Pfeffer’s descriptions, are not adversely af- fected because he dealt with taxa on the generic and familial levels. Works that would be affected are those based on material taken from localities where 2. bacidifera is known to occur: the equatorial waters of the Indian and eastern Pacific Oceans. Reports that may include specimens of B. bacidifera are by Hoyle (1904), Massy (1916), and Robson (1921, 1948). Hoyle’s (1904) description of B. abyssicola from the Panama re- gion gives no hint that would help in determining the identity of his specimen. The illustration is not particularly diagnostic either, but the second left arm (one of the five arms shown) looks as though it has distinct trabeculae; these are connected at their ends by a membrane and are not long. The tentacles look very long in the illustration and when their measurements are reduced to natural size (88 mm) they are 1 mm longer than the longest tentacle of B. abyssicola from the Antarctic; they are also somewhat longer than the tentacle of B. bacidi- fera. The long tentacles may be due to preservation or to an illustrator’s error. No other specific features can be seen in the illustration and it is not possible to determine with certainty to which species of Bathy- teuthis Hoyle’s specimen belongs. Massy’s (1916a) brief description of two small B. abyssicola from the southern part of the Bay of Bengal is not detailed enough to in- dicate if it should apply to B. bacidifera, and no illustrations are given. Trabeculae would be developed even in these small specimens, but Massy did not mention them. The arm suckers have 4, 5, and 6 teeth, and the suckers on the very short clubs have 4 teeth. In the absence of trabeculae, these features may refer the specimens to B. abyssicola, but this cannot be stated with certainty, because Chun’s and Pfeffer’s descriptions imply an ontogenetic occurrence of trabeculae in older specimens; Massy could have believed that her specimens were too small to have trabeculae. Robson’s (1921) supposed bathyteuthid from the Indian Ocean must be ignored, because no information whatsoever can be gleaned from the description or the specimen (see Historical Résumé). The Arcturus captured 11 small specimens in the eastern tropical Pacific in the region of the Galapagos and Cocos Islands. Robson 52 U.S. NATIONAL MUSEUM BULLETIN 291 (1948) considered 10 of them to be B. abyssicola,; the remaining speci- men (7mm ML) he called Bathytheuthis sp. primarily on the basis of different body proportions. Robson recognized that considerable varia- tion exists in specimens referred to B. abyssicola, but he did not illus- trate or adequately describe his specimens. Therefore, it is impossible to know if these specimens should be referred to B. abyssicola or to B. bacidifera. In conclusion, only one specimen mentioned in the literature prob- ably can be referred to B. bacidifera: Chun’s specimen of 18 mm ML from Valdivia Sta. 221, 04°05’S 73°24’K, in the Indian Ocean. OcCURRENCE IN RELATION TO PHYSICAL PARAMETERS.—The specimens of B. bacidifera all were captured in open nets that were calculated to have fished at depths ranging from 683 meters to 1550 meters. The depths of capture for the Dana specimens have been calculated by the method suggested by Bruun (19438) ; since no wire angles were taken during Dana tows, Bruun calculated that the fishing depth is one- third the length of the wire for less than 1000 meters of wire out and one-half the length of the wire for more than 1000 meters of wire out. The Dana specimens were taken as follows: two at 750 meters, one each at 1000 and 1050 meters, three at 1250 meters and one at 1550 meters. Three ELltanin specimens were taken in an Isaacs-Kidd Midwater Trawl (IKMT) that fished around 683 m, and a fourth came from 1373 m. It is extremely unfortunate that so few tows were taken dur- ing the period that the /'/tanin was working southward along the east- ern Pacific boundary. It is also unfortunate that Robson’s eleven Arcturus specimens from the Galapagos region and Hoyle’s two Albat- ross specimens from the Gulf of Panama are not specifically identifiable from the descriptions. The temperature-salinity relationships of B. bacidifera are plotted on figure 44. The zone of captures falls within the envelope of the Eastern Pacific Equatorial Water Mass. The range of the depths of capture corresponds to the band of the salinity minimum (34.55%. to 34.61%.) where the temperature ranges from just above 3° C in deeper water to nearly 6° C in the shallower portion of the range. The density values increase with depth from sigma-t=27.20 to 27.60. In addition, the oxygen concentration of the water layer between 200 to 2000 m is extremely low, between 0.1 ml/L and 2.0 ml/L; it represents a very broad oxygen minimum layer. The lowest values (0.1-0.5 ml/L) occur at the shallowest depths (ca 200-700 m) so B. bacidifera, with depth- of-capture values of 0.47 to 1.47 ml/L, lies just below the oxygen mini- mum layer. The narrow band along the eastern boundary of the tropical Pacific water mass is an area of high organic productivity (fig. 60). B. bactdi- BATHYPELAGIC SQUID BATHYTEUTHIS He fera apparently inhabits the water layer immediately beneath the zone of high productivity, an adaptive advantage commensurate with its anatomical adaptions for a relatively sluggish, upper bathypelagic existence. Chun’s largest specimen (18 mm ML) from the Indian Ocean, which may be B. bacidifera, was captured in the Indian Ocean Equatorial Water Mass in a plankton net that was towed vertically from 2000 meters. Granting that Bathyteuth?s normally lives below 500 meters, the temperature-salinity values (Tressler, 1963; Fell, 1965) indicate that Chun’s specimen came from temperatures and salinities as low as 2° Cand 34.7% at 2000 m to as high as 10° C and 35.1%. at. 500 m. The oxygen concentration in shallower portions of this zone ranges from less than 0.5 ml/L at around 500 m to 1.0 ml/L at 1000 m; in deeper portions it rises to about 2.5 m1/L at about 2000 m. If the distribution of B. bacidifera in the Indian Ocean Equatorial waters is governed by similar physicochemical parameters as this species in the Eastern Pacific Equatorial Water it would be found below about 500 m. Oxygen values in the two oceans (ca 0.5-1.5 ml/L) coincide between 750-1500 m and temperatures (ca 3.8°-6° C) coincide between 1000 and 1750 m. Salinities are high in Indian Equatorial water so there is no overlap of values, but sigma-t values of 27.20 to 27.60 in the eastern Pacific are found between 750 and 1500 m in Indian equatorial waters. Further consideration of the distribution of B. bacidifera is given in the main section on distribution. Bathyteuthis berryi Roper, 1968 PLATES 11-12 a-F Bathyteuthis berryi Roper, 1968, p. 169, pls. 5-7. Draenosis.—Protective membranes on arms present, well developed and fleshy proximally, no free trabeculae; suckers on arms extremely numerous, sucker rings with 10-14 protuberances; gills long and broad. Descrietion.—The mantle is very plump and robust; it is bullet- shaped in outline (pl. 11). The widest part of the mantle is about at the midpoint, and the mantle width is 50% of the mantle length. The mantle opening is slightly narrower and the margin bears low, ventro- lateral lobes. The mantle remains broad for much of its length, then tapers and terminates posteriorly in a broad, bluntly rounded tip. The fins are short, rounded, and widely separated posteriorly by the blunt tip of the mantle (pl. 11). Anterior insertions are very broadly separated by the dorsal surface of the mantle. Anterior and posterior fin lobes are about of equal dimensions. 54 U.S. NATIONAL MUSEUM BULLETIN 291 List of Material ML, Depth, Sex milli- Ship Sta. Location Date meters } meters HouoryPe: M 49 V 8714 33°14’45’""N 118°37'20’’W 7 V1I63 1200 PARATYPES: M 23 V 10540 29°05’04’’N 118°12’00’’W 6 IV 65 1300 M 20 V 10377 =33°25’00’’N 118°50’45’’"W 24 I165 1100 F 19 V 10976 = 32°35’00’’N 120°35’06’/’W 17 II 66 1300 OTHER MATERIAL: 19 V 41169 31°40'25’’N 120°22’42’’W 31 VII 66 1300 17 V 9905 29°28'14’’N 119°02'58’’W 8 VIII 64 800 16 V 9661 33°08/20’’N 119°12’385’’W 14 IV 64 1000 16 V 8700 33°15'30’"N 118°33'45""W 25 VI 64 800 14 V 9056 33°12'42" N -118°32/15/ Ws 14 X63 1100 13 V 10973 32°37'45’’N 120°24’30’’W_ 16 II 66 850 12; V 10730 33°27'40’N 118°52'50”W 26 IX 65 1000 11 V 10973 32°37'45'’N 120°24’30’’W 16 II 66 850 10 V 11189 32°25’06’’N 118°08’20’’W 4 VIII 66 5-900 9 V 8349 33°26/38’"N 118°54/15/’W 7 XII 62 1000 1 Estimated depths of capture. All specimens were captured by a 10’ Isaacs-Kidd midwater trawl. 2 L refers to larval or juvenile specimens of undetermined sex. peter kp ka Crees ees The funnel is very large, prominent, and long; it extends anteriorly to a point in line with the anterior borders of the eye openings. The pos- terior part of the funnel and the collar are bound to the posteroventral and posterolateral surfaces of the head with integument and gelati- nous, subcutaneous tissue. The funnel groove is very shallow, almost nonexistent. A small pore lies at the base of the funnel groove where the dorsal part of the funnel fuses with the head. Cartilaginous funnel components of the locking apparatus are simple and elongate with a smooth, shallow, median sulcus. The mantle com- ponent is a simple, straight ridge that articulates with the funnel lock. The dorsal member of the funnel organ is an inverted, roughly Y- shaped structure with short, broad limbs. A spatulate papilla extends anteriorly from the anterior limb of the Y-structure. The ventral pads have straight, diagonal anterior borders, nearly parallel sides and narrow, rounded posterior borders. A thin, broad funnel valve is present. The head is long and narrow; it is flattened dorsoventrally. The eyes are very large and are directed anterolaterally. The borders of the eye openings are circular; they lack an optic sinus. A minute, stubby olfac- tory papilla occurs on each side of the posterolateral surface of the head just anterior to the nuchal region. Nuchal folds and crests are absent. BATHYPELAGIC SQUID BATHYTEUTHIS 55 The nuchal component of the mantle-nuchal locking apparatus is long and narrow with three longitudinal grooves: a narrow, shallow, median groove and two broader, deeper, lateral grooves. The median groove complements the low, thin, median ridge of the mantle com- ponent of the locking apparatus, The lateral grooves receive the rolled edges of the rhachis of the gladius that partially make up and provide support for the mantle component. The head narrows considerably anterior to the eyes where it forms the base of the brachial crown. A single, small, simple photophore is embedded in the chromatophore layer of the subcutaneous tissue at the base of each of the dorsal three pairs of arms. Photophores are characterized by a darkly pigmented ring that is generally broadest posteriorly and a central mass that is much more lightly pigmented. Photophores in the holotype are embedded and not easily seen, but the photophores in the juvenile and larval specimens contrast more with the background pigmentation, are slightly raised, and are more readily seen. The arms are long, slender and drawn out into attenuate tips (pl. 11). All arms are of nearly equal length in the holotoype, so the arm for- mula for adults is 4=3=2=1. A moderately deep web joins the bases of the four pairs of arms. The depth of the web decreases from the dorsal to the ventral pairs, and no web occurs between the fourth arms. Low aboral swimming keels occur on the dorsal three pairs of arms, with those of the third arms the best developed. The web between the third and fourth arms extends distally along the fourth arms as the lateral membrane or “tentacular sheath.” Protective membranes occur on all arms. They are particularly well developed at the basal portion of the arms where they are thick, fleshy, and ruffle-like (pl. 114). The thickened ruffles diminish quickly distal to the bases of the arms, and the protective membranes extend distally as low, even keels. Protective membranes on the fourth arms are con- siderably less developed than on the dorsal three pairs. No distinct or separate trabeculae support the protective membranes. The oral surfaces of the arms are covered with extremely numerous small to minute suckers (pl. 114). The proximal suckers are quite small, and widely spaced; they originate in a single row, then soon increase in size and split off into two widely separated rows. About one- third of the way out the arms, the suckers become smaller and more closely packed and the rows become irregularly arranged so that occa- sionally 3-4 suckers occur across the oral surface of the arms. On the distal one-third to one-fourth of the arms, the suckers grade smaller, become very closely packed, and are exceedingly numerous. On the 56 U.S. NATIONAL MUSEUM BULLETIN 291 ventral arms, the sucker rows do not become so irregular and the suckers are not nearly so numerous as on the dorsal three pairs of arms. Suckers extend to the extreme distal tips of all the arms. About 275 suckers occur on each of the dorsal six arms of the holo- type. The dorsal arms may have very slightly fewer, and the ventral arms have around 150 suckers each. A specimen 23 mm ML (Velero 10540) has 175-185 suckers on arms I-III and 110 on arms IV. A specimen 12 mm ML (Ve/ero 10730) has 62-75 suckers on arms I-III and 48 suckers on arms LV. Inner sucker rings from the arms bear 10-14 very low, small, rounded or subtriangular, knoblike teeth (pl. 11c-r). Outer rings are made up of concentric rows of tiny chitinous bumps or pebbles. Both tentacles are missing from the holotype, probably having been lost during capture. In fact, the tentacles are broken off all but one specimen that is available at this time. The specimen (Velero 10976) is a juvenile 19 mm in mantle length. The single complete tentacle is long (ca 20 mm) and robust. The club is short (ca 4.5 mm), unexpanded, simple; no keels or membranes are present and no discrete divisions into manus, carpus or dactylus exist (pl. 11s). About 7 to 8 rows of very small suckers are distributed across the distal half of the club; fewer rows occur proximally. Between 150 and 200 suckers are present at this stage. The extreme distal portion of the club is reduced to a small, papilla-like tip with a few minute bumps that are probably precursors to future suckers. Inner sucker rings from the 19 mm speci- men (Ve/lero 10976) are extremely small, and it is difficult to deter- mine the dentition. Rings varied from being nearly smooth or slightly scalloped to having a few minute, low, subtriangular teeth as shown in plate 11G, although the figure may exaggerate the size of the teeth. The outer ring consists of concentric, pebbled rows. The buccal membrane is broad and heavily rugose. The seven buc- cal lappets are long and bear 4 to 6 small suckers; the inner chitinous rings of the suckers have around 10 small, low, papilla-like teeth (pl. 11H). The connections of the buccal membrane attach to the dorsal oral edges of arms I, IT, and IV and to the ventral oral edge of arm III. The beaks are small and are darkly pigmented only on the rostra (pl. 12c, p). Lamellae are lightly pigmented and fragile. The ros- trum of the upper beak is strong, sharply curved, and hooklike. The jaw angle is nearly a right angle. The rostrum of the lower beak is short and blunt ; the jaw angle is obtuse. The radula has seven transverse rows of teeth and two rows of lateral platelets (pl. 124). The rhachidian tooth has a broad base and a broad, bluntly pointed median cusp. The concave lateral borders of the cusp BATHYPELAGIC SQUID BATHYTEUTHIS 57 are very slightly irregular with minute denticles. The first lateral tooth has a straight, blunt cusp about as long as the rhachidian cusp; the concave lateral border has a few small, irregular denticles. The second lateral tooth has a long, straight pointed cusp. The third lateral has a very long, thin, scythe-shaped cusp. The marginal platelets are irregularly shaped oblong structures. The rhachis of the gladius is very long and slender; it makes up 60% of the total length of the gladius (pl. 128). No median ridge occurs along the gladius, but the lateral edges are rolled under to form strong rodlike supports to the rhachis. These rods extend posteriorly, decreasing in diameter, nearly to the tip of the gladius. The rhachis is U-shaped in cross section. The anterior tip of the rhachis is very thin and flaplike. The vane is very broad, thin, membranous and paddle-shaped. The rounded posterior border of the vane curves ventrally and forms a shallow cuplike terminus. The gladius lacks a conus. Most of the pigmentation is bleached out of the holotype, but the smaller specimens exhibit the maroon coloration typical of Bathyteuthis. The holotype is a mature male with fully developed sperma- tophores in Needham’s sac. A hectocotylus is lacking. Spermatophores are about 8 mm in total length; the sperm mass occupies about 72% of the total length, the cement body about 8% and the ejaculatory apparatus about 20% (pl. 128, F). The cement body is an elongate, vase-shaped structure with a flaring lip or collar where it joins with the narrow base of the short, bell-shaped end of the spiral filament. About one-third to one-fourth of the sperm mass is slightly pigmented. The gills are long and broad; counts and measurements were made on the gills of six specimens ranging from 16-49 mm ML. Gill fila- ments number from 19-21. The index of gill length to mantle length ranges from 43 to 53 (mean 48.6), that of gill width to mantle length from 7.5 to 15.7 (12.5), and that of gill width to gill length from 17 to 30 (25.5). Hovoryrr.—University of Southern Cailfornia. U.S.C. Hancock collections, AHF cephalopod type number 10. Tyre Locarrry.Catalina Basin, 10.9 miles SSW of West End Light, Catalina Island at 33°14’45’’N 118°37’20’°W. Velero Sta. 8714, about 1200 m., June 1963. Distripution.—Bathypelagic in the waters off Southern California. Erymo.tocy.—The specific name berryi is given in honor of Dr. S. Stillman Berry who has contributed a lifetime of study to malacology and teuthology. Discussion.—B. berry? was not discovered and described until after 58 U.S. NATIONAL MUSEUM BULLETIN 291 this manuscript was completed. The present description, an amplifica- tion of the original (Roper, 1968), and pertinent parts of the original discussion are inserted here for the sake of completeness. R. E. Young has compiled the capture data for the specimens of B. berryi that are currently available from the Velero collections in southern California waters. The specimens were distributed in 100 m increments as follows: 1 from 100-200 m, 1 from 300-400 m, 3 from 800-900 m, 5 from 1000-1100 m, and 5 from 1100-1200 m. The 1200 m level is the maximum depth sampled during the program. The two individuals from shallow water were taken in tows that had been preceded by tows that had fished at depths greater than 1000 m. It is reasonably safe to assume that these specimens were contaminants that had stuck in the net from the preceding deep tows. It seems likely also that only the shallow segment of the population of B. berryi was sampled at depths of 800-1200 m and that the bulk of the population lives below 1200 m. The three species of Bathyteuthis may be distinguished on the basis of several features, the most prominent of which are presented in the following list : Character abyssicola bacidifera berryt Free trabeculae absent present absent Protective mem- present absent present branes Arm suckers ! few (100) numerous (150) extremely num- erous (275) Sucker ring denti- 8-18, truncate 18-34, truncate 10-14, subtri- tion (Arms) angular Arms short, blunt short, blunt long, attenuate Gills short, narrow long, broad long, broad Spermatophore 68-72; 6-8; 20-25 63-66; 17-20; 72; 8; 20 proportions 2 15-20 Tentacles and short long missing from clubs material 1 The numbers in parentheses represent the approximate number of suckers on each of the six dorsalmost arms from specimens of about the same size (49 mm ML). 2 The size of the sperm mass, cement body, and ejaculatory apparatus respectively expressed as a per- centage of the total length of the spermatophores. The most striking and easily recognized character of bacidifera is the presence of long, finger-like trabeculae on the arms that have no interconnecting protective membrane. This feature is apparent even on the smallest larva available (6 mm ML) and readily separates the species. Both abyssicola and berryi possess thick, fleshy protective membranes. Considerable variation exists in the membranes, but they are always present and connect unmodified trabeculae. BATHYPELAGIC SQUID BATHYTEUTHIS 59 B. berryi is most readily distinguished from abyssicola by the extreme abundance of suckers on the arms and by long, wide gills. (The significance of gill size is discussed in a later section.) The arms of the holotype of berryi are 5-7 mm longer than the arms of abyssicola of the same mantle length (49 mm), and they are more attenuate. This trend holds in all specimens available. When material with tentacles in tact becomes available, differences in the clubs may be found. Although it is difficult to demonstrate quantitatively with the limited number of specimens on hand, the mantle of berryi appears to be slightly more plump that that of abyssicola. Comparison of Bathyteuthis abyssicola and Bathyteuthis bacidifera Morphological Comparison The most striking and readily observable difference between Bathy- teuthis bacidifera and B. abyssicola is the possession of long, free, finger-like trabeculae on the arms of B. bacidifera. B. abyssicola lacks this feature entirely, although its protective membrane differs from most other oegopsids by being thick, fleshy (semigelatinous), and unsupported by strong trabeculae. The protective membrane of abyss- icola exhibits considerable individual variation; the border may vary from straight and smooth to gently undulating or scalloped, but a comblike row of trabeculae does not occur. The free trabecular rods of bacidifera are well developed on the smallest specimens observed (6 mm M1L), so there is no danger of confusing even the larvae of the two species. The erection of a separate species might be warranted on the basis of this feature alone, but other less spectacular specific characters do exist. The clubs of bacidifera are longer and bear more suckers than do those of abyssicola of the same size. For example, the clubs of the holotype of bacidifera (87 mm ML) and of an abyssicola of 57 mm ML are nearly equal in length, but bacidifera has about 615 suckers on the ciub while abyssicola has around 525 suckers. An analysis of club lengths is presented in the following section. The gills of bacidifera are measurably larger than those of abyss7- cola and they tend to have a greater number of filaments. Geographic variation exists in the gill size of abyssicola from different areas but little overlap occurs with bacidifera. A detailed analysis of gill size appears in the section on geographical variation. Within species, sucker dentition varies slightly with the size of the specimen and with the size of the suckers on an individual specimen. 321-534 O—69——_5 60 U.S. NATIONAL MUSEUM BULLETIN 291 Between species, however, there are greater, consistent differences in sucker dentition; bacidifera has a greater number of teeth on the arm suckers than does abyssicola. Suckers of abyssicola of 28 mm ML have about eight protuberances around the aperture. Of these, the four distal ones are long, truncate, widely set teeth, and the four proximal ones are small, blunt, widely set knobs. On specimens of 38 mm ML the number of teeth increases to 9-14 (usually 12) with the distal 4-7 teeth long, truncate to rounded, and widely spaced and the proximal teeth knoblike. Smaller suckers from the same specimen have fewer teeth. In specimens 56 mm ML the arm suckers bear from 8-18 teeth depending on the diameter of the ring. The larger rings generally average 10-14 widely spaced protuberances with long, truncate to rounded distal teeth and short, blunt proximal knobs. Occasionally a medium-sized ring will have more very closely packed teeth than normal (up to 18). B. bacidifera of comparable size to B. abyssicola has slightly smaller sucker apertures but a greater number of protuberances. The larger sucker rings of the holotype (87 mm ML) have 18-26 (average 22) teeth. The teeth on the distal half are short, truncate to rounded, and closely packed, and they grade proximally into small, very closely packed knobs. The teeth on some rings are so closely packed that theit lateral edges are in contact. Medium-sized teeth have about 16 pro- tuberances. Exceptionally closely packed teeth and teeth on medium- sized rings tend to be rounded and knobby, not truncate. Rarely, teeth will be about twice as wide as normal or about two times longer than broad. During ontogeny teeth are added rapidly to the rings of bacidifera, but during comparable growth in abyssicola the number of teeth be- comes stabilized (pls. 10 r-H; 9 B-», F, G; 4 c-G; 5 G4). This should serve warning about placing too much emphasis on sucker dentition without considering the age (size) of the specimens. The dentition of suckers from the tentacular clubs and the buccal lappets does not seem to differ so noticeably between the two species. Again, dentition varies but both species have about the same number of points on the sucker rings; both club suckers and buccal suckers have 8-12 teeth and knobs. Table II summarizes the diameters of arm, club, and buccal lap- pet suckers for the two species of Bathyteuthis. The suckers of abyssi- cola tend to be slightly larger than those of bacidifera. BATHYPELAGIC SQUID BATHYTEUTHIS 61 TasBLe II.