H DISCOVERY REPORTS VOLUME XXXII -jOO Ul'l^^ DISCOVERY REPORTS Issued by the National Institute of Oceanography VOLUME XXXII CAMBRIDGE AT THE UNIVERSITY PRESS 1964 PUBLISHED BY THE SYNDICS OF THE CAMBRIDGE UNIVERSITY PRESS Bentley House, 200 Euston Road, London, N.W. 1 American Branch: 32 East S7th Street, New York 22, N.Y. West African Oifice: P.O. Box 33, Ibadan, Nigeria Printed in Great Britain at the University Printing House, Cambridge (Brooke Crutcliley, University Printer) CONTENTS SALPA FUSIFORMIS CUVIER AND RELATED SPECIES (published 4 October 196 1) By P. Foxton Introduction P^S^ 3 Acknowledgements .... Materials and Methods Synonymy and Previous Descriptions Description of the Species . Key to the Species. Morphological Characters of Taxonomic Importance Distribution Summary References Appendix Plates I, II 4 4 6 8 IS IS 27 29 30 31 following page 32 THE NATURAL HISTORY AND GEOGRAPHY OF THE ANTARCTIC KRILL {EUPHAUSIA SUPERB A D^A^.4) (published 9 November 1962) By James Marr Introduction 37 Acknowledgements 4° Early Records and Recent Literature 4° Scope of the Observations 4^ Material and Data S3 Methods in the Field S3 Identification of the Species S5 Laboratory Methods S^ Distribution in Outline S7 Physical Environment "4 Pack-ice ^4 Surface temperature 7S The Larval Stages ^^ Vertical distribution °° The developmental ascent 97 Deep and shallow horizontal dispersal 99 Diurnal vertical migration i°5 Some biological implications of the developmental ascent and its influence on the distribution and conservation of the euphausian population 118 The developmental ascent in relation to the local distribution of krill in the Ross Sea . . 123 %0, / oQ? ^v?ot| ^A^ ss. CONTENTS The Older Stages Economic and ecological importance . Sizes and quantities eaten by baleen whales Influence on the foetal growth rate of whales Patchiness ..... Habits and behaviour . Vertical distribution and migration Absolute density .... Food AND Nets Faecal pellets .... Spawning and hatching Speed of the Cold Bottom Water Causes of Patchiness Impact of the Swarms on the Ocean Pastures Happening of the Whales upon the Swarms Impact of the Whales on the Swarms . Intimate Structure of the Swarms Growth as Individuals and Swarms Moulting Reaction of Larvae, Adolescents and Adults to Ship Vertical Distribution and Migration Reconsidered Corrections Applied to the Catch-Figures . Horizontal Distribution, Growth and Dynamics of Dispersal Introduction and plan of presentation . The eggs and larvae from the vertical samples The larvae, adolescents and adults from the oblique and horizontal samples Distribution of Surface Discoloration and its Relationship to the OF THE Surface Population The Larval Populating of the East Wind-Weddell Surface Stream as an Local Distribution. The South Georgia Whaling Grounds The Coastal Topography of Antarctica and its Influence on the Conservation of the Surface Population Euphausia superha and E. triacantha ....... The Feeding Migrations of the Baleen Whales. Their Possible Origin Review OF Dynamics of Distributional Control Summary Appendix. Tables 62-4 References Plate III page 126 126 138 H7 148 154 157 170 172 176 176 212 215 240 241 243 245 251 256 258 268 278 284 284 289 358 Gross Distribution Annual Event AND History Depletion and 423 424 428 429 435 442 444 following page 464 412 412 413 [Discovery Reports, Vol. XXXII, pp. 1-32, Plates I-II, October, 1961] SALPA FUSIFORMIS CUVIER AND RELATED SPECIES By P. FOXTON CONTENTS page Introduction 3 Acknowledgements 4 Material and Methods 4 Synonymy and Previous Descriptions 6 Description of the Species 8 S. fusiformis Cuvier .......... 8 S. aspera Chamisso lo 5. thompsoni sp.nov 1 1 S. gerlachei sp.nov. 13 Key to the Species 15 Morphological Characters of Taxonomic Importance ... 15 The number of fibres per muscle band ....... 15 The arrangement of the body muscles 21 The external character of the test 22 The width of the body muscles in S. thompsoni and 5. gerlachei . . 23 The development of the stolon in S. thompsoni and S. gerlachei . . 25 Distribution 27 Summary 29 References 3° Appendix 3^ Plates I, II following p. 32 SALPA FUSIFORMIS CUVIER AND RELATED SPECIES By P. Foxton (Plates I-II and Text-figs, i-io) INTRODUCTION THIS paper represents the first part of a study of the salps of the Southern Ocean, and is a contribu- tion to the National Institute of Oceanography's studies of the distribution of oceanic organisms and the factors that govern them. Previous studies, which have described the distributions and Ufe-histories of species of Chaetognatha (David, 1955, 1958) and Euphausiidae (Baker, 1959; Marr, in press) show that even in an area of relatively simple oceanic circulation with well-defined water masses, such as the Southern Ocean, there is more regional differentiation in plankton distribution than had hitherto been supposed. Further- more, these regional differences become more apparent when it is possible to ascertain the range of variation between species, subspecies and local races. The salps commend themselves to a special study because, apart from the work of Michael (19 18) and Yount (1958), little attention has been paid to the relationship between their life-history and their environment in spite of the fact that they were one of the earliest planktonic groups to be described. In the Southern Ocean the salps are of especial interest because they occur at certain seasons in dense concentrations or swarms and, as they are primarily herbivorous, must play an important role in grazing down the phytoplankton, competing where distributions overlap, with such other key herbivores as Euphausia superba. The common salp of the subantarctic and antarctic zones of the Southern Ocean has in the past been assumed to be Salpa fiisiformis aspera, a variety or subspecies, characterized by a serrated test, of the type S.fusiformis. Confusion in the literature and discrepancies between published descriptions and specimens from the Southern Ocean made it necessary at the outset to examine in detail the taxonomy of the salp common in our collections and to establish the validity or otherwise, not only of previous identifications, but of aS". fusiformis aspera as a subspecies, form or extreme variant of S. fusiformis. Fortunately a large collection of material was available for study mainly from the Southern Ocean but also to a lesser extent from the Southern Indian, North Atlantic and South Atlantic Oceans. I have also, by the kindness of workers in other laboratories, been able to extend the geographical range of specimens examined into areas not represented in the plankton collections of the National Institute of Oceanography. The results of this study show that the differences between S . fusiformis and the variety S.fusiformis aspera are greater than had been previously supposed. Within what I shall refer to as the ' fusiformis group ' there has been a confusion of four species, three of these being hitherto variously known as S. fusiformis aspera. Morphological differences and features of geographical distribution allow the four species S. fusiformis, S. aspera, S. thompsoni and S. gerlachei to be recognized. Two of these species, S. thompsoni and S. gerlachei, are limited in their distribution to the Southern Ocean and it is with the ecology of these that a later paper will be primarily concerned. It was felt, however, that it 4 DISCOVERY REPORTS would be useful at this stage to present the systematic results in advance, particularly since there has been such confusion and divergence of opinion regarding the relationship between S. fusiformis and S. fusiformis aspera. ACKNOWLEDGEMENTS This study would have been severely restricted in scope were it not for the specimens so kindly made available to me by other workers. It is with pleasure that I record my indebtedness to Professor T. Tokioka, Seto Marine Biological Laboratory, Japan; Dr Leo Berner, Jnr, Scripps Institution of Oceanography, La JoUa, U.S.A. ; Dr J. L. Yount, University of Florida, Gainesville, U.S.A. ; Dr J. H. Eraser, F.R.S.E., Department of Agriculture and Eisheries for Scotland, Marine Laboratory, Aberdeen; and Dr E. Beyer, Institutt for Marine Biologi, Oslo, Norway. I should also like to thank Dr E. S. Russell, E.R.S., Director, Marine Biological Association, Plymouth, who kindly allowed me to participate in a cruise of R.V. 'Sarsia' where I was able to examine material from the Bay of Biscay. I am particularly grateful to Dr J. H. Eraser, E.R.S.E., for his comments on the original typescript and to my colleague, Mr P. M. David, who has given me helpful advice throughout this study. I wish to thank Baron Marc de Selys Longchamps for permission to reproduce two figures (fig. za, h) from the Reports of the Belgica Expedition, and the Editor, Commonwealth Scientific and Industrial Research Organization, Melbourne, for permission to reproduce two figures (fig. i a, b) from. H. Thompson's Pelagic Tunicates of Australia. Eigures i and 2 have been prepared for publication by Mr A. Style. MATERIAL AND METHODS This paper is based for the most part on material collected by ships of the Discovery Committee and subsequently the National Institute of Oceanography. As the ultimate aim of the work was a quanti- tative appraisal of an animal's distribution in time and space in relation to its life-history, I have used plankton samples taken with a standard net fished in a standard manner. In this paper only a small part of the total data — that relating to the systematic results — will be used. Most of the material was collected with a i -metre stramin net towed obliquely (NiooB) or hori- zontally (NiooH) at stations which were selected, from a vast number available,, to give as com- plete a geographical coverage of the Southern Ocean as possible. It was soon realized that the subantarctic and antarctic data should be compared with that from other oceans, and so where lines of stations have extended north of the subtropical convergence (indicated by a dotted line in Text-fig. 10, p. 28) use has been made of samples collected at them irrespective of the gear used. Such samples include those taken with the 2-metre stramin net fished obliquely (TYEB) or horizontally (TYEV) and use has also been made of material collected in the North Atlantic with the Isaacs-Kidd Midwater Trawl. While not relevant to this paper, full descriptions of the i -metre and 2-metre nets and their method of use may be found in Kemp, Hardy and Mackintosh (1929), while the Isaacs-Kidd Mid- water Trawl is described by Isaacs and Kidd (1953). I have, in addition, been able to examine specimens collected with a variety of gear from areas not represented in the Discovery Collections. These include specimens from the eastern and central Pacific, Japan, Bay of Biscay, and Iceland. I have also been able to augment the Discovery Pacific stations with material collected quite independently by the Norwegian 'Brategg' Expedition of 1947-48. All the measurements on which this study is based were made on specimens preserved in 5-10% neutral sea-water formalin. As the period of preservation prior to examination varies from a few months to 30 or so years some shrinkage can be assumed to have taken place. A few specimens, how- MATERIAL AND METHODS 5 ever, which were originally examined alive, or just a few minutes after preservation in 1951, showed no significant change in size when re-examined in 1959. It thus seems safe to assume that the effects of shrinkage due to preservation, at least in so far as they affect the morphological characters used in this study, are negligible. The number of aggregate and solitary individuals of each species in each sample was counted while the size composition of each form was determined by length measurements. The stage of development of the embryo or stolon was assessed by detailed examination of the oozoid ( = solitary) and blasto- zooid (= aggregate) with a low-power stereomicroscope. These measurements and counts result in data that will form the basis of a later paper on life-history and distribution. The present work is concerned with detailed structure in so far as it relates to the taxonomy of the animals, and this depends on the detailed examination and measurements to be described. In many previous works devoted to the description of collections of salps measurements of length are given often without the points of reference from which the measurements were made. This seems to be of importance, because while some authors distinguish between measurements that include the prolongations that often characterize the test, others do not, even though such extensions of the test may represent a large percentage of the total length. While strictly speaking the length of a specimen should be the distance between its anterior and posterior extremities such a measurement seems unsatisfactory when applied to the Salpidae. Firstly, the test is very variable in form not only between species but within species on account of age, natural wear and damage in the net, and, secondly, the test frequently becomes separated from the body n very ri h samples presumably through damage in the net, since salps are on the whole fragile animals. It thus seems more realistic to use measure- ments of length based on the body (or mantle; see Yount, 1954) rather than the test. In the present study the following standard measurements are used : total length, the distance between the anterior and posterior extremities of the test; body lefjgth, the distance between the mouth and the atrial opening; nucleus length, the nucleus (or alimentary canal) is typically ovoid in the species to be considered and so the measurement taken is the length of the major axis. Throughout this paper all lengths quoted will refer to body length unless otherwise stated. It will be realized from the relative shape of the solitary and aggregate stages that in the former total length and body length approximate to each other while in the latter they may be very different. Lengths in all cases were measured to the nearest millimetre on a clear plastic graduated scale. With larger specimens, particularly of the solitary form greater than 50-60 mm. in length, the measurement is not assumed to be more accurate than ±2 mm., owing to the extreme elasticity of the body especially in specimens that have become separated from their tests. However, this in no way invalidates con- clusions drawn from such data which are always grouped for purposes of comparison into 5 or 10 mm. size groups. The detailed examination of specimens was assisted by the use of aqueous stains. Toluidine blue (Yount, 1954) or methylene blue in dilute solution allowed the surface of the test (p. 22) to be seen in detail, and in some cases made visible denticulations on old worn tests that might well have been classed as smooth. Rose Bengale (Yount, 1954) injected either through the mouth or the atrial aperture into the body proved a useful method of seeing the arrangement of the body musculature, particularly in small specimens. A low-power stereomicroscope proved adequate to see the structure of individual muscle bands and except in very small specimens it was possible to count the number of constituent fibres in each muscle (p. 1 5). Where necessary the widths of individual muscle bands (see p. 23) were measured with a graduated eyepiece. 6 DISCOVERY REPORTS The scheme of nomenclature followed in this paper is based on that described by Yount (1954) in which the test is described in terms largely formulated by Stiasny (1926), while the muscle notation follows that of Streiff (1908). SYNONYMY AND PREVIOUS DESCRIPTIONS It is proposed in this section to describe only those works that are relevant to a discussion of the synonymy of Salpa fusiforniis and its related species, and in particular to the status of the serrated forms. For a detailed account of the taxonomy of the Salpidae in general, reference should be made to Metcalf (1918) or to Thompson (1948), who deals with the pelagic Tunicata as a whole. It should be remembered that in the early nomenclature of the group, solitary and aggregate forms of the same species often acquired different names. For example, Cuvier (1804) gives the name S. fusiformis for the aggregate form of the species, while Chamisso (1819) calls the solitary form of the same species S. runcinata. With the realization, largely as a result of the work of Chamisso (1819), that the two forms were morphologically different stages in the life-history of a single species, the one sexual, the other asexual, the two names became linked by a hyphen. Thus the species became known as S. runcinata-fusiformis Chamisso-Cuvier (Krohn, 1846; Traustedt, 1885; Herdman, 1888). Sub- sequently it was considered by Ritter (1905) that the first name in the couplet should be that having priority and so the specific name became S. fusiformis -runcinata Cuvier-Chamisso. While many authors followed the lead of Krohn (1846), others, notably Apstein (18946), foreshadowed modern usage in dropping the second name of the couplet, so making the species S. fusiformis Cuvier. In the following account of previous literature the nomenclature used is that of the author under discussion. Herdman (1888) described from the Challenger collections the solitary form of a new species— S. echinata — which differed from S. runcinata-fusiformis Cuv.-Cham. in having a serrated test and in M. (body muscles) IV-IX being parallel. He failed, however, to recognize the aggregate form of the species, although it can be concluded from his remarks (pp. 76-7) with regard to variation in specimens of the aggregate form of S. runcinata-fusiformis, from the geographical location of his stations, and from his figures (PI. VI, figs. 5-10) that he had a mixture of species in which such differences as I have taken to be specific were regarded as mere variations within the one species — S. nmcinata-fusiformis. Apstein (18946) in tropical material found two specimens of a solitary form that agreed with Herdman's (1888) description of S. echinata. However, on the basis of an examination of the stolon of one of the specimens, buds of the aggregate form were found and figured (fig. 14) similar to the aggregate of S. fusiformis but differing in the slight lateral separation of M. IV and V. On the basis of their general similarity Apstein concluded that his specimens were variants of S. fusiformis that occurred only in the solitary form, and so called them S. fusiformis var. echinata (p. 15). It should be noted that both Apstein (fig. 14) and Herdman (fig. 7) figure specimens of the aggregate form with a lateral separation of M. IV-V, and in neither case is any significance attached to this morphological difference. Ritter (1905) described specimens similar to those of Herdman (1888) and Apstein (18946) which he called S. fusiformis-runcinata form echinata. Figures 14-15 (p. 68) are the best to that date of the solitary form, and show the serrations of the test in detail and the body musculature with M. IV-IX parallel. Ritter's specimens were larger than those of the type S. fusiformis-runcinata and so he con- sidered echinata to be a 'well-marked style or form of the species associated with age'. From a large collection, which included antarctic specimens, Apstein (1906) described (pp. 249-50) and figured (PI. XXVI) well-serrated aggregate (fig. 4) and solitary (fig. 6) forms of S. fusiformis form echinata. Unlike previous descriptions of the serrated form the solitary is shown with M. VIII-IX medio-dorsally joined, while the aggregate is figured with a well-serrated test and M. IV-V not joined SYNONOMY AND PREVIOUS DESCRIPTIONS 7 laterally. Apstein's data covered a wide geographical range, and since the serrated form tended to occur in the colder waters (p. 269, figs. 5-6) he concluded that the form echinata was a cold-water variant of the type occurring in both aggregate and solitary form. Apstein, however, remarks (p. 250) that the differences between the form echinata and the type show great variation. This, together with the fact that the collection was from a wide geographical area, including subantarctic and antarctic material, suggests that there was a confusion of species. It will be shown later (p. 22) that the specimen of the solitary form (Text-fig. 6) is typical of a species limited in its distribution to the Southern Ocean, while the aggregate form (Text-fig. 4) shows characters associated with a species that occurs only in waters north of the subtropical convergence. Farran (1906) in an oft-quoted paper described the serrated form in Irish waters from specimens of the aggregate stage. His statement that the musculature was the same as that of the type refers to specimens of the aggregate stage only, but it has frequently been quoted as referring to the solitary form as well (Thompson 1948 p. 158). The serrated form of S. fusiformis has subsequently been described from collections made by most expeditions, and references are to be found in Apstein (1908); Ihle (1910, 191 1, 1912); and Murray and Hjort (1912), and others. Ihle (191 1, p. 587) used for the first time the name S. fusiformis aspera to designate this serrated form. From a small collection of material collected by the Belgica Expedition van Beneden and Selys- Longchamps (191 3) described in a rarely quoted paper a new variety of S. fusiformis which they called S. fusiformis Gerlachei. Their description, based on one specimen each of the solitary and aggregate forms, is most thorough and detailed, and a comparison is made between their specimens and Herdman's (1888) and Ritter's (1905) descriptions of 5. echinata and S. fusiformis-runcinata. The solitary form of S. fusiformis Gerlachei differs from S. echinata in having M. VIII and IX joined and having a highly serrated test with a medio-dorsal serrated ridge. Some of the differences they describe, such as the lack of a 'cephalic projection' in the aggregate form, are probably variations attributable to wear or damage but the major characters of muscle arrangement and test structure they found incompatible with the published descriptions of S. echinata and S. fusiformis-runcinata. They suggest (p. 119) that these differences are characteristic of a form, variety or species endemic to the part of the Antarctic from which they came, although they agree that on such limited material the possibility of individual variation cannot be ruled out. Metcalf (191 8) in his classic paper on the Salpidae of the Philippines concludes that the specimens of S. fusiformis aspera in his collection showed a complete intergradation between smooth and the most spinose test. His figure of the solitary form is of interest because it seems to combine the test characters of Ritter (1905, fig. 14) based on specimens from the San Diego region, and the musculature of Apstein (1906, fig. 6) based on antarctic specimens — a combination not found in any of the specimens to be described in this paper. Sewell (1926, p. 76) reviews briefly previous records of S. fusiformis aspera and their interpretation and concludes that ' it seems impossible to justify any separation, even as a distinct form, of those with spinose test from those with a smooth one, though it is possible that Ritter and Farran are right in their belief that the spinose character is a mark of age'. This view is favoured by Thompson (1948) while Yount (1954) goes further and groups all the variants together as S. fusiformis. The works described above are those relevant to the problem of the serrated form of S. fusiformis and the various conclusions of the authors cited may be summarized as follows. (i) New species: Chamisso, 1819; Herdman, 1888. (2) A subspecies (of which there may be others): van Beneden & Longchamps, 191 3. (3) Cold-water form; Apstein, 1906. 8 DISCOVERY REPORTS (4) Old-age variant : Ritter, 1905; Farran, 1906. (5) No recognizable variety or form: Sewell, 1926; Yount, 1954. It is perhaps not surprising in view of these differing conclusions that the overall picture of the geographical range of S . fusiformis extends from 60° N. to 65° S. (Thompson, 1948). Few, however, of the above papers are based on adequate subantarctic and antarctic material and it is significant that where specimens were available, well-serrated animals are described in which the solitary form has M. VIII and IX joined dorsally (Apstein, 1906; van Beneden, 1913), while in some cases specimens of the aggregate form are described in which M. IV and V are joined laterally (Herdman, 1888; van Beneden, 1913). It will be realized from the above that previous authors have attached varying taxonomic importance to different morphological features. When any one of these features, such as the external character of the test, has been considered on its own, such differences as occurred, particularly in those collections of material that cover a wide geographical range, have invariably been attributed to individual variation within a single species. The data to be presented here will show that it is possible by con- sidering variation in a complex of characters to distinguish four species. These species have been variously confused in the literature of the group. DESCRIPTION OF THE SPECIES Salpa fusiformis Cuvier, 1804 (Text-fig. la, b) Salpa maxima variety Forskal, 1775. Salpa fusiformis Cuvier, G., 1804, Ann. Mus. Hist. Nat., An. xii, Tome 4, pp. 360-82, pi. 68; Apstein, 1894^, 1901, 1906; Farran, 1906; Streiff, 1908; Ihle, 1910, 1935; Sigl, 1912; Bomford, 1913; Metcalf, 1918; Sewell, 1926, 1953; Thompson, 1948; Yount, 1954 (part); Bernard, 1958; Fagetti, 1959. Salpa runcinata Chamisso, 1819; Brooks, 1893. Salpa maxima Meyen, 1832. Biphora depressa Sars, 1829. Biphora tricuspidata Sars, 1829. Salpa runcinata-fusiformis Krohn, 1846; Traustedt, 1885; Herdman, 1888; Apstein, 1894a. Salpa fusiformis-runcinata Ritter, 1905; Ritter and Byxbee, 1905 (part). Salpa fusiformis fusiformis IhXc, 191 2; Stiasny, 1926. Salpa fusiformis f . typica Michaelson, 191 5. Solitary Form (Text-fig. la) (based on the examination of fifty-three specimens) External appearance. Elongate, with a slightly convex anterior and a squarely cut off posterior. In section the anterior is subcylindrical with a flattened dorsal surface, while the posterior is triangular. The mouth and the atrial apertures are terminal. Test. Smooth, firm and transparent. Thin, except where thickened to form characteristic ridges. From the posterior dorsal edge, starting at two spinose processes of the test there are the left and right dorsal limiting ridges which, proceeding forward, soon divide into outer and inner branches. Lateral ridges, also originating in posterior processes, form the edges to the dorsal surface. A ventro-lateral ridge runs below each dorsal edge while posteriorly there is a well-marked keel-like mid-ventral ridge which, directed forward, divides to form two ventral limiting ridges. Body muscles. Nine body muscles. M. I-III and M. VIII-IX converge and fuse medio-dorsally. j M. IV-VII are parallel. DESCRIPTION OF THE SPECIES 9 Muscle fibres. Mean number 29-4± 1-4 for M. IV, based on forty-eight specimens. See pp. 15- 21 for explanation. Alimentary canal. Compact, oval nucleus. Stolon. Proceeds forward, turning left at about the level of M. Ill, to proceed back emerging from the test in front of the nucleus. Length. Up to 45 mm. Text-fig. I. S. fusiformis (a) Solitary form, ventral view. Length 20 mm. (b) Aggregate form, dorsal view. Length 20 mm. (From Thompson, 1948.) Aggregate Form (Text-fig. ib) (Based on the examination of forty-eight specimens) External appearance. Body barrel shaped with conical anterior and posterior protuberances of the test which are typically long in relation to the body, giving it a slight asymmetry* and an elongate, fusiformis appearance. Test. Smooth, firm, transparent with no obvious ridges. Body muscles. Six body muscles. M. I-IV and M. V-VI fuse dorsally, thus forming two groups. M. IV-V approach and join laterally. Muscle fibres. Mean number 51 -8 ±1-4 for the total of M. I-VI, based on 45 specimens. See pp. 15-21 for explanation. Approximate mean number of fibres per muscle of 8-6. Alimentary canal. Compact, oval nucleus with its major axis lying to the right. Pale green in living specimens. * Note. In S. fusiformis and in the three other species to be described the protuberances or processes of the test are asymmetrical, the anterior being to the right, the posterior to the left, or vice versa depending on the zooid's original position in the stolon. 