—Range and average of diameters of inner sucker rings from B. abyssicola and B. bacidifera [Measurements in millimeters] Arm suckers Club Buccal Species ML range; mean suckers suckers abyssicola 56 0.16-0.20; 0.19 0.12-0.14 ~ abyssicola 38 0.14-0.20; 0.168 0.12 0.12 bacidifera of 0.12—-0.20; 0.145 0.08-—0.10 0.08—0.10 abyssicola 22-28 0.12-0.18; 0.155 0.08-0.10 ~ bacidifera 26 0.10-0.16; 0.13 0.06 ~ abyssicola 10-19 0.10-0.16; 0.125 0.10 0.06—0.10 bacidifera 17 0.10-0.10; 0.10 = "3 Morphometric Comparison The plots for mantle width in relation to mantle length show the same proportional increase in both species of Bathyteuthis (fig. 1). The two largest specimens of bacidifera are 37 mm in ML; one has a mantle width of 19 mm while the other is 15 mm. The narrower width falls well within the range for abyssicola of about the same ML; the greater width is one or two mm wider than would be expected in abyssicola of the same size, but the specimen, the holotype (pl. 6), is a fully ripe female swollen with eggs. With the material at hand bacidifera does not seem to differ greatly from abyssicola in mantle width. 30 1 B. abyssicola @ Mole ° Femole +B. bacidifera 4 30 ML.mm Ficure 1.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. mantle width. The plot for the length of the head of abyssicola and bacidifera shows narrow limits in specimens of less than 30 mm mantle length. Above 30 mm the head length becomes slightly more variable, but in general it continues the trend established in younger specimens. The 62 U.S. NATIONAL MUSEUM BULLETIN 291 head lengths of all sizes of bacidifera lie within the range of points for abyssicola (fig. 2). A comparison of the plots of head width against mantle length shows that the head of small bacidifera may be slightly narrower than that of abyssicola (fig. 3). In larger specimens the widths of the heads of the two species converge. The lengths of the fins may differ in larger specimens of the two species of Bathyteuthis. Only four specimens of bacidifera measure between 19 and 37 mm ML; these have basal fin lengths that range up to 20% greater than the basal fin lengths of abyssicola in the same size range (fig. 4). The plot of fin width against mantle length for abyssi- cola indicates that the fin width becomes more variable beyond a mantle length of 30 mm (fig. 5). The comparable plot for bactdifera shows that the two largest specimens (37 mm ML) and the specimen of 19 mm ML have fin widths that lie near the upper limit of fin width for the same size abyssicola. No significant difference is noted in the plot for fin width vs. fin length (fig. 6). Tentacle length is a difficult character to evaluate because the elastic tentacles are subject to greater expansion and contraction during fixa- tion than any other body part. A fair approximation of true tentacle length, however, may be made when a large number of specimens is measured and plotted as in the case of abyssicola from the Antarctic. When the plots for tentacle length vs. mantle length are compared, it is noted that the tentacles of the few specimens of bacidifera tend to be longer than the majority of abyssicola (fig. 7). Specimens of bacidifera 12 mm in ML have tentacles about 30% longer than the tentacles of abyssicola. At a mantle length of 37 mm the tentacle length of bacidi- fera is about 18% greater than in abyssicola. B.abyssicola @mMale OFemole +B. bacidifera Ficure 2.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. head length. BATHYPELAGIC SQUID BATHYTEUTHIS 63 1 B. abyssicola @Male OFemale +B. bacidifera 70 20 30 ML mm 0 50 0 Ficure 3.—Bathytheuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. head width. ° B. abyssicola eMale O Female +B. bacidifera Ficure 4.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. basal fin length. 1 e B. abyssicola eMale OFemale +B. bacidifera 0 ML,mm e Ficure 5.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. fin width. 64 U.S. NATIONAL MUSEUM BULLETIN 291 5 e B. abyssicola eMale OFemale +B. bacidifera 20 25 35 5 10 15 EW. sam Ficure 6.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of fin width vs. basal fin length. 80 2 B. abyssicola @Male OFemale +B. bacidifera 0 70 ML,mm Ficure 7.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. tentacle length. BATHYPELAGIC SQUID BATHYTEUTHIS 65 Comparison of club length vs. mantle length indicates that the clubs of bacidifera tend to be about 25-30% longer than those of abyssicola of the same size (fig. 8). The proportion of club length to tentacle length overlaps in the two species but the plot for bacidifera occurs in the upper range of values of abyssicola (fig. 9). A specimen of bacidifera of 15 mm ML would have a tentacle length of about 20 mm and a club length of 4.5 mm, while a specimen of abyssicola with the same tentacle length would have a mantle length of 19 mm and a club length of 3.5 mm. B._abyssicola @Mole oFemale + B.bacidifera Ficure 8.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. club length. oO e 15 ° ° ° ° ° Oo ° CL. $9 tH 200 0 mm oo + e+ o 00 ©, %.6 #0 1 SBe° o eoe of 0 & ° rae eo “Ge $ e ge 2 oe ° 8. abyssicola O; qos. Bkcatle ° +B. bacidifera Figure 9.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of tentacle length vs. club length. 66 U.S. NATIONAL MUSEUM BULLETIN 291 No difference in the lengths of the arms occurs between the two species (figs. 10-13). The predominant arm formula in each species is 4.3.2.1. In abyssicola of 37 mm ML the fourth arms average about 15.5% longer than the first arms; in bacidifera of the same size the fourth arms are 16% longer than the first arms. B. abyssicola e@Moale oFemale +B. bacidifera Ficure 10.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. Arm I length. 20: B.abyssicola e@Male OFemole 5 +B. bacidifera 0 0 ML, ma 0 50 60 70 Figure 11.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. Arm II length. BATHYPELAGIC SQUID BATHYTEUTHIS 67 In summary, little difference appears to occur tn the overall propor- tions of the two species. The larger specimens of bacidifera have fins that are approximately 20% greater in basal length and less so in width than the average for abyssicola of comparable sizes. These differences, 20. IL, mm 8 B. abyssicola @Ma'e 5 OFemale B. bacidifera 10 20 30 40 50 60 ML,mm Ficure 12.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. Arm III length. 25 20 B. abyssicola 5 eMale ° oFemale +B. bacidifera 9 ML,mm 30 Figure 13.—Bathyteuthis abyssicola, B. bacidifera: scatter diagram of mantle length vs. Arm IV length. 68 U.S. NATIONAL MUSEUM BULLETIN 291 however, may be too close to the limits of the range of variation of fin length and width for abyssicola, and because of the small sample size of bacidifera, may not be significant. Differences in proportions are noted with the lengths of the tentacles and especially of the clubs; the club is 28-30% longer in bacidifera than the average for abyssicola of the same mantle lengths. Therefore, both species exhibit nearly the same characteristics of proportional growth. Key to the species of Bathyteuthis 1. Protective membranes on arms low to well developed, fleshy, with straight te gently scalloped borders; trabeculae not free, enlarged, or elongate . . . 2 Protective membranes reduced or lacking; trabeculae free, elongate, rodlike; arm suckers numerous, rings with 18-34 protuberances; gills long, broad. B. bacidifera Roper, 1968 2. Arm suckers relatively few, ee with 8-18 protuberances; arms short, blunt; gillsshort, narrow. . . . . .B. abyssicola Hoyle, 1885 Arm suckers extremely numerous, rings with 10-14 protuberances; arms long, attenuate; gillslong, broad. . . . .... =. +. . .B. berryi Roper, 1968 Geographical Variation in Bathyteuthis; Interspecific and Intraspecific Variation of Taxonomically Important Characters Buccal Suckers The Bathyteuthidae and the Ctenopterygidae are the only families of Oegopsida known to possess suckers on the oral surface of the lap- pets of the buccal membrane. It is primarily (though not exclusively) because of the possession of this character that Bathyteuthis and Ctenopteryx had been brought together in the past under the Bathy- teuthidae. Both Naef (1923) and Grimpe (1925) considered that the buceal suckers, four rows of arm suckers, and many rows of club suckers are primitive characters within the Oegopsida, and they placed the Bathyteuthidae first in the suborder immediately adjacent to the Myopsida. The large number of specimens of Bathyteuthis available in this study makes it possible to examine the occurrence of buccal suckers over the geographic ranges of the species and to determine the value of these suckers as a taxonomic character at the species level. 1. Bathyteuthis abyssicola—Antarctic B. abyssicola from the Antarctic exhibits a variety of combinations of suckers on each of the buccal lappets. An individual may have 0, 1, 2, or 3 suckers on each lappet. Very seldom do all seven lappets of an individual bear the same number of suckers, except when 0 suckers occur. For instance, a specimen may have 0 suckers on two lappets, 1 BATHYPELAGIC SQUID BATHYTEUTHIS 69 sucker on each of three lappets, and 2 suckers on the remaining two lappets. Clearly a large number of combinations of 0-3 suckers on 1 to 7 lappets is possible. Tables III and IV show the combinations that are known to occur in B. abyssicola from Antarctic waters. At the outset of this study, it appeared that the distribution of suckers on the buccal lappets was entirely random, but as larger num- bers of specimens were examined it seemed that a certain degree of accuracy was attained in predicting the sex of specimens by noting the presence (combination) or absence of suckers on the lappets. (The male of abyssicola has no differentiated hectocotylus.) A sample of 123 specimens, 69 females and 54 males, displays six combinations of numbers of suckers on the buccal lappets (Table III) ; the table records the combinations of numbers of suckers only and does not consider the number of individual lappets that bear a particular number of suckers. Twelve of the females, 17.4% of the female sample, had 0 suckers on any of the lappets, while 53.7% of the males lacked buccal suckers; but 47.8% of the females and 33.4% of the males had the 0, 1 combination of suckers, in which 1 sucker is present on one to six of the lappets. The 0, 1, 2 combination included 13.1% of the females and 9.3% of the males. Fourteen females (20.83%) but only one male (1.8%) had a combination of 1, 2 suckers. One male had 1 sucker on each lappet and one female had a combination of 0, 2, 3. In comparing sucker combinations of males and females it is noted that nearly two-and-a-half times as many males as females have 0 suckers on the buccal lappets. The combinations 0, 1 and 0, 1, 2 are predominated by females in a ratio of two females to one male, although the 0, 1 combination is much more abundant. One-third of the total sample has 0 suckers, and 41.5% has only 1 sucker on some of the lappets and none on others. Therefore, about 75% of the sample population has 0 or 0, 1 suckers. About 12% of the sample is represented in the 0, 1,2 and 1, 2 categories. In general males tend to have lappets that bear fewer suckers; about 87% have 0 or 0, 1 suckers, and only about 11% have some lappets with 2 suckers. About 65% of the females have 0 or 0, 1 suckers and the remaining 35% have combinations that include 2 suckers. Table IV represents a sample of specimens that have 0, 1 suckers in all possible combinations of lappets from six lappets with 0 suckers and one lappet with 1 sucker to one lappet with 0 suckers and six with 1 sucker. Females have a broad range of combinations that. includes all possibilities. About 30% have six lappets with 0 suckers and one lappet with 1 sucker and 25% have the reverse order. The combina- tions of 2 and 5 and 3 and 4 are nearly evenly distributed in the re- maining 45% of the sample. The situation in males is striking: 757% 70 U.S. NATIONAL MUSEUM BULLETIN 291 of the sample has only one lappet with 1 sucker and six lappets with 0 suckers; 16.6% has more lappets with 0 than with 1 sucker, and only 8.3% has six lappets with suckers, one without. Therefore, over 90% of the males have more lappets with 0 suckers than with 1 sucker. In contrast, females have no predominance of numbers of lappets with 0 or 1 suckers; there are as many lappets without suckers as there are with suckers. Applying these figures to Table ITI, it is found that 13.5 (=75%) of the 18 males with 0, 1 suckers would have only one lappet with 1 sucker and six lappets with 0 suckers. This further strengthens the indication that males have fewer lappets with suckers and, in addi- tion, they have fewer suckers. Figure 14 gives the plots of combinations of buccal suckers for males and females against mantle length. The graph is a visual reinforcement of the points brought out in the preceding discussion. In addition, it shows a weak trend toward an increase in numbers of suckers (and number of lappets with suckers) with increasing mantle length. The 0 and 0, 1 categories are represented by a broad range of sizes for males 0,2,3 0,1 00eg0 f oogego veg 0eg = #0 8 gosh ogo @0 0 e gee @e0e Bageegeegoe 8 es g e 30 4 ML,mm Ficure 14.—Combinations of suckers on buccal lappets; Bathyteuthis abyssicola, Antarctic Ocean. BATHYPELAGIC SQUID BATHYTEUTHIS 71 and females. Males predominate in the 0 category where 13.8% (7.4% of the total male sample) are less than 20 mm ML and the remainder are grouped in the 22 to 49 mm range with 55.3% (29.6%) concen- trated between 30-40 mm. About 50% (28.7%) of the females in this category range from 39-61 mm and 25% are less than 20 mm. In the 0, 1 category males range from 14 to 52 mm in mantle length, and 83% (27%) of them are fairly evenly distributed through the 22-45 mm range. Females range from 12 to 57 mm in this category with 30% (14.5%) at less than 20 mm and 33% (16%) between 42 and 49 mm. In the 0, 1, 2 group 80% (13.8%) of the males are between 35-40 mm, the remaining, 1 specimen, is 49 mm. Females range evenly between 25 and 58 mm mantle length. The 1, 2 category contains only one male which is 75 mm in mantle length (the largest B. abyssicola on record). Females in the 1, 2 category range from 19 to 53 mm; only one speci- men is smaller than 31 mm; 71% (17%) are larger than 43 mm. There- fore, in the sample of 123 specimens there is a trend toward increasing numbers of suckers with increasing mantle lengths. In conclusion, it is apparent that considerable variation exists in the occurrence and numerical distribution of suckers on the buccal lappets. There are, however, trends in occurrence and distribution that are associated with sex and size (age) of the specimens. TaBLE III.—Number of suckers on the buccal lappets of B. abyssicola from Antarctic waters. Sample number 123; 69 females, 54 males Sucker Number Percent Number Percent Ratio Percent combi- Q Q J J O:0 of total nation population 0 12 17.4 29 53.7 1:2.4 33.3 0,1 33 47.8 18 33.4 1:0.55 41.5 0, 1, 2 9 To. 5 9.3 1:0.55 11.4 2 14 20.3 al 1.8 1:0.07 122 1 0 0 1 1.8 - 8 0, 2, 3 1 1.4 0 0 1:0 8 TaBLe IV.—Frequency of occurrence of 0, 1 suckers on combinations of lappets; B. abyssicola, Antarctic Number Number lappets, lappets, Number Percent Number Percent O suckers 1 sucker Q Q J fos 6 1 6 30 9 iD 5 2 2 10 1 8.3 4 3 2 10 1 8.3 3 4 3 15 0 0 2 D 2 10 0 0 1 6 5 25 1 8.3 €2 U.S. NATIONAL MUSEUM BULLETIN 291 2. Bathyteuthis abyssicola—Kastern Pacific Seven specimens of B. abyssicola from the eastern Pacific exhibit five combinations of numbers of buccal suckers. The smallest specimen, 13 mm ML, has 1, 2 suckers while the remaining specimens (16-30 mm ML) have 2, 3; 3, 4; 3, 4, 5; or 4, 5 combinations of suckers (fig. 15). The indication is that suckers are added very rapidly to the lappets during growth, following the general trend of B. bacidifera from the same waters. Slight overlap of buccal sucker combinations occurs be- tween abyssicola from the eastern Pacific and from the Southern Ocean. When overlap does occur between populations, a significant difference exists in the size of specimens that share a number-combina- tion. B. abyssicola of 13 mm MIL from the eastern Pacific has the 1, 2 combination of suckers; several Antarctic specimens have this combi- nation (12% of the total sample), but only one is less than 31 mm ML. Three eastern Pacific abyssicola, 18, 19, and 22 mm in ML, have a 2, 3 combination and a single Antarctic specimen, 51 mm, has a 0, 2, 3 combination; this specimen has the highest combination of suckers of all the specimens from Antarctic waters. The 20-30 mm size range in abyssicola from the eastern Pacific has combinations of 3, 4, and 5 suckers; the same size group from the Antarctic has 0; 0,1; and 0, 1, 2 (2. specimens) combinations. 3. Bathyteuthis abyssicola—Atlantic The combination of buccal suckers of 20 specimens of abyssicola from the Atlantic are plotted on figure 15. Again, the trend toward in- creased numbers of suckers with increasing size of the animal is ap- parent. The 11 specimens below 23 mm ML (9 below 20 mm) have the combinations 1, 2; 1, 2,3; or 2,3; the smallest size within each category increases with increasing numbers of suckers. Nine specimens range in size from 33 to 49 mm ML; two-thirds of these have combinations of 3, 4, 5, 6, and 7 suckers and five of these six have no fewer than 4 suckers on each lappet. B. bacidifera and abyssicola from the eastern Pacific apparently attain larger numbers of suckers at smaller sizes than does abyssicola from the Atlantic. That is, the 3, 4, 5 combination is found on a 30 mm ML specimen in the eastern Pacific and a 38 mm’‘specimen from the Atlantic; the 4, 5 combination is found on a 25 mm bacidifera and a 22 mm abyssicola from the eastern Pacific and on 45 and 47 mm specimens of abyssicola from the Atlantic. Half of the Atlantic abyssicola are 20 mm or less in ML and none of the specimens exhibit the 0; 0, 1; or 0, 1, 2 combinations. In Ant- arctic abyssicola 86% of the sample population falls into these three ‘ategories and 759% into the 0 or 0, 1 groups; 15.5% of the 0 and 0, 1 specimens are under 20 mm while no specimens under 25 mm occur BATHYPELAGIC SQUID BATHYTEUTHIS 73 in the 0, 1,2 group. Both populations have specimens in the 1, 2 group: only one Antarctic specimen is less than 31 mm MEL (19 mm) ; only one Atlantic specimen is greater than 19 mm ML (39 mm). Except for the slight overlap in the 1, 2 group these two populations of ebyssicola are distinct in the number-combinations of suckers on the buccal lappets. 4. Bathyteuthis bacidifera—EKastern Pacific B. bacidifera is represented by nine specimens that exhibit six com- binations of numbers of suckers on the buccal lappets (fig. 15). (This does not include the combinations of numbers of lappets with a par- ticular number of suckers.) There is a tendency toward increased numbers of suckers in larger specimens; only the three larger speci- mens (24 and 37 mm ML) have combinations that include 4 or 5 suckers (4, 5 and 2, 3, 4). Only one of the small specimens has 3 suckers, while the remaining individuals have combinations 0, 1; 0, Tor os The smaller specimens (less than 20 mm ML) overlap somewhat with abyssicola from the Antarctic except that bacidifera has no representatives with 0 suckers and abyssicola has none under 20 mm Suckers 4 34,5 3,4 Oo Female O Female @ Male @ Male ¢ Juvenile 0 Female @ Male Juvenile 0,1 20 ML 40 60 20 ML 40 60 20 40 60 B. bacidifera B. abyssicola B. abyssicola East Pacific Atlantic Figure 15.—Combinations of suckers on buccal lappets of Bathyteuthis. 74 U.S. NATIONAL MUSEUM BULLETIN 291 with 0, 1, 2 and none at all with 1, 3. Antarctic abyssicola is not rep- resented by any specimens that have 4 or 5 suckers on their lappets; only one specimen had as many as 3 suckers. Therefore, bacidifera appears to possess a greater number of suckers on buccal lappets than does abyssicola from Antarctic waters. The preceding discussion about the suckers on the buccal lappets of Bathyteuthis points out that variability exists in this character not only between species and populations but also between individuals of the same population. The general trend is toward an increase in numbers of suckers with increase in size and for populations in lower latitudes to have a significantly greater number of suckers on the lappets. B. abyssicola from the Antarctic exhibits the least tendency toward increased numbers of buccal suckers with increased size, and in addition very seldom ever has more than two suckers on any single lappet and often has no buccal suckers at all. In contrast, larger speci- mens of abyssicola from the Atlantic may have combinations of 4, 5, 6, and 7 lappet suckers, and the smallest specimens available have at least some suckers. In the eastern Pacific populations of abyssicola and bacidifera the data are less complete, but the trend is indicated. Significance of Buccal Suckers Possession of suckers on the buccal lappets apparently is a special- ization to an environmental pressure and not a primitive character as suggested by Naef; the character is variable enough to meet particular narrower demands of the environment. It is suggested here that the specialization is an adaptation associated with the acquisition of a par- ticular type of food in the deep sea habitat of Bathyteuthis. Contrary to Chun’s suggestion (1910), I feel that characters asso- ciated with the digestive system of oegopsids (other than basic struc- tures) are poorly suited for use in determining relationships of fam- ilies. Structures and procedures for obtaining food would seem to be necessarily susceptible to adaptation to various environmental require- ments. Recent cephalopods show a wide range of adaptations of the mechanical means for obtaining particular foods and of the physio- logical processes required for handling these special foods. Buccal suckers may be one of these adaptations. Bathyteuthis is a true deep-sea cephalopod, exhibiting a number of modifications to the deep-sea habitat. As such, it must be derived from a shallower living form. If suckers on the buccal lappets do re- fiect the primitive condition in the Oegopsida as suggested by Naef, it seems more probable that they would be found on a generalized, shallow-living form rather than on a specialized, deep-sea form. They are not found in shallow-water forms. The possession of suckers on BATHYPELAGIC SQUID BATHYTEUTHIS 15 the buccal lappets by the Bathyteuthidae would be an example of the retention of this primitive character in a deep-sea form only, with no trace of the character in other groups. Naef (1923) believed the buccal suckers to be a primitive character possessed by Protodecapus, a prototype. He placed the Metateuthoidea at the base line leading to Metateuthoidea Oegopsida and Metateu- thoidea Myopsida; the Myopsida and Oegopsida separate as equal entities. At the base of the bush of recent Oegopsida, Naef split off the Bathyteuthidae (including Ctenopteryw) and placed it closest to the Myopsida. Then, at the same point he derived the Cranchiidae, Chiroteuthidae, Brachioteuthidae, and Joubiniteuthidae in one line at the top of the order, the Ommastrephidae and Thysanoteuthidae on another, and finally all the rest of the oegopsids in the middle between the Chiroteuthidae and Bathyteuthidae. Since the forerunner of recent oegopsid families was supposed to have had buccal suckers and since the Bathyteuthidae has buceal suckers, the character had to be lost just prior to the appearance of all the other groups of Oegopsida. Naef’s scheme, while possible, is not the only explanation that can be offered. Several possibilities exist: (1) Buccal suckers could be present if the Bathyteuthidae arose from the base of the myopsid stem soon after the myopsids and oegopsids separated. (2) The Bathyteu- thidae could have split off very early from the oegopsid main line, and the Myopsida could have arisen from the base of the Bathyteuthidae. Both of these situations could explain the presence of buccal suckers either as a primitive character retained from the Protodecapus meta- teuthoid stem and lost to the remaining Oegopsida, or as a new charac- ter that appeared after the separation of the remaining Oegopsida. (3) Another possibility allows the buccal suckers to be an independently derived, convergent character. This situation would not require a closer relationship of the Myopsida and Bathyteuthidae than can be rec- onciled on the basis of existing information. The first two suggestions would imply an immediate and rapid divergence of the groups in structure and habitat: the Myopsida to a neritic and sublittoral (epibenthic) existence and the Bathyteuthidae to a bathypelagic existence. Naef’s suggestion allows for the separation of the suborders prior to the loss of buccal suckers to all oegopsids except Bathyteu- thidae. The idea of convergence of the buccal suckers is compatible with the very specialized nature of both of these groups. Number of Arm Suckers Bathyteuthis from different geographical areas show differences in the numbers of suckers on the arms. Figure 16 shows the numbers of suckers on each of the arms against mantle length for specimens of B. abyssicola from Antarctic, Atlantic, and eastern Pacific waters and of 321-534 O—69—6 76 U.S. NATIONAL MUSEUM BULLETIN 291 bacidifera from the eastern Pacific. Figure 17 shows the average num- ber of suckers on each arm of Antarctic abyssicola and of bacidifera. 