10 DISCOVERY REPORTS Embryo. Single embryo, behind M. V at its point of lateral fusion with M. IV. Length. Up to 25 mm. Distribution of S. fusiformis. Recorded by other workers in nearly all oceans (see Thompson, 1948). Present data demonstrate that the southern limit to its distribution is the region of the sub- tropical convergence. Data are insufficient to define its northern limit but it has been recorded at a station in 63° 51' N., 16° 20' W., and also round the N. Cape of Iceland in September 1957 (Fraser, 1959)- Salpa aspera Chamisso, 1819 (PI. I) Salpa aspera Chamisso, A. von, 18 19, De Animalibus quibusdam e classe Vermium Linnaeana, Fasc. I, De Salpa, 24 pp., PI. i (Berolini). Salpa echinata Herdman, 1888. Salpa runcinata-fusiformis Herdman, 1888 (part). Salpa runcinata-fusiformis var. echinata Apstein, 1894a. Salpa fusiformis var. echinata Apstein, 18946; Apstein 1906 (part), 1908; Ihle, 1910. Salpa fusiformis-runcinata form echinata Ritter, 1905; Ritter and Byxbee, 1905. Salpa fusiformis-runcinata Ritter and Byxbee, 1905 (part). Salpa fusiformis form aspera Ihle, 191 1; 1935; Metcalf, 1918; Sewell, 1926; Thompson, 1948. Salpa fusiformis Yount, 1954 (part). Solitary Form (PI. la and c) (Based on the examination of nine specimens) External appearance. Similar to S. fusiformis, with more convex anterior. Posterior squarely cut off with pronounced spinose processes. Test. (PI. I c) Serrated. Thin except for the thickened ridges which are similar in number and disposition to S. fusiformis. Ridges characterized by denticulations of the test which give a serrated appearance. Posteriorly the serrated ridges extend to pronounced spinose processes of the test. As in S. fusiformis there are ten ridges defined as follows : (a) Four dorsal-limiting ridges, (b) Two dorsal-lateral ridges, (c) Two ventro-lateral ridges, (d) Two ventral ridges which originate posteriorly as a single mid-ventral ridge. I shall refer to this arrangement of serrated ridges as ' simple '. Body muscles. Nine body muscles. M. I-III converge and fuse dorsally. M. IV-VII are parallel. M. VIII-IX may approach but do not fuse medio-dorsally. Muscle fibres. Mean number 100-7 for M. IV, based on nine specimens. Alimentary canal. Compact oval nucleus. Stolon. Proceeds forward to about the level of M. VII where, turning left, it doubles back fol- lowing the edge of the nucleus and emerging through the test on the right side of the nucleus between M. VIII-IX. Length. Up to 95 mm. Aggregate Form (PI. lb) (Based on the examination of twenty-nine specimens) External appearance. Body barrel shaped with conical anterior and posterior protruberances of the test. The latter show great variation in both their shape and length. None of the specimens showed the ' fusiform ' appearance so characteristic of S. fusiformis. DESCRIPTION OF THE SPECIES n Test. Serrated. Denticulations irregular in their distribution. Body muscles. Six body muscles. M. I-IV and M. V-VI fuse dorsally to form two dorsal muscle groups. M. IV-V approach laterally but do not fuse. Muscle fibres. Mean number 194-1 for the total of M. I-VI, based on twenty-nine specimens. Alimentary canal. Compact oval nucleus. Embryo. Single, as in 5. fusiformis. Length. Up to 60 mm. Distribution of S. aspera. Present data show its occurrence north of the subtropical convergence in Atlantic, Pacific and Indian Oceans. Yount (1954, pp. 303-4) has drawn attention to the similarity between S. maxima Forskal and the serrated form of S. fusiformis {= S. aspera) and suggests that these species may have been confused in the literature. Specimens of the aggregate form of S. maxima examined in the course of the present study have differed from S. aspera in having a most pronounced bulbous thickening of the test, ventral to the nucleus; while the solitary form is quite distinct in the absence of dorsal fusion of M. I-III, M. I-IX being in fact more or less parallel (see Thompson, 1946, Pis. 68 and 69). Spinose or serrated specimens of S. maxima have not occurred so far in the collections examined and the serrated specimens of this species figured by Yount (ibid. p. 301, fig. 13 and p. 302, fig. 14) I consider likely to be S. aspera. Salpa thompsoni sp.nov. (PI. II) [?] Salpa antarctica Meyen, 1832. Salpa echinata Herdman, 1888 (part). Salpa runcinata-fusiformis Herdman, 1888 (part); 1910. Salpa fusiformis var. echinata Apstein, 1906 (part). Salpa runcinata-fusiformis var. echinata Herdman, 1910 (part). Salpa fusiformis aspera Mackintosh, 1934; Hardy and Gunther, 1935. HoLOTYPE. A specimen of the solitary form 71 mm. long taken in a i m. net towed obliquely (N looB) between 89 m. and the surface at Discovery station 2144, 48° 04-3' S., 101° 07-2' E. on II December 1937. Registration number. British Museum (Nat. Hist.), i960. 10. 18. 5. Body proportions. Total length 75 mm., body length 71 mm., nucleus length 15 mm. Muscle-fibre number. 89 fibres in M. IV. Stolon. Stolon with two fully differentiated blocks of buds : Block I. 144 buds, blastozooids 3 mm. long Block 2. 230 buds, blastozooids i mm. long. No evidence that previous release of buds has occurred (see p . 25 ). Stolon extends forward to level of M . V. Allotype. A specimen of the aggregate form 20 mm. long taken in a i m. net towed obliquely (NiooB) between 91 m. and the surface at Discovery station 1906, 53° 54-6' S., 63° 58-1' W. on 30 November 1936. Registration number. British Museum (Nat. Hist.), i960. 10. 18.6. Body proportions. Total length 34 mm., body length 20 mm., nucleus length 5 mm. Muscle-fibre number. 188 fibres for the sum of M. I-VI counted along the right side. Embryo. Single oozoid present. Paratypes. Ten specimens of the solitary form taken in a i m. net towed obliquely (NiooB) from 122 m. to the surface at Discovery station 2574, 47° 59-9' S., 19° 56-2' E. on 3 February 1939. Registration number. British Museum (Nat. Hist.), i960. 10. 18.7. 12 DISCOVERY REPORTS Ten specimens of the aggregate form taken in a i m. net towed obliquely (NiooB) from lOO m. to the surface at Discovery station 2087, 44° 22-1' S., 23° 31' E. on 17 November 1937. Registration NUMBER. British Museum (Nat. Hist.), i960. 10.18. 8. Solitary Form (PI. Ila and c) (Based on the examination of ninety-two specimens.) External appearance. Elongate. Similar to S. aspera but more heavily serrated. Test (PI. lie). Heavily serrated, firm, with thickened areas forming pronounced serrated ridges. Number and disposition of lateral and ventral ridges is similar to S. aspera but dorsal surface is more complex owing to a number of less pronounced ridges which I shall call secondary ridges to distinguish them from the more obvious principal ridges. Although not always immediately obvious the secondary ridges become particularly evident when stained with methylene blue (see p. 5). The principal ridges are as follows : (a) inner and outer dorsal limiting ridges, (b) a dorsal-lateral ridge, (c) a ventro-lateral ridge, (d) posteriorly a single keel-like mid-ventral ridge which divides to form left and right branches or ventro-limiting ridges. The secondary ridges are as follows: (e) a medio-dorsal ridge extending for a greater part of the length of the body, (/) a secondary dorsal-limiting ridge which runs between and parallel to the inner and outer dorsal- limiting ridges, (g) a secondary dorso-lateral ridge which runs inside and parallel to the main dorso-lateral ridge, (h) connecting the inner and outer branches of the dorsal-limiting ridges is a short transverse ridge at about the level of M. VHI. I shall refer to this arrangement of serrated ridges as ' complex '. Body muscles. Nine broad body muscles similar in their disposition to S. fusiformis. M. I-HI fuse medio-dorsally. M. IV-VH are parallel. M. VHI-IX fuse medio-dorsally. Muscle fibres. Mean number 93-2 ±2-8 for M. IV, based on 92 observations. Alimentary canal. Compact, oval nucleus. Stolon. Similar to 5. aspera but it extends anteriorly much further before doubling back on itself. Length. Up to 120 mm. Aggregate Form (PI. lib) (Based on the examination of one hundred and twenty-six specimens.) External appearance. Body barrel shaped with conical anterior and posterior protuberances of the test. Great variation in body form. The anterior and posterior processes may be long, giving a typical ' fusiform ' shape, or shortened. Test. Serrated, firm. From the anterior protuberance a dorsal serrated ridge runs posteriorly to form the right dorsal-lateral edge of the test. At about the level of the atrial orifice this serrated ridge divides to form a right ventral-lateral ridge which reconnects with the dorsal-lateral ridge at about the level of the nucleus. On the left side of the test there are left dorsal and ventral-lateral ridges which originate in the posterior protuberance. In well-serrated animals the disposition of these ridges is such as to give, as Herdman (1888, p. 76) remarked, an 'angular or somewhat prismatic appearance' to the test. Body muscles. Six broad body muscles similar to S. fusiformis. M. I-IV and M. V-VI fused to form two dorsal groups. M. IV-V converge and join laterally. DESCRIPTION OF THE SPECIES 13 Muscle fibres. Mean number 1777 ±3-4 for the total of M. I-VI, based on 126 specimens. Alimentary canal. Compact, oval nucleus. Embryo. Single, in the same position as S. fusiformis. Distribution of S. thompsoni. Restricted to the Southern Ocean. Does not occur north of the subtropical convergence. Circumpolar in moderate latitudes. In high latitudes restricted to Atlantic, Indian and Australian sectors. Salpa gerlachei sp.nov. (Text-fig. 2) Salpa runcinata-fusiformis var. echinata Herdman, 1910 (part?). Salpa fusiformis var. gerlachei van Beneden and de Selys Longchamps, 191 3. HoLOTYPE. A specimen of the solitary form 56 mm. long taken in a i m. net towed horizontally (N 100 H) 5 m. belowthesurfaceat Discovery station 1654, 7S°43-6' S., 176° 59-4' E. on 25 January 1936. Text-fig. 2. S. gerlachei sp.nov. («) Solitary form, lateral view. Length 27 mm. (i) Aggregate form, dorsal view. Length 17 mm. (From Van Beneden, 1913.) Note. The aggregate form shown here is atypical in lacking an anterior protuberance of the test ; probably an artefact due to wear. Registration NUMBER. British Museum (Nat. Hist.), i960. 10.18. i. Body proportions. Total length 63 mm., body length 56 mm., nucleus length 11 mm. Muscle-fibre number. 48 fibres in M. IV. Stolon. Stolon with one fully differentiated block of buds consisting of 174 blastozooids, each I mm. long. Evidence that one block, at least, has already been released (see p. 25). Stolon extends for- ward to level of M. II-III. Allotype. A specimen of the aggregate form 23 mm. long taken in a i m. net towed obliquely (N looB) from 109 m. to the surface at Discovery station 1296, 69° o6-6' S., 131° 42-6' W. on i March 1934- Registration NUMBER. British Museum (Nat. Hist.), i960. 10. 18.2. Body proportions. Total length 34 mm., body length 23 mm., nucleus length 6 mm. 14 DISCOVERY REPORTS Muscle-fibre number. 127 fibres for the sum of M. I-VI counted along the right side. Embryo. Single oozoid present. Paratypes. Five specimens of the solitary form taken in a i m. net (N lOoB) which fished between 440 and 190 m. at Discovery station 1257, 67° 52-4' S., 129° 27-5' W. on 11 January 1934. Registration number. British Museum (Nat. Hist.), i960. 10. 18.3. Ten specimens of the aggregate form taken in a i m. net towed obliquely (NiooB) from 109 m. to the surface at Discovery station 1296, 69° o6-6' S., 131° 42-6' W. on i March 1934. Registration number. British Museum (Nat. Hist.), i960. 10. 18.4. Solitary Form (Text-fig. zd) (Based on the examination of sixty-seven specimens) External appearance. Elongate. Similar to S. thompsoni. Test. Heavily serrated, firm. Pattern and number of serrated ridges identical with those in S. thomp- soni, i.e. 'complex'. Body muscles. Nine thin body muscles. M. I-HI converge and join dorsally. M. IV-VH parallel. M. VHI-IX converge and join dorsally. Area of contact of the muscles that join dorsally is much smaller than in S. thompsoni. Muscle fibres. Mean number 48-8 ±1 -8 for M. IV, based on 67 observations. Alimentary canal. Compact, oval nucleus. Stolon. Similar to S. thompsoni. Length. Up to 75 mm. Aggregate Form (Text-fig. 2b) (Based on the examination of fifty specimens) External appearance. Body barrel shaped with conical anterior and posterior protuberances of the test. Similar to S. thompsoni. Test. Serrated. Disposition of the serrated ridges is the same as in S. thompsoni. Body muscles. Six thin body muscles. M. I-IV and M. V-VI fuse dorsally to form two muscle groups. M. IV-V converge and join laterally. Muscle fibres. Mean number 128-8 ±3-2 for the total of M. I-VI, based on 50 specimens. Table i . A summary of the features characterizing the four species of the 'fusiformis ' group Maximum External Mean body length character Arrangement of number Width of {mm.) of the test the body muscles of fibres muscles Distribution s. fusiformis Aggregate 25 Smooth M. IV-V laterally 51-8 -J" joined North of S.T.C.* Solitary 45 Smooth M. VIII-IX joined 29-4 s. aspera Aggregate 60 Serrated M. IV-V separate 194- 1 ~\ Solitary 95 Serrated M. VIII-IX paral- 100-7 North of S.T.C. (simple)f lel or nearly so s thompsoni Aggregate 60 Serrated M. IV-V laterally joined 177-7 Broad i8-9:it Middle latitudes of Southern ocean. Circumpolar Solitary 120 Serrated (complex)f M. VIII-IX joined 93-2 Broad 2i-2:iJ Northern limit— S.T.C. s gerlachei Aggregate 33 Serrated M. IV-V laterally joined 128-8 Narrow 27-6:iJ High latitudes of the Pacific sector of Southern Solitary 75 Serrated (complex)f M. VIII-IX joined 48-8 Narrow 35-8:i| Ocean * S.T.C.= subtropical convergence. f For explanation of the terms ' simple ' and 'complex' see pp. lo, 12. J Ratio of body length: muscle width, see p. 23. DESCRIPTION OF THE SPECIES 15 Alimentary canal. Compact, oval nucleus. Embryo. Single as in S. thompsoni. Distribution of 5. gerlachei. Confined to the high latitudes of the Pacific Sector of the Southern Ocean. Not circumpolar in its distribution. KEY TO THE SPECIES (The diagnostic characters are summarized in Table i) Aggregate 1. Test smooth; total fibre number 40-61 (mean 51-8). S. fusiformis Test serrated; total fibre number more than 113. 2 2. Muscles IV and V approach but not in contact laterally. S. aspera Muscles IV and V in contact laterally. -, 3. Muscle bands broad; total fibre number 140-235 (mean 1777) (ratio of body length:muscle width, 18-9: i). S. thompsoni Muscle bands narrow; total fibre number 1 13-159 (mean 128-8) (ratio of body length : muscle width 27-6:1). S. gerlachei Solitary 1. Test smooth; muscle fibres of M. IV 19-40 (mean 29-4). S. fusiformis Test serrated; muscle fibres of M. IV 36-130. 2 2. Muscle bands VIII and IX parallel (or nearly so); test ridges 'simple' (see p. 10). S. aspera Muscle band VIII and IX joined; test ridges 'complex' (see p. 12). 3 3. Muscle bands broad; fibres of M. IV 70-130 (mean 93-2) (ratio of body length : muscle width, 21-2: i). iS. thompsoni Muscle bands narrow; fibres of M. IV 36-71 (mean 48-8) (ratio of body length : muscle width 35-8: i). S. gerlachei MORPHOLOGICAL CHARACTERS OF TAXONOMIC IMPORTANCE It is proposed in this section to deal with those aspects of the morphology that have proved to be of importance in understanding the taxonomy of S. fusiformis and its related species. For a detailed account of the morphology of the Salpidae in general reference should be made to the monographs of Ihle (1937-58). In the present study variations in the structure and arrangement of tw^o features, the external test and the body musculature, have been found useful in distinguishing the different species, but it is not suggested that these same characters will necessarily be of equal significance in the other genera. The Number of Fibres per Muscle Band The muscle bands so characteristic of the Salpidae are composed of muscle fibres arranged parallel to each other and to the main axis of the muscle, and in many species the number per muscle can be counted using a low-power binocular microscope. Little significance has been attached to muscle- fibre number and only a few workers have considered it to be of taxonomic importance. Apstein (1906) distinguished between the aggregate forms of S. mucronata {= T. detnocratica) and S. flagellifera {= T. longicauda) on the basis of the number of fibres per muscle which was found to be charac- teristically different in the two species. Tokioka (1937) described from Japanese waters Thalia demo- cratica var. orientalis using the form of the test and the high number of fibres per muscle to distinguish it from the solitary form of T. detnocratica. Sewell (1953) discusses T. democratica and its related forms in some detail and gives significantly different fibre counts for the type and for T. democratica i6 DISCOVERY REPORTS var. orientalis. He concludes (p. 28) that in the soUtary form of T. democratica there is a definite tendency for the number of fibres per muscle to increase with age although his data (p. 32) show no such increase in the aggregate form nor in the solitary form of T. democratica f. orientalis (p. 36). Further data for specimens of both type and variety of T. democratica from waters of Algeria are given by Bernard (1958) which confirm the results of Tokioka (1937) and Sewell (1953). Berner (1954) has drawn attention to the marked difference in fibre number in Ritteriella picteti and R. ambionensis, two species which otherwise appear very similar. Berrill (1950) has discussed fibre number in general and has shown (p. 592) that the aggregate form of a species not only has fewer muscle bands than the solitary form but that they are individually narrower owing mainly to differences in the number of constituent fibres and in part to differences in the diameter of individual fibres. His data show not only that the solitary form may have three to four times as many fibres per muscle as its respective aggregate form but, what is more important, that there may be a considerable difference in the counts for different species. The following are some of the data given for individual muscle bands : S. fusiformis solitary, 30-33 fibres aggregate, 7 fibres T. democratica solitary, 10 fibres aggregate, 3 fibres S. maxima solitary, 55-65 fibres aggregate, 20 fibres Furthermore, in every species examined by Berrill (p. 593) the number of fibre rudiments first discernible in the embryo or bud is also the final number to be found in the largest solitary or aggre- gate individual; in other words there is no increase in number with age. In the course of checking Berrill's counts, specimens from the Southern Ocean, thought to be iS. fusiformis aspera, were examined and it was immediately apparent that both aggregate and solitary forms had at least three times more fibres per muscle band than the number given for S. fusiformis by Berrill. From the subsequent examination of specimens from as wide a geographic range as possible it was obvious that on the basis of fibre count alone three of the four species within the fusiformis group could be distinguished. The data for the aggregate and solitary forms of S. fusiformis, S. aspera, S. thompsoni and S. gerlachei are shown in Text-figs. 3 and 4, in which the fibre count for each specimen is plotted against its body length. Before discussing these results it is necessary to consider the treatment of the data, and the reasons for using the number of fibres in M. IV in the solitary form and the sum of the counts of M. I-VI in the aggregate form. Specimens of the solitary form of all species, except S. fusiformis, in which all the nine body muscles were sufficiently well preserved to enable each to be counted, were comparatively rare. Thus while it was realized, as the data in Table 2 show, that the number of fibres varies slightly from muscle to muscle and even from side to side, it was decided to count the number of fibres in one muscle only of each specimen. Body muscle IV proved convenient for this purpose as it is easy to locate and counts were made on the dorsal surface as near to the middle line as possible. A further reason for using one muscle is that S. thompsoni has a very high fibre count, ranging from 70 to 130 fibres per muscle, which would make the counting of all nine muscles in each specimen unduly tedious and quite unnecessary for specific identification. For the aggregate form total counts, that is the sum of the number of fibres in each body muscle (counted along the same side if possible), are used because, as the data in Table 2 show, there is considerable variation in fibre number from muscle to muscle. While such variation is insufficient to mask any difference based on single muscle counts between S. fusiformis and the other species, it is great enough to make the differentiation of 5. thompsoni and S. gerlachei obscure. MORPHOLOGICAL CHARACTERS OF TAXONOMIC IMPORTANCE 17 In Text-fig. 3 a and b individual counts for both forms of S. fusiformis and S. aspera are compared while the mean values are given in Table 3. Compared to the rather limited data for S. aspera the fibre counts for S. fusiformis show relatively little scatter about the mean of 51 -Si 1-4 for the aggregate and 29-4± 1-4 for the solitary form. It is also evident, for all species, but particularly in the case of S. fusiformis, that there is no increase in the number of fibres per muscle as the animal grows in size. The number of observations for S. aspera, particularly of the solitary form, are limited and so the means of 194-1 for aggregate and 1007 for solitary forms must be considered as approximate. Even so there is no overlap between the counts for S. fusiformis and S. aspera and the difference between Table 2. Individual muscle-fibre counts in the solitary and aggregate forms of three of the four species showing (a) the variation in fibre number between muscles, (b) the variation in fibre number between the left and right sides of the same muscle in Salpa fusiformis and (c) the variation in the number of muscle fibres from species to species. (S. aspera is not included because counts of individual muscles are not available for the solitary form) Body Body muscle length (mm.) A St. f I II III IV V VI VII VIII IX Tota Solitary S. fusiformis 3097 40 Left 36 29 30 30 31 28 36 34 28 282 Right 31 31 31 27 32 34 33 37 28 284 S. thompsoni 1392 35 Left 60 68 68 82 82 77 83 70 70 660 S. gerlachei 2226 65 Right 35 31 38 40 34 29 28 35 36 306 Aggregate S. fusiformis 3097 22 Left II 5 10 9 9 II — — — 55 Right 12 7 7 9 9 II — — — 55 S. thompsoni 2020 20 Right 37 28 28 27 29 36 — — — 185 S. gerlachei 2226 17 Left 26 22 19 19 20 21 — — — 127 Table 3. The variation in the mean number of muscle fibres in the four species, with the standard deviation and number of observations for each species. Standard deviations have not been given for S. aspera because of insufficient data for the solitary form Mean number of Standard muscle fibres deviation SoHtary (M. IV) 294 4-8 100-7 — 93-2 13-8 48-8 7-3 Aggregate (M. I-VI) 51-8 4-9 1941 — 1777 187 128-8 II-3 S. fusiformis S. aspera S. thompsoni S. gerlachei Number of observations 48 9 92 67 S. fusiformis S. aspera S. thompsoni S. gerlachei 45 29 126 50 them as shown in Text-fig. 3 and from the data of Table 3 is most striking. Text-fig. 3 also illustrates the great difference in the maximum size attained in the two species, at least as represented by the specimens available, S. aspera at its maximum being two to three times larger than S. fusiformis at comparable stages of maturity. Text-fig. 4 shows the data for S. thompsoni and S. gerlachei. Comparing Text-fig. 4 with Text-fig. 3 it will be seen that S. fusiformis in both its forms has fewer fibres per muscle than any of the other species, and does not attain the relatively large size of the other species. The data plotted in Text-fig. 4 for both forms of S. thompsoni and S. gerlachei show considerable scatter in the observations. Never- i8 DISCOVERY REPORTS theless, the data fall into two well-defined groups, one for S. thompsoni with a mean value of 177-7 ±3*4 for the aggregate and 93-2 ±2-8 for the solitary, and the other for S.gerlachei, with means of 128-8 ±3-2 and 48-8 ± 1-8 for aggregate and solitary, respectively. The difference between the species is even more obvious from Text-fig. 5 based on the data of Table 4 which gives the percentage frequency of occurrence of the various muscle fibres for the two species, and, for comparison, the data for S.fusi- formis. Text-fig. 5 shows that in the solitary form S. thompsoni and S. gerlachei are quite distinct. (a) 300. f2 200 O a: 100 O I50 0 . AGGREGATE ° - 0 0 0 0° 00 0 Oq 0 0 0 80 0 ■ 0 g 0 0 00 0 ; •• : .Ur«w*f • 0 S ospera - • S fusiformis 10 20 30 40 BODY LENGTH (mms) 50 60 > (b) : O . SOLITARY 100 0 0 0 0 0 0 0 0 0 50 ■ 0 '^J^''"' ' 0 S, aspero • S, fusiformis 10 20 30 40 120 50 60 70 80 90 100 110 BODY LENGTH (mms) Text-fig. 3. Variation in the number of muscle fibres in S. fusiformis and S. aspera; (a) the aggregate form (M. I-VI, see text) and (b) the solitary form (M. IV), plotted against body length. Although there is overlap between the data for the solitary forms of S. gerlachei and S. fusiformis there can be little confusion between them when other features (such as the character of the external test) are considered and particularly since they occur, as will be shown later, at the opposite ends of the range of distribution of this group of four species (see Text-fig. 6). The data for the aggregate forms of S. thompsoni and S. gerlachei in Text-figs. 4a and 5 show that the diflFerence between them is not so marked as it is in the solitary form. Nevertheless, two distinct groups can be seen with means of 177-7 ^^'^ 128-8 for S. thompsoni and S. gerlachei, respectively, (Table 3). It will be noted both from Table 3 and Text-fig. 5 that S. fusiformis with a mean of 51-8 fibres for M. I-VI cannot be confused in its aggregate form with either 5. thompsoni, S. gerlachei or S. aspera. (a) MORPHOLOGICAL CHARACTERS OF TAXONOMIC IMPORTANCE 300, ^ 200 a: CO u_ o UJ CO i lOo I50 !2 lOO (b) == CO § 50 AGGREGATE O O „ °o o o o 0S° °o ° OO ° 'oieggooo o 8 8°%, ^^ 808 oo @@LooP^fi°°88°^° 8o| 8^ •DoS*Q°°8t« ° o®oo ® o ° °ooi •: •.»:'! oOo ^ o o c O S. thompsofii • S. qerlachei lO 20 30 40 BODY LENGTH (mms) 50 SOLITARY o o cfco go o o o"o I (P<5 o 0 o ° ^°9°^ 8 § o § ,®o^o°o88oOooo *: .iM'TrJ:* O S. fhompsoni • S. qerlachei O lO 20 30 40 _J 1_ -J L- 60 I20 19 50 60 70 80 90 lOO IIO BODY LENGTH, cmms) Text-fig. 4. Variation in the number of muscle fibres in S. thompsoni and S. gerlachei; (a) the aggregate form (M. I-VI) and (6) the solitary form (M. IV) plotted against body length. Table 4. The percentage frequency of occurrence of muscle fibre counts in the solitary and aggregate forms of Salpa fusiformis, S. thompsoni and S. gerlachei ; based on the data plotted in Text-figs. 3 and 4. Fibre no. 10- 20- 20 30 30- 40 70- 80 40— 50— 60- 50 60 70 Solitary (M. IV) S. fusiformis Number of specimens 2 26 24 i — — — — % 3-8 49-1 45-4 i-Q — — — — " S. gerlachei Number of specimens — — 5 40 23 3 2 — 6-8 54-8 31-5 4-1 2-7 — 5. thompsoni Number of specimens • — ■ — — — — — 25 36 47 % ______ 18-5 26-6 34' 80- 90- 100- no- 120- 130- 90 100 no 120 130 140 13 10 9-6 7-4 I 07 3 2-2 Fibre no. 40- so- so 60 31 S. fusiformis Number of 1 2 specimens % 26-7 69-0 S. gerlachei Number of — — specimens 0/ /o S. thompsoni Number of — — specimens 0/ . /o 60- 70 4-4 1 10- 120 120- 130- 140- 150- 160- 170- 180- 190- 200- 210- 220- 230- 130 140 ISO 160 170 180 190 200 210 220 230 240 Aggregate (M. I-VI) — 7 17 27 25 ■S 20-7 32-9 30-5 — — — II 7-3 — 16 29 34 31 16 13 S 7-0 IO-2 18-5 21-6 19-7 IO-2 8-3 3'2 I I 0-6 0-6 3-2 DISCOVERY REPORTS ISO 50 NUMBER OF FIBRES Text-fig. 5. The percentage frequency of occurrence of muscle-fibre counts in the aggregate and solitary forms of 5. fusiformis (black solid), S. thompsoni (sohd line) and S. gerlachei (broken line). 125 100 S 75 25-, 30 • • • • • - ° .5 ° : 0° ••!• t 1. • •• • - • :: • • " 839 a - A^ a 9 A r r 1 1 < 1 1 35 40 50 55 DEGREES SOUTH bO 65 70 Text-fig. 6. Individual muscle-fibre counts of specimens of the sohtary form of S. fusiformis (A), S. aspera{o), S. thompsoni (•) and 5. gerlachei (a) plotted according to the latitude of the station at which they were taken. Note: observations for S. fust- j formis and S. aspera at stations north of 30° S. are not included. The data for S. aspera are insufficient to show the percentage frequency of occurrence as has been done for the other species. However, when the data plotted in Text-fig. 3 are compared with those of Text-fig. 4 it is seen that in fibre number S. aspera Hes roughly within the same range as S. thomp- soni. It is thus impossible to distinguish these two species on the basis of their muscle-fibre number but it will be remembered (Table i, pp. 14 and 15) that they can be distinguished in both aggregate and solitary form by differences in their muscle arrangement, and in addition, in the solitary form, by external diiferences in serrations of the test. As mentioned above there is overlap between the fibre counts for some of the species, as, for instance, between the solitary forms of S. fusiformis, S. thompsoni and 5. gerlachei (Text-fig. 5) and it might be argued that 5. gerlachei, characterized by fewer muscle fibres and thinner muscles (see MORPHOLOGICAL CHARACTERS OF TAXONOMIC IMPORTANCE 21 p. 23) than S. thompsoni but otherwise identical in all other morphological features, represents an extreme of variation. On this assumption one might expect there to be some sort of gradation in muscle fibre number between the limits of the animal's distribution. In Text-fig. 6 the data (for stations south of 30° latitude) for the solitary form of each species have been plotted against the latitude of the station at which they were taken, and it is apparent that no such gradient exists in the character of muscle-fibre number from north to south, the change from ' many fibres ' to ' few fibres ' being particularly abrupt. It will be noted that S. aspera on rather limited data has similar counts to S. thompsoni and so, as mentioned above, this character cannot be used to distinguish these two species. The Arrangement of the Body Muscles Unlike the external character of the test great reliance has been placed by many previous workers on muscle arrangement for differentiating salp species, and Streiff (1908) and others have given detailed accounts of the body, atrial and oral muscles of most species. While the arrangement of the atrial and oral muscles is both complex and difficult to study, the body muscles are relatively simple and easy to see, so making them useful as taxonomic characters particularly as their arrangement within a species is more or less constant. Sewell (1926, p. 67) has remarked that large variations do occur both with regard to the number of muscle bands and the connexions between them, and warns that the muscles are delicate and so their arrangement can be altered by rough handling. As an example Thompson (1948, PI. 73, fig. 3) shows the variation that can occur in the connexions of the body musculature of the solitary form of S. cylindrica. Nevertheless, in most species muscle arrangement may be combined with other characters, and it remains one of the most useful taxonomic features (Yount, 1954). In three of the four species described, S.fusiformis, S. thompsoni and S. gerlachei, there is the same basic arrangement of body muscles. In the aggregate form there are six dorsal body muscles which curve laterally to the ventral surface where they are interrupted. The six muscles are disposed in two groups with dorsal fusion of M. I-IV and M. V-VI ; and lateral fusion of M. IV and V. The solitary form has nine body muscles which like the aggregate are interrupted ventrally. There is dorsal fusion of M. I-III and M. VIII-IX, while M. IV-VII are parallel to each other. S. aspera (PI. I) has the same basic muscle arrangement as that just described except that in the aggregate form (PI. I b) M. IV-V converge but do not join laterally and in the solitary form (PI. la, c) M. VIII-IX do not join but like M. IV-VII are parallel or nearly so. It will be shown in the section on distribution that S.fusiformis and S. aspera are sympatric and so on occasions are taken together in the same net sample. From the foregoing it will be realized that setting aside other differences such as test character and muscle-fibre number there is no striking difference between S. fusiformis and S. aspera, except in the arrangement of body muscles in the solitary form, and as the two species can occur together this has contributed to the confusion, particularly with regard to the aggregate form of the two species, described in the historical section. Herdman (1888), for example, differentiated the solitary form of S. echinata on the basis of serrated test and M. VIII-IX being parallel, but failed to associate it with specimens of the aggregate form in his collection in which M. IV-V did not join laterally, regarding the latter as variations of S. runcinata {= S. fusiformis). There has m fact been little significance attached to the lateral separation of M. IV-V in the aggregate form, such separation being regarded as a mere variant (Herdman, 1888) or a character associated with old age (Ritter, 1905) : the latter in spite of Apstein's illustration (18946, PI. II, fig. 14) of a small blastozooid (4 mm. long) with M. IV-V laterally separate taken from the stolon of a solitary form in which M. VIII-IX were parallel. 22 DISCOVERY REPORTS There seems little doubt, however, in view of the evidence presented in this paper, that previous records of the ' asperate ' form of S. fusiformis from temperate or tropic seas are of the same species as the one that I have renamed S. aspera, because in every paper in which such serrated specimens of the aggregate form are figured they are shown with a muscle plan in which M. IV-V do not join laterally (for example, Herdman, 1888, PL VI, fig. 6; Apstein, 1906, PI. XXVI, fig. 6; Yount, 1954, fig. \2b). Furthermore, where a serrated solitary form is figured it is shown with M. VIII-IX parallel or nearly so (Herdman, 1888, PI. V, figs. 1-4; Ritter, 1905, fig. 14; Yount, 1954, fig. iia). There are, however, in the literature, three notable exceptions. Herdman (1888, PI. VI, fig. 5) figures a serrated specimen of the aggregate form in which M. IV-V join laterally, Apstein (1906, PI. XXVI, fig. 6) shows a serrated solitary form with M. VIII-IX joined medio-dorsally, while Metcalf (1918, fig. 80) figures a different view of what appears to be a similar specimen to that of Apstein. Herdman's and Apstein's figures, however, are of specimens from the Antarctic and so from the data presented in this paper conform not to S. aspera but to S. thompsoni. The figure in Metcalf 's paper is not original, being based on that of Ritter (1905, fig. 14) and it seems not unlikely that in it are combined the characters of test given by Ritter (for Californian specimens) with the plan of muscle arrangement given by Apstein (1906, PL XXVI, fig. 6) for an antarctic specimen. Such an association of characters has not appeared in any of the specimens examined in this study. The External Character of the Test The presence of serration or denticulations on the external surface of the test of some species of salps has been noted by many workers and has in some cases been used as a character for the separa- tion of new species, subspecies or variants; for example, T. democratica var. orientalis Tokioka (1937). S- fnaxima var. tiiberculata (Metcalf, 1918) and S. fusiformis aspera (Ihle, 191 1). Sewell (1926, 1953) discusses the occurrence of such denticulations at some length both in the Salpida and Pyrosomida and concludes that, owing to the complete intergradation between smooth and serrated forms, they are of little importance as characters for systematic differentiation, and attributes the variability of this feature to the great degree of plasticity inherent in the Salpida (see also Metcalf, 1918, p. 5). The evidence of Yount (1958) would seem to confirm this view and in the case of S. fusiformis serrated and smooth forms are listed as the one species. The results of the present study are contrary to the conclusions of Sewell and Yount, at least in so far as they relate to S. fusiformis and its related species. It was noted at the outset without the use of any special technique that two basic test forms could be recognized, one smooth and the other serrated and while this character alone does not allow all the four species to be distinguished it can be combined with other features and used for their identification. Detailed examination of the external surface of the test shows, as Stiasny (1926) has described, the presence of areas of thickening which give it a ridged appearance. In species with serrated tests the denticulations occur along these ridges, the arrangement of which in the solitary form in particular follows a definite pattern, the finer detail of which is easily visible when the test is stained with toluidine blue or methylene blue (see p. 5). As the serrations are a feature of the external surface of the test they are subject to wear, and with increased age tend to become smooth, particularly in the aggregate form. Staining, however, allows the remnants of the serrated areas to be detected in specimens which to the casual observer might appear smooth. The aggregate form of S. fusiformis (p. 9) has a smooth test devoid of any serrations and so can be distinguished from the three serrated species. It is not possible, however, in the aggregate form to differentiate the serrated species from each other on the basis of the external test, since the pattern of arrangement is less definite and possibly more subject to variation. Of the species, the two which MORPHOLOGICAL CHARACTERS OF TAXONOMIC IMPORTANCE 23 occur in the Southern Ocean, S. thompsoni and S. gerlachet, are typically more heavily serrated than 5.<7i'/)^a, having the arrangement of serrated ridges shown in PL II 6 and Text-fig. 2b. In the aggregate form of S. aspera (PI. lb) the serrations appear to be irregularly scattered over the test with a tendency to be well developed on the anterior and posterior processes. The solitary, like the aggregate, form of S. fusiformts has a smooth test with a pattern of thickened ridges similar to that of S. aspera (PI. Ic), except of course for the absence of serrations on all ridges, and is identical with the descriptions given by Herdman (1888) and Ritter (1905). A basically similar arrangement of serrated ridges is found in S. thompsoni (PI. II a, c) and S. gerlachet (Text-fig. 2 a) except for a more elaborate pattern on the dorsal surface where the secondary ridges, and particularly the presence of a medio-dorsal ridge Cridge no. 9 in PI. lie), distinguish these two species from S. aspera. It is thus possible, using the external character of the test of the solitary form, to distinguish between S. fusiformts, S. aspera and S. thompsorti or S. gerlachei. Other characters, however, must be used to distinguish between S. thompsoni and S. gerlachei. It has often been suggested that serrations are associated with age. Both Ritter (1905) and Farran (1906), for example, conclude that in S. fusiformis aspera serrations occur in the larger, older speci- mens. Apstein (18946), on the other hand, describes serrations in the smaller, younger specimens of Thalia democratica. Sewell (1953), however, considers that 'such a condition of the test may occur at any age up to mid-life, after which they tend to be eroded away'. In the present study it has been possible to examine specimens of at least one serrated species, S. thompsoni, in every stage of its development and so it is possible to make some generalizations about the occurrence of denticulations. Very young aggregates (3-4 mm. long) while still part of the stolon of the solitary form have a thin poorly developed test which shows no external sign of serrations. The smallest released free-living aggregates, 5 mm. in length, however, have a well-developed test characteristically serrated. It thus seems reasonable to infer that when a block or chain of aggregates is released from the solitary form there is very rapid test development which in S. thompsoni and S. gerlachei and probably also S. aspera is accompanied by the formation of serrated ridges. In larger, older specimens as Sewell (1953) described, the denticulations tend to wear away but in the largest specimens of the aggregate form of each serrated species examined some serrations were quite obvious. The solitary form, while still an embryo within the oozoid (aggregate), has a poorly developed test with no indication of denticulations, even though it may, before its release, attain a size of 9-10 mm., having all the morphological features of the adult, including a rudimentary stolon, already developed within it. In free-living solitary individuals in the 10-15 ^^- ^i^^ range the characteristic pattern of serrated ridges can be seen, so that as with the aggregate form one can infer that serrations develop very rapidly on the release of the embryo. Larger specimens often show evidence of a worn test, but in all cases examined it was possible to see the pattern of serrations and so distinguish, for example, S. aspera from S. thompsoni or 5. gerlachei. The Width of the Body Muscles in Salpa thompsoni and S. gerlachei It will be realized from the previous sections that the only precise taxonomic difference between S. thompsoni and S. gerlachei is in the number of fibres per muscle band. Other characters, however, which are less easy to define in precise terms, make it possible readily to distinguish these two species. One such character is the difference in relative width of the muscle bands in specimens of approxi- mately the same size. Contained in the preliminary analysis of the samples collected by R.R.S. 'Discovery IF in the Southern Ocean are reports of the occurrence of specimens of ' S. fusiformis aspera ' with ' narrow or thin muscles ', such comments being restricted to those samples taken in high latitude Pacific waters. ,. DISCOVERY REPORTS In this respect it is interesting to note that Herdman (1910, p. 21) records a specimen of the soUtary form taken at 67° 5' S., 179° 30' E. which had muscle bands narrower than those of his other speci- mens. Subsequent examination of the Discovery material has shown such specimens to be S. gerlachei and it thus seemed worthwhile to measure the width of the muscles in S. thompsoni and S. gerlachei and see whether there was any significant difference between them. For this purpose a micrometer eyepiece was used and a measurement was made of body muscle IV in both the aggregate and solitary form of each species. The data are shown in Text-figs. 7 and 8. It must be remembered that muscle width, unlike the muscle-fibre number, which remains constant throughout the life of the animal, 4-5 40 3-5. ' JO i ■ 25 u CO i20 1-5. 10 OS. SOLITARY I S thompsoni I S geriochei O O o o o' o ,0 o, 8 o o ^ 08 o, 'o o o o© o On „ O O O OOg Oq o 00 00 00 Oo g o „o O O Oo , . t. • • • qOOoo^ °* • • oo o" oo • • • ".» o o o o o t ••I o o o o o o o o o o o o 00 °ooOo 8 0.0 10 20 25 30 35 40 45 50 55 faO BODY LENGTH. Onms 65 70 75 80 65 90 95 100 Text-fig. 7. Variation in the width of M. IV with body length in the soUtary forms of 5. thompsoni and S. gerlachei. 40 Q 5 35 30. 25 20 15 10 OS AGGREGATE O 8 o -o o O 0° °0 0^0 §0 «oo o o o o ° Q ° |8|S .a^-r* o o o o OnO O S. thompsoni • S geriochei 10 15 20 25 BODY LENGTH onms) 30 35 40 45 50 55 60 Text-fig. 8. Variation in the width of M. IV with body length in the aggregate forms of 5. thompsoni and S. gerlachei. MORPHOLOGICAL CHARACTERS OF TAXONOMIC IMPORTANCE as increases as the animal grows in size. As previously mentioned (p. 5) body length cannot be measured in the larger specimens with great accuracy and this no doubt contributes to the considerable scatter of the data plotted in Text-figs. 7 and 8. Even allowing for this scatter, particularly in the data for S. thompsoni, it is obvious from Text-fig. 7 that in the solitary form S. gerlachei has much thinner muscle bands than 5. thompsoni. Text-fig. 8, in which data for the aggregate form of each of the two species has been plotted, show that there is again diflFerentiation between S. thompsoni with wide muscle bands and S. gerlachei with thin muscle bands, although the difference is not so marked as it is in the solitary form (see Text- fig. 7). From Text-fig. 8 it is obvious that there is considerable scatter in the observations which makes muscle width in the aggregate form a character of doubtful value for diflferentiating between the two species, even though the data for S. gerlachei consistently fall in the lower part of the graph. Table 5. The variation in muscle width in Salpa thompsoni and S. gerlachei expressed as the ratio body length: muscle width, together with the number of observations. Mean ratio of body length: muscle width Number of observations Solitary S. thompsoni S. gerlachei 21-2:1 35-8: 1 Aggregate 57 53 S. thompsoni S. gerlachei 18-9:1 27-6:1 33 57 The results plotted in Text-figs. 7 and 8 may be expressed as numerical ratios of body length and muscle width (Table 5) when it is seen that there are significant diflFerences between the ratios for the solitary forms of the two species. As mentioned above, owing to variability in the data, the ratios for the aggregate form cannot be considered to be reliable. The Development of the Stolon in Salpa thompsoni and S. gerlachei A consideration of stolon development and budding in the solitary form is outside the scope of this paper and will be described in a later work. There are, however, certain differences between S. thomp- soni and S. gerlachei, both in the size of the chains of buds and the size of the parent solitary individual at the time of release which, with the morphological differences already described, lend support to the view that they are in fact valid species. These differences will be discussed briefly here. In the stolon of a mature solitary individual, that is, one in which a chain of aggregates (blastozooids) is about to be released, three zones or blocks of bud formation can be distinguished, the distal two being fully differentiated into a number of segments or buds which can be counted. When the oldest distal block, which I shall refer to as block i , reaches a certain size it is shed in its entirety through a hole in the test so leaving a section of empty tube in the test. In the species to be described the hole formed by the release of a block is sealed off by a tissue ' strand ' of granular appearance which remains as evidence that one block of buds at least has been released. The section of empty tube in the test which was occupied by block i is quickly filled by the growth of block 2 which thus becomes the distal block. It is probable that this block is subsequently released and replaced by block 3 which may in turn on its release be replaced by another block. It will be realized from the foregoing that while it is possible, from the absence of a strand of granular tissue in the test, to deduce with certainty that no blocks have been released it is not possible to say, in the presence of a strand, whether the 26 DISCOVERY REPORTS distal block is block 2, 3 or 4. Nevertheless, the presence or absence of this granular strand is a valuable indication of the relative maturity of the solitary form. The smallest specimens in which it occurs provide evidence of the minimum body size at which chains of buds can be released, while the number of buds present in the distal block is an indication of the potential number of aggregates that can be released by one solitary individual at any one time. Data have been accumulated from the examinationof 287 specimens of 5. f^offz^ww? and 57 specimens of S. gerlachei and the results are given in Table 6 and Text-fig. 9. Table 6 compares the mean number of buds in block i (no strand) and block 2 (strand) with the body size of the specimen on which the 300. s qertach ei Size group (mm.) Block I 80 100 o 20 40 60 8O BODY LENGTH Cmms) Text-fig. 9. Budding in 5. thompsoni and S. gerlachi. The mean number of buds produced in the first (shaded) and second (unshaded) blocks of S. thompsoni and S. gerlachei plotted against body length. Full explanation in text. Table 6. The relationship between body length and budding in Salpa gerlachei and S. thompsoni. For full explanation see text. Total no. of 20-3030-40 40-50 50-60 60-70 70-80 80-90 90-100 loo-iio specimens S. gerlachei Mean number of buds 49-0 27-0 28-0 — — — — — — — Number obsv. 2 2 2 — — — — — — 6 Block 2 Mean number of buds — i49'2 198-6 2236 229-0 240-0 240-0 — — — Number obsv. — ^ 5 20 14 8 3 i — — 51 iS. thompsoni Block I Mean number of buds — 133-2 133-6 142-2 126-0 141-6 — — — — Number obsv. — 17 57 65 53 13 — — — 205 Block 2 Mean number of buds — — — 174-6 i8i-2 207-2 2170 297-2 262-0 — Number obsv. — — — 3 18 31 19 7 4 82 Mean number of buds in block i. S. gerlachei: 380 (6 obs.). Mean number of buds in block i. S. thompsoni: 134-8 (205 obs.). count was made, the data being treated in size groups of 10 mm. interval. The data for S. gerlachei are limited to specimens taken in a period from November to February and to eliminate seasonal differences, if any, the S. thompsoni specimens are also from stations in these months. In plotting the data in Text-fig. 9 it has been assumed that the distal block in those specimens with a strand is in fact block 2 ; it is obvious from the results that this is probably not a valid assumption, particularly for S. thompsoni; nevertheless, it in no way invalidates the argument that follows. The data in Table 6 and Text-fig. 9 show that there are a number of obvious differences between S. thompsoni and S. gerlachei both in the number of buds initially produced — block i — and in the size of the solitary form when this block is released. MORPHOLOGICAL CHARACTERS OF TAXONOMIC IMPORTANCE 27 In S. thotupsom the first fully differentiated block of buds occurs in the 30-40 mm. size range and may be released, as indicated by the presence of a strand, at 50-60 mm. although release may not occur until the animal is 70-80 mm. long. The number of buds in block i is remarkably