1. Bathyteuthis abyssicola B. abyssicola from areas other than the Antarctic tends to have more suckers than the Southern Ocean form. The numbers of suckers for Atlantic and eastern Pacific abyssicola fall between those of bacidifera and Antarctic abyssicola (fig. 16). In specimens below about 17-20 mm ML considerable overlap occurs in numbers of suckers, but in larger specimens a trend toward more suckers exists in Atlantic and eastern Pacific specimens. In fact, specimens of abyssicola from the eastern Pacific closely approach bacidifera in numbers of suckers. Since the number of observations on the two forms is small, however, I shall not place much emphasis on the apparent differences other than to mention that geographical variation exists in this character. Number of Suckers 180 12 90: 30 xX Bbacidifera B, abyssicolg: @E Pacific © Antarctic @ Atlantic 50 10 30 | il BUMS Wl IV Ficure 16.—Number of suckers on the arms of Bathyteuthis abyssicola from the Antarctic, Atlantic, and eastern tropical Pacific Oceans and of B. bacidifera. BATHYPELAGIC SQUID BATHYTEUTHIS 77 2. Bathyteuthis abyssicola vs. B. bacidifera The difference in numbers of suckers is most pronounced between abyssicola from Antarctic waters and bacidifera from eastern Pacific equatorial waters. B. bacidifera has a greater number of suckers on each of the arms except in very young specimens (fig. 17). Larvae of both species at about 6 mm ML have nearly the same number of suck- ers, but during the next 3-6 mm of growth bacidifera adds suckers more rapidly than abyssicola, so that the first three pairs of arms have from one-third to nearly one-half more suckers, and the fourth arms have nearly twice as many (Table V). With continued increase in mantle length the divergence in numbers of suckers increases gradually in the first three pairs of arms from about 40% more to 75% more suckers in bacidifera than in abyssicola. The fourth arms of bacidifera 140 120 100 Average Number Suckers Arms I-IV 80 60 40 @ Babyssicola o B. bacidifera 20 Arm!| tl Wl Iv 5.6/6 9—12/ 9-12 14-15/17 18-19/19 23-29/26 33-39/37 49:-75/ — Size Group, ML,mm Ficgure 17.—Average number of suckers on the arms of Bathyteuthis abyssicola (Antarctic) and B. bacidifera. Size groups are divided into mantle lengths of B. abyssicola/B. bacidifera. 78 Arm II III TE III IV II III re ii II III IV It III IV II III U.S. NATIONAL MUSEUM BULLETIN 291 TaBLE V.—Comparison of numbers of arm suckers B. abyssicola, Antarctic ML size group, milli- meters 5. 6 9-12 14-15 18-19 23-29 33-39 49-75 Range 16 19 18 14 23-43 25-45 25-48 20-35 38-50 44-49 41-57 35-46 56-60 64-69 63-71 54-63 57-87 60-96 68-91 53-72 66-91 82-96 73-102 68-92 91-101 104-120 97-119 79-96 Average (34) (38) (39) (28) (44) (47) (49) (41) (58) (67) (67) (59) (75) (82) (78) (64) (82) (89) (88) (78) (98) (110) (110) (87) ML size group, milli- meters 6 9-12 B. bacidifera Range 13 20 21 15 41-50 49-54 50-60 40-48 70 75 78 56 80-85 90-95 92-105 69-75 115-130 120-130 130-133 82-88 128-170 142-153 131-158 90-103 Average (45) (52) (56) (53) (83) (93) (99) (72) (123) (125) (132) (85) (144) (147) (145) (95) no specimens Times greater 81 1.05 He 4 1.07 1.32 1.37 1.43 1.89 1.59 1.59 1.59 1.37 1.43 1.39 1.48 1,22 1.64 1.52 1.69 1.33 1.76 1.65 1.65 1.22 exhibit less of a proportional increase than the other arms, but they maintain between a fourth and a third more suckers than in abyssicola of the same size group. Not only is the average number of suckers greater in bacidifera, but the ranges of values between the two species show virtually no overlap, so that the species are quite easily separable on numbers of arm suckers alone. BATHYPELAGIC SQUID BATHYTEUTHIS 79 In addition, the largest specimens of abyssicola (49-75 mm), the largest of which is two times longer in mantle length than the largest bacidifera, do not approach the largest known bac¢difera in range or average of suckers; the equivalent number of suckers would be found in specimens of bacidifera one-half to one-third the size of the largest abyssicola from the Antarctic. 3. Bathyteuthis berryi B. berryi was discovered after the above analysis was completed. The suckers on the arms of this species far outnumber those of abyss7- cola and bacidifera. The following figures for berryi can be compared with those in Table V: ML, millimeters Arm No. of suckers 12 I-III 62-75 IV 48 23 I-III 175-185 IV 110 49 I-III 275 IV 150 Gill Size A conspicuous difference in the size of the gills is apparent between abyssicola from the Antarctic and bacidifera. B. abyssicola has small gills while bacidifera has larger, more voluminous gills. A character such as gill size, however, may be a phenotypic expression of environ- mental conditions rather than a genotypic difference. Antarctic and eastern Pacific equatorial waters differ considerably in environmental conditions, particularly in oxygen content, which in this context would probably most influence gill size. The oxygen content of eastern Pacific equatorial waters is significantly lower than that of Antarctic waters. In an attempt to evaluate the extent and signficance of the differences observed between bacidifera and Antarctic abyssicola, it is necessary to examine the gill dimensions of specimens of abyssicola from the same locality as bacidifera and from localities other than the Antarctic and the eastern Pacific. Although the sample sizes are not large for the Atlantic and eastern Pacific populations of Bathyteuthis, they do pre- sent some interesting trends. Size of the gills may be determined by length, width, volume, and number of gill filaments. In the current study gill volume has been omitted. 1. Gill filaments The number of gill filaments for specimens of bacidifera and of abyssicola from the Antarctic, Atlantic, and eastern Pacific is plotted against mantle length in figure 18. The plot for abyssicola shows an 80 U.S. NATIONAL MUSEUM BULLETIN 291 Number Gill Filaments @B. bacidifera B. abyssicola © Antarctic @ E. Pacific or Atlantic 10 20 30 40 50 60 70 80 ML Ficure 18.—Number of gill filaments in Bathyteuthis bacidifera and B. abyssicola plotted against mantle length. increase in number of gill filaments up to about 30 mm ML; above 30 mm the points level off at an apparently maximum number of fila- ments (20-23). The three populations of abyssicola overlap in number of gill filaments and are not divisible into separate components. The plot for bacidifera rises sharply in the first 12 mm of ML; be- tween 17 and 37 mm the increase in numbers is gradual, rising from 23 to 26 filaments. Specimens of bacidifera of 9-12 mm ML have 20-23 filaments; abyssicola does not attain 20 filaments until a mantle length of about 22 mm, and only two specimens had 23 filaments (31 and 55 mm ML). In any size group there is a difference of 2-3 filaments between the minimum number in bacidifera and the maximum in abyssicola. East- ern Pacfiic abyssicola is inseparable from other populations of this species in number of gill filaments, and no apparent overlap of values exists between the two species. 2. Gill length Figure 194 shows gill length against mantle length. The points for abyssicola from the Antarctic fall well below those for bacidifera, while the points for abyssicola from the Atlantic and eastern Pacific he between these values. Atlantic specimens have gills nearly the same length as Antarctic specimens, but eastern Pacific specimens have gills intermediate in length between Antarctic specimens and bacidifera. Figure 20a plots the mean of the gill length to mantle length indices of the four populations. The gills of Antarctic and Atlantic abyssicola are about one-third as long as the mantle and gills of eastern Pacific abyssicola are just over 40% as long; B. bacidifera has gills nearly one-half as long as the mantle. BATHYPELAGIC SQUID BATHYTEUTHIS Sl Gill Length, mm x 0° us ° on X B.bacidifera X& ¥ e B,abyssicola go ° ° Sha fa $a antic @E. Pacific 10 20 30 40 50 60 70 75 ML Ficure 19.—Bathyteuthis bacidifera and B. abyssicola: a, gill width vs. mantle length; b, gill length vs. mantle length. 3. Gill width The plots of gill width against mantle length (fig. 196) indicate a distinct separation between the two species. The points for eastern Pacific abyssicola lie just above those for Antarctic specimens, but, still, these gills are distinctly narrower than those of bacidifera. The means of the gill width to mantle length indices for abyssicola, plotted on figure 204, lie between 5.5% and 7.4%, but the mean index for bacidifera is 12.4%. Although the gills of eastern Pacific abyssicola tend to be slightly wider than those of other specimens, the difference is not nearly so great as between the populations of abyssicola and of bacidifera. The gills of bacidifera average nearly 2.3 times wider than the gills of Antarctic abyssicola and 1.7 times wider than eastern Paci- fic abyssicola. Therefore, gill width is significantly greater in bacidi- fera than in abyssicola. 4. Gill width: gill length index Perhaps a better expression of actual gill size is the index of gill width to gill length. Figure 20c plots the average values for the popu- lations of Bathyteuthis; the data are summarized in Table VI. B. bacidifera has a mean gill-width to gill-length index that is consider- ably greater than that for abyssicola. That is, the width of any sized gill of bacidifera will be about one-fourth of its length, but in abys- sicola it will be only about one-sixth of the gill length. 82 U.S. NATIONAL MUSEUM BULLETIN 291 | M«““a YM | MM’Mm’umq©mMa{a Wa WIM t GC’ x tl. E: _abyssicola B bacidifera Ficure 20.—Bathyteuthis abyssicola and B. bacidifera: a, index of gill length to mantle length; 5, index of gill width to mantle length; c, index of gill length to gill width. B. bacidifera has larger gills than B. abyssicola, but in one dimen- sion (gill length) the eastern Pacific form of abyssicola approaches bacidifera. This is not totally unexpected since the Eastern Pacific Equatorial Water Mass is extremely low in oxygen content and certain species may have to adapt to the lowered oxygen concentrations (see following discussion and Ebeling and Weed, 1963; Marshall, 1960; Walters, 1961). The slight increase in gill length in eastern Pacific abyssicola, however, is not accompanied by an increase in number of gill filaments or a more favorable gill-width to gill-length ratio. Meristic characters might be preferred to morphometric features as being more stable and more indicative of differentiation. In this case, however, the meristic and morphometric characters combine to lead to the conclusion that the observed differences in gill size between abyssicola and bacidifera are specific and are not only a phenotypic expression of environmental conditions. Furthermore, the differences within populations of abyssicola are also genetic and are maintained within the geographic boundaries of the populations. BATHYPELAGIC SQUID BATHYTEUTHIS 83 TaBLE VI.—Gill indices of the populations of B. abyssicola and B. bacidifera ! B. abyssicola B. bacidifera Antarctic Atlantic E. Pacific E. Pacific Character 2 Range Mean Range Mean Range Mean | Range Mean GL: ML 28-40 (32.4) 29-40 (34.4) 38-45 (41.4) | 42-58 (47.5) GW : ML 4-7 (5.5) 5-7.2 (6.2) 6.7- (7.4) 8.3- (12.4) 8.3 55 GW : GL 12.5— (16.9) 13.3- (17.1) 16.4— (17.6) | 18-33 (25) 22 A 20 1 The dimensions of gills from B. berryi 16-49 mm in ML are appended here. Gill filaments number 19-21. Indices—GL : ML=43-53 (48.6); GW : ML=7.5-15.7 (12.5); GW : GL=17-80 (25.5). 2 G@L=gill length; GW=gill width; ML=mantle length. Significance of Gill Size In the order Octopoda the number of gill filaments is commonly used as a taxonomic character, but in the suborder Oegopsida the size of the gills has not previously been examined for taxonomic or biologi- cal significance. Robson (1925, p. 1337) presented a study of gill size in the deep-sea Octopoda. He showed that, in general, the gills of deep- sea octopods are more or less reduced in area, in number of filaments, and by atrophy of the inner demibranch. Although Robson reached no definite conclusions, he suggested that the reduction in respiratory sur- face in deep-sea octopods was related to lowered metabolism in the low temperatures of the depths. Voss (1967) also considered that reduced gill area is due to reduced metabolic requirements. Differences in gill size have been recorded for several groups of meso- pelagic and bathypelagic fishes. Marshall (1960) and Walters (1961) have noted that the surface area of gills in bathypelagic fishes is much less than in mesopelagic species. Three species of the bathypelagic genus Gonostoma exhibit a trend toward reduction in gill surface with the shallowest living species having the greatest gill area. The sugges- tion for these fishes is that the deeper living species tend to have lower metabolisms and to lead less active lives. Walters considered that the bathypelagic giganturids have a metabolic level about one-third that of coastal fishes. Ebeling and Weed (1963) reported a clinal tendency in the length of the gill filaments in three geographically distinct populations of the melamphaid fish, Scopelogadus mizolepis. Gill filaments on the’ first arch of S.m. mizolepis from the Sargasso Sea were short, those of S. m. bispinosus from the Gulf of Panama were long, and those of S. m. mizolepis from the Indo-Pacific and off West Africa were intermediate in length. Contrary to Marshall’s findings in Gonostoma spp., Ebeling 84 U.S. NATIONAL MUSEUM BULLETIN 291 and Weed were unable to correlate differences in gill area in Scopelog- gadus with depth of capture. However, there was a correlation between gill size and the oxygen content in the areas where the subspecies live. Oxygen concentration in the Sargasso Sea habitat of S. m. mizolepis with short filaments is relatively high, ranging 3-6 ml/L, while the oxygen content of the eastern tropical Pacific habitat of S. m. bispino- osus with long filaments is very low, from less than 0.1 to 2 ml/L; much of the 200-2000 m depth range of S. m. bispinosus corresponds to the oxygen minimum layer. Specimens of another melamphaid, Poro- mitra megalops, also from the oxygen-poor tropical eastern Pacific, have a greater gill surface area than do specimens from the North Atlantic. Ebeling and Weed suggested that the increase in gill surface is an adaptation toward more efficient utilization of oxygen in regions of low oxygen concentration. An interesting parallel exists between Scopelogadus and Bathyteu- this. B. bacidifera has gills with greater dimensions and more filaments than abyssicola. The only specimens of bacidifera that are available to this study come from the eastern tropical Pacific within the boundaries of the oxygen-poor eastern Pacific Equatorial Water Mass. Oxygen concentrations were determined at each station so an accurate estimate of oxygen values at depths of capture is available (Table VII). The oxygen values range from 0.47 ml/L at 750 m to 1.47 ml/L at 1550 m, and they correspond to the midportion of the oxygen minimum layer which has its lowest values somewhat shallower (200-700 m). Large gill size in bacidifera, therefore, may well be an adaptation to these very low oxygen concentrations. The various populations of abyssicola show differences in gill size also. B. abyssicola has the smallest gills in the Antarctic; it has larger gills in the eastern Pacific; it has intermediate-sized gills in the At- lantic. Oxygen content of Antarctic waters is very high, ranging from 4.0 to 5.0 ml/L in the depth range that abyssicola inhabits; oxygen TaBLeE VII.—Ozygen, temperature, salinity, and sigma-t values at Bathyteuthis stations in the eastern tropical Pacific (cirea 07° N 80° W) Station Depth, Temperature number meters Oxygen ml/L oC; Salinity °/ 5. Sigma-t D 1203 XVI 750 AS 5.92 34.56 20.20 D 1208 VIII 750 09 5.58 34.56 250) D 1203 XIII 1000 87 4.84 34.59 27.40 D 1208 XVI 1050 .96 4.70 34.58 27.41 D 1209 III 1250 1.12 4.2 34.60 27.47 D 1203 VI 1250 Leaked 3.9 34.58 27.50 D 1203 XII 1250 1.21 3.8 34.61 Ziad D 1208 XIV 1550 1.47 S00 34.60 27.56 D 1209 I 1750 BATHYPELAGIC SQUID BATHYTEUTHIS 85 values nearer the surface and in greater depths may exceed these values. The few specimens of abyssicola captured in the eastern Pacific were taken in the same stations as the deeper specimens of bacidifera where oxygen values ranged from 1.12 ml/L to 1.47 ml/L. Oxygen concentrations in the eastern tropical Pacific do not exceed 2.0 ml/L until depths greater than 2000 m (Sverdrup et al., 1942; Wooster and Cromwell, 1958). The Atlantic specimens of abyssicola came from a broad range of oxygen concentrations with extremes of 2 ml/L and 6 ml/L, but generally the values were 3-5 ml/L. The correlation between oxygen concentration and gill size in Bathy- teuthis, as in Scopelogadus, prompts speculation that larger gills pro- vide more efficient means of obtaining oxygen. The possibility may be further strengthened by the fact that when the two species of Bathy- teuthis occur in the same area of extremely low oxygen concentrations they both possess longer, wider gills. Relationship of Ctenopteryx to Bathyteuthis Shortly after its description by Appelof in 1890 Ctenopteryx was united with Bathyteuthis in the Bathyteuthidae by Pfeffer (1900). Most later authors accepted this designation (e.g., Chun, 1910; Pfeffer, 1912; Naef, 1923; Grimpe, 1922, 1925; Thiele, 1935) but generally implied that the two genera were not closely related. To emphasize the distinctiveness of the genera, Grimpe (1922) erected the subfam- ilies Bathyteuthinae and Ctenopteryginae. Allen (1945) felt that even subfamilial distinction was insufficient so she withdrew Ctenopteryx from the Bathyteuthidae and elevated it to the family Ctenoptery- gidae based solely on fin structure and body proportions. Allen’s act constitutes the latest revision of the higher taxa that encompass Cteno- pteryx and Bathyteuthis. On the basis of the material at hand, it is necessary to reevaluate the systematic positions and relationships of these two genera. A brief description of sonie taxonomic characters of Ctenopteryx will help to determine the degree of similarity with Bathyteuthis. Some of the characters compared may be only specific, but since the species of Ctenopteryx are poorly understood, it is advisable to use whatever characters are pertinent to this discussion. The funnel-mantle locking apparatus is the simple, straight ridge and groove type. The funnel component is broad posteriorly and nar- rows anteriorly ; the groove is deep and narrow. The funnel component of Ctenopteryx is broader posteriorly, but in general the locking ap- paratuses of the two genera exhibit no significant differences. 86 U.S. NATIONAL MUSEUM BULLETIN 291 The tentacles of Ctenopterya are long and slender. The tentacular stalks are naked. In proportion to the length of the tentacle, the clubs are short and unexpanded; they bear no distinct carpus, manus, or dactylus. Eight to fourteen rows of minute, closely packed suckers cover the oral surface of the club. A thin, narrow keel extends along the dorsel aboral border of the club. Protective membranes are lack- ing. The clubs of Ctenopteryx and Bathyteuthis are similar in basic design, but Ctenopteryx tends to have a few (up to four) more rows of suckers. The connectives from the buccal membrane of Ctenopteryx attach to the dorsal oral edges of arms I and II and to the ventral oral edges of arms ITT and IV. This arrangement is in contrast to the dorsal, dorsal, ventral, dorsal arrangement in Bathyteuthis, Ctenopteryx has 12-15 suckers in two rows on each of the buccal lappets. These are considerably more numerous and are slightly larg- er than the buccal suckers of Bathyteuthis. The fins of Ctenopterya are subterminal, and in adults they are near- ly as long as the mantle. The fins consist of long, muscular supports connected by a thin web. In Bathyteuthis, too, the fins are subterminal, but they remain small and simple throughout life. The suckers on the arms of Ctenopterya originate at the bases of the arms in 1 or 2 rows. In arm pairs I-ITI the suckers increase to six rows on the distal half; arms IV retain two rows throughout. Bathyteuthis has a similar increase to four rows. The suckers on the arms are minute in both genera. Swimming keels and protective membranes are rudi- mentary in both genera. Dentition on the sucker rings is usually a specific character; in Ctenopteryaw, however, the sucker rings may be distinct at the generic level, since they lack true teeth entirely. They bear only roughly scal- loped borders. Bathyteuthis has sucker rings that bear truncate to rounded teeth. The funnel organ in Ctenopterya is very large. The dorsal member has an inverted U- or V-shape with long, broad limbs. A distinct papilla protrudes anteriorly from the apex. In Bathyteuthis the dorsal mem- ber of the funnel organ is noticeably smaller, but since a great deal of intraspecific variation exists in the structure of the funnel organ, no strong significance can be placed on this difference alone. The gladius of Ctenopteryx has a long, narrow rhachis that is deeply V-shaped in cross section. The narrow sides of the rhachis are straight, and the lateral edges are tapered, not rolled under. A heavy, median ridge extends the length of the gladius; the ridge tapers posteriorly along the vane and terminates just before the end of the gladius. The vane is thin, broad, and rounded. A conus is lacking, but the postero- BATHYPELAGIC SQUID BATHYTEUTHIS 87 lateral edge of the vane folds under to form a shallow, cup-shaped terminus of the gladius. In contrast, the long narrow rhachis of Bathyteuthis is C-shaped in cross section; its lateral edges are rolled under to form longitudinal rods, and it lacks a median ridge. The lateral rods of the rhachis taper along the vane and terminate before the end of the gladius. The vane is tissue-thin and marked with series of concentric lines. A conus is lacking in Bathyteuthis, too; the posterolateral edges of the vane barely turn under to form a broad, spoonlike, tissue-thin end to the gladius. Therefore, the gladii of the two genera exhibit marked dif- ferences. The eyes of Ctenopteryx are huge and are directed laterally. A large, strip-photophore is located on the ventral half of the bulbus (in some species, at least). The eyes of Bathyteuthis are large and globular; they are directed anterolaterally. Photophores on the eye are unknown. The muscular bridles that connect the anterior dorsal wall of the funnel to the ventral surface of the head are narrow and thin in Ctenopteryx. In Bathyteuthis the bridles are broad, thin bands em- bedded in the posterior depression of the funnel groove. Some Ctenopteryx have a huge, median, visceral photophore. Vis- ceral photophores are lacking in Bathyteuthis, but it has six simple photophores at the bases of the dorsal three pairs of arms. Ctenopteryx has no corresponding light organs. The female reproductive systems show marked differences between the two genera. The oviducal glands of Ctenopterya are small and in- conspicuous; they are flat, round glands that le within the viscero- pericardial coelom. The oviducal glands of Bathyteuthis are large, conspicuous, swollen structures that lie in the mantle cavity dorso- lateral to the nidamental glands. The nidamental glands of both genera are similar in structure and positon. The most striking feature of Ctenopteryx is the possession of “accessory nidamental glands.” No other oegopsid, including Bathyteuthis, is known to have these glands. Accessory nidamental glands are found in sepioids and myopsids. In Ctenopteryx the supposed accessory nidamental gland is a single, median structure that lies on the wall of the nephridial coelom anterior to the nidamental glands and ventral to the kidneys. It is flat, lobate, glandular, and unsculptured; it bears little resem- blance to the paired, swollen, striated accessory nidamental glands of myopsids and sepioids. This review of the characters of Ctenopteryx reveals that