constant in each size group with a mean, based on 205 specimens, of 134-8 buds. It will be noted that subsequent blocks, as shown by the data from specimens with strands (block 2, in the figure) have a significantly greater number of buds than block i , which suggests that the animal's capacity to produce buds increases as it grows. This view is further supported by the very high mean values of 297-2 and 262-0 buds per block for animals in the 90-100 mm. and loo-i 10 mm. size range; these blocks, as mentioned above, probably being blocks 3 or 4 — in other words those produced subsequently to the release of block 2. In S. gerlachei, block i is fully segmented at a body size of 20-30 mm. compared with 30-40 mm. in 5. thompsoni, while it is released at a size of 30-40 mm. compared with 50-60 mm. in the other species. Block i in S. gerlachei has strikingly fewer buds than 5. thompsoni with an average of 38-0 buds per block, although it should be noted that this is based on only six specimens. Subsequent blocks have about the same mean number of buds as in specimens of S. thompsoni of comparable body length. The data thus show that compared with S. thompsoni, S. gerlachei produces initially in block i significantly fewer buds which are released when the solitary form is relatively much smaller. These differences in asexual development might be attributed to environmental change, the specimens being geographical races of the same species. In view, however, of the differences in morphological character that are apparent in both aggregate and solitary forms it is assumed that they are specific differences. DISTRIBUTION Any consideration of the distribution of the four species described in this paper is limited by the geographical range of the stations from which samples were available for study. As mentioned previously (p. 5) most of the stations are from the Southern Ocean and only a relatively few lines of observations extend north of the subtropical convergence. These latter stations have been supple- mented by material made available by other workers which, however, as a consequence are rather scattered in geographical location. Nevertheless, in spite of the relative paucity of data from areas other than the Southern Ocean, an attempt has been made in Text-fig. 10 to show the occurrence of the four species on a world chart using all the stations from which material has been personally examined. Text-fig. 10 shows the positions at which the various species have been recorded, each species being indicated by a different symbol. Also indicated in the figure is the mean position of the subtropical convergence based on Deacon (1937). With regard to this convergence it should be remembered that although as a physical-chemical feature it is well defined it is more variable in position than the antarctic convergence ; and, furthermore, its position in some areas, particularly the eastern Pacific, is largely conjectural. From Text-fig. 10 it is obvious that the region of the subtropical convergence represents a faunistic boundary, at which two of the species, S. fusiformis and S. aspera, reach the southern limit of their distribution. For S. thompsoni it is the northern limit to its range. Records of these species either north or south of what is apparently their normal range do occur, but in Text-fig. 10 the mean position of the subtropical convergence is shown and if its contemporary position could have been plotted for each crossing by a line of stations its appearance as a boundary would be even more clearly defined. Text-fig. 10 shows that both S. fusiformis and 5. aspera have been taken north of the subtropical convergence in Pacific, Atlantic and Indian Oceans. It is not possible from th^ data available to define the northern limit to the distribution of either species, ahhough in the present material, S. fusiformis has been recorded as far north as 63° 5 1 ' N., 16° 20' W. while S. aspera has been recorded at 37° 24' N. 4-2 DISTRIBUTION 29 123° 23' W. It will be seen from Text-fig. 10 that S. thompsoni is typical of other Southern Ocean plankton in being circumpolar in its distribution (Baker, 1954). S. gerlachei, on the other hand, is restricted to the Pacific, occurring with one exception in a sector between 65° S. and the ice edge roughly between 175° E. and 80° W. The distribution of this species compared to other antarctic species is thus remarkable in not being circumpolar. An apparent anomaly in distribution occurs in the region of 150° W. where there are a number of records of S. gerlachei to the north of 65° S. Speci- mens of the solitary form were taken in 1951 at stations 2743 and 2744, and as an independent check, I was able to examine specimens taken on the 1949 Brategg Expedition where S. gerlachei occurred at two stations which were nearly coincident in position to the 'Discovery' stations. Hydrological observations in this area (Deacon, 1937, p. 31 ; Midttun and Natvig, 1957) suggest that there may be a northerly outflow of surface water from the Ross Sea similar to that of the Weddell Sea current in the Atlantic. Such a current system might account for the northward extension to the animals' distribution that is a feature of this area. With the exception of these observations, S. gerlachei ranges from the ice edge to about 65° S. and so occurs principally in an area of the West Wind Drift but possibly extending in the south into the East Wind Drift. How this species has evolved and is able to maintain itself in this area is not clear, since no known hydrological feature appears to serve as a boundary between its distribution and that of S. thompsoni. Waters with the same characteristics of temperature and salinity occur in other sectors of the Southern Ocean and yet S. gerlachei occurs only in the Pacific. It is also remarkable that S. gerlachei is not taken either in the Drake Passage or in the Bransfield Strait where there is interchange of water between the Pacific and the Atlantic (Deacon, 1937). This area particularly the Bransfield Strait is in fact frequently rich in salps, but at all the stations from which specimens have been examined, only S. thompsoni occurred. With regard to the boundary between S. thompsoni and S. gerlachei it will be remembered (Text- fig. 6, p. 20) that the change in fibre count from one species to the other is abrupt, particularly when it is realized that in Text-fig. 6 the occurrences of S. thompsoni at positions south of 65° are without exception from sectors other than the Pacific. Mixing of the two species is of course possible and at station 1295 specimens of the solitary form of both species were taken in the same haul. They were, however, easily recognized using the characters described and presumably such occurrences can be attributed to localized water mixing. While a consideration of the quantitative distribution of S. thompsoni and S. gerlachei is outside the scope of this present paper, and will be described later, it is worth noting that both species are taken at certain times of the year in great numbers as, for example, at station 1261 where 66,000 S. gerlachei were estimated to have been taken in a 20 min. (NiooB) haul which compares with the richest hauls of S. thompsoni taken in more northerly waters. SUMMARY 1. Within the 'fusiformis group' there has been a confusion of four species. Three of these have hitherto been variously known as Salpa fusiformis aspera to distinguish them from the type S. fusi- formis Cuvier. 2. Using a complex of morphological characters which includes the number of fibres per muscle band, the arrangement of the body muscles, and the external character of the test the four species can be identified. They are A^./ww/ormwCuvier, 5. aspera Chamisso,5'.fAoOTp^om'sp.nov., and iS.^er/acAezsp.nov. 3. Additional evidence based on variations in muscle width and stolon development supports the view that S. gerlachei should be distinguished from S. thompsoni and regarded as a separate species. 4. The southern limit to the range of distribution of both S. fusiformis and S. aspera is the sub- tropical convergence. The data do not allow the northern limit to be defined. 30 DISCOVERY REPORTS 5. S. thompsoni has its northern Umit of distribution at the subtropical convergence. It occurs in subantarctic and antarctic waters and is circumpolar in distribution. 6. S. gerlachei is an antarctic species restricted to high latitude waters of the Pacific sector and is remarkable in not having a circumpolar distribution. REFERENCES Apstein, C, 1894a. Die Salpen der Berliner Zoologischen Sammlung. Archiv. Naturgesch. Jahrg. 60, Bd. i, pp. 41-54. 18946. Die Thaliacen der Plankton-Expedition. Ergeb. der Plankton-Exp. der Humboldt- Stiftung. Bd. II. 1901. ' Salpidae, Salpen'. Nordisches Plankton, vol. i. Kiel and Leipzig. 1906. Salpen der deutschen Tiefsee Expedition. Wiss. Ergebn. Valdivia, vol. xii. Jena. 1908. Die Salpen der deutschen Siidpolar Expedition. Dtsch. Siidpol. Exped. vol. ix. Berlin. Baker, A. de C, 1954. The circumpolar continuity of antarctic plankton species. Discovery Rep. vol. xxvii, pp. 201-18. 1959- Distribution and Life History o/Euphausia triacantha Holt and Tattersall. Discovery Rep. vol. xxix, pp. 309-40. Beneden, E. Van and M. de Selys-Longchamps, 1913. Tuniciers. Caducichordata [Ascidiaces et Thaliaces). Expedition Belgica, 1897-99, Zool. Ill, pp. 3-122. Bernard, M., 1958. Systematique et Distribution Saisonniere des Tuniciers Pelagiques d' Alger. Rapports et Proces-Verbaux des Reunions, vol. xiv (Nouvelle serie), pp. 21 1-3 1. Berner, Leo., 1954. On the previously undescribed aggregate form of the Pelagic tunicate Ritteriella picteti (Apstein) 1904. Pacific Sci. vol. viii, no. 2, pp. 121-4. Berrill, N. J., 1950. Budding and development in Salpa. J. Morph. 87, pp. 553-606, 16 figs. BoMFORD, T. K., 1913. Some salps taken by R.I.M.S.S. 'Investigator' in the Bay of Bengal and Andaman Sea. Rec. Ind. Mus. vol. IX, Miscellanea, p. 243, part iv, no. 16. Calcutta. Brooks, W. K., 1893. The genus Salpa. With a supplementary paper by Maynard M. Metcalf. Johns Hopkins Univ., Mem. II, pp. 1-396, pis. I-LVII. Chamisso, A-VON, 1819. De Animalibus quisbusdam e classe Vermium Linnaeana. 1815-1818. Fasc. i. De Salpa, pp. 1-24, pi. I. Berolini. CuviER, G., 1804. Memoire sur les Thalides et sur les Biphores. Aim. Mus. Hist. Nat., An. xii. Tome 4, pp. 360-82, pi. 68. David, P. M., 1955. The distribution of Sagitta gazellae Ritter-Zahony. Discovery Rep. vol. xxvii, pp. 235-78. 1958. The distribution of the Chaetognatha of the Southern Ocean. 'Discovery' Rep. vol. XXix, pp. 199-228. Deacon, G. E. R., 1937. The hydrology of the Southern Ocean. 'Discovery' Rep. vol. xv, pp. 1-124. Fagetti, E., 1959. Salpas Collectadas Frente A Las Castas Central Y Norte De Chile. Rev. Biol, mar., Valparaiso, vol. ix, nos. 1-3, pp. 201-8, pi. VII. Farran, G. p., 1906. On the distribution of the Thaliacea and Pyrosoma in Irish waters. Fish. Ireland Sci. Invest. 1906. Pt. II, App. I. Dublin. FoRSKAL, Petrus, 1775. Descriptiones Animalium Avium, Amphibiorum, Piscium, Insectorum, Vermium; quae in itinera orientali observavit, pp. 112-17 (Salpa). Eraser, J. H., 1959. Plankton investigations from Aberdeen in i<)$'j. Ann. biol., Copenhague, vol. xiv (1957), pp. 29-30. Hardy, A. C. and Gunther, E. R., 1935. The plankton of the South Georgia whaling grounds and adjacent waters, ig26-ig2y . Discovery Rep. vol. XI, pp. 1-456. Herdman, W. a., 1888. Report on the Tunicata collected during the voyage of H.M.S. 'Challenger' during the years 1873-76. Part III. 'Challenger' Reports, Zoology, vol. xxvii, pp. 1-163. 1910. Tunicata. Nat. Antarctic ('Discovery') Exped. 1901-4. Nat. Hist. vol. v, pp. 1-26. Ihle, J. E. W., 1910. Die Thaliacean der Siboga Expedition. Siboga Expeditie, Monograph 56d. Leiden. 191 1. Ober die Nomenklatur der Salpen. Zool. Anz. vol. xxxviii, pp. 585-9. Leipzig. 191 2. Desmomyaria. Das Tierreich, vol. xxxii. Berlin. 1935- Desmomyaria. Handb. Zool. (Kuckenthal), Bd. V, 2. 1937-58. Salpidae. Bronn's Klassen. Band 3, Suppl. Tunikaten, Abt. 2, Buch 2, Lief. 2, 3, 4; pp. 69-401. Isaacs, John D. and Lewis, W. Kidd, 1953. Isaacs-Kidd midwater trawl. Scripps Institution of Oceanography. S.I.O. Ref., 53-3, pp. 1-18. Kemp, S., Hardy, A. C. and Mackintosh, N. A., 1929. The 'Discovery' investigations, objects, equipment and methods. Discovery Rep. vol. i, pp. 141-232. Krohn, a., 1846. Observations sur le generation et le diveloppement des Biphores (Salpa). Ann. Sci. nat., ser. 3 (Zool.), vol. vi, pp. 110-31. Mackintosh, N. A., 1934 Distribution of the macroplankton in the Atlantic sector of the Antarctic. Discovery Rep. vol. ix, pp. 65-160. REFERENCES 31 Metcalf, M. M., 1918. The Salpidaea: a taxonomic study. Bull. U.S. Nat. Mus. no. 100, vol. ii, part 2, pp. 5-193, pis. I-XIV. Meyen, F. J. F., 1832. Beitrdge zur Zoologie, gesammelt auf einer Reise um die Erde. Nova Acta Leop. Carol., vol. xvi, PP- 363-422. MlCH.\EL, Ellis L., 1918. Differentials in behaviour of the two generations 0/ Salpa democratica relative to the temperature of the sea. Univ. Calif. Pub. Zool. vol. .win, no. 12, pp. 239-98, pis. 9-11, i fig. in text. MiCHAELSON, W., 1915. Tunicata. Beitr. Meeresfauna Westafr. vol. I, p. 319. Hamburg. MiDTTUN, Lars and Natvig, Johan, 1957. Pacific Antarctic Waters. Scientific results of the 'Brategg' Expedition, 1947-48, no. 3, pp. 1-130. Murray Sir J. and Hjort, J. 1912. The Depths of the Ocean. London. RiTTER, W. E., 1905. The pelagic Tunicata of the San Diego region excepting the Larvacea. Univ. Calif. Pub. Zool. vol. 2, no- 3. PP- 52-113. pis- i-iii- RiTTER, W. E. and Byxbee, Edith S., 1905. The pelagic Tunicata. Reports on the Scientific Results of the expedition to the tropical Pacific, in charge of Alexander Agassis, by the U.S. Fish Commission steamer ' Albatross' from August i8gg to March igoo. Mem. Mus. Comp. Zool. Harvard, vol. xxvi, no. 5, pp. 195-214. Bars, Michael, 1829. Bidrag til Soedyrenes Naturhistorie. Bergen. Se\\'ell, R. B. Seymour, 1926. The salps of Indian seas. Rec. Indian Mus. vol. 28, part 2, pp. 65-126. 1953- The pelagic Tunicata. Brit. Mus. (Nat. Hist.) John Murray Exped. 1933-34. ^ci. Reps. 10 (i), pp. 1-90. Sigl, M. a., 1912. Die Thaliacean und Pyrosomen des Mittelmeeres und der Adria. Denkschr. Akad. Wiss. VVien, vol. LXXXVIII. Stiasny, G., 1926. Vber die Testa der Salpen und ihre systematische Bedeutung. Pubb. Staz. zool. Napoli, vol. 7, pp. 385-456. Streiff, R., 1908. Vber die Muskulatur der Salpen und ihre systematische Bedeutung. Zool. Jahrb., Abt. f. Syst., Geog. u. Biol. Tiere, vol. xxvii, no. i. Jena. Thompson, H., 1948. Pelagic Tunicates of Australia. Commonwealth Council for Scientific and Industrial Research, Melbourne. 196 pp., 75 pis. ToKiOKA, T., 1937. Notes on Salpas and Doliolutns occurring on the Pacific coast of middle Japan. Annot. zool. jap., vol. xvi, no. 3, pp. 219-36. Traustedt, M. p. a., 1885. Spolia atlantica, i; Bidrag til kundskab om salperne. Danske Vid. Sehk. Skrift. vol. 6, part 2, PP- 337-400. pis- i-ii- YouNT, J. L., 1954. The taxonomy of the Salpidae {Tunicata) of the central Pacific Ocean. Pacific Sci. vol. viii, no. 3, pp. 276-330. 1958- Distribution and ecologic aspects of central Pacific Salpidae (Tunicata). Pacific Sci. vol. xii, no. 2, pp. 11 1-30. APPENDIX Stations at which one or other of the four species have been recorded. Unless otherwise stated all are stations worked either by R.R.S. 'Discovery II' or R.R.S. 'William Scoresby' (prefixed by W. S.) and their positions and other data are given in the relevant station lists (Discovery Reports, vols. I, III, IV, XXI, XXII, XXIV, XXV, XXVIII). Station data for material collected by the 'Brategg' Expedition are given in Midttun and Natvig, 1957. Salpa fusiformis 428, 429, 432-435. 673, 677, 687, 690, 696, 699, 701-708, 710, 711, 713-715, 841, 845-847, 873, 877, 968, 969, 1608-1612, 1736, 1740, 1744, 1756, 1760, 1801-1803, 1806, 2027, 2086, 2148, 2150, 2350, 2380, 2418, 2419, 2453, 2483, 2524, 2526, 2527, 2576, 2683, 2689, 2690, 2691, 2727, 2728, 2731, 2817, 2818, 2889. Full data for the following stations are unpublished : Date Lat. Long. 3095 (night) 22. V. 54 44°49'N. 16° 13' W. 3096 (night) 23- V. 54 44° 40' N. 16° 20' W. 3097 25- V- 54 44°44'N. i6°4o'W. 3297 15- ix- 55 32°2l'N. 09° 28i' W 3502 28. ii. 57 32° 11' N. 65° I2'"W. 3508 2. iv. 57 33° 26' N. 75° 27' W. 367s 26. iii. 58 28° 02' N. i6°55'W. 3680 I. iv. 58 30° 13*' N. i5°59'W. 369s 7. iv. 58 33° 26' N. i5°oi'W. 3700 9. iv. 58 36°37'N. 14° 09' W. 32 DISCOVERY REPORTS Material from other sources : Date 'Sarsia' cruise P6/158. St. 13 13. v. 58 14 14. V. 58 15 14. V. 58 (Dr J. H. Fraser) Haul E 57/752 10. ix. 57 (Dr L. Berner) 9. ii. 57 4. vi. 58 (Professor J. L. Yount) 9. iv. 53 (Professor T. Tokioka) 3. iii. 36 6. ii. 51 Salpa aspera 677, 690, 693, 696, 705, 706, 714, 872, 873, 878, 1612, 1748, 1804, 1806, 2146, 2452, 2457, 2483, 2527,2683, 2728, 2731, 2820. Lat. Long. 47° 05' N. S°S^'^- 47° 03' N. 6° 30' W. 47° 20' N. 6° 23' W. 63° 51' N. 16° 20' W. 28° 58I' N. 118° 21-3' W 37° 34' N. 123° 33' w. 20° 34.6' N. 157° 32-3' w Seto. Sirahama Seto. Sirahama ial from other sources: Date Lat. Long. (Dr L. Berner) II. V. 50 32° 04' N. ii8°i2'W. 14- V. 57 32° 26' N. 119° 21' W. 4. vi. 58 37°34'N. 123° 33' w. Salpa thompsoni 168, 171, 176, 197, 198, 199, 201-203, 210, 229, 565, 566, 666, 668, 669, 671, 716-718, 727, 729, 731, 733, 737, 739, 741, 760, 820, 825, 829, 883, 884, 886, 887, 903, 904, 906, 911, 920, 921, 943, 946-948, 951, 958, 959. 964. 971, 1061, 1077, 1098, 1229, 1233, 1235, 1236, 1276, 1295, 1318, 1320, 1326-1333, 1335, 1366, 1442, 1446, 1513, 1515, 1519, 1524, 1545, 1547, 1549, 1550-1552, 1630, 1642, 1792, 1794, 1808, 1827, 1892, 1906, 1912, 2002, 2020, 2087-2091, 2094, 2097, 2100, 2131, 2137, 2144, 2154, 2160, 2177, 2204, 2206, 2209, 2212-2214, 2217, 2220, 2238, 2241, 2258, 2261, 2274, 2277, 2280, 2294, 231 1, 2313, 2340, 2457, 249s, 2531-2533, 2574, 2575, 2585, 2620, 2735- 2737, 2741, 2769-2771, 2809-281 1, 2825, 2826, 2828, 2830, 2835-2837, 2842, 2869, 2875, 2876, W.S. 400, W.S. 401, W.S. 403. Material from other sources : (Dr F. Beyer) 'Brategg' St. Date Lat. Long. 7 16. xii. 57 59° 25' S. 9o°oi'W 2. i. 48. 64° 30' S. 128° 00' W ao 4. i. 48 61° 36' S. 138° 25' W 28 17. i. 48 58° 00' W. 150° 05' W Salpa gerlachei 1220 1222, 1237, 1239, 1242-1246, 1254-1263, 1269-1271, 1282, 1283, 1286, 1293-1297, 1301, 1306, 1307, 1309, 131 1, 1654, 1665, 1667, 2225, 2226, 2228, 2235, 2237, 2240, 2243, 2244, 2246, 2247, 2271, 2743, 2744, 2833. Material from other sources: Date Lat. Land. 66°25'S. 89°55'W. 6i°53'S. 150° 27' W. 6o°55'S. 150° 28' W. 6o°o5'S. i5o°28'W. (Dr F. Breyer) 'Brategg' St. 15 20. xii. 47 24 16. i. 48 25 17. i. 48 26 17. i. 48 i J-i AJ'i -tiK^siggA .ft .(.mmj8 ilJ iteiol) .nnol viiiiii .'.:-a(n 81 libgn^J -<(bb6 ,.mia,if_ xfjgn-jj iuo;) vJatv 's^.so':; ■.th TO aii' .11 bhs T .8iM -;'5i'W. W. ^i'rotcsso Other PLATE I \a Salpa aspera A. Solitary form, dorsal view (body length 85 mm.). B. Aggregate form, dorsal view (total length 31mm., body length 18 mm.). ,2683, C. Solitary form, postero-dextro-dorsal view, semidiagrammatic. Redrawn from Ritter (1905). (i) mid-ventral ridge, (2) ventro- lateral ridge, (3) dorso-lateral ridge, (4) outer dorsal limiting ridge, (5) inner dorsal limiting ridge. Note : precise details of the oral and atrial muscles are not shown in Pis. I and II. ^. 119° 21' W. 37 123' 33' W. -i« -27. 729, 731, 733, 737, • 9S8, 959. 964. ■ ,:- vt'^, 1513. , -, :v-2, 2020, : 17, 2220, 2238, '. 273s- \v. , I'ioi, 1^06, 1307, 1220 1222, 1237 . *J" > J ' J /' '-' \ 2744, 2833. 1309, Material fr. (D. 26 4" DISCOVERY REPORTS, VOL. XXXII PLATE I KwoM 17 rfjgn -birn (s) ,38b-';' 'Jaia'aoeiJ ( 1 ) ,.iybii gnittmi(-Isa-iob la.trri (f.) ,3j{hri ieiaJst-oaiob (j) .agbi . ' .,,.;. .1 l-r-jTsI oaiob m.bdi.^js (d) ,rj«bh gnitimil-JBeioh i-jhk; .^ itfil-riTtri-j-; CK) ,3gbn gntJiinJI-lh-ioft ' '•'';br'0'>9> ^ •Ml''-' ^ -' ■ ■~.nal ybod ,.rntii |.f (JjSOUilUtgsib'niSfc ,.73(7' (l:fj:;)f, PLATE II Salpa thompsoni sp.nov. A. Solitary form, dorsal view of the holotype (body length 71 mm.). B. Aggregate form, dorsal view of the allotype (total length 34 mm., body length 20 mm.). C. SoUtary form, postero-dextro- dorsal view, semidiagrammatic. (i) transverse ridge, (2) mid- ventral ridge, (3) dorso-lateral ridge, (4) inner dorsal-limiting ridge, (5) outer dorsal-limiting ridge, (6) secondary dorso-lateral ridge, (7) secondary dorsal-limiting ridge, (8) ventro-lateral ridge, (9) medio-dorsal ridge. DISCOVERY REPORTS, VOL. XXXII PLATE II [Discovery Reports. Vol. XXXII, pp. 33-464, Plate III, November 1962] THE NATURAL HISTORY AND GEOGRAPHY OF THE ANTARCTIC KRILL {EUPHAUSIA SUPERB A DANA) By JAMES MARK CONTENTS Introduction page 37 Acknowledgements 40 Early records and recent literature 40 Scope of the observations 48 Material and data 53 Methods in the field 53 Identification of the species 55 Laboratory methods 56 Distribution in outline 57 Physical environment 64 Pack-ice 64 Surface temperature 75 The larval stages 88 Vertical distribution 88 The developmental ascent ......... 97 Deep and shallow horizontal dispersal 99 Diurnal vertical migration 105 Some biological implications of the developmental ascent and its influence on the distribution and conservation of the euphausian population . 1 1 8 The developmental ascent in relation to the local distribution of krill in the Ross Sea 123 The older stages 126 Economic and ecological importance 126 Sizes and quantities eaten by baleen whales 138 Influence on the foetal growth rate of whales 147 Patchiness 148 Habits and behaviour 154 Vertical distribution and migration 157 Absolute density 170 Food 172 Faecal pellets 176 Spawning and hatching 176 The act of spawning 176 Time of spawning 177 Depth of spawning 178 Localities of spawning 1 89 Time of hatching i99 Depth and localities of hatching 199 Influence of the sinking shelf water on the liberated eggs . . . 205 Deep spawning reconsidered ........ 208 Conclusions and future research 210 Speed of THE cold bottom WATER 212 Causes of patchiness 215 Effect of animal exclusion 216 Effect of pack-ice 218 The patches as swarms 219 Impact of the swarms on the ocean pastures 240 Happening of the whales ltpon the swarms 241 Impact of the whales on the swarms 243 36 CONTENTS Intimate structure of the swarms P<^S^ 245 Sex ratio ^45 Relative size of the sexes 248 Developmental condition of the sexes 250 Growth AS INDIVIDUALS AND SWARMS . . • • • •251 As individuals .... 251 As swarms 252 Moulting 256 Reaction of larvae, adolescents and adults to ship and nets . . 258 Powers of the individuals to evade 258 Capacity to act in unison 202 Scattering of swarms by the ship 264 General inferences to be drawn 265 Vertical distribution and migration reconsidered .... 268 Corrections applied to the catch-figures 278 Horizontal distribution, growth and dynamics of dispersal . . 284 Introduction and plan of presentation 284 The eggs and larvae from the vertical samples 289 The eggs 289 The deep living NaupUi, Metanauplii and First Calyptopes . . 298 The shallow living First, Second and Third Calyptopes . . .308 The shallow living First, Second and Third Furcilias . . . .323 The shallow living Fourth, Fifth and Sixth Furcilias . . . .331 Total eggs and larvae and total surface larvae 342 Gross distribution of total eggs and larvae 349 Growth and dispersal in the surface drift 340 The larvae, adolescents and adults from the oblique and horizontal samples 358 The massed surface larvae .....■••• 35° The small whale food 374 The staple whale food 384 Gross distribution of total euphausian surface population . . . 4°8 Distribution of surface discoloration and its relationship to the gross distribution of the surface population 412 The larval populating of the east wind-weddell surface stream as an ANNUAL event 4^2 Local distribution. The SOUTH GEORGIA WHALING grounds . • .413 The COASTAL topography of ANTARCTICA AND ITS INFLUENCE ON THE DEPLETION AND CONSERVATION OF THE SURFACE POPULATION . . -423 euphausia superb a and e. triacantha 424 The feeding migrations of the baleen whales. Their possible origin and history 42° Review of dynamics of distributional control 429 Summary 435 Appendix. Tables 62-4 442 References 444 Plate III follomng page ^6^ EDITOR'S NOTE This report is a comprehensive study of an important organism in relation to its environment and the treatment and interpretation of the data are those of the author. 37 THE NATURAL HISTORY AND GEOGRAPHY OF THE ANTARCTIC KRILL (EUPHAUSIA SUPERB A DANA) By James Marr (Plate III and Text-figs. 1-157) We always can tell a likely spot to look for whales by the colour of the water. Whales feeds upon insects which swarm in good whale water in myriads, making the water look quite thick and dark brown. When we see the water have this appearance we keep a good lookout for fish. If there is no food in the water you may be sure you won't see no fish, unless they happen to be passing on their way up or down the country, or are on the search for good feeding ground. It's my opinion a whale can go a long time zvithout meat, but when they do feed they swallow a tremendous quantity. We often take bucketsful of whales' food out of their throats and mouths when cutting out the whalebone. Whales' food consists of small red insects or animalcules, or whatever you may choose to call them, of a regular, uniform reddish colour, and spindle-shaped, tapering away to the tail. They are found principally in the Arctic and Antarctic Seas, where they exist in enormous numbers. They don't exceed an inch in length, yet they are the principal food of these great fish. From the remarks of John Gravill, captain of the Arctic whaleship 'Diana' of Hull in 1866-7, quoted in the diary of Charles Edward Smith (1922), M.R.C.S., surgeon of the 'Diana' and one of the survivors of this tragic voyage. INTRODUCTION DURING the long course of the investigations conducted by the former Discovery Committee, and continued latterly by the National Institute of Oceanography, special attention was given to the study of the biology and distribution of the Antarctic whale food, or the krill as it is commonly known, a name derived from the Norwegian 'kril', meaning fry or very young fish. Over the years, from an immense area girdling the polar continent, material accumulated, covering in the end the many diverse phases of the life of this swarming pelagic prawn from the egg to the adult state. The results of its analysis, representing, in field and laboratory, the labour of some twenty years, are now presented in the following survey of the biology and zoogeography of a species, economically and ecologically the most important of Antarctic macroplankton animals, and in sheer mass of living matter, it has been said, perhaps amounting to all the rest of the Antarctic macroplankton combined (though see p. 135). In the title I have preferred the wider Geography to the more fashionable Zoogeography, my chief concern in this treatise having been to obtain all along a broad understanding of the principles underlying the distribution and movements of these southern shrimps over their whole world range. I make no excuse for Natural History. It covers a wide field, expressing euphoniously and in plain English what we understand of the life and environment of animals. Although for long ousted by the classical Ecology, today, as Walford (1955) has said, it is 'now at last again coming into its own, though perhaps under a diflFerent label '. Many in fact (Hedgpeth, 1957) still regard ecology as an unnecessary synonym for natural history. A knowledge of the distribution and movements of the whale food, and above all what is controlling them, is an essential step towards a fuller understanding of the distri- 38 DISCOVERY REPORTS bution and movements of the whales themselves, of much indeed that is of major concern to those who engage in whaling. However wise, far-sighted, or effective the provisions of the international control may have been there can be few who would say that the Antarctic whale fishery has yet been established on a permanent footing. There must be many in fact who would doubt it. I take this opportunity, therefore, to express the hope that, purely academic as much of the information presented here must necessarily be, some of it at least, when the existing whaling regulations come annually under review, may provide some possible matter for reflection in the deliberations that ensue. For today, with a rapidly expanding world population, millions, hitherto backward and dispossessed, seeking everywhere for mounting quantities of food and raw materials, the prospect of the decline alone, not to mention the destruction, of the vital produce of Antarctica can hardly be contemplated without the gravest concern. As Morgan (1956), in his recent survey of world fisheries observes, the earth's population, already in 1954 standing at 2500 millions, may well, if it maintains its present rate of increase, reach some 4000 millions by 1980, the need to find food for this growing press of humanity presenting perhaps the most searching problem the world has ever had to face. Indeed by a.d. 2100 Parkes (1955) envisages a horde of 10,000 millions averaging out at 200 to the square mile over the whole earth's surface, including the Antarctic continent. The mounting impact of whaling upon the Antarctic whales has for long been recognized as a serious problem, now perhaps reaching alarming proportions. An early warning of its impending consequences was uttered by Hjort, Jahn and Ottestad in an important paper published in 1933, and already in 1937 we find Laurie calling attention to the increasing plight of the great blue whales upon which for long the brunt of the onslaught had been falling. ' The stock ', he wrote, ' is already seriously depleted and further hunting on the same scale bids fair to make Blue whales so scarce that they will cease to be a source of profit to the industry and so diminished in numbers that the stock even if completely protected may take many years to recover.' More recently Ruud (1955) called attention to the overfishing of fin whales that seemed to be taking place between 1946 and 1953 and Mackintosh (1959a, 19596), summing up the position today, notes that it is generally admitted that the stock of blue whales at least has been severely reduced^ and that while there have been certain indications of a decline, now steepening, in the fin whale population,^ doubts have been expressed as to whether it is actually taking place, there being as yet no final proof that it is. Ottestad (1956) takes a more serious view, computing that if the present day rate of killing is not substantially reduced, the Antarctic fin whale population will be so depleted by 1962 that an annual catch of as little as 10,000 would be seriously imperilling its existence. And this he says is taking an optimistic view. Continued concern for the condition of the blue and fin whale stocks was expressed also by the Scientific Sub-Committee of the International Commission on Whaling at its annual meeting in 1959. The situation, however, is constantly kept under review and so if wisdom is to prevail there may yet be time for the nations concerned, as Dawbin (1952) in his recent survey of the fishery has said, to bring about a balance between the 'killing and the replacement rate of the whale populations', or as Wheeler (1934) wrote ' In the season 1932-3, 18,624 blue whales were killed, 807% of the total blue and fin whale catch. In the season just concluded (1958-9) the same percentage involving only 1191 blue whales was down to 4-4 {Norsk Hvalfangst-Tidende, 1959, no. 9). * Chittleborough (1959c, i960) has just called attention to the continued decline that appears to be taking place in the winter population of humpbacks hunted west of Australia. Dawbin (1960a), however, notes that the absence of certain adverse trends in catch-statistics for the northbound population of humpbacks passing New Zealand between 1947 and 1958 suggests that the stock in Australian-New Zealand Antarctic waters is still in a relatively sound condition. He does call attention, however (Dawbin, 19606), to certain possible adverse trends shown by the New Zealand catch in 1959, yet still does not seem to think they give cause for alarm, noting that a distinct rise in percentage of immature animals shot that year might be due to intensity of hunting early in the season, gunner selection and other factors. INTRODUCTION 39 SO long ago, reach 'a scientific basis for the attainment and maintenance of economic security'. Commenting on the 'workable' International Whaling Convention of 1946, Haden-Guest (1955) remarks, ' Biologists consider its provisions still very inadequate, both as regards the Antarctic and elsewhere, but it has at least prevented rapid extermination, and there is still time to ensure the pre- servation of nearly all whale species '.^ Indeed in his survey of the economic wealth of the Arctic, Haden-Guest hopefully puts the whales, both northern and southern, among his ' renewable resources '. It is for the whalers and whaling nations to show he is right. In compiling this report I have consulted the works of many authors, many of them not, or not strictly, scientific. They include the writings of explorers and sailors, sealers and whalers, historians, geographers and cartographers, ship's surgeons, contributors to official departmental publications and compilers of books on whaling, sealing, polar and deep-sea exploration, meteorology, ornithology, oceanography, marine biology and ecology, as well as sundry other references. I may sometimes have been critical, but only because in seeking to unravel a problem of enormous complexity I have had access to a body of data and material unparalleled in the annals of the investigation of any single plankton animal, if not indeed of plankton animals as a whole, and, like Totton (1954) in his classic account of our equally large collections of Discovery Siphonophores, have endeavoured 'to build firmly on sure foundations '. As Deacon (1957) has written, the emphasis in marine biology is moving towards the studies of populations — towards learning more about the distributions of the known species in relation to their physical and biological environments, and towards learning more about the factors which make some parts of the oceans more productive than others. It is a long task, even for the smaller, slow moving organisms which are easily caught. There are usually many stages between the eggs and the adults and when these have been discovered a large number of net hauls made over a full range of seasons, places and depths have to be sorted and the different stages counted and measured. Only then can provisional hypotheses be sought to explain the distributions in relation to spawning and feeding habits, depth of water, temperature, water movements and other relevant factors. The more active animals which escape large towed nets present greater difficulties, and ideas as to their size and distribution seem to depend to a large extent on the type of gear used and the speed at which it can be towed. It is with such questions that this report is largely concerned and it is along these broad lines that the teeming multitudes of these southern shrimps have been studied. I would recall too the words of the late Dr Stanley Kemp uttered (Kemp, 1934) over a quarter of a century ago. In the close interrelations which exist between animals and plants and the chemistry and movements of sea-water, the whole system may be likened to an extremely complex machine : it is the way in which this machine works, and the great seasonal changes in its operation, that we are seeking to understand. Knowledge of this kind, to be obtained only by the slow accumulation of data and material and its subsequent treatment in the laboratory, is not merely of scientific interest ; it is fundamental to the solution of almost all economic problems in marine biology, including those of whales. At a recent symposium on marine and freshwater biology in CaHfornia Dr T. H. Bullock (1958) called for a wider outlook on marine ecology, urging his audience that the presentation of balance sheets and flow charts for whole ecological systems — of tide pools, estuaries, seas, and ultimately the entire ocean — should be a matter for concerted effort and not left, as it is today, to the labours of a few, working parochially in somewhat circumscribed fields. I am not an ecologist. I have presented the ecological side of this problem in the simplest of terms. Nevertheless, I believe that when the Antarctic balance sheet is presented, as in the long run it must one day be, the krill will emerge 1 Laws (1960 a) has expressed the same view, noting that without the timely restraint imposed by this convention depletion would have been 'much more rapid and catastrophic' than it has actually been. 40 DISCOVERY REPORTS supreme in this southern field, and later, when the final answer to Dr Bullock's massive challenge has been found, feature large, perhaps very large, in the ecological budget for the oceans as a whole. In a work such as this there must inevitably be imperfections. The hypothesis, for instance, that the Antarctic bottom water is playing an important part in the distribution of the early larval stages (pp. 205-8) may to many seem untenable. It is at least, however, a possibility and, in common with other interpretations of our data, such as for instance my views on vertical migration, perhaps equally debatable, points a broad path along which future exploration may travel. As Petersen (1915) said on publishing the first hypothetical maps of the bottom fauna of the North Atlantic, ' I venture to make this attempt, well knowing how imperfect it must be, because it will have to be made sooner or later, and the sooner it is made, the sooner, I trust, will it be repeated to better eff^ect '. It sometimes happens, although not very often, that one whose business is cast among certain animals, who is familiar with them yet does not profess to study them closely, can express in a few simple words much of what we have to say later in our scientific publications after years of patient research. The simple lines I have quoted from John Gravill for instance, an Arctic whaleman, cover broadly it will be seen, and with a remarkable economy of English, much of what appears in this and many other papers on whales and whaling, and could well be used, just as they are, as an end-piece to this report or in the long summary that goes with it. ACKNOWLEDGEMENTS To the widely scattered members of the former Discovery Committee's staff, both sailors and scientists, whose sustained eifort in field and laboratory led to the enormous accumulation of material and data upon which this report is based, the following pages are dedicated in grateful acknowledgement and appreciation of their seamanship, comradeship and devotion in the Antarctic seas. I am greatly indebted to Professor Sir Alister Hardy, F.R.S., and Dr G. E. R. Deacon, C.B.E., F.R.S., F.R.S.E., for reading through the long manuscript and giving a great deal of help and much valuable advice. From an early stage too I have had many helpful discussions with Dr N. A. Mackintosh, C.B.E., who, with the late Dr Kemp, was largely responsible for planning the voyages of our vessels and ensuring the necessary spatial and seasonal coverage of the circumpolar sea that has contributed so much to the results we have obtained. I have to thank also Mr A. Style for the artistry and clarity he has brought to the final execution of my text-figures and Mr E. C. Mansell who completed so competently the long task Mr Style began. EARLY RECORDS AND RECENT LITERATURE Although it remained unnamed until Dana's original descriptions of 1850 and 1852 it is clear that Euphaiisia superba was known to many of the eighteenth and nineteenth century voyagers, explorers, sealers and whalers, who provided our earliest knowledge of the existence and extent of the southern continent and discovered so many of its off-lying islands and island groups. It may have been seen by Cook (1777) in February 1775^ and, although practically nothing has come down from them in the literature, must certainly have been well known to the South Shetland sealers (Allen, 1899; Bruce, 1920; Gould, 1 941) from about the end of 1819 onwards. It may well in fact have been familiar at a still earlier date to the fur and elephant sealers who (Murphy, 1948) first began to visit South Georgia shortly after its rediscovery by Cook in 1775 and who as Matthews (1931) relates used occasionally to 1 The discoloured, 'uncommonly white', water sighted then from the 'Resolution', near Candlemas Island in the South Sandwich group, a phenomenon that so alarmed the officer of the watch that thinking it to be shoal water, he tacked the ship instantly, could well it seems (p. 1 50) have been caused by a swarm of E. superba. Cook himself suggests it was probably due to a shoal of fish. EARLY RECORDS AND RECENT LITERATURE 41 hunt the krill-eating right and humpback whales in its sheltered bays and inshore waters. It is first definitely mentioned in January 1820 by Bellingshausen (1945) who refers to its abundance in the Antarctic seas and its importance as the food of penguins. The British sealer, Richard Sherratt (1821), noted it as the food of both seals and penguins while Weddell (1827), during his historic southern voyage of February 1823, reported several instances of surface discoloration strongly suggestive of those that are now known (p. 149) to be caused by dense surface concentrations of the krill. Webster (1834), surgeon of H.M. Sloop 'Chanticleer', refers to its great abundance in the South Shetland area and describes how gorged adult penguins feed their chicks with it by regurgitating the contents of their stomachs. He describes too the mortality among the ' shrimps ' that drift into the hotter parts of the steaming littoral of volcanic Deception Island and are cast up in large numbers on the shore. The logs of John Biscoe (1831), John Balleny (1839) and John McNab (1839),! Balleny's second mate, contain several references to surface discoloration in circumstances that again suggest the sighting of krill patches, McNab's entry on the day of the discovery of the Balleny Islands, 'water very much discoloured, and whales in shoals', being perhaps the most suggestive. Jacquinot (1842), surgeon of the French corvette ' Zelee ', records it as the food of penguins and is the first to mention it as the diet of the southern baleen whales and of the Crabeater seal, adding that it was well known to the right whalers of his time. He is the first to recognise it to be a euphausian, a new species he thought of Thysanopoda, and the first to call attention to its occurrence in immense shoals ('bancs') giving a reddish tint to the sea. He records its 'extreme abundance ' (in 1838) in the western Weddell drift region and recognised it was this abundance which attracted large numbers of whales, right, humpback and blue, to this locality. Wilkes (1852) also notes it as the food of penguins, records its abundance near the Antarctic mainland south of Australia in January 1840 and, like Jacquinot, states it was this no doubt that drew the whales in such large numbers to these high latitudes. It was one of Wilkes's officers. Lieutenant Totten of the U.S.S. 'Porpoise', who in February 1840 collected^ the first specimen oiEuphaiisia superba to be scientifically described, and the original coloured drawing of it by Dana (1855, PI. 43) is based on colour notes and a sketch by this officer.=^ Ross (1847) notes it as the chief food of the Crabeater seal. Dickson (1850), surgeon of H.M. hired barque 'Pagoda', comments on the 'innumerable whales ' and flocks of birds that were to be seen ofFEnderby Land and remarks on the ' muddy ' discoloration of the water that was encountered both there and in the Weddell drift, especially in the neighbourhood of the pack. The German sealer, Eduard Dallmann, ^ Published in extract in The Antarctic Manual (1901). 2 Jacquinot states that the French expedition collected several specimens in 1838, but except for his statement that it apparently belonged to a new species of Thysanopoda there is no description of them. ^ Banner (1954) notes that the type specimen originally described and figured by Dana is lost, that his description is 'broad' and does not mention many of the characteristics now considered important and that the figure itself is small and lacking in detail. On these grounds he proposes that the name Euphausia superba Dana be regarded henceforth as a nomen dubiunt, unavailable for future use, and that the matter should be referred to the International Commission on Zoological Nomen- clature. I hesitate to think that such a drastic step will be necessary, for although Dana's description is perhaps broad his figure at least is unmistakably that of E. superba. Lieutenant Totten's specimen was collected some time in February 1840 in 66° 05' S, 157° E, a position not far from the Balleny Islands, and both in colour and form Dana's original drawing presents a remarkably lifelike picture of this very large euphausian for which much of the credit must be given to the accuracy of the collector's observation, colour notes and sketch. It is of an animal 2 in. (about 50 mm.) long, and, although Dana does not say so, from the slender shape of the carapace and the powerful development of the expodites of the thoracic limbs, it is clearly, as Bargmann (1937, PI. i) shows, that of a fully mature aduh male. I am perfectly satisfied, therefore, that Dana's drawing can only be referred to E. superba and that his well-chosen name should stand. From its size alone, as well as position of capture, this early figure could not in fact be referred to any other species. It is of historical interest, however, that in his Synopsis Generum Crustaceorum Ordinis ' Schizopoda\ in which his original brief description appeared, Dana (1850) gives the position of Lieutenant Totten's specimen as 'prope long, orient. 150° et lat. aust. 60°'. This in fact (p. 60, Fig. 5a) is well outside the northern range of the krill in this meridian, the round figure for the latitude evidently being a mistake which he put right (Dana, 1852) in his longer and better known description which appeared two years later. 42 DISCOVERY REPORTS quoted by Racovitza (1903), records its shoaling habit, the sea 'eine schmutzig braune Farbe', in the west Graham Land channels in 1873, Racovitza commenting that the discoloration logged by Dallman was evidently produced partly by the swarming krill and partly by ' les prairies de Diatomees ' on which the congregating Euphausia fed. Sars (1885) refers to the capture by H.M.S. 'Challenger' of 'numerous specimens' of larval and early post-larval E. superba,^ the largest of the latter only 17 mm. long, not far from the Antarctic mainland near 80° E, in February 1874. 'They were', he says, 'as usual, taken in the townet, at the surface of the sea'. This record is of much historical interest since it is the earliest reference to the remarkably slow growth-rate we have since found (p. 355, Fig. 107) is peculiar to the krill in the highest latitudes of its geographical range. In the same high latitudes, it is interesting to record, Vanhoffen (1903) reports equally backward euphausians, 'wenige der kleinen E. antarctica' , near Gaussberg in February 1902. Burn Murdoch (1894), who visited the east side of Joinville Island in the ' Balaena ' in 1 892-3 , refers to the stones and red shrimps found inside the penguins, and the penguins, red shrimps and stones found inside the seals, W. S. Bruce (1894), naturalist in the same vessel, noting the enormous numbers of ' finner whales ' that frequented these waters and the Euphausia that swarmed on the surface. Bull (1896) refers to its teeming abundance near the Balleny Islands and on the outskirts of the Ross Sea, stating that it appeared to sustain all higher animal life, baleen whales, seals and penguins, in these latitudes, and that in the ice, when disturbed by the passage of a vessel it would 'scatter for shelter in millions, reminding one of the animation in a disturbed ant-heap'. Cook (1900) refers to the 'full meal of shrimps' obtained by Adelie penguins from narrow leads in the solid pack far south in the Bellingshausen Sea, Racovitza (1900 a) to the 'immense bancs' of krill that were encountered there and to the overriding importance of the role they must play in the economy of Antarctic life. Elsewhere (Racovitza, 1903) he refers again to its great abundance in this region, notes the heavy toll of it exacted by penguins and Crab- eater seals and states that it appeared to be the food of the humpback whale. He calls attention too, in some notes quoted by Hansen (1908), to its great activity, predominantly surface habitat and, as Wilton (1908) repeatedly records, preference for the cracks and open spaces between the ice-floes rather than the poorly illuminated depths below them. These notes contain a further reference to its phytoplankton diet. To sum up, the value of these early observations lies principally in their repeated emphasis on the surface habitat of this species and its tendency to occur in shoals, in their repeated emphasis on its abundance and in that in aggregate they reveal its distribution to be circumpolar. Above all, since these early sightings, without exception, manifestly could not have been made in the dark, they reveal that these animals at times are massed conspicuously at the surface during the daylight hours, a phenomenon suggesting strongly that in so far as their vertical movements are concerned the krill do not always behave with the same regularity as some other euphausians are said to do. Since the beginning of this century a great deal of work has been published on the Antarctic krill, and although I have to refer to much of it more fully later a summary of the principal findings is appended here.^ Lonnberg (1906), referring to the diet of the blue, fin, right and humpback whales examined by Mr Erik Sorling at South Georgia in 1904-5, observes that 'the food consists of "kril", that is Euphausiids ', the humpback, according to Sorling, feeding exclusively on these animals. Sorling, then assistant taxidermist in the Natural History Museum in Stockholm, is thus the first to call serious * Originally described by Sars as E. antarctica but since shown (Holt and Tattersall, 1906; Coutiere, 1906; Tattersall, 1908) to be larval and early adolescent forms of E. superba. ^ The many references to E. superba that appear in Antarctic ornithology are not included here. For such as I have seen the reader is referred to the ecological part which begins on p. 1 26. EARLY RECORDS AND RECENT LITERATURE 43 attention to the major importance of the ecological role of the krill in this southern field, the earliest to be opened up by the modern whalers. He remarks in his field notes on its abundance in these island waters, where, he writes, it was eaten not only by whales but by many birds and fishes, caUing particular attention to the local abundance of the large krill-eating Nototheniid, Notothenia rossii, which already at that early date (see also p. 131) was being salted down in barrels and finding a ready market in Buenos Aires. Hansen (1908) gives the first descriptions, with excellent figures, of what are in fact the first four Furcilia stages, referring to them as larvae of Euphausia superba Dana, stages A, B, C and D. Tattersall (1908) shows that the E. antarctica of Sars (1885) and the E. glacialis and E. australis of Hodgson (1902) are invalid, all three in fact being developmental phases, larval, adolescent or adult, of E. superba} Risting (1912), referring to the feeding of the southern humpback, records the following passage in which the opening sentence foreshadows something of what has since been learnt of the distribution and movements of these southern shrimps. With the approach of the Antarctic spring great quantities of whalefood ' sprout ' up along the south polar ice-edge, and this food is carried by the currents towards the coasts of the great south polar island groups. From November and onwards nature has consequently spread a rich table to which the plankton-eating whales can proceed and find food in all luxuriance. In greatest numbers the humpback puts in an appearance.