Bathy- teuthis and Ctenopteryx share some features that may indicate rela- tionship: simple mantle-funnel locking apparatus; short, simple club with many (8-14) rows of minute suckers; short arms with more than 88 U.S. NATIONAL MUSEUM BULLETIN 291 two rows of small suckers (4-6 except on arm IV); suckers on the buccal lappets, etc. Although differences in these characters do occur between genera, they are not necessarily of a magnitude that require separation at the familial level. The different order of attachment of the buccal connectives to the ventral arms, however, is a major obstacle to the close relationship between Ctenopteryx and Bathyteuthis. A survey by Young and Roper (1968) pointed out that the arrangement of buccal connectives is an extremely stable character within the families of Oegopsida. The Bathyteuthidae with Ctenopteryx included would be the only oegopsid family that did not conform to this pattern. For this reason it is difficult to reconcile the differences in attachment of buccal con- nectives between Ctenopteryx and Bathyteuthis. Although the general shape of the gladius is similar in Ctenopteryx and Bocheewshen, some of the details of structure indicate that the similarities may be more superficial than earlier authors thought (e.g., Pfeffer, 1900; Naef, 1923). Unfortunately, no analysis of the nate of the detailed structure of the gladius in showing relation- ships in higher taxa is available. Without such an analysis for the Oegopsida it is impossible at this time to determine how much weight to place on the observed differences between Ctenopteryx and Bathyteuthis. A priori it would seem, however, that the differences in the structure of the gladii are basic and probably of familial significance. Naef (1923) was the first to recognize the gland in Ctenopteryx that he believed to be homologous with the accessory nidamental glands of sepioids and myopsids. He gave a comparison with the true accessory nidamental glands of myopsids and sepioids, and he concluded that the glands are similar in structure, except that they have become fused in Ctenopteryx. This may be so, but the gland that I have observed in Ctenopteryx bears no resemblance to the accessory nidamental glands of other decapods. It may be an entirely different structure, but further comparative studies are neces- sary before a final decision can be reached. In any case, the gland is lacking in Bathyteuthis, and the oviducal glands of the two genera are clearly dissimilar. Insummary, Ctenopteryx and Bathyteuthis share several distinctive features, one of which is unique among the Oegopsida. Other im- portant characters, however, show sharp contrasts between the genera. The first set of characters implies that Bathyteuthis and Ctenopteryx may be related forms, but the second set precludes a very close relation- ship. Bathyteuthis and Ctenopteryx are sufficiently distinct to warrant their separation into separate families. BATHYPELAGIC SQUID BATHYTEUTHIS 89 Familial Relationships of the Bathyteuthidae Few earlier authors have attempted to establish the relationships of the Bathyteuthidae with other families of the Oegopsida. The Bathy- teuthidae is usually placed in a diverse group that includes the Histio- teuthidae, Brachioteuthidae, Ommastrephidae (e.g., Hoyle, 1885, 1904, etc; Pfeffer, 1900, 1912; Thiele, 1935; Voss, 1956). Chun (1910) con- cluded that although the buccal connectives are similar to those of the Enoploteuthidae, Histioteuthidae and Ommastrephidae, the nearest relatives of the Bathyteuthidae cannot be given and that the family occupies a truly isolated position in the Oegopsida. Naef (1916, 1921, 1921a, 1922) listed the Bathyteuthidae at the beginning of the Oegop- sida next to the Myopsida. In his monograph Naef (1923) explained his reasons for considering the Bathyteuthidae primitive: primarily, the unique possession of suckers on the buccal lappets, and the 4-rowed arrangement of suckers on the arms and more than four rows on the club; secondarily, rhachis incompletely grown over by mantle muscle, fins loosely attached, neck folds weak, etc. Grimpe (1922, 1925) con- curred with Naef’s decision. It is not unique for the Bathyteuthidae that familial relationships have yet to be established. Most families of the Oegopsida stand alone in spite of the valuable works of Chun (1910), Pfeffer (1912), and Naef (1923). Familial relationships seem to exist for the Enoploteu- thidae and Lycoteuthidae (Voss, 1962), for Ommastrephidae and Thysanoteuthidae, and for Chiroteuthidae, Mastigoteuthidae, Proma- choteuthidae, and Joubiniteuthidae (see discussions in Roper and Young, 1968; Young and Roper, 1968). The Bathyteuthidae is a distinct family ; but is it more closely related to some families than to others? Several characters are useful at the familial level in the Oegopsida and some of the more important of these have been reviewed by Young and Roper (1968) and by Roper, Young, and Voss (in press). The funnel-mantle locking apparatus is one of the most stable familial characters in the Oegopsida, and it is the primary character that defines some families (e.g., Ommastrephidae and Thysano- teuthidae). Three types of locking apparatus exist: (1) Mantle and funnel components are separate and simple; a straight ridge on each mantle component locks into a straight, median suleus on each fun- nel component. Although some variability exists among families in relative dimensions of the apparatus, most oegopsid families have the simple type. These include the Enoploteuthidae, Onychoteuthidae, Lycoteuthidae, Architeuthidae, Histioteuthidae, Octopoteuthidae, Batoteuthidae, etc. (2) Mantle and funnel components separate and complex. This category includes the inverted-T and lazy-T locks of the 90 U.S. NATIONAL MUSEUM BULLETIN 291 Ommastrephidae and Thysanoteuthidae, the ovoid, ear-shaped locks of the Chiroteuthidae and Mastigoteuthidae, the ovoid, bowl-shaped locks of the Promachoteuthidae and Joubiniteuthidae, and the sub- triangular locks of the Cycloteuthidae. (3) Mantle and funnel com- ponets completely fused. This type is found only in the Grimaldi- teuthidae and Cranchiidae. The Bathyteuthidae, therefore, with its straight, simple locking apparatus, is aligned with the majority of oegopsid families and is distinct from the Ommastrephidae, Thy- sanoteuthidae, Chiroteuthidae, Mastigoteuthidae, Cranchiidae, etc. The sulcus of the funnel component in the Bathyteuthidae tends to be- come broader and shallower in the posterior part, and in this respect, the lock resembles that of the Histioteuthidae and Octopoteuthidae. Tentacular clubs provide stable familial characters among the Oegopsida. The typical club is short, expanded and somewhat flat- tened; it is generally divisible into carpus, manus, and dactylus; it generally has protective membranes, swimming keels, and a 4-rowed arrangement of suckers (and/or hooks). The Ommastrephidae (ex- cept lex), Thysanoteuthidae, and Lycoteuthidae are examples of the basic type. Several variations from the basic type occur. For instance, histioteuthids and gonatids have a few more rows of suckers on the manus or dactylus; architeuthids and neoteuthids have a small cluster of irregularly arranged suckers on the proximal part of the manus. Some families have distinctive clubs that bear no resemblance to the typical club. These include the Mastigoteuthidae, Promachoteuthidae, Joubiniteuthidae, Ctenopterygidae, Brachioteuthidae, Batoteuthidae, Octopoteuthidae, and Grimalditeuthidae. The Grimalditeuthidae lack tentacles; the Octopoteuthidae lack tentacles in adults (except Taningia); the Brachioteuthidae have expanded clubs with the normal 3-4 rows distally, but rows of suckers that extend proximally along the tentacular stalk. The Mastigoteuthidae, Promachoteuthidae and Joubiniteuthidae have distinctive clubs that are very long and unexpanded; they bear many rows of minute, closely packed suckers. The Ctenopterygidae have short, simple clubs with about 14 rows of suckers. The clubs of the Bathyteuthidae are unlike those of most other oegopsid families; they are relatively short, unexpanded, and undiffer- entiated into carpus, manus, and dactylus. They bear 8-10 rows of minute, closely packed suckers. The extreme distal tip serves as the growing end where sucker proliferation occurs. This feature is shared with the Mastigoteuthidae and Promachoteuthidae, but presently too little is known about growth of the clubs and sucker proliferation in other oegopsid families to place emphasis on this similarity. The BATHYPELAGIC SQUID BATHYTEUTHIS 91 clubs of the Bathyteuthidae weakly resemble those of the Promacho- teuthidae, Mastigoteuthidae, or Joubiniteuthidae by having many rows of minute suckers, but the bathyteuthid clubs are very short and they have fewer sucker rows; in this respect they resemble the Ctenopterygidae. The connectives of the buccal membrane attach to the oral surface of the base of each arm at either the dorsal or ventral edge. The con- nectives attach dorsally to arms I and II and ventrally to arms III in all oegopsids. (Z'noploteuthis dubia Adam, 1960, is the only known exception; the connectives attach dorsally to all four pairs of arms.) The connectives to arms IV may attach dorsally or ventrally. The order of attachment is a constant feature within oegopsid families. Furthermore, families that show relationships by other characters also have the same order of attachment (e.g., Lycoteuthidae and En- oploteuthidae; Mastigoteuthidae, Chiroteuthidae, and Promachoteu- thidae). The Bathyteuthidae have buccal connectives that attach dorsally to the ventral arms. This character is shared with the Enoploteuthidae, Lycoteuthidae Histioteuthidae, Psychroteuthidae, Neoteuthidae, Architeuthidae, and Ommastrephidae. Most oegopsid families have a biserial arrangement of suckers or hooks along the arms. The Enoploteuthidae and Octopoteuthidae, which have mainly hooks, may have two to four rows of minute suckers at the arm tips. Species of Gonatidae generally have two median rows of hooks and two lateral rows of small suckers on the dorsal three pairs of arms. At the arm tips the hooks may be replaced by suckers. The ventral arms always have four rows of suckers. The tips of the arms in Gonatopsis octopedatus have many rows of small suckers. The Joubiniteuthidae have five to six rows of suckers along the dorsal three pairs of arms and four rows along the ventral arms. The Ctenop- terygidae have five to six irregular rows. In the Bathyteuthidae the suckers originate on the bases of the arms in a biserial arrangement, then increase to three and four irregularly arranged rows. The number of rows of suckers in the Joubiniteuthidae, Ctenopterygidae and Bathyteuthidae is generally similar, but other characters weigh more heavily to preclude a close relationship. Therefore, multiple rows of suckers occur independently in a few families that are widely diverse on the basis of more stable familial characters. The gladius in oegopsids should show relationships between higher taxa, but no comprehensive study of oegopsid gladii has appeared in the literature, and, other than the gross features, no taxonomically important characters of the gladius have been delineated. In fact, in many species the gladius remains undescribed. Still, it is possible to find general similarities between gladii of some groups. Several fami- lies have a long, thin rhachis, a reduced or absent free vane, and a 321-534 O—69——7 92 U.S. NATIONAL MUSEUM BULLETIN 291 conus (usually long) formed by the ventral infolding and fusion of the lateral edges of the vane. Diverse families with this type of gladius include the Chiroteuthidae, Mastigoteuthidae, Joubiniteuthidae, Bato- teuthidae, Grimalditeuthidae, Brachioteuthidae, some Cranchiidae, etc. Another type of gladius, like that of the Ommastrephidae, is heavily constructed with a thick rhachis, a long, narrow (or absent) vane, and a short, heavy conus. Lycoteuthids, enoploteuthids, and histioteuthids have a type of gladius that tends to have a short free rhachis, a fairly broad, convex or hourglass-shaped vane, and an open conus. The Onychoteuthidae have a short, free rhachis, a long, narrow vane, and a long spikelike extension of the conus. The paddle-shaped bathyteuthid gladius, with its long, free rhachis, broad, thin vane, and flat, open “conus,” is unmatched in simplicity of shape and struc- ture; at the present stage of knowledge about teuthoid gladii, the bathyteuthid gladius cannot be closely associated with the pen of any other family. One of the most distinctive features of the Bathyteuthidae is the occurrence of suckers on the buccal lappets. This feature alone sets it apart from all other oegopsid families except the Ctenopterygidae. Ctenopteryx, of course, was included in the Bathyteuthidae for many years as a subfamily, primarily because of this shared character. In addition, Ctenopteryx has supposed accessory nidamental glands, a unique organ in the Oegopsida. Some sepioids and myopsids have buccal suckers and accessory nidamentals also, and Naef regarded them as primitive characters. The occurrence of the accessory nidamental glands in Ctenopteryx in conjunction with the presumed primitiveness of the buccal suckers, four rows of arm suckers, and more than four rows of club suckers, led Naef to place the Bathyteuthidae at the beginning of the Oegopsida next to the Myopsida. The question is not settled, however, for it may be argued that some of these characters are not primitive but are adaptive. Many rows of suckers on the arms and clubs and suckers on the buccal lappets may be adaptations to particular environmental requirements or, more specifically, to certain types of food organisms. The many-rowed condi- tion, or tendencies toward it, occurs in several diverse groups of oegop- sids: on the arms of gonatids, joubiniteuthids, enoploteuthids, and octopoteuthids, and on the clubs of mastigoteuthids, promachoteuthids, joubiniteuthids, gonatids, histioteuthids, brachioteuthids, and bato- teuthids. All of these groups are mesopelagic or bathypelagic or, in the case of gonatids and brachioteuthids, they are transitional forms that live closer to the surface and range down into the mesopelagic zone. An increased number of suckers on arms and tentacles would be BATHYPELAGIC SQUID BATHYTEUTHIS 93 advantageous in capturing and holding prey. Buccal lappets on an expanded buccal membrane fit neatly between the bases of the arms, and suckers on the lappets would provide an additional holding mechanism. Finally, it is not certain that the glandular structure that Naef de- scribed in Ctenopteryx is the homologue of the accessory nidamental glands of sepioids and myopsids. Naef perhaps realized this, for had he been more certain he surely would have made a greater point of it. Therefore, based on the characters presently used to indicate familial relationships, the Bathyteuthidae can be aligned with the relatively small group of families that have simple, straight locking apparatuses and buecal connectives attached to the dorsal side of arms IV. Within this group it shares a club that has (at least some) suckers in more than fours rows with the Histioteuthidae, Psychroteuthidae, Neoteuthidae and Architeuthidae; other similarities in the clubs, however, are lacking. The Bathyteuthidae and Ctenopterygidae share a number of features, e.g., suckers on the buccal lappets, more than two rows of suckers on the arms, more than four on the clubs, and simple, straight locking apparatuses, but they differ in the attachment of buccal con- nectives to the fourth arms. This difference is basic and probably pre- cludes a close relationship between the two families. => FR 7 a ey or oi a 7 7 - a ai riscrioma cae wrt ss sou ae ee ra ery Cp eg hie hyh. Seneca aces) gemma ely Mee tee vaiiasdettS a adil elon (Puna yg) ae dake Teeth Vairaaa pe stags fe ree pegs “bike, §6F4,.5% 28 ‘Beare deen s dh ee a ded retesteslietn dcr aatl\ od iahed, aan 7 WEE elominc re aly ke idtaliael, ol) piokagnRgee a inate aa: Lenole, weary ae eel Oee A, ae eve Linnean a BEV bc hte rey 0 egiimer shi) hasan cues a thts bitiioad’ aailty. eg Leere ties. 1 aaubaidy: ft 3 jen Lviean tr Gemietea eee jaihetawiigile ad re ohbatt aeagr give chlhielk tan leonard qu ool syhonte piri] im tiphie teyee ordi eee cy . fue. pens Fi F? Sg heats Lads = ner Bask, ebhe ee cethanney, acinar hive, GING hash s ie WE Bie ee! oe. eb te Pay ey ee b ear yl. septate ial, See se WP aude ta OE aed ceria tig) iid Con siath oer em Gia, eho 40 t= fe feos! lilet$y hv il meocpme- 6 eae = Fale aa ios Oy ett? Vet pie 7 ads am ty BPG opps Ee (reer? theme tiny basitats ho pout fae ube elgiiie big. sober: Hollins thgader lems ted Tete vice eS tO aan te = | ay hep se ige qeey eater ey 7) ssa wig aay _ tine nate el ey Lee T «bee Matqubletds att a ie ars iviw , ws es iene = ast “ee esl 2 = ove ' = - 6A *. gr) ane bs 9 ¥ = 4% - § ori Ge 4 ae oi) bei os - or) / = Aree a i , al - 7hll in R ww . - . _ - en \ 7 bys - td Abou, i A¢-vetiy? noe f i us le f 1) teesb i, 1h! Seam 2 (ome, <1! 69 70-79 Salinity 34.-—%o Figure 40.—Frequency of captures and individuals of Bathyteuthis abyssicola at salinity increments in the Antarctic Ocean. Hatched area is number of captures (n= 136); clear area is number of individuals (n=591); crosshatch and dots are breakdown of captures and individuals by .03°/oo. BATHYPELAGIC SQUID BATHYTEUTHIS 137 200 50 RRO EE TTF oo RRR RRR RY LLL LLL LL .00-.24 .25-.49 .50-.64 -65-.69 .70-74 -75-.79 80-82 83-85 .86-—89 Sigma-t 27.—- Ficure 41.—Frequency of captures and individuals of Bathyteuthis abyssicola at density increments in the Antarctic Ocean. Hatched area is number of captures (n= 136); clear area is number of individuals (n=591); crosshatch and dots are breakdown of captures and individuals by .02 units. <4.00 00-14 15-29 30.44 45-.59 60 84 85 99 >5.00 Oxygen ml/L Figure 42.—Frequency of captures and individuals of Bathyteuthis abyssicola at oxygen concentrations in the Antarctic Ocean. Hatched area is number of captures (n=88); clear area is number of individuals (n= 360). og a 9% 9% v8 1Z oP € t I I S[ENPIAIPUT [B04 Jo % = T €ST SSE 6h Col GS 8T &@ G G ST@NPIAIpul Jo Jaquinny o g¢ N g 8" lz 8Z EI eI ez 9°9 VP ¢ eT somnydvo e304 Jo % es T 9€ 8e HT 81 € 6 9 v Z sainydvo Jo Joaquin Ny 4 4 a Gu< [g4'-€2" ZL'-0L'] 69°-29"-99"-F9" 9-09" GS-08" BF-0F—«GE-08"—-6Z'-02" °°/.XX'pE ‘AIUITES 5 69 an P G Ge & 67 LE CE 9T 9g CT syenpra a “Ipur 1830} jo % 4 S G G 6G LIZ 88T 6 €& 6 sTenpia Z -Ipul jo Joquinn = ce < A 8 8 GT GT 9°9 83 OLT IT og soinqdvo [e404 Jo % m I I Z Z 6 ge 0G CT 1 — saanqdvo Jo Joquiny bp 200°G=.0'F 668-09". » .67'S—.0'S* «.66'c-.09'S ~ [.67'o-.0°S © 661-09 1] <6r1T-.0'T 66-09 «6% -.0 y , ‘ainyesaduay, s/DNpiipUr OOF pun saunjdnd gg ai a1ay? Uabhxo sof ‘sppnprarpur [Eg pup saunjdvo OE] asp asay) fizisuap pun ‘fjrurpos ‘ainzosadwaz 40,4] "UDIIQ I1jILDJUP ay? Ur sanjoa uabhzo pun ‘fAjisuap ‘Ajvuros ‘ainzouadwa) snoaunyjnwis 70 epooIsshqe ‘gq fo aUaldNIIQ—TIJTA ATAVL 138 139 BATHYPELAGIC SQUID BATHYTEUTHIS og vl VI ¥G 86 L'6 II FG g S]VNPTAIPUI [BI10} JO % G G G8 ZOT PE OF 88 I S[ENPIAIpU Jo JoquinNyy L’g Lg FG &% GT 1 v1 TT soinjdvo [v}0} jo % G g 1Z 0Z SI II ZI T seinjdvo jo raquinN oog< 66°-S8" [F8"-09° 69°-SF'] PP -0¢" 62°-ST FI'-00° 00°F T/Tut xx'p “UaSAXO 08 jig eI £6 6z 8'8 9°¢ oP 6 S[BNPIAIPUI [B}O} Jo % OT 6L TL1 Gg ELI og 1Z GZ g S[ENPIAIPU Jo JoquNnN 9 61 vo or 8ST ZS 6 og GT soinjdes [vyo} Jo % 8 9G 9T 6 L oI MU z soinjdvo jo raquinyy cs< [¢8'-€8" 28-08" 6L°-SL" PL-OL] 69°-G9° +9'-0¢° 6F-SZ = bZ-00" XX'JZ = }-vUIsIS ‘AzISUACT 10 321-534 O—69 140 U.S. NATIONAL MUSEUM BULLETIN 291 and 4.84 ml/L. In the 4.45-4.59 ml/L group 23% of the tows took 28% of the specimens while in the 4.60-4.84 ml/L group 24% of the tows captured 24% of the individuals. A sudden drop-off occurs at concen- trations higher than 4.85 ml/L; in the two increments 11.4% of the total successful tows took only 2.8% of the specimens. Tows that were taken at these concentrations were either in very shallow or very deep water, above and below the range of abundance of B. abyssicola. The peaks reflect in part, at least, the geographic positions of the captures. The high O, concentrations (above 4.70 ml/L) tend to come from the Atlantic sector; middle values generally come from the eastern Pacific, Drake Passage, and Peru Current region; captures at low concentra- tions come mostly from west of 120°W. Of course, values through the whole range of concentrations may be found in nearly every area, but they may not be at optimum depths or temperatures for B. abyssicola. The species is distributed primarily in low to medium O, concentra- tions (4.00 to 4.85 ml/L) in or below the O, minimum layer in the Antarctic, but O, content is not the primary factor in governing the distribution of B. abyssicola, because, where other factors are favorable (e.g., temperature, depth, salinity), specimens have been taken in very high or very low O, concentrations. This is especially true in the Atlantic sector of the Antarctic where the O, concentration is high at all depths. In other sectors of the Antarctic the very high O, concentra- tions are encountered only at relatively shallow depths; B. abyssicola has been taken occasionally at these depths where it is out of its normal temperature range as well. The capture-plots on the vertical sections and the frequency distribu- tions of oceanographic parameters together delineate the physico- chemical conditions under which Bathyteuthis abyssicola exists in the Antarctic Ocean. Geographic Distribution in Relation to Water Masses An understanding of the three-dimensional distribution of pelagic marine organisms requires the correlation of both horizontal plots of geographic range and vertical plots of relationships of bathymetric range with water masses. Ebeling (1962) has summarized previous works that attempted to delimit animal distributions by the physico- chemical and biological parameters of the environment. Very little has been done with the distributions of cephalopods, particularly the oceanic forms. Bruun (1943) presented the geograph- ical and vertical distribution of Spirula spirula but made no attempt to correlate distribution with parameters of water masses. In a later work based on Galathea material, Brunn (1955) reevaluated his previ- ous decisions about the vertical range of S. spirula; he concluded that BATHYPELAGIC SQUID BATHYTEUTHIS 141 the lower limit of distribution of this species is governed by the 10° isotherm that generally lies at 400 m (more or less) over the geographic range of Spiruda. Bruun’s earlier work had set the depth limit of Spirula at 1750 m. Thore (1949) examined the distributions of the Pana pelagic octopods, particularly of Japetella diaphana. The verti- cal distributions were determined principally by Bruun’s method of statistical analysis of depths of captures; the distribution of J. dia- pnana was defined by parameters of temperature, salinity, and pro- ductivity. Pickford (1946) was the first to study the distribution of a marine animal in relation to the T-S characteristics of water masses; she plotted the distribution of Vampyroteuthis infernalis on T-S dia- grams and determined that density is the common factor governing the distribution of this species. Most workers (e.g., Haffner, 1952; Bieri, 1959) have plotted only the points of capture on T-S diagrams, but Sund (1961, 1964), Ebeling (1962), and Ebeling and Weed (1963) have constructed T-S curves for each capture (over the determined vertical range of each species) and have plotted these in the water mass envelopes. This method more precisely defines the water column and the water mass in which the species were captured. Backus, Mead, Haedrich, and Ebeling (1965), using a different approach, have presented a statistical method of determining faunal boundaries between or within water masses. A large amount of material from a number of locations plus con- current oceanographic and capture data are needed to determine prop- erly the distribution of a species in relation to the physicochemical parameters of its environment. These requirements are easily met for Bathyteuthis abyssicola from the Antarctic Ocean, but for this species from the Atlantic and eastern Pacific and for B. bacidifera the data are less complete. Nevertheless, it is possible to make some fairly defini- tive statements concerning the distribution of Bathyteuthis. Antaretie Ocean Figure 43 is the T-S capture diagram for specimens taken by the Hitanin in Antarctic waters. Capture points rather than capture curves are used because of the uniform nature of the water masses in the Antarctic region. The vertical component of the plot represents Antarctic Circumpolar Water of high salinity and low temperature and Lower Deep Water of the same characteristics and at greater depths. Most of the points plotted below 1°C are from the Atlantic sector of the Antarctic Circumpolar Water Mass while the points above 1°C are from the Pacific sector and the Drake Passage. Five of the deepest captures from the Peru Current, representing Deep Water, are also included in the vertical component of the plot. The area where 142 U.S. NATIONAL MUSEUM BULLETIN 291 ANTARCTIC INTERMEDIATE ° . 