^ It is at this season lean, but later, as the southern summer advances, and it can feed without stint on its natural food — kril — it becomes rapidly fatter and fatter, and from February to April it acquires a layer of blubber so thick that one can very seldom find the like on the northern whaling grounds (Translation by M. A. C. Hinton, 1925). Liouville (191 3) refers to the Euphausia that provide the principal diet of the blue whales frequenting the South Shetlands and Bellingshausen Sea, 'le seul fond alimentaire pour Cetaces offert par le plancton de surface ', adding that other rorquals in these waters, fin and humpback, almost certainly are also euphausian feeders. Zimmer (1913), working on rather poorly preserved material, gives the first general account of the anatomy and discusses the function of the various structures. Raab (191 5), working with Meganyctiphanes norvegica and Euphausia krohnii, extends the researches of Zimmer, caUing attention to certain anatomical and histological differences, as well as similarities, that exist between the Antarctic krill and these northern forms. Clark (1919), in his account of the biological work of the ' Endurance ' expedition, notes the heavy toll of krill that was taken by Gentoo penguins on Elephant Island from April to August 191 6, concluding it must be plentiful in certain, if not all, parts of the circumpolar sea throughout the southern winter. The stomachs of the fin, blue and hump- back whales he examined at South Georgia all contained ' Euphausiae, with a mixture of Amphipods '. Hinton (1925) records that the fin, blue and humpback whales examined by Major G. E. H. Barrett- Hamilton at South Georgia in the season 1913-14 were feeding on 'kril'. Mackintosh and Wheeler (1929) refer to its enormous abundance in this locality and are the first definitely to show that for all practical purposes it provides the exclusive diet of the great southern Balaenopterids, the blue and fin whales. Rustad (1930), basing his conclusions largely on the aggregate findings of earlier expedi- tions, states that, together with Thysanoessa macrura, it is evidently the most abundant and widely distributed of the southern euphausians, and publishes a map, also based largely on these findmgs, which clearly shows its distribution is circumpolar. He calls attention to its importance as the food of whales, seals and birds, adding that further study of its life-history and 'habits' would be of great practical value to the whaling industry. Two large, rather damaged, specimens of a First Calyptopis, he mentions briefly but does not figure, might possibly, as Rustad suggests, be referred to Euphausia superba. Marshall (1930) reports on the stomach contents of a large number of whales taken in the northern part of the Ross Sea in 1928-9, finding that in every instance where it was fresh enough to be 1 Steuer (1910, 1911) refers to Hodgson's E. australis as Eucopia {Euphausia) australis. 2 The great abundance of humpbacks to which Risting refers may not be altogether real since it was this small species (Hinton, 1925; Bennett, 1931) that was the principal quarry of the small and under-powered whale-catchers of his time. 44 DISCOVERY REPORTS determined the food consisted of E. superba. Kemp and Nelson (193 1) report great quantities of this species in the surface waters of the deep, now submarine, volcanic crater they discovered between Cook and Thule Islands. In a long and important paper, to which many further references will be made, Ruud (1932) in his report on the euphausians of 'Vikingen's' expedition to the Weddell Sea, gives the first published account of the two-year life-cycle of the krill,^ a phenomenon then thought to be without parallel among other euphausians, and gives the first short descriptions, with figures, of the egg, First Nauplius, Metanauplius, First, Second and Third Calyptopis together with that of an early Furcilia stage which is clearly the First. He emphasises that the species is essentially a surface one, the Antarctic surface layer being the principal locus of abundance of both adults and larvae alike, and calls repeated attention to the importance of the Weddell Sea current as a carrier and distributing agent of the larvae in the Atlantic sector. From the monthly developmental condition of the ovaries of the females he examined and from the brief period of larval abundance this expedition encountered, he concluded that spawning was a ' spontaneous ', that is, not a protracted, phenomenon, which, during the season ' Vikingen ' was in the field, was confined to the first half of January. He found a remarkable scarcity of eggs and Nauplii in the plankton, a phenomenon leading him to suppose that the krill spawn close to or right under the Antarctic pack, where, hidden away and beyond the reach of our nets, the early larval development would take place. Rustad (1934) is the first to call attention to the later (February) spawning characteristic (p. 177) of higher latitudes near the Antarctic mainland and shows conclusively that in Thysanoessa macrura at least the early larval development takes place in deep water and not at the surface as postulated by Ruud. Hart (1934), referring to the stomach contents of adult and post-larval krill, states that the most frequently occurring remains are those of the diatoms Fragilaria antarctica^ and Thalassiosira antarctica. In post-larval specimens he often found entire examples of large Foraminifera (Globigerina sp.) and suggests they are possibly eaten for their calcium content. Mackintosh (1934) refers to its extreme patchiness and tendency to form shoals in the western Weddell drift region and notes that many of the big shoals have been seen at the surface both by night and by day. Disregarding samples containing over 1000 specimens and so eliminating what he has called ' the disturbing influence of heavy shoal catches ' he shows that Euphausia superba undergoes only a minor degree of diurnal vertical migration and suggests that the greater part of the population remains permanently near the surface, ' especially perhaps when forming shoals '. He calls attention to how readily the older and more active krill dodge out of the way of a surface net by day and shows that in the Falkland Dependencies this species is numerically among the most important of macroplankton animals, including it in his ' cold-water ' group which in this region may occur in large numbers anywhere south of the 3° isotherm. Hardy and Gunther (1935) describe its distribution on the South Georgia whaling grounds and give the first eye-witness account in modern times of the swarming habit there. Following Ruud (1932)^ they point out that it would seem likely that the older larvae ('Furcilias' and 'Cyrtopias') encountered in this locality in March, April and May are not of local origin but get carried there in the surface drift from some distant locus of spawning in the south. They are the first to give a practical demonstration of the remarkable patchiness of its distribution. From a wealth of material, Fraser (1936) gives the first comprehensive description, richly illustrated by figures, of every stage of the larval development from the egg to the last Furcilia, and it is on this classic account that more recent determinations of the eggs and larvae have been based. He discards the former division of the later stages of euphausian development into Furcilias and Cyrtopias, and this, together with his recognition that both are in fact Furcilias, of which in E. superba there are 1 Already noted, however, by my former colleague, Dr J. F. G. Wheeler, in a report to the Discovery Committee in 1930. ^ Revised by Hendey (1937) and now called Fragilariopsis antarctica. ^ And incidentally it transpires (p. 43), Risting. EARLY RECORDS AND RECENT LITERATURE 45 only six, constitutes perhaps his most important contribution (see also Fraser, 1937) to our knowledge of euphausian life-history. He is the first to show that spawning is a long drawn out phenomenon, lasting in the Falklands sector of Antarctica from November to March, and not the ' spontaneous ' outburst of Ruud, and the first too to suggest, from the deep occurrence of the Metanauplius, that the early larval development must take place at great depths and could not be the surface phenomenon postulated by that author. John (1936) places it in his ' southern group ', concluding from the observa- tions of 'Discovery II ' in the summer, autumn and winter of 1932 that its normal range is from the Antarctic coastline northwards to the northern edge of the pack-ice and a little beyond. He remarks on the concentration of larval and early adolescent krill encountered at the ice-edge by ' Discovery II ' during her circuit of Antarctica in 1932-3, and is the first to record how the larvae in some measure of abundance may be carried far to the north of the ice in cold water deflected from the south by the submarine ridge connecting Gaussberg with Kerguelen. Matthews (1937, 1938a, 19396) makes the first definite reference to this species as the food of the humpback, right and sei whales on the South Georgia whaling grounds. Bargmann (1937) describes the development of the male and female reproductive system, distinguishing seven clearly marked stages in each, and in a later paper (Barg- mann, 1945) uses these stages to work out the composition and growth-rate of the euphausian popula- tion as a whole. Extending the work of Ruud (1932) she shows that the females are consistently smaller and take longer to reach sexual maturity than the males and that the total period of growth from the egg to the adult occupies a minimum of 22 months in the male, and 25 in the female. She gives an excellent coloured plate of the adult male and female. Barkley (1940) gives a detailed description, with admirable figures, of the filtering mechanism by which E. superba ingests its food, and from an examination of some 1300 krill stomachs demonstrates conclusively that by far the most frequently occurring and abundant remains are those of the chain-forming diatom Fragilariopsis antarctica, although other organisms, of which he gives a long list, the more important of them small, spineless ('glatten') diatoms, are also ingested. Hart (1942) gives an excellent diagram showing the overriding importance of Euphausia superba in the Antarctic food chain and again refers to Fragilariopsis antarctica as an important constituent of its diet. Nishiwaki and Hayashi (1950) report that the blue and fin whales taken by the Japanese whaling fleets near the Balleny Islands and on the northern outskirts of the Ross Sea in 1947-8 were feeding, for all practical purposes exclusively, on Euphausia superba. Nishiwaki and Oye (195 1) studied 1640 krill samples collected from blue and fin whales' stomachs in the Ross Sea-Balleny Islands area in 1948-9. They call attention to the predominance of 'small' whale food (see p. 140) in these high latitudes and note that of 392 stomachs examined in December, 40 % were empty, and that of the stomachs with food in them 68 % contained little or scarcely any krill at all. The December scarcity of feeding-stujff these figures suggest, however, is not altogether surprising, for the Japanese were operating then in the West Wind drift, a region recently shown (Marr, 1956) to support only a meagre population of krill. Perhaps the most interesting of Nishiwaki and Oye's findings was that the stomachs of night-flensed whales were often empty, and they suggest tentatively that, since most of the whales worked up at night are those that are caught the afternoon of the day before, these animals in the main are early morning feeders and that possibly they feed only once a day. Dell (1952), working from Bargmann's plate, gives an admirable line drawing of the adult male. Sheard (1953), discussing the larval development of the Euphausiacea as a whole, defines the Furcilia phase as that part of the development during which 'the abdomen is differentiated in detail to act as a locomotor organ'. Thus for all euphausians he proposes that this particular phase, on the precise definition of which there has for so long been such widespread dis- agreement, should now include only three stages, Furcilia i in which the pleopods develop as non- functional rudiments, Furcilia 2 in which the pleopods, with their musculature, develop as functional 46 DISCOVERY REPORTS elements and Furcilia 3 in which the elaboration of the telson and ' final definition ' of the pleopods takes place. His Furcilias i and 2 are the same as Eraser's but his FurciUa 3 telescopes Eraser's Furcilias 3, 4, 5 and 6. In this report I follow Eraser throughout, regarding the lumping of his Eurcilias 3-6 into a single stage as an unnecessary over-simplification of the developmental path he elaborated with such care and precision. As Boden (1955), who also follows Eraser, has said, 'it is less useful to "lump" the furcilia stages than to "split" them', and this became repeatedly obvious to me when I began to work out the distribution and horizontal movements of the larval population, and again when I came to investigate the swarming habit to which (p. 226) the young krill are prone. MauchUne (1959) follows Sheard, except that he suggests it would be better if all furcilias with five setose pleopods, but retaining the primitive seven-spined telson condition of Eraser's Eurcilias 1-3, should be excluded from Eurcilia 2 and regarded as a separate stage, Eurcilia 3, his Eurcilia 4 telescoping Eraser's Eurcilias 4, 5 and 6. It will become abundantly clear, however, as we proceed that neither the Sheard nor the Mauchline plan can be used to follow the growth, distribution and movements of the larvae with precision and that neither will serve to demonstrate the finer points of the structure of the larval swarms. It would repeatedly happen in fact, to take a concrete instance involving distributional confusion, or misrepresentation, that charts portraying the distribution of the last larval stage of Sheard or MauchUne would at the same time purport to display the distribution of a developmental phase, with all the hard and fast morphological characters of Eraser's EurciUa 6, when it did not in fact exist in the plankton. Marshall (1954) calls attention to some striking anatomical diflterences between this voracious herbivore and certain omnivorous or rapacious Antarctic euphau- sians of essentially more catholic taste. Peters (1955), reporting on the examination of some 8000 whale stomachs by German expeditions from 1936 to 1939, notes that the percentage of empty stomachs in the pelagic catch is higher than that found by Mackintosh and Wheeler (p. 138) at South Georgia. South-west of Kerguelen, in the early part of December 1937, he records a maximum percentage empty of 897, the pronounced local scarcity of euphausians this high figure suggests again being attributable to the fact that the expedition concerned was operating in the krill-poor West Wind drift. Marr (1955) gives a condensed account of the southern krill referring briefly to its great ecological importance, its virtually nektonic behaviour, uneven distribution, swarming habit and predominantly surface habitat. Marr (1956) gives a preliminary account of its distribution during the modern pelagic whaling season showing that the principal concentrations are confined to high latitudes near the Antarctic mainland and elsewhere only to such lower latitudes as are affected by the current from the Weddell Sea. He calls attention to the marked asymmetry of its circumpolar dis- tribution, refers again to its congregating at the surface, and shows that east of 30° E the circumpolar West Wind drift is very sparsely populated. Zimmer (1956) gives a useful summary of the work of earlier students of this species, referring particularly to the more recent papers by Mackintosh (1934), Hardy and Gunther (1935), Eraser (1936), John (1936), Bargmann (1937), Barkley (1940) and Barg- mann (1945). Dall and Dunstan (1957) record euphausian remains, undoubtedly those of E. superba, in the stomach of a humpback caught off southern Queensland in July 1956 and suggest that the krill are not confined to the Antarctic, but may sometimes occur as bathypelagic inhabitants of lower latitudes to which they would get carried in the deep north-flowing Antarctic intermediate layer. As Jonsgard (1957) points out, however, this at first sight quite sensational find, without postulating any pathological condition affecting digestion, could readily be explained if the animal had been killed immediately on arrival from the southern feeding grounds after an extraordinarily rapid passage.! j^g emphasises that even in 'burnt' whales, in which the stomach contents have been reduced to an amorphous chyme, partially digested eyes, fragments of carapaces and other hard 1 At a steady 4-3 knots (Chittleborough, 1953) it could have done it in 21 days. EARLY RECORDS AND RECENT LITERATURE 47 exoskeletal parts of this euphausian can often be detected, adding that if Dall and Dunstan are correct, ' it is hard to beHeve that in the stomachs of the other 2000 humpback whales examined at Tangalooma there were found no remains of E. siiperba'. As I have noted elsewhere (Marr, 1957) all our evidence (p. 168) points to the feeding of the baleen whales as being a surface phenomenon and not a bathypelagic one as suggested here, and that of the many hundreds of observations we have in the Antarctic intermediate layer, covering all the known migrational routes of the humpback whale, there is not one that gives the slightest sign of a deep northward transport of the krill away from the Antarctic zone. But perhaps the most serious objection to the whole suggestion that E. superba at this deep level could be carried so far from its normal habitat is that the animals supposedly so carried away would find little or nothing to eat. Nemoto (1959) agrees with the views expressed by Jonsgard and me. Brodsky and Vinogradov (1957) record the frequent occurrence of 'large concentrations' of early developmental forms of E. superba (Calyptopes and Furcilias) near the Antarctic mainland in 120° E, from observations made some time during the summer and autumn of 1956, precisely when, however, not being stated. In February of the following year the Russians in the ' Ob ' (Beklemishev, 1958a; Beklemishev and Korotkevitch, 1958) claim to have found a cyclonic movement off Enderby Land between 60° and 64° S, in the centre of which there is deep water rising to the surface from below.^ This ' new ' water they say, poor in ' plantlets ', is continually displacing the existing surface water towards the periphery of the cyclone with the result that the whole area is one in which the phytoplankton is scarce. In the centre of this vortex they report a rich population of E. superba at the surface, its density, as revealed by Juday nets, being of the order of 50 per cubic metre,^ adding that patches of these animals seen on the surface were about 100 m. in diameter. They state that this dense local accumulation of euphausians had been carried to the surface from below by the upwelling deep water, but produce no evidence that there was in fact a deep population from which such a movement could have sprung. They state too, again, however, without supporting evidence, that the 'developing young forms of E. superba live mainly in the intermediate warm layer, and rise to the surface at the Antarctic divergence'.^ In a more recent reference to this and other regions of atmospheric generated upwelling which Russian expeditions have reported to be associated with cyclonic movements along the East Wind- West Wind boundary zone, Beklemishev (1959) goes much farther, suggesting there is a deep population of krill in the warm intermediate layer which rises to the surface in the upwelling centres of the cyclones along this boundary zone. ' Antarctic baleen whales ', he writes, ' live mainly on krill, E. superba, which inhabits the moderately warm layer, and rises to the surface (the o-io metre layer) at the Antarctic divergence'. Thus he concludes the 'irregularity of euphausiid distribution is not insignificant or accidental (although Marr affirms otherwise),* but is related to non-uniform upweUing'. We shall see later, however (pp. 157-70), from a wealth of bathymetric data, that there is no evidence that the krill, in its restricted meaning as food for the baleen whales, normally lives deep as Beklemishev suggests.^ We find on the contrary that the mass of the feed, larval, adolescent and adult, spends by far the greater part of its existence in the Antarctic surface layer, at or not very 1 It has been suggested (Ivanov and Tareev, 1959) that this and similar upwellings recently reported by Russian expeditions along the East Wind-West Wind boundary zone are generated by atmospheric cyclones centred over them (see also p. 217). 2 Beklemishev (1959), citing Moore (1950) on the marked ability of euphausians to avoid towed nets, raises this figure to 500. * Or, as it is more commonly known (p. 58, Fig. 4), the northern boundary of the East Wind drift. * I do not in fact affirm otherwise. Beklemishev seems to have misconstrued much of what I said in my advance note (Marr, 1956) on the summer distribution of the krill, or perhaps he was working from an indifferent translation. 5 In a later paper (Beklemishev, i960) he refers to this deep population as 'the later Furcilia larvae and young adolescent kriir, stages in fact that Eraser (1936) long ago demonstrated are massed, evidently throughout life, dominantly near the surface, or at any rate in the 50-0 m. layer. 48 DISCOVERY REPORTS far away from the surface itself, and that it is only as Nauplii and Metanauplii, which in the strict sense are not krill at all, that for a while (p. 97) it is strictly bathypelagic. Nor do we find these deep early stages, as older larvae, coming to the surface at the Antarctic divergence any more regularly than they do to the north or south of it. In fact it is rather the reverse (p. 310, Figs. 74 and 75, and especially p. 354, Fig. 106). Hustedt (1958) reports on a collection of diatoms from krill stomachs south of Kerguelen, adding many new records to Barkley's original list. In a brief account of the distribution of the plankton in the Indian sector of Antarctica Korotkevitch (1958) again mentions £. superba, noting it as one of the more typically occurring organisms of the plankton population of the waters bordering the continental land. In a recent paper on the distribution of whales in the Atlantic sector Arsenev (19580) states that with the approach of autumn and the formation of sea ice the krill descend to depths inaccessible to the whales, rising to the surface again with the approach of spring. The same appears in Kellogg and Whitmore (1957), who, in a reference to the immense swarms of E. superba found in Antarctic waters, state that the migration of the whales towards the tropics coincides with 'the chilling of the surface water and descent of their favorite food '. It will be shown later, however (p. 169), that there is nothing in our bathymetric data to suggest that such a migration takes place. He makes some extremely interesting observations, however (Arsenev, 19586), on the distri- bution of whales and their food, noting the ' high degree of regularity ' with which the biggest accumu- lations of whales are found where the krill, as observed from the decks of vessels, is itself in great abundance. He adds, of course, that this is not always the case, whales occasionally being encountered in large numbers where no feed is visible, and vice versa, and suggests they may be congregating where there is no apparent sustenance for them because their food is too deep to be seen. Nemoto and Nasu (1958) and Nemoto (1959) are the first to show that in high southern latitudes the baleen whales do not always feed exclusively on E. superba, Nemoto and Nasu, working on the Pacific side of Antarctica, reporting Thysanoessa macrura as a major addition to the diet of fins and humpbacks, while Nemoto, working in the same area, reports many of the sei whales there to be feeding on the swarming pelagic amphipod, Parathemisto gaudichaudii} In the same paper Nemoto surveys the food and feeding habits of the Antarctic and North Pacific baleen whales, showing what a striking difference there is between the catholic diet of the northern species and that of their southern relatives. In these northern waters, he records, the swarming Euphausia pacifica, Thysanoessa inermis, T. longipes, T. spinifera, T. raschii, Calanus finmarchicus, C. cristatus, C. plumchrus and Metridia lucens, all contribute conspicuously to the diet of one species or another, while fishes such as herring, anchovy, sardine, capelin, saury, Alaska pollock, sand eel, mackerel, Atka mackerel, and the squid, Ommastrephes sloaneipacificus, in one way or another are equally important as food. In view of the position in the Antarctic, where E. superba alone is so widely eaten, it is indeed a remarkable list.