2” ATLANTIC UPPER DEEP e . o” PACIFIC UPPER DEEP O ATLANTIC x DRAKE PASSAGE © PACIFIC O CONVERGENCE ZONE xx @ HUMBOLDT Ornt e ANTARCTIC CIRCUMPOLAR ND LOWER DEEP 3 a 5 6 ; 348 Ficure 43.—T-S capture diagram of Bathyteuthis abyssicola in the Antarctic Ocean. See text for details. the plot bends sharply toward lower salinity (below 34.70%.) is classi- fied as Upper Deep Water (Sverdrup et al., 1942), but the points actually represent captures made in the waters underlying the Antarc- tic Convergence zone; the water in this section of the plot is considered to be constituted primarily of Circumpolar Water in transition with the Deep Water that underlies the Subantarctic Upper Water. The three captures from the Peru Current represent Pacific Upper Deep Water, and the single capture at a very high salinity (34.80%) cor- responds to the Atlantic Upper Deep Water. Water between salinities of about 34.60%.—34.40%. comes from below the convergence zone (at the surface) and represents the mixture of water masses that occurs in this transitional region. The plots at the lower salinities and higher temperatures lie in the Antarctic Intermediate Water. The great ma- jority of captures and specimens comes from Circumpolar Water of maximum salinity and relatively high temperatures (1.25°-2°C) and BATHYPELAGIC SQUID BATHYTEUTHIS 143 from Circumpolar-Transitional water at lower salinity (around 34.65%.) and slightly higher temperature (2°-2.35°C). Although there are some areas in the Antarctic waters studied to date where B. abyssicola is more abundant than in other areas, it is reasonable to assert that this species is circumpolar in distribution ; when the remaining portion of the Antarctic Ocean is explored (the eastern half), probably no signficant changes will be required in the distributional pattern as it is now understood. The easternmost record of B. abyssicola in Antarctic waters is Hoyle’s type specimen from 46°16’S 48°27’E between Prince Edward and the Crozet Islands. No further captures are recorded across the Indian or southwestern Pacific-Antarctic Waters until the area of operation of the E/tanin southeast of New Zealand. The southernmost location for B. abyssicola is recorded by Hoyle (1912) at 71°22’S 18°15’W off Coates Land. E/tanin specimens were taken a little beyond 66°S in the Bellingshausen Sea during Cruise 13. Future cruises of Eltanin into higher latitudes should disclose the southern limit of dis- tribution of Bathyteuthis abyssicola. The limit will correspond, in part, to water that shoals above 1500-2000 m. Presumably this species inhabits the bathypelagic zone of the entire circumpolar ocean. Eastern Pacific Ocean The T-S capture diagram (fig. 44) for Bathyteuthis in the eastern Pacific Ocean along South and Central America reveals that the genus occurs in two distinct water masses. Captures of B. abyssicola that were made in the Peru Current between about 33° and 51°S fall along the T-S values of Antarctic Intermediate Water; the points at the greater depths (2000-3000 m) overlap with Pacific Upper Deep Water. The locations of captures of B. bacidifera and B. abyssicola in the eastern tropical Pacific are plotted in figure 45 where they lie in the Pacific Equatorial Water Mass. This relatively warm, saline water mass at- tains its characteristics by in situ advective mixing. In the area of the captures off Central America the thermocline is shallow and a pod of nutrient-rich deep water ascends to the euphotic zone and is responsible for the region’s high productivity. The T-S capture plots for both species are clustered in the 700-1750 m section of the Pacific Equatorial Water Mass envelope where the temperature ranges from 3°-6° C and the salinity ranges from 34.55-34.65%.. Records from the literature include Hoyle’s (1904) in the Bay of Panama and Robson’s (1948) around Cocos and Galapagos Islands. The Dana and Eltanin specimens and Robson’s (1948) records are plotted on the vertical sections of hydrographic parameters (figs. 46— 49). One of Robson’s specimens, a larva, was taken at 550 m where 144 U.S. NATIONAL MUSEUM BULLETIN 291 | eae PACIFIC OCEAN Ficure 44.—T-S capture diagram of Bathyteuthis in the eastern South Pacific Ocean. In Pacific Equatorial Water Bathyteuthis bacidifera is marked by “x” and B. abyssicola by ‘a. B. abyssicola in Intermediate Water is marked by a dot. the temperature is generally between 7° and 8° C; four captures, in- cluding the holotype of B. bacidifera, were made at 700-750 m close to the 6° isotherm. All remaining captures were made in water colder than 5° C, the majority between 4° to 5° C (fig. 46). Vertical sections from the type locality of B. bacidifera are plotted in figure 49. Oxygen concentrations in the eastern tropical Pacific are very low with mini- mum values of less than 0.5 ml/L occurring in the upper layers. The plot of captures against oxygen concentration (fig. 47) shows that the two species occur below the minimum concentrations that range from about 0.5 ml/L to 2.0 ml/L. The effect of these low concentrations is discussed elsewhere. The distribution of the two species of Bathyteu- this in relation to phosphate concentration coincides with the region of PO, maximum (fig. 48). The plots all occur at concentrations greater than 2.75 microgram atoms/liter. Phosphates are quickly diminished in the upper waters where high productivity takes place. BATHYPELAGIC SQUID BATHYTEUTHIS 145 TRANSITIONAL e B. bacidifera B.abyssicola: * previous records x DANA @B. berryi TRANSITIONAL 160° 120 W 80° Ficure 45.—Distribution of Bathyteuthis abyssicola, B. bacidifera, and B. berryi in the tropical eastern Pacific Ocean. Although presently there are no specimens to substantiate it, both species of Bathyteuthis may be expected to range farther westward in the Pacific Equatorial Water Mass borne by the deeper components of the North and South Equatorial Currents. Bathyteuthis is presently unknown throughout the remainder of the vast’ Pacific waters. Whether species will be found to inhabit the various Pacific water masses is difficult to predict ; Bathyteuthis is rare in salinities lower than 34.50%. It may be excluded from the eastern and western North Pacific Central Water Masses and the North Pacific Intermediate Water Mass because these masses have salinities lower than 34.50%. Perhaps Bathyteuthis occurs at least in the cooler, more saline waters of the southern eastern and western South Pacific Central Water Masses. Because Bathyteuthis is so widespread in the Antarctic, Atlantic, and to a lesser extent the Indian and eastern Pacific Oceans, in a variety of water masses, it would seem that a species should occur 146 U.S. NATIONAL MUSEUM BULLETIN 291 500 @) B.bacidifera @ DANA A ARCTURUS \ \ S Ficure 46.—Vertical section, 85°W; temperature °C. Capture points of Bathyteuthis abyssi- cola and B. bacidifera in the eastern Pacific. (Physical data in figures 46-49 from Wooster and Cromwell, 1958.) in the rest of the Pacific also. The diverse water masses in which Bathy- teuthis has been found, however, all have salinities higher than 34.50%. in the temperature range of Bathyteuthis (except in some of the Antarctic Intermediate Water); the salinities of the Pacific water masses (except Pacific Equatorial and Transitional) are all lower than 34.50%. at temperatures below 8° C. Possibly factors other than those of the physicochemical environment combine to limit the distributions of Bathyteuthis. Atlantic Ocean The distribution chart shows that B. abyssicola has been captured at widely scattered points throughout the Atlantic Ocean (fig. 50). The species is common in the eastern South Atlantic where exploratory fishing has been relatively light. The Dana made a number of captures during her northward passage; the Pillsbury caught several speci- mens during two cruises to the Gulf of Guinea; and the Chain has made two captures in the region. On the other hand, exploratory fishing has been very heavy in the western Atlantic, the Gulf of Mexico, and the Caribbean Sea, but B. abyssicola has been collected BATHYPELAGIC SQUID BATHYTEUTHIS 147 ee Ste ere Sores © ee 0 See, os. 2 82 ies @) B.bacidifera abyssicola ® DANA 85° W —O,, mi/L A ARCTURUS Ficure 47.Vertical section, 85°W; oxygen ml/L. Capture points of Bathyteuthis abyssicola and B. bacidifera in the eastern Pacific. only a few times. The IMS collections contain only one specimen from the Gulf of Mexico, which has been very thoroughly fished by the Oregon, and two specimens from the northern Florida Current. The Straits of Florida have been fished extensively by Gerda, but no B. abyssicola have been taken. The Vana, Gerda, and recently, the P2//s- bury and other vessels have worked the Caribbean Sea but this species has not been among the captures. Woods Hole vessels and the Dana and Pillsbury have extensively explored the western North Atlantic and Sargasso Sea, yet only one capture of B. abyssicola is recorded. Some specimens have been taken off northeastern America across the North Atlantic in the region traversed by the Gulf Stream. Only four larvae of B. abyssicola have been reported from the eastern North Atlantic (Massy, 1916; Joubin, 1920, 1924), an area that has been subjected to very extensive exploratory fishing. The rarity of B. abyssicol’ from the Gulf and Caribbean may be related to the relatively shallow sill depths that enclose the Caribbean and Mexican Basins, though more information is needed before defi- nite conclusions can be made. Along the Antilles Are from the Wind- ward Islands to South America the sill depth is less than 1000 m except 148 U.S. NATIONAL MUSEUM BULLETIN 291 Bigs we OO EIS OO Oe eo wig a2ad ig elo ie) oie ine) site's ‘werelieralazauere eee ee een cote are es = oe omen aoe veo 2.75 e e S166 S100 60e, @ B. bacidifera @ DANA A ARCTURUS oO 85 W — PO, ,Ag-at./L Figure 48.—Vertical section, 85°W; phosphate, microgram atoms/L. Capture points of Bathyteuthis abyssicola and B. bacidifera in the eastern Pacific. between Dominica and Martinique where a narrow channel drops to less than 1500 m. The deepest portals into the Caribbean are the Wind- ward Passage between Cuba and Hispaniola and the Anegada and Jungfern Passages between the Virgin and Windward Islands, and the Virgin Islands and St. Croix Island. The sill depth of these pas- sages does not exceed 1600 m. Similarly the sill across the Yucatan Channel into the Mexican Basin is about 1600 m; the portal into the Straits of Florida has a sill depth of 800 m. The water that flows through the Caribbean is a mixture of North and South Atlantic Water Masses. The surface waters are composed of western North Atlantic (Sargasso Sea) water in a ratio of 83-4 to 1 over South At- lantic waters; slightly deeper, the ratio falls to 2 to 1 (Sverdrup et al., 1942). Antarctic Intermediate Water and Upper Deep Water flow in over the sills; the deep water of the Caribbean is renewed from the North Atlantic through the Windward and the Anegada and Jungfern Passages. Below 500-1000 m the temperature and salinity conditions Srekeece eeccccen eee M5 4s Selehe re eee 0% 149 BATHYPELAGIC SQUID BATHYTEUTHIS (2561 {[]aMWOID pUe JaISCOAA WOIJ) DIOYSYO satu oy¥dIpuT stoquinu do} ‘S,T][-o8 SEpNIweT “led W1oyseo oYd UT piafiprong svynaykywg Jo adAqojoy ey jo yutod ainqdeD “J/suoje weigosW ayeydsoyd pur “J/fw ueshxo ‘5, oinqesoduiay ‘Mo78-.08 “SUOIIDES [VOIeA—'6F TUNA Diayipi20q-g yo Aytjo20] adhy y/yowB = ayoydsoud J/jw u2eBAxo ooot oos oe aunjosadway ool soiliw 00z Qea->s SPS WRU ISS RREREES VLSI PSs, SOL UI roe \ ARG Y Lp oe WO; SS. CRW WY Ue Yj GY CY Vi KZ DSSOWN b§ WX Gy, bs ZZ ¢ \ \ ? re Yi 4 eee et SL pegs KYB |) NP. * ve \ \ \ AS | / A ; ty eS te x ify Ly \ 7 ‘i \ ‘ , Wo. ff) / 5 — ts : S ernie PX oss], , if hey Ef? iS fs aN ; Soy PB 3 a. rp rake x \ 2 . ee a fae : 7 ED me ES ; UY, ooz—001 Z| ee TA XD me ; e ate ee YY = | 01 WUD: eT iE 2 pgted m . 5 S Cie Lec : wa OS ee DS - p Eco p = a ‘ a, > . io ; Co uvaA/zW/D'SWO ae SS RN eran & A fo 160 U.S. NATIONAL MUSEUM BULLETIN 291 broad belt of productive water extends across the North Pacific in cold- temperate and boreal regions; parts of the Bering Sea equal the Norwegian Sea in organic production. The Sargasso Sea, which con- stitutes much of the North Atlantic Central Water Mass, has the lowest values of productivity. South Atlantic Central Water, Indian Central Water, and North and South Pacific Central Water are also low in or- ganic production (Nielsen and Jensen, 1957). Vertical mixing and strong upwelling along the western coasts of the American and A frican Continents result in greater production than in western and central regions of the Atlantic and Pacific. The Benguela Current that flows northward along southwest Africa has areas of intense upwelling and high productivity, and in this way it is analogous with the Peru Current off western South America (Hart and Currie, 1960). Recent explorations in the Gulf of Guinea and along the south coast of West Africa by the Institute of Marine Sciences (Miami) and the Bureau of Commercial Fisheries leave little doubt that these waters are con- siderably more productive than has previously been thought. Therefore the shading on the chart in this area underestimates actual productiv- ity. The eastern Pacific equatorial waters are highly productive, and a belt of relatively high production extends westward along the equator and attenuates between 160° and 180°W. A similar, though less pro- nounced, westward extension of productive waters occurs in the eastern tropical Atlantic, although it is bisected by a tongue of low-productiv- ity water borne eastward by the Equatorial Countercurrent (Nielsen and Jensen, 1957). Moderate to high productivity occurs around the near-shore waters of the Indian Ocean and in a narrow band along the divergence between the South Equatorial Current and the Equatorial Countercurrent at 8°-10°S. A positive relationship between productivity and distribution of Bathyteuthis is apparent when the locations of captures (fig. 50) are compared with the chart of organic production (fig. 60). The corre- spondence in the Antarctic Ocean is expected, but nearly every other capture of Bathyteuthis throughout the eastern Pacific, the Atlantic, and the Indian Oceans, as well, comes from a region of highly produc- tive waters. This is especially so along the west coast of South America, in the Gulf of Panama, off southern California, in the North Atlantic, and off West Africa. The captures in the midequatorial Atlantic, off northeastern South America, in the Gulf of Mexico, off the southeastern U.S., and in the Indian Ocean are in areas of moderate productivity. No captures are available from the regions of lowest productivity ; the rarity of Bathy- teuthis in the western Atlantic and Gulf of Mexico and its apparent absence from the Caribbean Sea and Central North Atlantic may well BATHYPELAGIC SQUID BATHYTEUTHIS 161 be attributed to the low to minimum productivity of these regions. This condition would account for the discrepancies between the poor success of the extensive exploratory fishing efforts in the West Atlantic and the high success of the small exploratory fishing effort in the East Atlantic, especially in the tropical region. Moreover, this may be an example of the exclusion of a species from an area that has the proper physicochemical oceanographic conditions (e.g., the Caribbean) but lacks proper biological conditions. The comments in the preceding oceanographic section concerning dis- tribution in the water masses that presently are not known to contain Bathyteuthis are also applicable to the discussion of distribution with productivity. Bathyteuthis may be expected to inhabit the oceanic re- gions that combine the optimal oceanographic conditions and high productivity. In addition, Bathyteuthis may well occur in regions where more influential factors, e.g., productivity and temperature, are optimal and outweigh suboptimal conditions of less important factors. For instance, Bathyteuthis may occur in North Pacific Inter- mediate Water where the salinity generally is below the normal range (34.5%) but temperatures are within range and productivity is high. Areas of low or minimum productivity, e.g., the Central Water masses, also generally do not have oceanographic conditions that are suitable for Bathyteuthis. Conditions within the acceptable range for Bathyteu- this are met in some transitional areas (e.g., bordering the southern Central Water masses of the three oceans) where productivity is mod- erate. Otherwise, it appears that Bathyteuthis does not occur in re- gions of poor organic production regardless of the physicochemical properties. Each water mass has diagnostic biological characteristics (e.g., in organic production and faunal composition) as well as unique tem- perature-salinity characteristics. The T-S relationships, and other physicochemical phenomena such as current systems, vertical mixing, enrichment of surface waters with nutrient salts, and insolation, may characterize or index the resultant biological elements or productivity and food supply that are the primary limiting factors. Ebeling (1962) and Marshall (1963) have emphasized that these biological factors have significance as isolating mechanisms in the deep-sea fauna. Ebeling found that distributions of species of J/elamphaes generally follow the water masses and that when they depart from this it is to follow the contours of productivity. Moreover, four species that occur in Central Water masses of low food supply are dwarf species, while seven species found in productive waters are giants. And within one species, Jelamphaes janae, which has a disjunct distribution, the adults in the eastern Pacific are considerably larger than adults in the 162 U.S. NATIONAL MUSEUM BULLETIN 291 Indian Central where productivity and food supply are lower. Also, the bathypelagic fish faunas of the productive subantarctic, subarctic, transitional, and eastern equatorial regions exhibit a high degree of endemism; these faunas do not overlap into the sterile Central Water masses. In his review of speciation of deep-sea fishes, Marshall (1963, p. 189) concluded that earlier colonizers lived in more productive parts of the ocean: in those parts most resembling the environments of their shallow-water ancestors, which were probably fishes from subtropical and tropical regions. These more favorable parts would be the waters over the upper reaches of the continental slope and those in equatorial oceanic regions. The least productive parts of the tropical ocean, particularly those underlying the great central gyres, would have been colonized last. Concerning the colonization of temperate and polar waters, Marshall (1963, p. 191) stated that the temperature barriers are less imposing than the great contrast between growing and dormant seasons that exist in higher, and particularly polar, latitudes. Bathypelagic species that inhabit open ocean regions in the tropics must be adapted to the uniformity and stability of continuous growing seasons; their meta- bolic and reproductive rhythms would be unsuited to the unstable, cyclic productivity of the high latitudes. But, great upwellings occur along eastern boundaries of warm oceans, and the “. . . dominant physical features of upwelling regions is their irregularity even under normal conditions and this is paralleled with a constantly changing biological picture” (Hart and Currie, 1960, p. 285). Therefore, the productive, though fluctuating, conditions of upwelling regions offer an environment somewhat similar to that of higher latitudes. Species existing in upwelling regions may be preadapted to the fluctuating conditions of temperate and cold waters; such species could be the colonizers of the high-latitude waters (Marshall, 1963, p. 191). A more detailed analysis of the zoogeography of Bathyteuthis in the light of the preceding discussion must await more extensive ma- terial from the Atlantic, Pacific, and Indian Oceans. The informa- tion presently available, however, suggests that the physicochemical and biological environmental conditions enumerated above (partic- ularly organic production), acting as integrated isolating mecha- nisms, have molded the distributional patterns exhibited by the popu- lations of B. abyssicola and B. bacidifera. If we assume that bathy- pelagic cephalopods emerged and dispersed in a manner similar to some other invertebrates and to deep-sea fishes, then the ancestors of the species of Bathyteuthis could have arisen in the warm, produc- tive seas. B. abyssicola has perhaps taken advantage of possible pre- BATHYPELAGIC SQUID BATHYTEUTHIS 163 adaptation to fluctuating conditions and has dispersed to the produc- tive regions of higher latitudes and has differentiated into distinct geographic populations. The more restricted B. bacidifera, on the other hand, apparently does not share this degree of adaptability and is limited to equatorial waters of the eastern Pacific Ocean (and possibly the Indian Ocean). Summary of Distribution With Physicochemical and Biological Factors B. abyssicola inhabits several different water masses and is capable of crossing water-mass boundaries particularly to follow zones of high productivity. Within particular water masses, however (e.g, Antarctic Circumpolar, Pacific Equatorial, and Atlantic Central). L. abyssicola exhibits geographical populations. B. bacidifera on the Ge J/& eee ae ee, 8 Ficure 61.—Summary of T-S-density relationship of Bathyteuthis captures in the Ant- arctic, Atlantic, Pacific, and Indian Oceans. The boxes outline areas of occurrence. Ab- breviations as follows: 1, Indian Ocean A, Antarctic Intermediate (Pacific) n, North Atlantic Ocean c, Circumpolar Water Mass (majority of s, South Atlantic Ocean captures) p, Pacific Equatorial Water c, Circumpolar Water Mass (few captures) 164 U.S. NATIONAL MUSEUM BULLETIN 291 other hand, appears to be restricted to the Equatorial Water Masses of the Pacific and possibly the Indian Ocean; no equatorial water mass occurs in the Atlantic, and despite heavy exploratory fishing efforts in equatorial regions, no specimens of this species have been captured. Wherever Bathyteuthis is found, it occurs within waters of relatively high productivity that have similar oceanographic characteristics. Figure 61 presents the T-S-density relationships of Bathyteuthis cap- tures from Antarctic, Pacific, Atlantic, and Indian Oceans. In the Antarctic the great majority of captures (small box) falls between 1° to 2.5° C, 34.65%. to 34.75%, and sigma-t=27.70 to 27.85. The few re- maining Antarctic captures (large box) range from 0° to 3° C, 34.30%. to 34.75%, and sigma-t=27.20 to 27.90. In the Pacific Equatorial Waters (P) the captures were made between 3° to 6° C, 34.55%. to 34.65%0, and sigma-t= 27.25 to 27.55. In the Atlantic (N,S) most cap- tures came from conditions between 2.5° to 4.5° C, 34.85%. to 34.97 %o, and sigma-t = 27.70 to 27.87. Other scattered captures from shallower South Atlantic Central Water (S) range from 4° to 7.5° C, 34.50%. to 34.75%, and sigma-t = 27.00 to 27.50. The few captures in the Indian Ocean (I) range from 2.5° to 7.5° C, above 34.70% to 34.95%, and sigma-t= 27.30 to 27.80. B. abyssicola occurs over a broad range of oxygen concentrations from about 1.50 ml/L to greater than 5.25 ml/L, while B. bacidifera occurs in lower concentrations from less than 0.50 ml/L to about 2.75 ml/L. In either case, specimens occur in or, most often, below the oxygen minimum layer. Finally, Bathyteuthis is most closely associated with areas where high productivity takes place in the surface layers (fig. 60). Captures come from the layer of maximum or high phosphate concentrations. The high productivity, of course, is responsible for the oxygen mini- mum layer. In conclusion the species of Bathyteuthis may be considered steno- thermic (1°-5° (-7°) C), stenohaline (34.50%.-34.95%.), stenopycnic (sigma-t=27.00-27.20-27.90), and euryaerobic (1.5-5.25 ml/L). (B. bacidifera is oligoaerobic (0.5-2.75 ml/L)). They are true bathype- lagic cephalopods that seldom approach within 1000 m of the bottom (usually more than 2000 m) or within 500 m of the surface. Specimens that occur in less than 750 m are generally juveniles or larvae and are in waters that have suitable temperature and salinity values (because of high latitude locations or areas of upwelling). But, above all, the limiting factor that has the greatest effect upon the distribution of Bathyteuthis is organic productivity. BATHYPELAGIC SQUID BATHYTEUTHIS 165 Vertical and Regional Distribution Calculation of Maximum Depth of Capture and Vertical Range The shallowest depth of the vertical range of a species is easily de- fined by the shallowest tows that took specimens, when an adequate number of samples have been taken. The depths where most of the population lives are indicated by the frequency of captures and num- bers of specimens at particular depth increments. The lower limit of the vertical range of deep-sea, pelagic species is extremely difficult to establish with open nets because of the possibility (or probability) that the specimens in deep tows were captured while the nets were be- ing set or retrieved. Often the lower limit of a species is set near the depth below which a sudden, sharp decline occurs in numbers of cap- tures and specimens. Frequently, however, no sharp break exists and in general this approach is unsatisfactory. Bruun (1943, p. 21) devised a method for determining the vertical range of Spirula spirula taken during the Dana expeditions. First he calculated the number of Sp/ru/a that were caught during a standard- ized tow (S-200 hours) in the depth-layers (in terms of meters of wire out) where Spirula appeared to be abundant. To determine if the specimens that were taken in nets fished at depths greater than the zone of abundance actually came from those depths, Bruun first calculated the number of Sp/ru/a that would be expected to be caught during the time that the deep-fishing nets passed through the layer of abundance; then he compared this figure with the actual catch per standard tow. When the actual catch was smaller than the number that would be expected while the net passed through the zone of abundance, the specimens probably were not taken at the set-depth but at shallower depths while the net was being set or, more likely, hauled. Pickford (1946) and Thore (1949) used the same method for determining the vertical distributions of Vampyroteuthis and pelagic octopods. Since the specimens of B. abyssicola taken on Eltanin have been cap- tured in open 3-meter Isaacs-Kidd Midwater Traw] nets, some of which fished at depths in excess of 3000 m, a means of determining actual depths of capture and true vertical range must be applied. More than one hundred 3-meter IKM'T tows (and numerous 1-meter IK MT tows) were made at depths of less than 500 m during the cruises cur- rently being studied (through Cruise 15); no specimens of B. abyssi- cola have been taken in these tows. This is a good indication that the species does not normally occur shallower than 500 m in the Antarctic. (A survey of captures through Cruise 24 has revealed that a total of only three tows shallower than 500 m have captured B. abyssicola. 166 U.S. NATIONAL MUSEUM BULLETIN 291 20 60 Ne: 100 140 Figure 62.—Vertical distribution by 250-meter increments of total sample population of Antarctic Bathyteuthis abyssicola captured by standardized 2-hour tows. These were taken under very unusual oceanographic conditions, ap- parently during a period of extreme upwelling at the localities—see below.) The major block of specimens comes from 500-2500 m. Only a few individuals have been taken from 2500-3000 m (fig. 62). The follow- ing calculations are designed to indicate if the few specimens from tows that fished below 2500 m actually were captured in these depths. Total number of 2-hour tows taken between 500-2500 m=215 Total number of specimens captured between 500-2500 m=565 Therefore, average number of specimens captured per 2-hour tow between 500— 2500 m=2.63 Total number of 2-hour tows taken between 2500-3000 m=21 Total number of specimens captured between 2500-3000 m=33 Therefore, average number of specimens captured between 2500-3000 m=1.57 BATHYPELAGIC SQUID BATHYTEUTHIS 167 In the overall productive layer (500-2500 m) 21 two-hour tows would have been expected to catch 21 X 2.63 specimens=55 specimens. The actual catch (88 specimens) in the 2500-3000 m layer was only 60% of the expected. Further, to determine if the specimens were captured in the deep layer or in the productive layer while the nets were set and hauled through it, the factor of time must also be considered : the average time required for a tow below 2500 m is 5.3 hours, with two hours spent towing at depth and 3.3 hours spent setting and hauling. Thus, it re- quires 3.3 hours to cover 5000 m (2500 down, 2500 up) giving an aver- age set-haul rate of 1500 m per hour. Since no B. abyssicola are caught in the upper 500 m (1000 m up and down), about 0.7 hours can be sub- tracted from the travel time leaving 3.3—0.7=2.6 hours during which the net is in the productive zone between 500 and 2500 m. (Another approach: the net has to pass through 4000 m of productive zone (2500 —500= 2000; 2000 X2=4000) which requires 4000+ 1500 or 2.6 hours.) If the average catch is 2.63 specimens per two-hour tow in the productive area, then 3.42 specimens would be expected to be caught in 2.6 hours while the net passed down and up througlt the zone of abundance (2.6 X2.63=6.85; 6.85+2=3.42). Since only 1.57 speci- mens were captured per two-hour tow below 2500 m, probably no speci- mens were captured below 2500 m, i.e., B. abyssicola does not normally live below 2500 m. Since the 2250-2500 m zone shows some fall-off in the graph (fig. 62), the same method can be applied to the tows taken below 2250 m in an effort to determine if the lower limit of distribution is shallower than 2500 m: Total number of 2-hour tows between 500-2250 m=196 Total number of specimens captured between 500-2250 m=510 Average number of specimens captured per 2-hour tow between 500-2250 m=~=2.60 Total number of 2-hour tows below 2250 m=40 Total number of 2-hour tows between 2250-2500 m=19 Total number of specimens captured below 2250—86 Total number of specimens captured between 2250-2500 m=53 In the layer below 2250 m forty 2-hour tows averaged 2.15 specimens per tow. In the layer between 2250 and 2500 m nineteen 2-hour tows averaged 2.79 specimens per tow. In the productive layer 40 tows would be expected to catch 104 specimens (40 x 2.60=104). The actual catch was 86 or 82% of the expected catch. In the productive layer 19 tows would be expected to catch 19 2.60=49.4 specimens. The actual catch was 53 or 9% greater. Again the time factor of the nets passing through the productive layer should be taken into consideration. The same times required for 168 U.S. NATIONAL MUSEUM BULLETIN 291 the below-2500 m calculation also apply to the below-2250 m calcu- lation. The nets have to pass through 3500 m of productive zone (2250—500=1750; 1750 2=3500 m). This requires 2.3 hours. There- fore, 2.99 specimens would be expected to be caught while the nets were traversing the zone where B. abyssicola is abundant. This figure is greater than those representing the total catches below 2250 m and also the limited zone of 2250-2500 m, although the latter catch ap- proaches the theoretical figure. Taken as a total unit below 2250 m, 2.15 specimens may be sufficiently fewer than the theoretical 2.99 specimens to indicate that B. abyssicola does not live in numbers below 2250 m. On the other hand, the 2.70 specimens value for the zone 2250— 2500 m may approach too closely the 2.99 value to be significant for establishing the lower limit of vertical distribution of B. abyssicola within this zone. With the sources of error inherent in any system using open nets, this is as far as the calculations should be pursued. A more conservative and perhaps more accurate approach can be taken by assuming the IKMT captures specimens only while being hauled and not while being set. The IKMT probably does not filter a large volume of water while being set because the rate of pay-out of cable may be nearly as great as the speed of the vessel; water may pass through the net at only one or two knots. This would be sufficient speed to capture the lethargic and less mobile forms, while allowing the more active or more perceptive animals to escape. Based on the morphology of B. abyssicola and on observations of living specimens, it is clear that this animal is not an extremely rapid, active swimmer ; neither is it completely lethargic and immobile. Possibly, some speci- mens may be captured even at the low filtering speed during the setting of the net. Therefore, the following calculations represent minimum values. The assumption that the IKMT fishes only while being hauled will give greater significance to the few specimens taken in the deeper tows by reducing by 1%4 the number of specimens which are assumed to be captured in the productive levels during set and retrieval of the trawl. At depths greater than 2500 m the time spent fishing in the shallower productive zone would be 1.3 hours instead of 2.6 hours, dur- ing which time 1.71 specimens would be expected to be captured. Since 1.57 specimens per 2-hour tow were captured below 2500 meters this value probably still does not indicate that B. abyssicola regularly lives below 2500 m. The result is different when considering the zone below 2250 m and especially the block between 2250 and 2500 m. Assuming that the net fishes only during haul-in, the catch would amount to 1.5 specimens. Since all tows below 2250 m averaged 2.15 specimens, it is possible that 0.65 specimens could actually be captured below 2250 meters. In the BATHYPELAGIC SQUID BATHYTEUTHIS 169 restricted zone between 2250 and 2500 m the average was 2.79 speci- mens per 2-hour tow. Therefore this zone may account for 1.29 speci- mens per tow. These different approaches to the determination of the deeper levels of distribution lead to the conclusion that B. abyssicola does not occur in appreciable numbers below 2500 m and that the lower range of occurrence for this species is in the region of 2250-2500 m. A closer estimation than this is not warranted because of the nature of the sam- pling gear and the variability involved in determining depth of cap- ture primarily from wire angle. A further test of the method is to examine the individual tows taken below 2500 m where an apparently large number was captured. For each tow the number of specimens is calculated that would be expected during the time it took to pass through the productive layer. Four tows that fished below 2500 m caught an apparently large number of specimens: . aor vi Ti ” itanth nee BL Aver age No. . N meee sation number vours in specimens expected in ; specimens productive expected productive number s Q 2 captured zone per 2 hrs. zone 1323 4 23 2. 63 Saul 1288 + o10 2. 63 4.3 867 7 320 2. 63 4.6 782 6 5. 0 2. 63 12 The two tows that caught more specimens than would be expected should be compared with other IKMT tows taken in the same area at the same time. In the immediate vicinity of station 1323 four IKMT tows to the productive zone produced 6, 9, 11, and 15 specimens, all well above the average. This is an area of high concentration of B. abyssicola, and it indicates that the estimated catch here is too low and that most (or all) of the four specimens from 1323 were taken above 2500 m. With station 867 the picture is not quite so clear because other tows taken within the immediate area were all made at less than 1300 m where relatively few specimens would be expected. Station 858, taken at the same longitude and 2° farther south, was the closest tow that fished at a depth where B. abyssicola is abundant, and it caught twelve specimens, considerably more than the average. Therefore, it is prob- able in this case, too, that apparently high numbers taken below 2500 m were actually taken in the shallower zone of abundance. Figure 62 gives the total number of specimens captured by standard 2-hour tows with depth. The bulk of the population occurs between 1000-2500 m. (A shght decline in numbers occurs between 1500- 170 U.S. NATIONAL MUSEUM BULLETIN 291 20 40 60 NO. Ficure 63.—Antarctic. A, Average number of specimens captured per tow for total tows (hatched area) and percentage of successful tows (clear area) by 250-meter increments. B, Total number of 3-meter IKMT tows (clear area) and number of successful tows (hatched area) by 250-meter increments. 1750 m.) Figure 63a gives the average number of specimens captured per total tows grouped by 250 m intervals. The average catch, as well as the total catch, drops off between 1500-1750 m then reaches a peak at 1750-2000 m. The percent of successful tows taken at each depth incre- ment is also plotted on the graph. The plot generally follows the plot for captures; that is, the greater the catch and the greater the average per tow, then the greater the number of successful tows (or vice versa). Even the tows below 2500 m were relatively successful, undoubtedly because they passed through the thick layer of abundance. Figure 638 shows the relationship between total number of tows at each depth interval and the number of successful tows. A great number of tows was made between the surface and 1000 m, but only a few below 500 m caught B. abyssicola. Fewer tows were made below 1000 m but they were considerably more successful (also compare with figure 68 for percent of successful tows). BATHYPELAGIC SQUID BATHYTEUTHIS 171 Vertical Distribution by Size The sizes (ML) of 563 specimens of B. abyssicola are plotted against depth of capture in figure 64. The plot shows a tendency for the smaller specimens to be located at shallower depths and the larger specimens to be concentrated at greater depths. Although no sharp break occurs on the overall chart, the division between the shallow and deep concentrations hes at about 30 mm ML. This length also corre- sponds with the size above which all males are fully mature and all females are either well advanced in maturity or are fully ripe. The following discussion determines the degree to which the size-segments of the sample population are concentrated at difference depths. Thirty-six percent (204 specimens) of the total population is com- posed of specimens less than 30 mm in ML; the remaining specimens range from 30 mm to 75 mm in ME. Eighty-three percent (167 speci- mens) of the small specimens was captured between 500 and 2000 m; 111 of these (55% of the small population) came from 500-1300 m. The major concentration, representing 46.3% of the small population (94 specimens), lies between 800 m and 1300 m; 40 specimens (20%) range between 500 m and 1000 m, and 8.4% lie between 500 m and 750 m. Thirty-six specimens (17%) come from nets that fished below 2000 m; the points are scattered and it is probable that the bulk of these came from the shallower tows, perhaps from the zone of con- centration (800-1300 m) or from a possible secondary concentration at 1700-2000 m. If the specimens below about 2000 m were propor- tionally redistributed according to the calculations below they would he within the layer of abundance. The greatest proportion (72.5%) of the individuals in the segment of the population above 30 mm was captured below 1500 m; 519% of the large population is concentrated between 1650 m and 2200 m, with 13% below 2200 m. The remaining 27.5% of the large population is distributed between 625 m and 1500 m in the following proportions: 3.3% between 625-750 m; 9% between 625-1000 m; 18.6% between 1000-1500 m. The graph (fig. 64) shows some specimens plotted below 2500 m; but the calculations made in the preceding section indicate that few B. abyssicola live below 2500 m; therefore, the specimens deeper than 2500 m should be proportionally redistributed in the layer of abundance between 1650 m and 2500 m (see calculations below). Between 1500 and 2200 m are 187 specimens that range from 30 mm to 50 mm in ML; this group represents 52% of the large population and 33% of the total population. Tables IX—XI summarize the distributions of specimens in the sample population by size at the various depth increments. A total of 490 specimens lies between 1000 m and 2500 m; these account for 321-534 O—69——12 NATIONAL MUSEUM BULLETIN 291 U.S. 172 “(3x0 928) SUOI}IPUOD [ENSNuN JopUN SMO} OM} ATUO UT UdYL} 919M SI9JOUI CG/ ULY} sso] Jo syydap ye sUdUTIDads aZI¥] DYT, *(3X9} 99s) DDUepuNqe JO 9U0Z JOMOT[PYS JY UI payNqrIysIpar oq P[NOd sJoJaI QOSZ MOTEq asoyi Jo ysow ‘oinjydvd jo syidep pejdeI1109 Ye JOU pue syidop popsose1 ye poqord o1v sjuIog *(yZuUs] apURUT) azIs Aq vjoIISSKqo Siyynakyjwg IyoIeJUY JO UOTNQIIYsIp [ed1IOA—'Fg AUNTY wu ‘HIONTT JTLNVW ce oe SZ 0% Sl Ol g 09 cS os SY OV ajluaanr o ajowiajo ajpwe 58. cies eodve 000@ ei x@ eheg w'H 0. ©) © i 0. %, 0 0 ; e ° qd ° e OOSL OooL 00S BATHYPELAGIC SQUID BATHYTEUTHIS ies TaBLe IX.—Bathymetric distribution of the total sample population by size group; 563 specimens of B. abyssicola from the Antarctic Ocean Percent of Less than 30mm; Greater than 80 mm; Depth, Total number total popula- number and percent number and percent meters of specimens tion of total population of total population 500-750 29 5.1 17=3% 12=2.1% 750-1000 44 7.8 24=4.2 20=3.6 1000-1300 112 20.0 70=12.4 42=7.6 1000-1500 150 26.6 83= 14.7 67=11.9 1500-2000 169 30.0 43=7.6 126=22.4 2000-2500 Lit 30.4 36=6.4 185= 24.0 563 36% 64% TaBLE X.—Bathymetric distribution of greater than 380mm ML size group; 360 specimens Depth, meters Number of specimens Percent 625-750 12 3.3 625-1000 32 8.9 1000-1500 67 18.6 625-1500 99 27.5 1500-1650 30 8.3 1650-2200 184 51.1 2200-2500 47 13.1 TaBLeE XI.—Bathymetric distribution of less than 30 mm ML size group; 203 specimens Depth, meters Number of specimens Percent 500-750 Wi 8.4 500-1000 41 20 500-1300 alata 55 800-1000 24 11.8 800-1300 94 46.3 1000-1300 70 34.5 1000-1500 83 41 1500-2000 43 21 2000-2500 36 Wet 87% of the total sample population of B. abyssicola. The distribution is fairly even by 500 m increments. Of the total population 60.4% (340 specimens) occurs between 1500 and 2500 m; of this segment 77% (261 specimens) are greater than 30 mm in ME aid only 23% (79 specimens) are smaller than 30 mm MIL. Only 13% of the total popula- tion occurs between 500-1000 m, over half of which are individuals smaller than 80 mm ML. These data confirm that small B. abyssicola tend to live at shallower levels than larger specimens. Furthermore, an examination of all 174 U.S. NATIONAL MUSEUM BULLETIN 291 tows shallower than 750 m shows that 7 of the 8 specimens greater than 30 mm were taken in one tow and the remaining specimen was taken in an adjacent tow. These two unusual captures (1133 and 1137) were made south of 66°S in the area of upwelling along the Antarctic Divergence. Of the 11 specimens below 30 mm ML taken in less than 750 m 10 are larvae and juveniles 21 mm ML or smaller. Nine of these were taken in two tows in the Peru Current; the other two came from tows in the convergence zone. Another tow that captured 6 specimens above 30 mm fished at depths that varied from 823-1300 m; these have been plotted at the shallower depth but could just have easily come from 1300 m; this tow (812) also came from a southern locality, 64° 45’S, where upwelling occurs. One other shallow tow (1383) to 914 m accounts for 10 more of the shallow specimens above 30 mm mantle length; this station was located at 55°S 149°W in the tongue of frigid water that extends northeastward from the Ross Sea, possibly an area of strong vertical movement. Therefore, all except 6 specimens larger than 80 mm that were captured in less than 1000 m came from a few tows that fished in areas of unusual conditions, particularly upwell- ings. Further, the great majority of tows in less than 1000 m caught only small specimens or mostly small specimens with only one or two larger individuals. The calculations in the previous section indicate that B. abyssicola probably does not live below 2500 m. Therefore, the 35 specimens that came from nets that fished at depths greater than 2500 m may have come from the shallower depths, presumably from the zone of abun- dance. A method for the proportional redistribution of the deepest living specimens has been devised. The size group above 30 mm had 23 speci- mens captured deeper than 2500 m. It is assumed that they were dis- tributed in the 2000-2500 m zone in numbers proportional to the plotted distribution ; 112 specimens are plotted in the zone: 65% (73) from 2000-2200 m, and 35% (39) from 2200-2500 m. Therefore, about two-thirds (or 15) of the specimens from below 2500 m can be added to the 2000-2200 m zone, bringing the total number of specimens in that zone to 88 and in the overall zone of abundance (1650-2200 m) to 184,a4% increase. Specimens in the less than 30 mm size group are evenly distributed in the 2000-2500 m range; 12 specimens are plotted below 2500 m. (Actually, specimens below 2000 m are relatively sparse and it 1s possible that that depth is the lower limit of vertical range for small specimens.) It might be tempting to redistribute all the small speci- mens from greater than 2000 m into less than 2000 m, but it is safer for the present to redistribute only those from greater than 2500 m. BATHYPELAGIC SQUID BATHYTEUTHIS 175 Since no real clustering occurs between 2000 and 2500 m, the 12 speci- mens from greater depths can be divided evenly through the zone, giving an additional 2.4 specimens to each 100 m increment. There- fore, the number of specimens in the 2000-2200 m range is increased by 4.8 (or 5) to 15 and in the 1650 to 2200 m overall zone of abundance to 50. The total of both size groups for the 1650-2200 m layer is 234 specimens or 41.6% of the total population; 78.7% of this number are greater than 30 mm ML and 21.5% are less than 30 mm ML. These represent 33% and 8.9% of the total population respectively. Depth of Capture in Relation to Depth of Ocean Several of the earlier specimens of B. abyssicola, including the holotype taken by the Challenger, were captured in bottom trawls. This, and the “peculiar structure” of B. abyssicola, led Hoyle (1885, p. 272; 1886, p. 169) to conclude that this species was a bottom dwell- ing form that collects “nutritive matters from an oozy bottom.” Sub- sequent captures in vertically hauled plankton nets, however, soon left little doubt that Bathyteuthis is a pelagic squid. A scatter diagram (fig. 65) was constructed to determine if any re- lationship exists between the depth of the sea and the depths inhabited by B. abyssicola. The pattern shows in general that with greater ocean 0 (Depths X 1000) ! 2 3 4 5 Distance from Bottom, m Ficure 65.—Relationship between ocean depth and the distance from the bottom of captures of Bathyteuthis abyssicola in the Antarctic Ocean. 176 U.S. NATIONAL MUSEUM BULLETIN 291 depths the captures were made farther from the bottom, which in- dicates that this species inhabits a particular zone in the sea in rela- tion to the surface and not to the bottom. The points along the left and upper edges of the scatter diagram beginning at 500 m above the bottom and 3000 m depth represent the maximum depths of capture for B. abyssicola. The points on the right. and lower edges of the scattergram represent shallowest cap- tures. Most captures were made in water between 3000 and 5000 m deep, and most ranged from 1400 to 3800 m above the bottom. Only one capture occurred where the bottom was less than 2500 m deep: the tow was made 1100 m above the bottom at a bottom depth of 1650 m along the edge of the Peru Current off southern Chile. The bottom descends very steeply in that area, and possibly the capture was made over deeper water a short distance from the point of the sounding or the specimen was carried inshore by currents. Two factors must be considered. Because of the possibility of the midwater trawl fouling, it is generally not set to fish close to the bottom, so the indication that B. abyssicola does not often occur closer than about 1000 m from the bottom may merely be a reflection of cautious trawling techniques. In addition, tows deeper than 3000 m are rare from the Z7tanin because they require such a long time to complete. The chart of geographical distribution of Bathyteuthis (fig. 50) shows that captures are generally made well offshore over deep water. Captures that are plotted relatively close to shore are in areas where the bottom drops off steeply, e.g., along the western coasts of con- tinents (South and Central America, West. Africa, Southern Cali- fornia). If B. abyssicola is found inshore in shallow water it is probably because currents have swept it out of its normal deep habitat. This is undoubtedly the case with the single specimens taken by the Pelican IT and the Silver Bay in the northern Straits of Florida. In regard to the Straits of Florida another possibility exists. Some evi- dence indicates that specimens of normally deep-living species in the Caribbean and Atlantic are forced up into the relatively shallower water over the 800 m sill between Florida and the Bahama Banks. In any case, B. abyssicola normally is not associated with the bottom in any direct way. Aspects of Regional Distribution and Relative Abundance 1. Regional Occurrence Bathyteuthis abyssicola is common throughout the Antarctic, but is not equally abundant in all areas. Furthermore, within particular areas its abundance is dependent upon its proximity to the Antarctic Con- vergence Zone. Table XII gives a breakdown of successful tows and BATHYPELAGIC SQUID BATHYTEUTHIS 177 average catch by areas; Table XIII gives the number and percent of successful tows at various depths by area. Figures 66 through 68 are graphic representations of the data. The plots of distribution on the vertical sections along selected meridians also illustrate the points of the discussion (figs. 21 to 36). The term Antarctic Convergence is used in this section to refer to a general area of the ocean that underlies the Antarctic Convergence Zone and is not used in the restricted sense in reference to the oceanographic phenomenon at a point on the surface. “Total tows” refers to all the tows made at depths greater than 500 m, because 100 3-meter IK MT tows were made at depths of less than 500 m throughout the Antarctic during the period covered by this study, and not a single specimen of B. abyssicola was captured. (Three exceptions have occurred since Cruise 19 and these are discussed elsewhere. ) 2 46° 8 24 6 8 2 4 24 6 8 500 WL WL. LLL WLLL. 7 $3axr7vmo = Wil 2000 \ WN N N : YY NN \ \ NS (ZLLL LA CLALL. LLL Ficures 66.