^ And in so far as the north as a whole is con- cerned it is far in fact from complete, for it omits a number of species Nemoto records as occasionally or fortuitously swallowed and could of course be much extended if plankton and fishes from other boreal regions were included. SCOPE OF THE OBSERVATIONS In the fourteen years prior to the outbreak of the war the Royal Research Ships 'Discovery', 'Discovery II' and ' William Scoresby ', operating more or less continuously in far southern waters from 1926 onwards, together accumulated a vast body of data on the Antarctic plankton as well as on that of the warmer seas through which they passed on their passages to and from the whaling grounds. ' Of even greater interest is the recent Japanese discovery of a small, evidently subspecifically distinct, race of 'Pigmy Blue Whales' frequenting the warm West Wind region between Kerguelen and Heard Island (Ichihara, 1961). The small krill on which these animals were reported feeding has been identified by Mr Nemoto as Euphausia vallentini. ^ See also Sakiura, Ozaki and Fujino (1953), vvho give a shorter yet still distinctly cathoHc list. SCOPE OF THE OBSERVATIONS 49 For the first four years, from 1926 to 1929, the investigations were concentrated mainly in the Falklands sector of Antarctica, notably round South Georgia, for it was there, from the land-based whale-catchers of the time, that most of the whaling then took place (Discovery Reports, Station Lists, 1929 and 1930). From 1929 onwards, with the rise and spread of the pelagic industry (Hjort, Lie and Ruud, 1932), the observations began to be extended farther and farther afield until as time went on they eventually came to cover practically the entire Antarctic Ocean, reaching northwards in many instances from the polar continent itself, or from very close to it, far into Subtropical waters (Discovery Reports, Station Lists, 1932, 1941, 1942, 1944, 1947, 1949 and 1957). Only in the Weddell Sea south of 65° S (see Figs, i and 4) and between 30° and 60° W, and again in the Pacific sector south of 70° S, and between 90° and 150° W, are there significant gaps in the otherwise closely knit pattern of our circumpolar coverage. In both areas the pack-ice has for long been known to be exceptionally compact and impenetrable, the narratives of the early explorers. Cook (1777), Bellingshausen (1945),^ D'Urville (1842), Ross (1847), Larsen (1894), Nordenskjold (1905), Bruce (1906), Filchner (1922), Shackleton (1919) and Wordie (19216), showing that both in the Pacific sector and Weddell Sea it has existed as a permanent obstruction to navigation from the late eighteenth century onwards. Even James Weddell's much extolled high southern record of February 1823, when with the 'Jane' and 'Beaufoy ' he sailed into the great gulf which now bears his name, and, reaching 74° 15' S, found there and to the south ' a clear and navigable sea ', need not be taken, as it generally is, to indicate that that year was phenomenal and that the whole of the Weddell Sea lay open.^ Much of Weddell's southerly route, as his track-chart shows, lay approximately along the 30th meridian west longitude and his northerly route between 35° and 40° W. Both inward and outward passages, therefore, were made well to the eastward of where the hard core of the obstruction has always been found to lie. The most we can safely conclude in fact is that in February 1823 the Weddell Sea was phenomenally open between 30° and 35° W, and, as Debenham (1959) also suggests, possibly has never been so open since. It may be remarked too that where so many have failed it is extremely unlikely that either Morrell, who has been so widely discredited, in the 'Wasp', or Johnson, in the 'Henry', ever reached the high latitudes Morrell (1832) claims that they did between meridians 40° and 48° W. Even if Morrell, who claims to have sailed over so much of what is now known to be the continental land, did in fact reach 70° 14' S in these parts in March 1823, as Roberts (1958) suggests is probable, I think he must have done so well to the east of 40° W. Mill (1905), and Wordie (1921a), in his bathymetric survey of the Weddell Sea, both cast serious doubt on the claims that Morrell has made, while Kling (1922), who in June 1912 sledged over the sea ice towards 'Morrell Land', or 'New South Greenland ' as it has also been called, from the beset and drifting ' Deutschland ', is equally sceptical. But perhaps the most damning of all historians is Fricker (1900), who, drawing attention to the audacity and ' travellers tales ' with which Morrell adorned his accounts and his frequent departures into the ' realm of plausible fable ', notes with regret at the beginning of this century that the New South Greenland of the ' Henry ' together with other equally palpable inventions of this romancing sealer 'still haunt our charts'. They continued to haunt them in fact until 1930 when John Bartholo- mew and Son, Ltd, Edinburgh, published a South Polar Chart based at that time on the latest informa- tion available. One must, however, in fairness quote Gould (1929), a leading defender of Morrell though sceptical of much that he claims, ' ... we can at least be charitable to the memory of a man who has, I suggest, received far more than his due share of posthumous defamation'. The only major 1 I refer here to the recent English translation of Captain Bellingshausen's original report which was published in St Petersburg in 183 1. ^ Hobbs (1939) has made the surprising and groundless suggestion that this voyage never even took place, that it was in fact a fake and that Weddell's historic account can be relegated to the 'realm of fiction'. so DISCOVERY REPORTS I 120 Fig. I. The circumpolar observations, the open circles showing stations falling between mid-December and May, the solid black circles stations falling between June and mid-December. SCOPE OF THE OBSERVATIONS S' penetration of the Weddell Sea west of 35° W of which there is authentic record was achieved by Larsen in the 'Jason ' in December 1893. He reached 68° 10' S, not through the heart of the obstruc- tion, however, but along the east coast of Graham Land where the current (p. 314, note i) runs fast and tends to carry the ice away. Ellsworth (1938) reported open water here extending for 300 miles due south of Dundee Island in November 1935. These parts of the Pacific sector and of the Weddell Sea have repeatedly defied our efforts to explore them and it may well be that as in the past they will always remain inaccessible to vessels except perhaps to the modern ice-breakers now beginning to be increasingly used in the exploration and development of the southern continent. Even these powerful vessels, however (Ronne, 1952; Capurro, 1955; U.S.N. Hydrographic Office, 1956 and 1957a; Lakteonov, 1957; Guozden, 1959), so far have made little impression on them.^ 40° 38° ° 36° 34° 32-1 5d 5^ 0 - 0 56' 0 0 °:°°:?^^*.° ^ ° J ZjV 1 n * 0 °0 * k □ 0 0 » **' ■> 40" 38° 36° 34° 32 — = Fig. 2. Local observations. The South Georgia whaHng grounds. Fig. 3. Local observations. The Bransfield Strait region. The most recent account of the enduring obstruction that manifestly exists in the heart of the Weddell embayment is given by Fuchs (1958), who, describing the voyage of the 'Theron' in 1955-6, when, beset near 68° S, 25° W, she narrowly escaped freezing in, records the following interesting passage: 'All the time the combination of south-easterly winds and a south-westerly current was moving the whole melange of pack ice, bergs, and ship to the west and north. I became convinced from our drift of fifteen to twenty miles each day that the ice of the central Weddell Sea is always turning in a clockwise direction, a certain quantity of floes and bergs being constantly delivered to the open water along the northern edge of the ice while the remainder continues to rotate as a hard core for many years'. Shackleton (19 19) speaks in similar terms, referring to the great quantities of ice that sweep along the continental coast from the east under the influence of the prevailing current, filling up the ' bight of the Weddell Sea as they move north in a great semicircle ', and, like Fuchs, suggesting that some of this ice ' doubtless describes almost a complete circle '. The pack of the central Weddell Sea in short presents an insuperable obstacle to southing vessels. I saw something of it in January 1932 and wrote the following account of it (Marr, 1933) shortly afterwards : ' It was by far the heaviest ice that any of us had yet seen, stretching solidly to the south- ward without a single crack through which a ship might force a passage. It had apparently been 1 Less than a year ago, however, American ice-breakers, working in the eastern part of the Pacific sector, did in fact succeed in penetrating to an ice-bound coast in 100° W (Anon., 19606). 3-2 52 DISCOVERY REPORTS subjected to enormous pressure for great floes had been rafted up and were piled high in in- describable confusion. We realized then the nature of the forces which had overwhelmed the Endurance' . In Fig. I I have plotted the majority of the stations where various types of plankton net suit- able for the capture of the krill or its larvae were fished and subsequently examined.^ Unavoidably a number of stations have had to be omitted, notably in localities such as the South Georgia whaling grounds and the Bransfield Strait where there has been particularly heavy crowding. In the former locality alone 385 closely spaced observations should appear. These together with the equally closely spaced observations in the Bransfield Strait are shown in Figs. 2 and 3, only a selection of them being reproduced on the circumpolar scale. For the rest it is apparent that apart from the gaps already mentioned the Antarctic Ocean proper, taking that as meaning everywhere southward of the Antarctic convergence, has been substantially covered, while the warmer Subantarctic and Subtropical waters to the north, although on the whole less intensively fished, have not by any means been neglected.^ Throughout the summer and autumn months, that is to say from December to May, but particularly in January, February and March, much of the Antarctic Ocean, as Mackintosh and Herdman (1940) show, is relatively ice-free, or at any rate covered by navigable pack, and at one time or another during these months our vessels have reached high latitudes, on more than one occasion carrying their observations up to the shores of Antarctica itself. From June to November on the other hand, that is to say throughout the winter and spring months, the pack extends as an impenetrable barrier far to the north of the continental land^ with the result that our observations then are restricted in most sectors to relatively low latitudes. The striking disparity between the southern limits of the two sets of observations is shown in Fig. i in which open circles are employed to indicate the summer-autumn ones and solid black circles the winter-spring ones. Only in the Falkland sector between 30° and 60° W do the southern limits of the black and open circles coincide. The chief reason for this is that it is here (Herdman, 19486) that we find the main out-thrusting of pack-ice from the Weddell Sea. There is always a great deal of heavy ice about, presenting even in summer an effective barrier to southing vessels. Broadly, around the greater part of Antarctica, the average distance separating the southerly limits of the summer-autumn and winter-spring observations is from 300 to 400 miles. The gap is most conspicuous, however, in the Atlantic-Indian Ocean sector between 30° W and 60° E where the summer-autumn observations extend for as much as 600-800 miles farther south than the winter-spring ones ; and this again springs from the northerly encroachment of the ice-sheet that in winter covers so much of the Weddell stream. Thus it transpires, although our observations are spread over every month of the year, a vast area 1 This chart includes 84 stations made by the Norwegian (' Norvegia ' and ' Vikingen ' ) expeditions that visited the Antarctic between 1927 and 1931 and 63 stations made by the British-Australasian (B.A.N.Z.A.R.E.) expedition in 1929-31 (Johnston, 1937). It includes too the post-war observations of R.R.S. 'Discovery 11' in 1950-1 (Herdman, 1952). ^ In a brief reference to the development of world oceanography over the last 30-35 years Zenkevich (1958) unaccountably overlooks this massive coverage of the southern seas (already shown in many published papers), stating that in common with other parts of the southern hemisphere ' systematically acquired data on the distribution of plankton ' from Antarctic waters were very limited. ^ Continental at least in the sense that the vast ice-sheet that covers it is continental. Doubts have been expressed as to whether the whole of the underlying rock is continental, it having been suggested it may comprise the separate parts of an archipelago. Recent seismic and gravimetric measurements by Russian I.G.Y. expeditions, however (Shumskiy, 1959), show that East Antarctica at least is ' a typical continent submerged by the weight of a thick ice sheet to a depth of several hundred metres, and in some parts to below sea level. Only in West Antarctica is there the probability that a chain of mountainous islands is joined to the East-Antarctic Continent by the over-lying ice sheet'. See also Anon. (1960a) and Robin (i960). A channel does it seems cut through the supposed continent from the Ross to the Bellingshausen Sea (Bentley and Ostenso, 1961). SCOPE OF THE OBSERVATIONS 53 of the Antarctic Ocean, involving much of the richest of the whaUng grounds, has for six vital months, from June to November, never been visited by our vessels, nor in the nature of things is it ever likely to be. Indeed since the highest latitudes can only be attained in January, February and March, it is clear that much of the region south of the 65 th parallel, and probably everywhere south of the Antarctic Circle, an enormous coastal belt that will be demonstrated later to be of major importance to the whaling industry, is virtually closed to navigation for all but three months of the year — an accident of geography that will always present a problem to Antarctic oceanographers. MATERIAL AND DATA The material consists of our enormous collection of samples of the whale food, together with its eggs and larvae, which, over a period of 15 years, has been amassed since these investigations began, the principal data, in addition to copious field notes on habits and behaviour, including counts and determinations of the eggs and larval stages, and counts, measurements and determinations of sex and developmental condition of the older animals collected in plankton hauls and from whales' stomachs. I have drawn freely too upon the comprehensive reports by Fraser (1936) and Bargmann (1945), the very large series of egg and larval stage determinations, and of larval and adolescent measurements, published by the former, and the measurements and developmental condition of the 8029 adolescent and adult krill dissected by the latter having provided a major contribution to the data, without which the picture presented here of the development, life-history and distribution of this species would be very far from complete. At one time or another the majority of the former Discovery Committee's scientific staflF took part in the measuring, counting and sexing, etc., of the older krill taken in our stramin nets. Of the many however who have contributed to this work, particular mention must be made of our assistant, Mr W. F. Fry, who, while serving in R.R.S. 'Discovery II' worked continuously on krill during the two voyages she undertook between 1935 and 1939, analysing in the course of his task nearly 3000 stramin net samples. It was he too, handling some hundreds of thousands of specimens, who counted, measured and sexed the vast majority of the krill we gathered, providing much of the basic data from which the development of the older stages of this species has been worked out and contributing much to our understanding of its habit in the sea. In all, the catches of 7339 vertical, and of 5122 horizontal and oblique nets have been examined — a total of 12,461 samples. The vast majority of the vertical samples were analysed before the war, 2253 by Fraser, 1505 by Fry and 3124 by me. A further 457 were analysed by Mr Peter Foxton while serving in R.R.S. 'Discovery II' during her first post-war commission of 1950-1. METHODS IN THE FIELD Descriptions of the several types of plankton net employed in these investigations and of the manner in which they were used at sea have already appeared in the Discovery Reports, notably in papers by Kemp and Hardy (1929), Hardy and Gunther (1935), Mackintosh and Ardley (1936), Ommanney (1936), John (1936) and Marr (1938). Of the several townets mentioned in these accounts three require some further consideration here since it was through them that the great bulk of the material was obtained. They are the fine silk 70-cm. diameter net hauled vertically and referred to as the N70V, the coarser stramin loo-cm. diameter net towed either horizontally or obliquely and referred to as the N looH or N looB (or sometimes simply as the stramin net) and the so-called Young Fish Trawl or TYF, a large stramin net mounted on a 200-cm. diameter ring which was generally fished obliquely, but occasionally vertically, and referred to as the TYFB or TYFV. In deep oceanic water a routine series of hauls with the N70 V reaching as a rule down to a maximum depth of 1000 m. was S4 DISCOVERY REPORTS made as follows: 50 m. to the surface, 100-50111., 250-100 m., 500-250111., 750-500111. and 1000-750 m., the net, except for the uppermost of the series, being closed at the end of each haul. Not infrequently the N70V series was extended downwards to include the 1500-1000 and more rarely the 2000-1500 m. layers. In shallow water less than 1000 m. deep the net was fished as above down to the maximum depth permitted by the soundings. The towed nets, the coarser Nioo and TYF, were fished over the stern in the wake of the ship through a fairlead on the port quarter. In the early days of the investigations it was the practice to tow the N 100 horizontally, a flight of three such nets being so disposed on the same warp that one would fish at or just below the surface (0-5 m.), the other two at approximately 50 and 100 m. respectively. Both nets fishing at the deeper levels were as a rule closed before hauling. A radical change in towing practice took place after 1927, the treble horizontal flight with the NiooH being abandoned in favour of a single towing with an open net hauled obliquely to the surface from a depth of about 100 m. This practice resulted in considerable saving of time on station and proved to be almost as good a method of sampling the surface and sub- surface population of the whale food as the more elaborate three-level horizontal towings had been in earlier years. The routine surface horizontal net with which the investigations originally began was reintroduced in 1932 when it became the practice from time to time to fish it on a separate line some three to four fathoms long through the starboard quarter lead while the surface (roo-o m.) oblique net was coming in. This practice became increasingly adopted from 1933 onwards, becoming routine in 1935. At a still later stage, beginning with the upper subsurface strata in 1931, the systematic examination of the whole bathymetric horizon from the surface down to 1000 m., involving the loo-o, 250-100, 500-250 and 1000-500 m. layers,^ was undertaken by means of flights of from two to three, occasionally as many as five or six, oblique loo-cm. diameter stramin nets, all but the uppermost rigged for closing and suitably disposed on the same long warp. The large Young Fish Trawl or TYF, although used also in the upper strata, was chiefly employed to explore the great depths below 1000 m., where it was thought the krill, if present at all, would be so widely scattered as to elude capture altogether by a smaller net. As already mentioned it was used obliquely, and on rarer occasions vertically, and, except in the upper 250-0 m. layer, was closed at the end of each haul. Depths of 150 m. or less were determined by Kelvin Tube, greater depths by depth gauge. The fine-meshed N70V, as Fraser (1936) has shown, proved the most suitable instrument for the capture of the very young stages of the krill, more especially the eggs, Nauplii, Metanauplii, Calyp- topes and early Furcilias, and it is from the analyses of the catches of this net that our knowledge of the distribution and movements of these stages has been mainly derived. The later Furcilias and older krill including fully grown adults also appear in the catches of the vertical net but, the larvae apart, only in negligible numbers. For the capture of these older stages the horizontal or oblique stramin net has proved the more eflicient apparatus. Owing to the patchy distribution of E. superba and its tendency to gather in surface swarms, there is a large element of chance in sampling the populations and special problems arise when comparing samples from different nets. In view of the varying circumstances in which our nets were fished, in certain instances corrections have to be applied to the catch-figures to allow such comparisons to be made. These matters are discussed on pp. 57-9, and especially on pp. 278-84. ^ I am speaking here in terms of nets fished at various levels falling within the broad limits of these several vertical horizons. IDENTIFICATION OF THE SPECIES 55 IDENTIFICATION OF THE SPECIES For descriptions and figures of E. superba and of its eggs and larvae the following authorities may be consulted, those shown in italics contributing both descriptions and figures, the others descriptions only or documented notices of its occurrence. Dana (1850) Hansen (1913) Rustad (1930) Dana (1852 and 1855) Tattersall (1913) Ruud (1932) Sars (1883) Zimmer(i9i4) Rustad {ig24) Sars (1885) Hansen (1915) Eraser (1936) Hodgson (1902) Coutiere (1917) John (1936) Holt and Tattersall (1906) Tattersall (1918) Bargmann (1937) Coutiere (1906) Hansen (1921) Barkley (1940) Tattersall (1908) Tattersall (1924) Dell (1952) Hansen (1908) Illig (1930) Zimmer (1956) Zimmer (1913) The principal papers, in which the most comprehensive descriptions and best figures of the krill or its larvae may be found, are those by Sars (1885), Tattersall (1908), Hansen (1908), Hansen (191 3), Ruud (1932), Eraser (1936), John (1936) and Bargmann (1937). In practice the older stages of this euphausian are readily recognised in the plankton not only by their great size and distinctive red coloration, but also by the fact that as a general rule they are captured in such large numbers that the sheer reddish mass of them swamps or blankets the rest of the catch. The older stages of the larvae, from the First Furcilia onwards, and the adolescents both show this distinctive coloration and they too, notably the larvae, are generally taken in such immense numbers that they appear as a red cloud in the samples. In general the stages just mentioned can readily be distinguished in fixed samples (i) by the characteristic manner in which the slender abdominal segments, in a low downwardly directed arc, taper gracefully away to the rear, and (2) by the exceptionally long development of the setae of the thoracic limbs, a development conspicuously more pronounced in this than in any other member of the genus. Without having to resort to dissec- tion the adult male can be recognised at once by the large eyes, heavy chitinization, the some- what slender and reduced appearance of the enveloping carapace and powerful development of the expodites of the thoracic limbs. The eyes of the adult female are distinctly smaller than those of the adult male (Bargmann, 1937, PI. i) and when gravid the former, heavily distended with eggs, is always obvious. The eggs, Nauplii, Metanauplii and the three Calyptopis stages are also comparatively easy to pick out. In 1930, during the first voyage of R.R.S. 