—a, c, Total number of 3-meter IKMT tows (clear area) and number of successful tows (hatched area) at 250-meter depth increments in the Atlantic sector of the Antarctic Ocean (a) and in the Peru Current (c). 3, p, Percentages (clear area) of total number of 3-meter IKMT tows (hatched area) that caught Bathyteuthis abyssicola at 250-meter increments in the Atlantic sector of the Antarctic Ocean (s) and in the Peru Current (p). 178 U.S. NATIONAL MUSEUM BULLETIN 291 24 6 8 Ficure 67.—a, c, Total number of 3-meter IKMT tows (clear area) and number of successful tows (hatched area) at 250-meter increments in the Drake Passage (a) and Drake Passage Convergence (Cc). 3B, p. Percentages (clear area) of total number of 3-meter IKMT tows (hatched area) that captured Bathyteuthis abyssicola at 250-meter increments in the Drake Passage (B) and Drake Passage Convergence (D). In the Atlantic sector of the Antarctic Ocean 12 successful tows were made from a total of 34 tows; the tows, 35.2% successful, averaged 0.82 specimens per total tow. Under the Convergence Zone only three tows were attempted; all were successful and averaged 1.67 specimens per tow (Table XII). Only 1 of 11 tows between 500-1000 m in the Atlantic sector was successful , while 4 of the 8 tows between 1000— 1500 m were successful. The few deeper tows attained 100% success in the 2000-2250 m range, but this was offset by the unproductive shallow tows (Table XIII, fig. 66a, 8). The combined success of Atlantic tows was 40%. In the Drake Passage 16 (55%) of the 29 tows were successful in catching B. abyssicola for an average of 1.65 specimens per total tow. Fifteen of eighteen tows (83.5%) in the Drake Passage Convergence BATHYPELAGIC SQUID BATHYTEUTHIS 179 4 6 50% NN Ss N Mf a \\ / TY ewe Ficure 68.—a, c, Total number of 3-meter IKMT tows (clear area) and number of successful tows (hatched area) at 250-meter increments in the South Pacific (A) and South Pacific Convergence (Cc) regions of the Antarctic Ocean. B, p, Percentages (clear area) of the total number of 3-meter IKMT tows (hatched area) that captured Bathyteuthis abyssicola at 250-meter increments in the South Pacific (B) and South Pacific Convergence (D) regions of the Antarctic Ocean. were successful and averaged 4.85 specimens per total tow, a consider- able increase over the area to the south. In the Drake Passage no cap- tures were made in eight tows to less than 1000 m, whereas four of seven tows at the same depths in the Drake Passage Convergence caught B. abyssicola. No tows were made between 1000-1250 m in the Drake Passage Convergence, but every tow (12) below 1250 m was suc- cessful; success was slightly less complete below 1250 m in the Drake Passage (figs. 674-p). Therefore, in the convergence zone the tows were more successful and caught more specimens. The combined suc- cess of the two Drake Passage areas was 66%. The South Pacific region produced 42 successful catches (65.6% ) from 64 tows. The average catch was 3.28 specimens per total tow. Forty-three tows in the South Pacific Convergence yielded 30 successes or 69.8% of the total for the region, only slightly higher than in the 180 U.S. NATIONAL MUSEUM BULLETIN 291 TaBLe XII.—Number of successful 3m IK MT tows and average catch of B. abyssi- cola in different regions of the Antarctic Ocean Total speci- Average Number Number mens number/ Average suc- -unsuc- Percent from success- number/ cessful cessful Total suc- 2-hr. ful total Area tows tows tows cessful tows tow tows S Atlantic (Scotia Sea) 12 22 34 35.2 28 2.34 0.82 Atlantic Convergence 3 = 3 100.0 5 1.67 1.67 Drake Passage 16 13 29 55.0 48 3.0 1.65 Drake Passage Con- vergence 15 3 18 83.5 87 5.8 4.85 South Pacific 42 22 64 65.6 203 4.78 3.28 South Pacific Con- vergence 30 13 43 69.8 151 5.04 3.52 Humboldt (Peru 12 i 13 92 57 4.75 4.38 Current) N of Atlantic Conv. 1 6 a 14.3 2 2.00 0.29 N of Pacific Conv. 3 5 8 60 17 2.66 2.12 TaBLeE XIII.—Percentage of successful 3 m IKMT tows at 250 m increments in different regions of the Antarctic Ocean *Depth, DP DPC SP SPC SA H NCA| NCP meters mee) seme Ul a J a) ae ee) Lh = RG 500— 750 OF B TE Bara Fi Oe eae Se Oe ae Oman 0 50 11 42 0 100 0 0 751-1000 OF SSDS By Se sel om GaGa 8 On OF nr GAO xee0 0 60 27 73 17 3 = 50 1001-1250 2 OO) HOE SET DADA cD Queen Oia eee DO Raw 100 = 50 78 50 100 = = 1251-1500 OOO 4S = OS 9 HOF PD 2s ON ON 2 Oar ao = 100 100 84 50 = 67 a 1501-1750 2. 0) 35,016 2) 3-013) «2.100. 2000) tae co 100 100 75 100 60 a 0 a 1751-2000 5 PO ORO) 2 ah eS? OP ee AZ RON One ON (Omeee. 100 100 90 100 20 100 = = 2001-2250 4° DNS OLS. A OU SOS Ol wale OS Oi ees 80 100 84 100 100 0 = 50 2251-2500 3 Be Oe 2-0 OKO OSs" 0% ks SOF Ors re 75 100 78 == = 100 = 0 2501-2750 O 2OricO Ww? 62.)206 20) ed, (Ou MOsOr pOm tetas 0 = 50 a 100 100 = 50 2750-3000 2 210i 0-2 = 0:0" OO 2a OO! Oso neo 100 = 1004 = 0 100 a = *The figures in the upper row of each increment represent the number of successful (s) and unsuccessful (u) tows; the lower figure is the percent of success. DP = Drake Passage; DPC = Drake Passage Convergence; SP=South Pacific; SPC =South Pacific Convergence; SA=South Atlantic; H=Humboldt Current; NCA= North of Atlantic Convergence; NCP=North of Pacific Convergence. BATHYPELAGIC SQUID BATHYTEUTHIS 181 region to the south. The average catch, 3.52 specimens per total tow, was also only slightly higher. The major differences are seen in the distribu- tion of successful catches with depth; in the South Pacific tows between 500-1250 m were less successful than tows to the same depths in the South Pacific Convergence (fig. 68a—p). Below 1250 m the percentage of success was generally quite high. So each region in the South Pacific had nearly equal success in averaging about the same catch, although the values for the convergence region were slightly higher. The two areas had 68% success of the total tows. Thirteen tows were made in the Humboldt (Peru) Current, and twelve of them (92%) succeeded in capturing B. abyssicola, The 12 tows produced 57 specimens for an average of 4.38 specimens per total tow. All depths to which tows were made yielded 100% success except the 2000-2250 m zone where the sole unsuccessful tow was made (fig. 66c, p). The number of tows in the region of the Peru Current is not large, but the degree of success attained and the number of specimens captured excels the performance of tows in the regions of the Southern Ocean. In summary, the Drake Passage and the South Pacific regions had about the same overall success of captures (66-67%) but the distribu- tion of successful catches and number of specimens between regions was disproportionate: 83.5% successful with 4.85 specimens per total tow for the Drake Passage Convergence versus 55% and only 1.65 specimens per total tow for the Drake Passage. The South Pacific and South Pacific Convergence regions both yielded many successful tows (65.6-69.8% ) that averaged 3.28 to 3.52 specimens per total tow. The Atlantic sector was not fruitful in comparison; only 40% of the tows were successful. All tows (8) in the convergence zone were successful, but they averaged only 1.67 specimens per tow. A meager 35.2% of the tows in the Scotia Sea region produced a scant 0.82 specimens per total tow. Therefore, up to a 6-fold difference in abundance of B. abyssicola exists between different areas of the Antarctic. The Drake Passage Convergence and the Peru Current are the areas of peak abundance followed by the nearly equal South Pacific regions. The Drake Passage and the Atlantic Convergence regions yield about the same low aver- age catch. (The data for the Atlantic Convergence are so few that the picture may change if more material becomes available.) The Scotia Sea, in relation to other areas, is nearly a desert for B. abyssicola. 2. Influence of a Few Exceptional Tows During the Z/tanin cruises under discussion (3-6; 8-10; 13-15; see Appendix) a total of 236 3-meter IKMT tows was made below 500 m in the depth range of B. abyssicola, of these, 1384 (57%) were 182 U.S. NATIONAL MUSEUM BULLETIN 291 successful in capturing this species (100 tows shallower than 500 m were entirely unsuccessful). The captures averaged 2.5 specimens per total 2-hour tow, and they ranged from 1 to 24 specimens per tow. Several tows caught large numbers of B. abyssicola,; these have a pro- nounced effect on the overall catch figures. Sixteen 2-hour tows that caught 9 or more specimens captured a total of 210 specimens. This means that 11.8% of the 134 successful 2-hour tows accounted for 35.2% of the total sample. All but two of these stations were located in regions of the Antarctic Convergence. The specimens are grouped at the peaks of occurrence on the vertical distribution chart, and, in fact, they are responsible for these peaks. Only one capture was in less than 1000 m (914 m) and only two were deeper than 2100 m (2150 m, 2269 m). Table XIV compares the total number of specimens captured at depth intervals with the number taken in exceptionally successful tows and indicates the influence of these tows on the distribution of the total sample population. The major effect of the exceptional tows is apparent when they are deleted from the curve of the bathymetric distribution. This results in the smoothing out of the curve so that the peaks of abundance shift from sharp, bimodal peaks at 1000-1500 m and 1750-2250 m to a single major peak at the 1750-2000 m range with very slight secondary peaks above and below this depth incre- ment (fig. 69). The large number of specimens from relatively few tows does not mean that the concept of the vertical distribution of the total population has to be altered; instead it indicates that B. abyssicola has a patchy distribution, both horizontally and vertically with peaks of abundance located under the region of the Antarctic Convergence. TaBLE XIV.—The influence of exceptionally successful tows on the total sample population A B C D Total Number number specimens from specimens/2- tows with >9 Percent of Depth, meters hour tow specimens total A minus B 500-750 30 0 0 30 750-1000 51 13 25.0 38 1000-1250 97 68 MOE, 29 1250-1500 75 PH 36.0 48 1500-1750 58 0 0 58 1750-2000 129 58 44.9 al 2000-2250 (2 35 48.6 37 2250-2500 53 9 17.0 44 2500-2750 19 0 0 19 2750-3000 14 0 0 14 BATHYPELAGIC SQUID BATHYTEUTHIS 183 Ficure 69.—Influence on curve of vertical distribution of 16 tows that caught 9 or more specimens. A, curve for total sample population; B, curve exclusive of the 16 exceptional tows. Antarctic. 3. Seasonal Abundance Although seasonal data are incomplete, there is some indication of seasonal fluctuation in the abundance of B. abyssicola. Of the 16 tows that captured more than 9 specimens (totaling 210 speci- mens), 12 were taken during October, November, and December (5, 5, and 2 tows respectively), 3 were taken in August and Septem- ber, and 1 was taken in July (well north of the convergence) (Table XV). No such large captures were taken between April and July. Furthermore, a breakdown of successful to unsuccessful tows by months indicates that more tows were successful during the austral late spring and early summer months. During October, November, and December there were 74 successful tows and 40 unsuccessful tows, a success rate of 65% or nearly 2 to 1. During August and September there were 34 successful tows and 22 unsuccessful tows, for 60% suc- cess, or 1.5 to 1. In April through July there were 20 successful and 27 unsuccessful tows for a success rate of 42% or 0.74 to 1. These winter values, however, may be more a factor of locality than of season, be- cause one cruise is primarily responsible for the reverse of the suc- cessful to unsuccessful ratio, Cruise 8 took place during April and May 184. U.S. NATIONAL MUSEUM BULLETIN 291 TaBLE XV.—Seasonal and regional occurrence of the 16 tows that caught 9 or more specimens per 2-hour tow Elt. Sta. Number Depth, In Convergence Number Month specimens meters region 99 VII 10 1210 No 1201 VIII 20 1120 Yes 1262 Vill 14 1230 Yes 1269 IX 24 1248 Yes 248 x 12 1370 Yes 1287 x 9 2269 No 1324 xX 15 1958 Yes 1327 xX 9 2060 Yes 1328 xX 11 1775 Yes 846 XI 12 1866 Yes 847 XI 9 1991 Yes 858 XI 12 2099 Yes 1364 XI 11 1848 Yes 1383 XI 13 914 Yes 354 XII 14 2150 Yes 382 XII 15 1280 Yes in the Scotia Sea region with the poor success to unsuccessful ratio of 6 to 20 or only 23% successful. This region has been shown to be sparsely populated with B. abyssicola, so the low success may be related to area. On the other hand, Cruise 9 covered the same general area in August and September and it scored an 8:7 success ratio (538%), still fairly low in comparison with other areas, but it may indicate an increase in abundance during the austral spring. Unfortunately, no further data are available for this region. The remaining winter tows were made in the Drake Passage and the South Pacific and these amounted to 14 successful tows and 7 unsuccessful tows for a 66% success rate, equally as successful as spring and summer tows. It appears then that B. abyssicola can be caught with about the same frequency of success in most regions of the Antarctic during most of the year. But the data suggest that more individuals are present during the austral spring and summer or at least that they tend to congregate in patches in the region of the convergence. This suggests a correlation with the spring-summer peak of high organic productivity in the surface waters of the Antarctic Ocean. No evidence exists to suggest that there are significant differences in the size of B. abyssicola over the seasons. Specimens of all sizes and all stages of maturity have been taken throughout the year; often specimens that range from larvae through juvenile and maturing stages, to ripe and spent individuals occur in the same trawl-haul. Few tows that catch more than 2 or 3 specimens have all individuals of the same size or stage of maturity, unless they are shallow tows which tend to have predominantly larvae and juveniles. The 16 tows that BATHYPELAGIC SQUID BATHYTEUTHIS 185 caught more than 9 specimens per 2-hour tow had the following composition in order from greatest to least abundance: five tows had ripe, maturing, juvenile, and spent stages that ranged in size from 14-56 mm ML; three tows had ripe, maturing, and spent individuals 25-56 mm ML; five tows had nearly equal numbers of juvenile, matur- ing, and ripe specimens 14-48 mm ML; two tows had ripe and matur- ing specimens 26-54 mm ML of which the ripe were 214 times more numerous; one tow had even numbers of juvenile and larval specimens that ranged from 8-21 mm ML. 4. Size-Frequency Distribution The composition of the sample population from the Antarctic by size is represented in the frequency distribution histograms (figs. 70 and 71). The standard for size, mantle length, has been plotted in 3 mm increments for 598 individuals for the total sample population (fig. 70) and 287 males and 284 females for the plot by sex (fig. 71). The plots for the populations are moderately skewed. Few small lar- Uy 3 Yj; 8 60 NO. 40 20 3 1 15 19 23927 35 39 59 63 67 71 5 ML,mm 43 47 51 ae Ficure 70.—Size-frequency distribution of total sample population of Bathyteuthis abys- sicola from the Antarctic Ocean; n=598. FEMALE n= 284 KAY MALE n= 287 Ficure 71.—Size-frequency distribution of males (n=287) and females (n=284) of Bathyteuthis abyssicola from the Antarctic Ocean. 186 U.S. NATIONAL MUSEUM BULLETIN 291 vae are present in the 4-11 mm size range, but the plot for larger larvae and juveniles (12-23 mm) constitutes a secondary peak in the curve, with a large number of individuals in the 12-15 mm category. The slight depression between 24-31 mm represents individuals in the maturing stage. Most males, particularly by 28-31 mm, are ripe, but females are only in the beginning of the maturing stage. The major peak in the sample population is composed of individuals from 32-47 mm in ML; in this range all the males are ripe and the females are maturing and ripe. At about 48 mm the population, composed of ripe males and ripe and spent females, decreases rapidly in numbers. There- fore, the portion of the sample population that occurs with the great- est frequency has individuals that are near or in the spawning conditions. Although the sex ratio of males to females in the overall population is 1:1, some marked differences occur within certain size groups. In general, there are a few more males than females in the juvenile and maturing stages up to the 28-31 mm category. There are several more raales than females in the 32-39 mm group where all the males are ripe and the females are maturing or ripe. A large difference occurs in the 40-43 mm increment where there are 47 males and 26 females; 1.e., only 36% of the specimens in this category are females. A sudden reversal of the predominance of males occurs above the 40-43 mm in- crement. Only a few more females than males exist in the 44-47 mm class, but above that the males drop off sharply in numbers so that in the 48-51 mm and 52-55 mm categories males represent only 27% and 10% of the specimens. The 56-59 mm and 60-63 mm size groups con- tain only females; no specimens occur between 64 and 75 mm, but the largest B. abyssicola recorded is a male of 75 mm mantle length. Such a preponderance of females over males in the larger sizes is curious. Apparently, males do not generally grow as large as females, although the single “giant” specimen at least partially contradicts this conclusion. Males mature early and some may have fully devel- oped spermatophores and packed Needham’s sacs at about 25 mm mantle length. All males above 28-30 mm are fully ripe. On the other hand, females generally are not ripe until about 40 mm and often not until 45-48 mm. Size differences between the sexes of other adult squid have been noted, primarily in loliginids (review in Roper, 1965). Adam (1952) noted that the males of Jlex illecebrosus coindetii were smaller than the females, and this is borne out in IMS specimens from the Gulf of Guinea. We have also observed, though not tabulated, that males of Ommastrephes pteropus tend to be smaller than females; none of the numerous large specimens are males. Garcia-Tello (pers. comm.) of the Estacion Biologia Marina in Monte Mar, Chile, informs me that BATHYPELAGIC SQUID BATHYTEUTHIS 187 the same phenomenon exists in the Humboldt squid, Dosidicus gigas. Finally, our observations on Architeuthis, as sparse as they are, indi- cate that the males mature at a much smaller size than the females, and apparently they do not attain the larger size of the females. All of these species are oceanic forms that live mostly in the upper water layers. Until now information on size differences between sexes in deep-sea forms has been entirely lacking. A priori it would seem that once the males of the forms mentioned have fully developed spermatophores packed in Needham’s sac, no biological necessity would exist for attaining a larger size, so long as small size did not render the males incompatible with the larger females. Another presently unempirical possibility is that large males of B. abyssicola are able to escape the net, while females, all fully ripe at the larger size, are unable to avoid capture. B. abyssicola, however, is gen- erally a poor swimmer; it is not equipped for rapid motion, and it is difficult to imagine that even the “sleek” males could be significantly more agile than the females to avoid being captured. 5. Comparison of Antarctic and Gulf of Guinea Tows A rough estimate of the relative abundance of Bathyteuthis in differ- ent parts of its range can be made by comparing trawl-catches between two different regions. Midwater trawl tows were made during the two cruises of the R/V Pillsbury to the Gulf of Guinea; this material gives the only available information that can be used to compare with the Eltanin tows. Unfortunately, no vessel has conducted a survey that adequately compares with the program on the Z/tanin. Therefore, the data from the Pillsbury are offered only as a preliminary indica- tion of relative abundance. Table XVI is a summary of tows and catches made by the Pillsbury in the Gulf of Guinea and the //tanin in the Antarctic. In brief, the Pillsbury tows that were set at depths (300 to 3000 m) where B. abyss?- cola might be caught were 17% successful; in the Antarctic 52% of the tows in the same depth range was successful. This figure includes 28 tows in less than 500 m where B. abyssicola normally does not occur in Antarctic waters. The captures in the Gulf of Guinea averaged 0.21 specimens per total tows, while those in the Antarctic averaged 2.3 specimens per total tows. (The difference is even greater if the 300-500 m column is omitted.) Again, the Pi/sbury data are meager in com- parison with the Z/7tanin data, but they reflect relative numbers of L. abyssicola in Antarctic and Guinean waters. Jtanin tows were three times more successful than Pillsbury tows and they captured 10 times the number of specimens. This may be a reasonable estimate be- cause, although no data from other vessels are available, a great amount 321-534 O—69——13 188 U.S. NATIONAL MUSEUM BULLETIN 291 TaBLE XVI.—Summary of 3-meter IKMT tows made by R/V Pillsbury and USNS Eltanin Depth 300- 500- 1000- 1600- 2000- 2500- Range 500 1000 1500 2000 2500 3000 Total Pillsbury: Total tows 11 10 4 1 1 2 29 No. success- ful 2 0 1 1 0 1 5 % successful 18 0 25 100 0 50 17 No. specimens 2 0 1 1 0 2 6 Average/tow 0.18 0 0.25 1 0 1 0.21 Eltanin: Total tows 28 90 44 47 36 18 263 *235 No. success- ful 0 25 32 39 28 12 136 136 % successful 0O 28 73 80 78 66 52 58 No. specimens 0 81 172 187 125 33 598 598 Average/tow 0 0.90 3.9 4.0 320 1.8 23 2.6 *Data exclusive of the 300-500 m range where B. abyssicola is not represented. of exploratory fishing with midwater trawls has been conducted in the Atlantic Ocean, and the records of Bathyteuthis are disproportionately small. The IMS Museum houses the collections of Gerda, Pillsbury, Oregon, Dana, etc., representing thousands of tows, and only a few specimens of B. abyssicola are among them. Furthermore, I have searched the collections of pelagic material at Woods Hole Oceano- graphic Institute, at the Museum of Comparative Zoology, and at the U.S. National Museum and have found only one or two specimens. The conclusion is that B. abyssicola is not abundant in the Atlantic; it is found mostly in regions of relatively high productivity (e.g., eastern tropical waters). Comparisons With Other Species; Abundance and Distribution Bathyteuthis abyssicola is the most abundant cephalopod that has been taken during the Z'ltanin cruises. Nearly 600 specimens are repre- sented in the material studied through Cruise 15. The species is so common in midwater tows that personnel returning from cruises show little enthusiasm about having caught so many specimens and about having been able to observe these bathypelagic animals alive in aquaria. This discussion considers the three species that follow B. abys- sicola in abundance. 1. Crystalloteuthis glacialis Chun, 1906 Crystalloteuthis glacialis has been reported only three times in the literature: originally by Chun (1906), in Chun’s Valdivia monograph BATHYPELAGIC SQUID BATHYTEUTHIS 189 (1910), and finally by Dell (1959). Dell gives the distribution of the species in the Antarctic as between 40°F and 143°E in 750-1710 m. The geographic range is now extended in Antarctic water throughout the range covered by the Z/tanin from about 25°W to 160°W, and it un- doubtedly extends throughout the entire Circumpolar Water Mass. Through Cruise 15 a total of 349 specimens of C. glacialis was taken, 231 by 3-meter IKMT and the remainder by other gear, principally the 1-meter IKMT. It is second to B. abyssicola in abundance in Ant- arctic waters. Table XVIII gives the breakdown of captures by 250 m depth increments. A total of 598 specimens of B. abyssicola was cap- tured during the same period, mostly by 3-meter IKMTs (Table XVII). Therefore, in total numbers, B. abyssicola is about 1.7 times more common than C’. glacialis, and in midwater trawl-hauls it is about 2.5 times more abundant. TaBLE XVII.—The composition of the sample population of Bathyteuthis abyssi- cola Hoyle, 1885, by 250 m increments; success of trawl hauls and average catches Average Average number number Number spect- spect- Depth speci- Total Number Number mens/ Percent mens/ captured, mens/ number success- wnsuc- success- success- total meters 2-hr. tow tows ful cessful ful tow ful tows 0-250 - 72 - 72 - - — 251-500 - 28 = 28 - _ - 501-750 30 45 10 35 3.0 22.2 0.6 751-1000 51 45 15 30 3.4 33.3 i 1001-1250 97 20 14 6 6.9 70.0 4.7 1251-1500 75 22 18 4 4.2 82.0 3.4 1501-1750 58 23 18 5 3.2 74.8 2.5 1751-2000 129 24 21 3 6.1 87.5 5.4 2001-2250 72 17 13 4 5.5 71.5 4.2 2251-2500 53 19 14 4 3.7 78.9 2.7 2501-2750 19 11 6 5 3.2 54.5 Lee 2751-3000 14 7 6 2 2.4 71.4 Li 3001-3250 ~ ~ - - - = = 3251-3500 - 3 - 3 = = = 598 336 134 Vertical distribution is plotted on the graph (fig. 72) and presented in Table XVIII. The bulk of the sample population is concentrated between 500 and 1000 m where about 57% of the specimens occur. Nearly 69% live between 250 m and 1250 m. Below 1000 m a sharp de- crease in number exists and about an equal number of specimens oc- curs in the remaining depth increments, making it difficult to estab- lish the lower depth limit. It could be as deep as 2000-2500 m, but probably the limit is much shallower, perhaps around 1000-1500 m 190 U.S. NATIONAL MUSEUM BULLETIN 291 Ficure 72.—Vertical distribution by 250-meter increments of the total sample population of Crystalloteuthis glacialis Chun, 1906. Antarctic. just below the sharp decrease in catch. The average number of speci- mens per tow for the total tows ranges from a minimum of 0.14 at the shallowest and deepest depth intervals to maximums of 1.58 and 1.78 in the 500-1000 m increments. Although considerable overlap exists in the vertical ranges of C. glacialis and B. abyssicola, the zones of maximum abundance are dis- tinct, the former species between 500-1000 m and the latter between 1000-2250 m. B. abyssicola, however, is associated with C. glacialis more than with any other pelagic Antarctic cephalopod. The two spe- cies frequently are caught in the same tow, but these generally are tows that have fished deeper than 1000-1500 m into the zone of abundance of B. abyssicola. Perhaps the most distinctive feature of the distribution of C. glaci- alas is that it is strictly limited in its northward extent by the location of the Antarctic Convergence. This is truly an Antarctic squid that is found extremely rarely north of the convergence, and then only as a straggler. It is found in the convergence zone but especially south of the convergence in the Circumpolar Water Mass. The details of the distribution must await a more thorough analysis, but at least the boundaries are clear. In fact, C. glacialis can be used as an indicator species; aboard H'ltanin the location of the Antarctic Convergence can 191 SQUID BATHYTEUTHIS Y PELAGIC BAT v0 oT O'S Te sg Vs ce el 8°0€ ELS 8g sé uorpjndod ANOY-G 1070} fo yuao1agq FTO cr'0 89°0 L¥'0 90 T19°0 Iv 0 C60 8L1'T 8oT $G'0 FTO $0} 7070} /suamads Laquunu aboiaapy evr V9E OCF SIT OLE L1@ LCS ose SLE G GY L9€ ov jnfssaoons quaalaq OT ol oT OF LT 8% ST LZ ag Le ot oe no} infssavons /suawmads LaQuenu abosaapy 9FZ 06 9€€ TES 092 9 T Me T T Le ¥ TI F ¢ Il 8 61 cl €T ST G ZY 8 8 GI 6 FG ol GT 8T G ZS LI FI ZY G GG 6 6 Si Li 0G FI 61 8% LI cP Le O08 9% 61 cP 19 TZ ST OT 8% ST CT 69 is GL 9 OT pnfssao jnfssavons smo} suawmiads Mo} LnoYy-g -onsun LaQUn Ay LIQUNU fo saqunu —_/suawmasads JOQUN AT 1210, jonpy LIQUNAT sayojpo abpsaap pun synoy jmp.) fo ssa0ons ‘sjuamasour w Ogs fq ‘9061 ‘unyg sIfeIoe{s styynoyzoy[eysAig fo uorvjndod ajdws ay) fo worrsodwos ayf[— TIITAX A@TaVL O00E-TS2z OGLZ-T0S% 00SZ-1SZ% 0SZZ-1002 O000Z-TELT OSLT-10ST OOST-TSal OSZI-T00T OOOT-TSZ 0SLZ-T0S 00S-TS% 0SZ-0 $19]9UL ‘painjd pa yidaq 291 NATIONAL MUSEUM BULLETIN U.S. 192 s'0 GZ Og 0'¢ Oe i OLT OTT OTT OTF uoupnjndod ANOY-g 1D}0} fo yuarsag c0'0 62'0 or 0 €F'0 42'0 ce0 921°0 L¥0 620 FIT $10} 1070} /suawmoads daqunu aboiaapy og g8 8°06 0°92 LZ 0'SZ ees L9G OSE 9°€% ynfssaoons Moz ynf{ssavons 1UdId19q OT OT CS 07% ST ot rT Ss ST aa SP /suawmads daqunu abniaay 82 9 iat 8T oI 61 LT LI cT 0€ T €€ ol 61 6 g¢ Lt re wid ud Od NS jnfssasonsun ynfssaoons smo} saqunu JaQuUunny LaQuUunny sayojv9 abpsaan pun L UT 61 LI FG €% GG 0G cP cP 8% ol 1970.1 98T N T ~oonwte 62 8T 1% 08 suawmeads fo daqunu /suawmads jonpoy 661 TZ GS o8 No} Lnoy-g JaQUunn 000E-192z OSLZ-10SZ 00SZ-TSZz 0S22—-1002 000G-TSLT OSLT-I0ST OOST-ISZT OSZI-TOOT OOOT-TS2 O¢Z-10S 00S-1SZ 0SZ-0 $ajaUe ‘paunjdvo yidaq SjnDYy ]mD4} fo ssavons ‘syuawmasour w Ogg fg ‘OIEI ‘unyD eqoid styynoyoryovig fo uownjndod ajdwos ay) fo uorjsodwoos ayJ,—X]X AIAV], BATHYPELAGIC SQUID BATHYTEUTHIS 193 be determined by the presence or absence of this species in the IKMT tows. While both the vertical and horizontal ranges of these two species overlap to a certain extent, C. glacialis, truly an Antarctic species, is more limited and apparently is unable to survive in the conditions that exist in the subantarctic regions. B. abyssicola, on the other hand, may occur together with C’. glacialis, but it possesses far greater flexibity in its tolerance to different oceanographic conditions and is much more broadly distributed. 2. Brachioteuthis picta Chun, 1910 Next in abundance is Brachioteuthis picta Chun, 1910, represented by 190 specimens (Table XIX, fig. 73). Specimens of Bathyteuthis abyssicola taken during the same period were 3.1 times more abundant than Brachioteuthis picta. The vertical ranges of the two species over- lap slightly between 500-1000 m. In Brachioteuthis picta 41% of the population inhabits the upper 250 m and 28% lives between 500- 1000 m; 80% lives between the surface and 1000 m. A rapid decrease in numbers occurs below 1000 m, and the remaining 20% of the sample population is spread evenly between 1000 and 2250 m. Undoubtedly Ficure 73.—Vertical distribution by 250-meter increments of the total sample population of Brachioteuthis picta Chun, 1910. Antarctic. NATIONAL MUSEUM BULLETIN 291 U.S. 194 £% ev £%G PT oP L'0 VIZ Lg 6% oss uounjndod LNOY-B 1270} fo juI0lagq €e°0 ceo ol'0 60°0 0¢€ 0 S00 99°0 sT'O FIO TT SMO} 7D} 0} /suaunsads Laqunu abouaapy ost Lit es eV 16 O'¢ 19 68 Lh 69 ynfssaoons quaddad O€ GT 0% Oe OT 0'OT 0% 0% PCT 0} ynfssavons /suawmads Laqunu abnor L 6 61 GT GG GG 0G 61 A Iv 96 L9 ynfssavonsun pnfssaovons LOQUN AT FG i Z j Tl = 61 dt FG £6 GG 02 cP cP 8% GL momNHMANANN sayojno abpiaan pun synvy 7Nv47 fo ssavons ‘sjuamasour UW OGE fig ‘LEg] ‘buaquueT snooreqyue snyeuor) fo wornjndod ajdwos ay) fo uorjtsodwoo ay, —XX AAV, smo} JaQuUnn soe 1210L, iat N € +HWDOoR OTN SO 18 suawmiads fo saqunu />NpPp OFT | € SH OOARONMS LL no} LNoy-% /suaunoads LOQUN A OO0E-TS2z OSLZ—-T0SZ 00SZ-1S2za 0SZZ—-T00G 0002-TS2LT OSLI-IOST OOST-IGZI OSZI-LOOT OOOT-IT¢L OSLZ-L0¢ 00S-T¢S3 0¢Z-0 $4aqaUl ‘painyd po yidaq BATHYPELAGIC SQUID BATHYTEUTHIS 195 many of these were captured while the nets were passing through the shallow layer of abundance, and the lower limit for Brachioteuthis picta is probably much shallower than 2250 m, perhaps around 1000 m. 3. Gonatus antarcticus Lonnberg, 1897 The only other pelagic cephalopod that has been caught in the South- ern Ocean in any quantity is Gonatus antarcticus Lonnberg, 1897. A total of 141 specimens was captured during the same period that 598 specimens of B. abyssicola were taken. Therefore, the relative abun- dance for specimens captured by all means is 4.2 specimens of B. abyssi- cola to each Gonatus antarcticus. G. antarcticus is primarily a shal- low living form (Table XX, fig. 74) ; 55% of the sample population was taken in the upper 250 m; 30% was distributed between 250- 1000 m. The sharp drop-off in catches between the 750-1000 m level and the 1000-1250 m level from 21.4% to 0.7% indicate that G. antarcticus does not normally live deeper than 1000 m and that the captures below that depth probably were made in the shallow zone of abundance. It seems curious that the most abundant cephalopod in the Antarctic Ocean is a bathypelagic species that apparently outnumbers its closest rival nearly 2 to 1. The three most common species besides B. abyssicola 377-090 Ficure 74.—Vertical distribution by 250-meter increments of the total sample population of Gonatus antarcticus Lonnberg, 1897. Antarctic. 196 U.S. NATIONAL MUSEUM BULLETIN 291 are all relatively shallow-living forms that have 70% to 85% of their numbers in less than 1000 m; only 18% of all B. abyssicola captured occur in less than 1000 m. Taken together the three common shallower living species outnumber B. abyssicola by 13%. Assuming that the trawling gear is mutually selective, the implication is that the shallow end of the vertical range of B. abyssicola overlaps with the deeper parts of the ranges of Crystallotheuthis glacialis, Brachioteuthis picta, and Gonatus antarcticus, and that B. abyssicola replaces these three species in the greater depths. Approximately 15 additional species of squid of the suborder Oegop- sida were collected in Antarctic waters through Cruise 15 and their numbers total less than 250 specimens. The most common of these, Alluroteuthis antarcticus Odhner, 1923, is represented by 69 specimens. Deeper dwelling forms, e.g., species of Mastigoteuthis, Chiroteuthis, and Histioteuthis, are represented by fewer than two dozen specimens each. Therefore, the dominance of Bathyteuthis abyssicola in the total known Antarctic cephalopod fauna, as well as among the bathypelagic inhabitants, is undisputed. It is a striking example of a species that occurs in small to moderate numbers over a broad geographical range but is so successfully adapted to a particular environment that it is the dominant species. 1 10. LL 12; nS 14. 15. BATHYPELAGIC SQUID BATHYTEUTHIS 197 Summary and Conclusions . A historical résumé traces the systematics and distributional rec- ords of Bathyteuthis abyssicola Hoyle, 1885. . The name B. abyssicola Hoyle, 1885, is shown to have priority over Benthoteuthis megalops Verrill, 1885. . A redescription of B. abyssicola from the type region, the Antarc- tic Ocean, is presented ; the species is illustrated in detail. . A study of morphometric features demonstrates the growth charac- teristics of B. abyssicola. . Two additional species of Bathyteuthis, B. bacidifera and B. berryi, are presented and illustrated from the eastern Pacific Ocean. . The taxonomic and morphometric characteristics of the species of Bathyteuthis are compared. . Distinct geographical variation exists between populations of B. abyssicola from the Antarctic, eastern Pacific, and Atlantic Oceans. . Characters that vary geographically include the number of suckers on the buccal lappets, the number of suckers on the arms, and the size of the gills. . Gill sizes are shown to be related to the oxygen content of the en- vironment; large gills may be a mechanism for more efficient respiration in areas of low-oxygen availability; e.g., in the east- ern tropical Pacific. The relationship between Bathyteuthis and Ctenopteryz is exam- ined and the family Ctenopterygidae is reinstated. The familial relationships of the Bathyteuthidae to other groups of oegopsids are surveyed; the family exhibits no close rela- tionships and holds a distinctive position within the Oegopsida. The oceanographic features of the Antarctic Ocean are reviewed. An analysis of the physicochemical parameters of the Antarctic Ocean (temperature, salinity, oxygen, and density) is presented on vertical sections along meridians 25°, 35°, 55°, 65°, 75°, 115°, 130°, 160° west and along 60° south latitude between 25° and 160° west. The distribution of B. abyssicola in relation to these parameters in the Antarctic Ocean is analyzed. The geographical distribution of B. abyssicola in relation to water masses in the oceans of the world is analyzed. 198 16. iis 18. 19: 20. 21. U.S. NATIONAL MUSEUM BULLETIN 291 The three-dimensional distribution of Bathyteuthis is governed in part by physicochemical characteristics, but the primary lim- iting factor in the distribution of this bathypelagic species is a biological phenomenon: organic productivity. High organic production accounts for the abundance of specimens from the Southern Ocean, Peru Current, Gulf of Panama, eastern Atlantic, ete. The vertical range of B. abyssicola in the Antarctic Ocean is deter- mined to extend from 500 to 2500 m with a peak of abundance between 1000 and 2250 m. Larvae and juveniles tend to live at shallower depths than adults. The occurrence of adults in the shallower end of the range is generally related to unusual oceanographic conditions, ¢.g., up- welling along the Antarctic Divergence. B. abyssicola, a true bathypelagic squid, lives at least 1000 m above the ocean floor. Although B. abyssicola occurs throughout the Antarctic Ocean it is not equally abundant in each region; its relative abundance is greatest in the Drake Passage Convergence, the southern Peru Current, the South Pacific and the South Pacific Con- vergence regions and least in the South Atlantic (Scotia Sea) region. The relatively few tows that caught exceptional numbers of B. abyssicola were made during the austral spring and summer seasons in the convergence zones. . The size-frequency composition of the sample population con- sists of a major peak at 31 to 47 mm mantle length and a second- ary peak at 11 to 23 mm ML. . The sex ratio of the total population is about 1:1, but males pre- dominate below 43 mm ML and females greatly predominate above 47 mm ML. . A comparison of 3-meter IKMT catches reveals that tows in the Antarctic were three times more successful and caught ten times the number of B. abyssicola than tows in the Gulf of Guinea. . B. abyssicola is the most common pelagic cephalopod captured in Antarctic waters ; the vertical ranges and the relative abundance of the three next most common species of oegopsids are com- pared with those of B. abyssicola. The ranges of these species overlap between 500-1000 m, but B. abyssicola replaces them at greater depths. BATHYPELAGIC SQUID BATHYTEUTHIS 199 References ABEL, O. 1916. Paliobiologie der Cephalopoden aus der Gruppe der Dibranchiaten. Gustav Fischer, Jena. 281 pp. ADAM, W. 1952. Cephalopodes. Résult. Sci, Expéd. Oceanogr. Belge dans les eaux cétiéres Africaines de l’Atlantique Sud, vol. 3, fase. 3, pp. 1-142. 1960. Cephalopoda from the Gulf of Aqaba. Contributions to the Knowl- edge of the Red Sea, no. 16. Bull. Sea Fish. Res. Station, Haifa, no. 26, pp. 1-27. ALLAN, J. 1945. Planktoniec cephalopod larvae from the eastern Australian coast. Rec. Australian Mus., vol. 21, no. 6, pp. 817-350. Backus, R. H., MEAp, G. W., HaepricH, R. L., and EBe.ine, A. W. 1965. The mesopelagic fishes collected during Cruise 17 of the R/V Chain, with a method for analyzing faunal transects. Bull. Mus. Comp. Zool., Harvard, vol. 134, no. 5, pp. 139-158. Baker, A. deC. 1954. The circumpolar continuity of Antarctic plankton species. Dis- covery Rep., vol. 22, pp. 201-218. BEEBE, W. 1926. The Arcturus Oceanographic Expedition. Zoologica, vol. 8, no. 1, pp. 1-45. Berry, 8. 8 1912. Some necessary changes in cephalopod nomenclature. Science, vol. 36, no. 932, pp. 6438-646. 1917. Cephalopoda. Australian Antarctic Exped. 1911-1914. Sci. Rep., ser. C, vol. 4, no 2, pp. 1-39. BIERrI, R. 1959. The distribution of the planktonic Chaetognatha in the Pacific and their relationship to the water masses. Limnol. and Oceanogr., vol. 4, no. 1, pp. 1-28. Brinton, BP. 1962. The distribution of Pacific euphausiids. Bull. Scripps Inst. Oceanogr., vol. 8, no. 1, pp. 51-270. Bropig, J. W. 1965. Oceanography. Jn Hatherton, T., ed., Antarctica. F. A. Praeger, New York, pp. 101-127. Bruun, A. F. 1943. The biology of Spirula spirula (L.). Dana Rep., no. 24, 46 pp. 1955. New light on the biology of Spirula, a mesopelagic cephalopod. Essays in the natural sciences in honor of Captain Allen Hancock, pp. 61-72. Univ. Southern California Press, Los Angeles. 1956. The abyssal fauna; its ecology, distribution and origin. Nature, vol. 177, pp. 1105-1108. 1957. Deep Sea and abyssal depths. Mem. Geol. Soc. America 67, vol. 1, pp. 641-672. BULLETIN HyDROGRAPHIQUE 1932. Bulletin Hydrographique pour l’année 1931 publié par le bureau du Conseil service hydrographique. Conseil permanent international pour l’exploration de la Mer. Bull. Hydrogr., Copenhagen, 1932, pp. 46, 98-99, 108-109. 200 U.S. NATIONAL MUSEUM BULLETIN 291 Butts, H. R., JR., and THOMPSON, J. R. 1965. Collections by the exploratory fishing vessels Oregon, Silver Bay, Com- bat, and Pelican made during 1956-1960 in the southwestern North Atlantic. Spee. Sci. Rep. U.S. Fish Wildl. Serv.—Fish. 510, 130 pp. CARCELLES, A. R. 1953. Catalogo de la Malacofauna Antarctica Argentina. Anales del Museo Nahuel Huapi, vol. 3, pp. 155-250. CHUN, C. 1908. Ueber Leuchtorgane und Augen von Tiefsee-Cephalopoden. Verhand. Deutschen zool. Ges., vol. 13, pp. 67-91. 1906. System der Cranchien. Zool. Anz., vol. 31, pp. 82-86. 1910. Die Cephalopoden. I. Teil: Oegopsida. Wiss. Ergebn. Deutschen Tiefsee-Exped. Valdivia, vol. 18, no. 1, pp. 1-410- CLARKE, A. H., JR. 1962. Annotated list and bibliography of the abyssal marine mollusks of the world. Bull. Nat: Mus. Canada 181, 114 pp. CLARKE, M. R. 1966. A review of the systematics and ecology of oceanic squids. Adv. Mar. Biol., vol. 4, pp. 91-800. CLOWES, A. J. 1933. Influence of the Pacific on circulation in the southwest Atlantic Oceans. Nature, vol. 131, pp. 189-191. 1938. Phosphate and silicate in the Southern Ocean. Discovery Rep., vol. 19, pp. 1-120. DEACON, G. E. R. 1933. A general account of the hydrology of the Southern Ocean. Dis- covery Rep., vol. 7, pp. 171-238. 1937. The hydrology of the Southern Ocean. Discovery Rep., vol. 15, pp. 1-214. 1937a. Note on the dynamics of the Southern Ocean. Discovery Rep., vol. 15, pp, 125-152. 1963. The Southern Ocean. Jn Hill, M. N., ed., The Sea, vol. 2. John Wiley & Sons, N.Y., pp. 1-554. DEFANT, A. 1961. Physical oceanography. Pergamon Press, London, vol. 1, 729 pp.; vol. 2, 598 pp. DELL, R. K. 1959. Cephalopoda. Rep. British Australian New Zealand Antarctic Exped., ser. B., vol. 8, no. 4, pp 89-106, figs 1-10. EBELING, A. W. 1962. Melamphaidae I. Systematics and zoogeography of the species in the bathypelagic fish genus Malamphaes Gunther. Dana Rep., no. 58, 164 pp. EBELING, A. W., and WEED, W. H., III 1963. Melamphaidae III. Systematics and distribution of the species in the bathypelagic fish genus Scopelogadus Vaillant. Dana Rep., no. 60, 58 pp. Ewine, M., and HEEZEN, B. C. 1956. Some problems of Antarctic submarine geology. Geophys. Monogr., vol. 1, pp. 75-81. Facer, E. W., and McGoweEn, J. A. 1963. Zooplankton species groups in the North Pacific. Science, vol. 140, no. 3566, pp. 453-460. BATHYPELAGIC SQUID BATHYTEUTHIS 201 FELL, J. W. 1965. Bionomics and physiological taxonomy of marine occurring yeasts. Ph. D. Dissertation, University of Miami (unpublished). 1967. Distribution of yeasts in the Indian Ocean. Bull. Mar. Sci., vol. 17, no. 2, pp. 454-470. FLEMING, R. H., and LAEVASTU, T. 1956. The influence of hydrographic conditions on the behavior of fish. FAO Fish. Bull., vol. 9, no. 4, pp. 181-196. FRASER, F. C. 1936. On the development and distribution of young stages of the krill (Euphausia superba). Discovery Rep., vol. 14, pp. 1-192. FRIEDMAN, S. B. 1964. Physical oceanographic data obtained during Eltanin Cruises 4, 5, and 6 in the Drake Passage along the Chilean Coast and in the Brams- field Strait, June 1962-January 1963. Tech. Rep. No. CU-1-64, NSF GA-27, Lamont Geological Observatory, 55 pp. FUGLISTER, F. C. 1960. Atlantic Ocean atlas of temperature and salinity profiles and data from the International Geophysical Year of 1957-1958. Woods Hole Oceanogr. Inst. Atlas Series, vol. 1, pp. 1-209. GoopRicH, H. S. 1892. Note on a large squid (Ommastrephes pteropus Stp.). Journ. Mar. Biol. Assoc. United Kingdom, vol. 2, pp. 314-321. GRIMPE, G. 1922. Systemtische Ubersicht der europiiischen Cephalopoden. Sber. Naturf. Ges. Leipzig, vols. 45-48, pp. 36-52. 1925. Zur Kenntnis der Cephalopoden-fauna der Nordsee. Wiss. Meeres- untersuch., vol. 16, no. 3, pp. 1-124. GUNTHER, FE. R. 1936. A report on the oceanographical investigations in the Peru Coastal Current. Discovery Rep., vol. 13, pp. 107-276. HAFFNER, R. E. 1952. Zoogeography of the bathypelagic fish, Chauliodus. Syst. Zool., vol. 1, no. 3, pp. 113-1388. Harpy, A. C., and GUNTHER, E. R. 1935. The plankton of the South Georgia whaling grounds and adjacent waters. Discovery Rep., vol. 11, pp. 1-456. Hart, T. J. 1934. On the phytoplankton of the southwest Atlantic and the Bellings- hausen Sea, 1929-1931. Discovery Rep., vol. 8, pp. 1-268. Hart, T. J., and Currig, R. L. 1960. The Benguela Current. Discovery Rep., vol. 31, pp. 123-298. Hoye, W. EB. 1885. Narrative of the Voyage of the Challenger Expedition. The Cephalo- poda. Rep. Voy. Challenger, vol. 1, part 1, pp. 269-274. 1885a. Diagnosis of new species of Cephalopoda collected during the cruise of HMS Challenger. II. The Decapoda. Ann. Mag. Nat. Hist., ser. 5, vol. 16, pp. 181-208. 1885b. Preliminary report on the Cephalopoda collected during the cruise of HMS Challenger. II. The Decapoda. Proc. Roy. Soc. Edinburgh, vol 13, no. 120, pp. 281-310. 202 U.S. NATIONAL MUSEUM BULLETIN 291 1886. Report on the Cephalopoda collected by HMS Challenger during the years 1873-1876. Rep. Voy. Challenger, Zool., vol. 16, no. 44, pp. 1- 246, 33 pls. 1886a. A catalogue of recent Cephalopoda. Proc. Roy. Phys. Soc. Edinburgh, vol. 9, pp. 205-267. 1886b. On the existence of Cephalopoda in the deep sea. Rep. British Assoc. Biol., 1886, p. 1064. 1904. Reports on the Cephalopoda. Reports on the dredging opera- tions .... Albatross. Bull. Mus. Comp. Zool., Harvard., vol. 43, no. 1, pp. 1-71, 12 pls. 1904a. A diagnostic key to the genera of recent dibranchiate Cephalopoda. Mem. Proc. Manchester Liter. Philos. Soc., vol. 48, no. 21, pp. 1-20. 1909. A catalogue of recent Cephalopoda. Second Supplement, 1897-1906. Proe. Roy. Phys. Soe. Edinburgh, vol. 17, no. 6, pp. 254-299. 1910. a at Ss: 7 a ee ee oe Janeen ete Pfam e -- Wipe tag alee, bath at = Th a et -§) A MUSEUM BULLETIN 291, PLATE 4 Bathyteuthis abyssicola, a-B, 57 mm ML, Elt. 354; c, E-1, x, 38 mm ML, Elt. 274; p, J, 28 mm ML, Elt. 1201: a, Tentacular club, left; B, distal tip of club. c-c, Inner sucker rings: c-p, Arm I; £, Arm IJ; r, Arm III; c, Arm IV. u, Outer sucker ring, Arm IV; 1, buccal sucker ring; J-x, club sucker rings. U.S. NATIONAL MUSEUM BULECETIN 2917 PEATE, 5 oOo @O©@ @©Ce sy ) Ge OO eo ee nie 65 a) VE Oo O@n2v z od Qo: oe G Le 5016. fey ey in “OTO) BS « * QO), © Oe ©« ‘0 Q) © ©... So ) © LL © © © OO) 6 Le. Te ee ice —— n Bathyteuthis abyssicola, larvae, A-p, G, 12 mm ML; E-F, 6mm ML; 4, 16 mm ML; 1, 29 mm ML: a, Dorsal view; B, Arm I, left; c, tentacular club, left; p, open mantle cavity; E, dorsal view; F, tentacular club, left; c-1, sucker rings. BULLETIN 291, PLATE 6 U.S. NATIONAL MUSEUM “nae, view. Bathyteuthis bacidifera, holotype, female, 37 mm ML, Elt. 34: a, Dorsal view; B, ventral U.S. NATIONAL MUSEUM BULLETIN 291, PLATE 7 Bathyteuthis bacidifera, a, paratype, and B-F, holotype, females, 37 mm ML, Elt. 34; G, paratype, female, 34 mm ML, Elt. 54: a, Open mantle cavity; B, funnel component of locking apparatus; c, mantle component of locking apparatus; p, nuchal cartilage; E, funnel organ and valve; F, pore of funnel groove; c, gladius. U.S. NATIONAL MUSEUM BULLETIN 291, PLATE 8 Bathyteuthis bacidifera, a, B, F, holotype, and c, p, E, paratype, females, 37 mm ML, Elt. 34: a, Brachial crown with buccal membrane expanded; B, right Arm I; c, upper beak; p, lower beak; E, radula; F, brachial crown showing connectives of the buccal membrane. U.S. NATIONAL MUSEUM BULLETIN 291. PLATE 9 Bathyteuthis bacidifera, holotype, female, 37 mm ML, Elt. 34: a, Tentacular club; 3, c, inner sucker rings, Arm I; p, £, inner and outer sucker rings, Arm I]; Fr, inner sucker ring, Arm III; c, inner sucker ring, Arm IV; n, buccal sucker ring; 1, J, inner and outer sucker rings, left tentacular club. U.S. NATIONAL MUSEUM BULLETIN 291, PLATE 10 Bathyteuthis bacidifera, juvenile and larval specimens: a, Dorsal view, male, 11.5 mm ML, D 1208 VI; z, left Arm I, D 1208 VI; c, open mantle cavity, D 1208 VI; p, left Arm I, male, 17 mm ML, D 1208 VIII; £, tentacular club; r, sucker from left Arm I, male, 11.5 mm ML, D 1208 VI; Gc, sucker from left Arm I, male, 17 mm ML, D 1208 VIJI; », sucker from left Arm I, female, 26 mm ML, D 1208 XIV. U.S. NATIONAL MUSEUM BULLETIN 291, PLATE 11 0° sete Sw aee a Se 0. eee a2, 208.050) 9 2,9,0 a? 2° oO © Saces Be) 092 © © OO,909 o@ 22 @ ee08° oe hau " O G2 e%e%y0% 0 Bathyteuthis berryi, holotype, a, B, p-H, male, 49 mm ML, Velero 8714, and a Paratype, c, 1, juvenile, 19 mm ML, Velero 10976. a, Ventral view; 8, Arm I; c, tentacular club; p-c, inner sucker rings from Arms I-IV; u, buccal sucker ring; I, inner sucker ring from tentacu- lar club. U.S. NATIONAL MUSEUM BULLETIN 291, PLATE 12 greet LY of it i 2 A y fal 4 Bathyteuthis berryi, a-r, holotype, male, 49 mm ML, Velero 8714: a, Radula; B, gladius; c, upper beak; p, lower beak; ©, spermatophore; F, enlarged section of spermatophore. B. bacidifera, G—u, paratype, male, 28 mm ML, Pillsbury 510: c, Spermatophore; u, en- larged section of spermatophore. U.S. GOVERNMENT PRINTING OFFICE: 1969 O—321—534 . = = ey pa “HUAI 3 908