'Discovery 11', Eraser (1936) definitely established the specific identity of the large euphausian eggs he found in the plankton by comparing them with others shed by gravid E. superba females kept in aquaria. He remarks that their relatively great size together with their characteristic milky or porcelain white colour in reflected light renders them, particularly when present in bulk, comparatively easy to detect, even with the naked eye. Their diameter — o-6 mm. when fresh and only slightly less when preserved — greatly exceeds that of the other euphausian eggs with which they might be confused and which are often present in the samples in considerable abundance. The structure of the egg-shell which is thin, delicate, transparent and unsculptured and the fact that the egg mass almost up to the point of hatching completely fills the shell are other useful aids to identification. There is no oil globule or other mechanism of flotation which suggests they may not be buoyant. In the samples I examined I found, as Eraser found, eggs in all stages of development up to one where the fully formed First Nauplius appeared to be almost 56 DISCOVERY REPORTS on the point of hatching. At this stage the nauplius, enfolded by its swimming appendages in the globular form of the shell itself, appears to fit somewhat loosely inside the shell and there appears to be some little space, far less pronounced, however, than that figured by Ruud (1932), between it and the enveloping capsule. Fraser (1936) states that the nauplius completely fills the egg capsule and it may be, therefore, that the looseness I have described results from some slight shrinkage of the nauplius away from the shell after fixation. In eggs in which naupliar development is far advanced, however, the connection between nauplius and shell is an extremely precarious one, so much so that if eggs in this condition be subjected to unnatural strains artificial rupturing of the shell can readily be induced. At Station 2594, where a very large haul of eggs was obtained, no fewer than 1080 empty eggshells were found in the sample, all apparently having become detached from fully developed nauplii, either as a result of jostling in the net or through sudden immersion in the fixative. Fraser notes that in general the eggs are opaque and rather densely granular. I have noticed, however, that in the earlier developmental stages there is a certain measure of translucency in trans- mitted light which gives to the egg a characteristic pale golden yellow colour, both translucency and colour being most marked in the unsegmented egg and becoming less marked as segmentation pro- ceeds. In a female that had recently spawned oflF Enderby Land in April 1932, 1 found a few unshed unsegmented eggs still clinging to the almost empty ovary. They were translucent and showed the same pale golden yellow colour typical of the early developmental stages found in the plankton. It is in fully segmented eggs, already beginning to diflferentiate the incipient buds of the naupliar append- ages, that the opacity which Fraser describes is most noticeable. I found only nine specimens of the First Nauplius, all, as were the two that Fraser found, from shallow coastal water. Of these one had evidently just been hatched, for it still had the swimming appendages folded about the body in the globular form illustrated by Fraser (1936, Fig. i a), while the others, their limbs being fully extended, had obviously been swimming about freely when caught. In the eight free-swimming specimens the dorsal aspect of the body is more perfectly oval than shown by Fraser (1936, Fig. ib) with just a faint indication of broadening at the posterior end. In colour all present the typical milky or porcelain white appearance characteristic of the egg itself and in length they range from 0-63 to 0-66 mm., the measurements of the First Nauplii Fraser found. I have nothing further to add to Fraser's description, which he remarks is that of a normal First Nauplius. The remarkable scarcity of this stage in the plankton upon which Fraser also comments, is discussed on pp. 98 and 205. Though somewhat less pronounced the characteristic whiteness of the egg and First Nauplius persists in the Second Nauplius, in the Metanauplius and up to the Third Calyptopis stage, and this, particularly when these stages are present in mass, renders them quite conspicuous in the samples. LABORATORY METHODS In analysing the samples from the vertical nets the larvae were picked out with the unaided eye and identified as to stage under a binocular microscope, small to moderate samples of from one to 300 or 400 individuals being counted in their entirety, larger samples fractioned as described by Mackintosh (1934, p. 70). In the larger stramin nets the gatherings of larvae, adolescents and adults often ran to thousands, tens of thousands and even hundred of thousands of individuals and with such samples fractional counting was always necessary. Throughout we used the method described by Mackintosh, in many instances, for the bulkier samples, after pre-fractioning by the jug- or bucket-ful. Samples of the order of from one to 500 or so, in the vast majority of instances were counted and measured outright. DISTRIBUTION IN OUTLINE 57 DISTRIBUTION IN OUTLINE Although the main section on the distribution does not follow until later it is necessary at this stage to anticipate it with a short account of its major aspects and their relationship to certain physical and geographical features of Antarctica and its surrounding waters to which repeated reference will be made in the text which follows. The principal water masses and surface movements of the circumpolar sea and the physical and geographical features of Antarctica with which the krill are associated are shown in Fig. 4, the hydrological boundaries^ and surface movements following Colbeck (1905), Deacon (1937, pp. 14-20, Fig. 4), Clowes (1934), Herdman (1959), Mackintosh (1946) and Nasu (1959), the mean position of the summer ice-edge following Mackintosh and Herdman (1940, Pis. Lxix and Lxx). It has long been known that euphausians on the surface in daylight are difficult to catch, that with rapid movements they can dodge out of the way of an approaching net or react by mass sinking when disturbed by the hull of a moving vessel. Tattersall (1924) seems to have been the first to call serious attention to this phenomenon and since then many authorities, notably Bigelow (1926), Mackintosh (1934), Hardy (1935), Einarsson (1945), Moore (1950), Barham (1957) and Zelikman (1958), have called further attention to how readily they can avoid capture either as individuals or by concerted action. At an early stage in my study of our material I found that our stramin nets, above all our surface stramin nets, were consistently yielding smaller, in most instances far smaller, gatherings of krill in the East Wind zone than they were in lower latitudes. I began at once, however, to suspect something wrong, for in these high latitudes the whales have for long been known to be plentiful and today yield a rich and profitable harvest to the whalers. Obviously then there must be enough for them to eat. This anomalous situation, abundance of well-fed whales, but an apparently indifferent food supply, first of all led me to suspect that in this summer sunlit zone our nets could not have been behaving with the same efficiency as they were in lower night-dark latitudes, and eventually led me to go into the whole question of the relative efficiency of our variously used apparatus and of the relative success with which in diflFerent places and under various conditions they were sampling the swarming krill. A full account of the results of this investigation is given on pp. 258-68. Meanwhile, it is necessary to call attention to two of the more important facts that have emerged, (i) Our nets on the surface in daylight, that is, our i-m. diameter (0-5 m.) stramin nets, are so easily avoided by the older animals once they have grown to about 20 mm. long that such data as they provide as to the presence or absence, abundance or scarcity, of these stages becomes for all practical purposes valueless. (2) The same nets fished obliquely from about 100 m. to the surface, although producing by and large quite adequate samples of the total surface population, do not in general by any means sample it with the same success, either by night or by day, as the stramin nets on the surface do when fished in darkness, or, if the escaping over 20 mm. population be ignored, in daylight as well.^ In the East Wind zone a very large number of the surface nets we fished were fished in the prevailing summer daylight of these high latitudes, and throughout this period of absence of darkness, or at best twilight, their gatherings were negative or consistently low; so low in fact that in most instances I was ^ Wexler (1959) proposes that the classic Antarctic Convergence of Deacon should now be referred to as the Antarctic Divergence, arguing, from a few widely scattered bathythermograph records, that it is a region of cold upwelling induced by horizontal divergence of the surface layer. His data, however, seem not to be enough for this to be conclusive and in any case Garner (1958), also using the bathythermograph, finds no evidence of such upwelling along what has for long been recognised as a convergence zone. ^ While there is good evidence that the larger and more active animals at times remain on the surface by day and at such times readily avoid the surface net, there is also evidence (p. 273) that some desert the surface in daylight and, sinking to slightly deeper levels, contribute further to the poverty of our daytime gatherings of the older stages of the surface population. S8 DISCOVERY REPORTS 50° 60 A.\.\ .iil.H.l.iil.CTilZ 30° I^ID^I i.j.j.i.i.i.Tz^^ic:!^ -WD NORTHERN BOUNDARY OF WEDDELL DRIFT -EWD NORTHERN BOUNDARY OF EAST WIND DRIFT PRINCIPAL SURFACE MOVEMENTS OF THE COLDER ANTARCTIC WATER CIRCUMPOLAR WEST WIND DRIFT -^ ICE EDGE^ — MEAN POSITION IN MARCH \.^.\A,\.\,\.).i.iA.j:znz. SOUTH AFRICA ! ' '.!■>.>-> w^ 120^ 150° W 180"E 150° Fig. 4. Map of Antarctica and the circumpolar sea showing physical and geographical features, hydrological boundaries and surface water movements, referred to in the text. It should be noted that the Antarctic convergence can be located more definitely than the boundaries of the Weddell and East Wind drifts, but all these boundaries are subject to variation or uncertainty. DISTRIBUTION IN OUTLINE 59 compelled to ignore them and to turn to the gatherings of the useful though less efficient oblique net before any reliable estimate of the actual abundance of the population in this, it seemed, krill-rich zone could be made. At many stations too in more northerly waters, where there are dark nights to work in, the surface net even in darkness was never fished, so that much of our data from both high and low latitudes does in fact come from the oblique net alone. In so far then as surface nets were concerned it was clear that in the higher latitudes, with their long or continuous hours of summer daylight, our sampling of the krill was being conducted at a distinct dis- advantage. In short it became obvious, having regard to the greater sampling power of the horizontal net when used on the surface in northern darkness, that the oblique net, upon which we had so largely to rely in the East Wind zone, was not providing us with a true picture of the abundance of the krill in this coastal stream, nor was it doing so in the lower latitudes where it alone was so widely used. In view of the manifest disadvantage under which the summer East Wind samples were collected, and of the lesser efficiency of the oblique vis a vis that of the night surface net, it became clear to me that a true and unbiased picture of the relative abundance of the krill in the circumpolar sea could not in fact be presented until these differences had been resolved. Briefly and simply I did this by taking the gatherings of the surface (0-5 m.) and oblique (loo-o m.) nets and comparing them at all stations where both nets night and day were fished simultaneously, finding that, taking the surface population as a whole (but ignoring the valueless data relating to the escaping over 20 mm. population provided by the surface nets in daytime), on the average the surface net was twice as efficient a sampler of the larvae and four times as efficient a sampler of the adolescents and adults as the oblique, the adolescents and adults referring to all individuals over 15 mm. long. In all instances, therefore, where oblique nets alone were used, or where, as in daylight surface nets, if used, could not, or virtually could not, sample the escaping (over 20 mm.) population, the gatherings of the oblique nets have been multiplied by 2 or 4^ as requisite in order to bring them into line with the higher orders of abundance repeatedly revealed by our surface nets. A correction has also been applied to all towed net samples to equate them to a 30-minute haul; for example the catch of the 20-minute oblique net fishing through loo-o m. is multiplied by 1-5 to the nearest whole number. Much of more general biological interest has also come out of this investigation, but as I have said presentation of full details has been deferred and only major findings mentioned here. I have deferred the matter deliberately so that the reader approaching the main distributional part which begins on p. 284 will come to it with the full facts of the sampling, and above all with the corrections applied to certain catch-figures, freshly stamped in his mind. In the meantime it should be noted that the further corrections explained on pp. 278-84 have been applied as necessary to the quantities shown in certain tables and figures, especially in the construction of Tables i, 2, 37, 39 and 47, and of Figs. 56, 5c, 13, 24, 25 and 26. They do not, however, bias such general conclusions as have been drawn either from these five tables or six figures. The same conclusions can be drawn, and in fact originally were drawn, ^ from data presented without them. Whatever value they may have elsewhere it is above all in the East Wind zone, with its virtually nightless summer, that their usefulness comes most to the fore, in so far as they serve to offset there the severe handicap under which so much of our material was collected. The gross or absolute distribution of this species, based on every occurrence we have recorded from the egg to the adult state, is shown in Fig. 5 a and the distribution of its major concentrations, based on gatherings of not less than 1000 in the stramin nets and of not less than 100 per 1000 m. vertical haul, in Fig. 5 b, the figures together presenting in sharp relief such regions as have repeatedly been 1 See p. 279, Tables 57-60. - In the course of many early and long since rejected attempts to tabulate our data and plot it on the charts. 4-2 6o DISCOVERY REPORTS Fig. 5a. Gross or absolute distribution of the Antarctic kriU, the positive observations as solid, the negative observations as open, circles. Based on all nets. DISTRIBUTION IN OUTLINE I ; , 1. 1 . 1, 1.), J,, M. /,;.;,/, lu ,),! ,i, i,i .1, i,i, i^,i,^,iii,^,^i,j. ,' . t 6o; 90 120 150° W 180°E 150° Fig. 5 b. Principal concentrations of the Antarctic krill. 62 DISCOVERY REPORTS 20 Fig. 5 c. Principal concentrations of the older Antarctic krill. DISTRIBUTION IN OUTLINE 63 recorded as rich in these euphausians and such others as, equally often, have been recorded as poor.i From this presentation of the data it will be seen that while the species is circumpolar as has always been supposed, and as Baker (1954) has now definitely shown, the northern limit of its occurrence however falling increasingly away to the south as one travels from the Atlantic side of Antarctica eastwards round the continent, in the major aspects of its distribution (Fig. 5 b) it exhibits a circum- polarity of marked asymmetry. For instead of being evenly distributed all round Antarctica approxi- mately between the same latitudes throughout as certain other circumpolar plankton animals (John, 1936; Baker, 1954; David, 1955) are known to be, it manifestly has on the one hand a high latitude, and on the other a low latitude distribution closely associated with the surface movements of the colder Antarctic waters. In fact it is clearly never found in any real measure of abundance except in (and, as will be shown later, under) two great surface streams — the west-flowing East Wind drift moving coastwise in high latitudes all round Antarctica and the great outflow of cold water to which this coastal stream gives rise as, moving westwards into the Weddell Sea, it strikes the projecting mass of Graham Land lying athwart its path and, as the Weddell current, or, as some prefer to call it, the Weddell drift, is deflected northwards and eastwards into relatively low latitudes. In brief it is enough for the present to observe (i) that the main concentrations of the krill, whatever the stage of its development, are confined almost exclusively to these two great currents and to the South Georgia and Bransfield Strait whahng grounds to which they penetrate, and (2) that the rest of the circum- polar sea, or West Wind drift as it has been called, the whole vast expanse of it eastwards of 30° E where the Weddell stream appears to lose its impetus, has little to offer as a feeding ground for the questing whales.^ Indeed but for the existence in it of certain strictly limited areas afl^ected by tongues of cold water deflected from the East Wind region by submarine ridges, the West Wind zone as far as the distributional problem is concerned could virtually be disregarded. A notable instance of such deflection is the movement associated with the Kerguelen-Gaussberg Ridge, the northerly encroach- ment of cold water to which it gives rise carrying with it minor concentrations of euphausians, almost exclusively, our data show, furcilias or early post-larval forms, far to thenorthof their normal latitudes. In so far as this particular deflection is concerned it may be remarked that along the Antarctic coast south of Australia the East Wind stream flows strongly, the Australian expedition of 191 1-14 (Davis, 1919) reporting it turning sharply to the north-west in 97° E where the projecting mass of Termination Ice Tongue 3 lay athwart the westerly flow. Other instances of East Wind encroachment into the West Wind zone will be noted, for example north of the Balleny Islands, north-east of the Ross Sea, west of Graham Land and in the Drake Passage. As an indicator species, above all as an indicator of the major movements of the great Antarctic surface streams, E. superba in fact may be regarded as supreme. The overriding importance of the East Wind- Weddell surface stream as a carrier of the krill, and the relative unimportance of the West Wind drift, is demonstrated again in Fig. 5 c which shows all gatherings of 100 or more of the older stages of the whale food* (euphausians over 20 mm. long) which (p. 147) provide the major portion of the diet of the whales. This figure illustrates too how in the depth of winter, with the ice-edge (Herdman, 1953) lying far to the north of its summer mean, searchers would fail to strike the krill in any substantial measure of abundance throughout a vast area of the 1 How poor in fact such regions are will be shown repeatedly in the main section on the horizontal distribution which begins on p. 284. * How little will again repeatedly be shown in the main distributional part. See also the monthly West Wind catch-figures given in the Appendix. ' Termination Ice Tongue (Cumpston, 1939) has now broken away. * All, that is, but for around South Georgia where our observations (Fig. 2) are too crowded to permit every such gathering to be shown on the circumpolar chart. 5^ DISCOVERY REPORTS circumpolar sea, a belt extending virtually all round Antarctica east about from 30° E to 40° W. They would strike it in fact only round South Georgia and along a narrow northerly strip of the Weddell drift, and elsewhere only south-east of Kerguelen where there is strong deflection from the East Wind zone. They would probably not, however, strike it between 30° and 60° E, for although this is a region affected by overflow from the Weddell stream such overflow it seems (p. 384) occurs principally in spring and not, on any major scale, at other times of the year. Between 30° and 95° E, it will be seen, the krill appear to be concentrated at or very near the ice-edge. This in fact is only an appearance that springs from the coincidence of the winter ice-edge with the northern limit of abundance of this species between those meridians. PHYSICAL ENVIRONMENT Pack-ice Among the many references to this species in Antarctic literature there are some stressing its abundance in the open sea, others that it is essentially a creature of ice-infested water, par excellence of the polar pack. The latter predominate. Borchgrevink (1895), for instance, referring to the multitudes that frequent the pack-ice near the Balleny Islands, naively declares that they are 'usually found to be swimming about in cavities in the ice-floes, evidently seeking a refuge from their enemies, the whales, which feed principally upon them '. Bull (1896) again, commenting on the scarcity of whales encountered by the 'Antarctic' in the inner ice-free zone of the Ross Sea in January 1895, states that this was rather to be expected since the small crustaceans on which these animals fed were most abundant 'under and between the floes '. In another reference to this ice-free zone he adds, ' Only a few whales are observed, from which we conclude that they prefer the ice, with its millions of red "shrimps" and molluscs'. Racovitza (1903) remarks on its great abundance in the pack-ice of the Bellingshausen Sea, Andersson (1905) on the muhitudes that occur at the surface 'besonders neben eisschoUen ', while Holt and Tattersall (1906), referring to the Ross Sea area, say, 'All were taken outside the barrier ice, and as Mr Hodgson seems to have fished the waters below the ice very thoroughly, it may be taken that E. superba is a creature of the open sea'.^ Trouessart (1907) refers to the Euphatisia 'qui forment de veritables bancs dans la mer libre et qui s'avancent rarement sous la glace', Menegaux (1907) and Liouville (1913) to its abundance near icebergs, Wilson (19076) to the vast numbers 'frequenting chiefly the pack-ice, the edges of the icefloes, or the foot of the ice cliffs which form the sea faces of the Barrier 2 snow plains ' and Gain (i 91 4) to the ' enormous bancs ' of this species ' frequentant le plus souvent la lisiere du pack, la voisinage des icebergs et celui des glaciers en bordure du continent '. Brown (1920) remarks on the local abundance of these animals so often betrayed by congregating birds, adding 'but their scarcity in ice-free waters is characteristic. Among the pack they seem more plentiful '. Risting (1922) also remarks on the prodigious quantities that frequent the ice-edge, ' hvori blaavalen fraadset, idet den gik frem og tilbake langs iskanten'. Mackintosh and Wheeler (1929) record its enormous abundance in the ice-free waters round South Georgia, adding that it is to be found in dense shoals usually in the neighbourhood of land. Ruud (1932) records precisely the opposite, noting its occurrence far from land and known banks, but always in close proximity to the pack. He continues, ' Our whalers always tried to follow the drifting ice, partly in order to be in water which was calm enough for handling the carcases and partly because the blue whale prefers to frequent 1 It should be noted, however, that Mr Hodgson's negative results below the ice were simply due to the fact that he was fishing in a part of the Ross Sea where E. superba is now known (p. 124) not to occur. ^ This must refer to the Ross Barrier and it could not have been E. superba that Edward Wilson saw swarming there. It must in fact (p. 124) have been E. crystallorophias. PHYSICAL ENVIRONMENT 65 the edge of the ice or even the larger lanes between the pack-ice. Evidently, therefore, there must be great quantities of krill out at sea among the icebergs and drifting floes, and not chiefly near land and coastal banks as suggested by Mackintosh and Wheeler. ... All whalers in the Antarctic ', he remarks later, ' know by experience that the whales are often to be met with near icebergs and packs, and that the krill is most abundant there'. Rustad (1934) has much the same to say, observing that the ' tendency of the whales to prefer the vicinity of the ice ', is obviously caused by a similar tendency in E. superba, while Bennett (193 1), referring to the diatoms with which the floes are so often stained, states that 'this fact is probably the clue to the abundance of whale food near ice'. Fraser (1936) concludes that the ice-edge is ' predominantly and almost exclusively the locus of adolescent krill and not of adult krill ', at any rate at the time of year (October) just prior to when spawning is believed to commence, adding later that the habitat of the late Furcilia stages too is also predominantly at the ice-edge. John (1936) also remarks on the abundance of larvae and adolescents and scarcity of adults at the ice-edge, Feltmann and Vervoort (1949) stating that the surface population as a whole is most abundant in the 'immediate neighbourhood of the drift ice'. Ommanney (1949) suggests too that it is along the edge of the pack that the richest fields of the krill occur. From the mass of data bearing on this controversial matter that has now been accumulated, it can be shown that while the foregoing statements are perhaps largely true of the particular areas or times of year to which they refer, they cannot be applied universally to the krill throughout its whole circumpolar range or throughout its whole developmental history. In presenting these data, using the material from our oblique (loo-o m.) and horizontally towed surface stramin nets, I have divided the surface population into three broadly differentiated developmental phases: (i) krill under 16 mm. long (vast majority larval), (2) krill 16-20 mm. long (very young adolescents), and (3) krill over 20 mm. long (older adolescents and adults). Since, however, the older stages of this large euphausian are so active (Mackintosh, 1934; Marr, 1955, 1956), avoiding the surface net, in some measure at least, even in darkness (p. 265) and by day (p. 259) with relatively enormous success,^ the daylight surface gather- ings of (3), the most successful escapers, have been omitted from the data presented. The relative abundance of these three developmental phases at the ice-edge and in the open sea is shown in Table i . In this, and in Table 2 which follows, stations counted as ' at the ice-edge ' include some in close proximity to it, within about half a mile or less, and some actually inside the ice-belt itself, stations ' in the open sea ' including some with the ice-edge possibly not very far away but not in sight. The vast majority however of those counted as in open water lie at distances ranging from ten up to several hundred miles away from the nearest ice. Looking at the figures broadly (Table i) we find little to suggest that the ice-edge is predominantly and almost exclusively the locus of late larval and adolescent krill as Fraser concluded, or to support the view, so widely held by the whalers and others, that the whale food as a whole is most abundant near the pack. Indeed, during the autumn months, especially in March and April, ^ we find conditions directly at variance with Fraser's findings. For then enormous numbers of the under 16 mm. class, principally Calyptopes, but including both early and late Furcilias, are present in the open sea, whereas at the ice-edge the numbers are conspicuously smaller, virtually no larvae whatsoever having been recorded there in March. In winter, it is true, the under 16 mm. group, now represented mainly by the late Furcilias with, in June, the last surviving Calyptopes, undoubtedly occurs in enormous numbers at the ice-edge, but it certainly does not predominate there, there being equally enormous * In so far as night evasion alone is concerned Moore (1950) goes as far as to say that with certain far smaller but highly luminescent euphausians the night-time surface density has occasionally appeared to the eye to be as much as 75 times greater than the corresponding density revealed by nets. - Our data for May are scarce and have been combined with the April data. 66 DISCOVERY REPORTS Table i. Relative abundance of E, superba at the ice-edge and in the open sea based on average catch- figures, data for the East Wind-Weddell surface stream in roman type, West Wind data in italics. Numbers of net hauls upon which annual averages are based are shown in parentheses {For corrections, see pp. 282-3) Month Sept. Oct. Nov. Seasonal average Dec. Jan. Feb. Seasonal average March April-May Seasonal average June July Aug. Seasonal average Under 16 mm. , ' > Ice-edge Open sea 16-20 mm. Over 20 mm. Number of net hauls 4041 24 1204 6 1451 4 2018 8 158 6 42 1 432 178 6343 3301 8923 6217 47 411 29 1213 3 662 9 780 12 10 5 6 4 1 4130 30 1883 10 3319 23 8313 6829 4714 4 5903 1 Under 16 mm. 3HS 18 (41) {38) 2749 15 (638) {313) Ice-edge Open sea Spring 484 ^9 1612 2 1139 23 1201 21 Summer 288 31 1 22 93 2 Autumn 125 12 32 2 665 6 1472 37 967 25 22 45 2 36 7 79 Winter Annual averages 16-20 mm. Ice-edge Open sea Ice-edge 1 201 21 '28) Open sea 967 25 (239) {156) Ice-edge Open sea Ice-edge Open sea 31 1 598 2 160 ^7 344 13 SI 12 1 162 26 710 761 22 1 147 611 927 18 10 654 1 217 569 33 480 22 727 9 1826 j5 1791 44 1566 22 1607 3 87 2 1058 3 329 92 20 152 6 Over 20 mm. Ice-edge Open sea 655 "74 17 17 (91) (1058) [57) (450) 8 6 IS 2 7 20 30 28 II 7 21 21 12 44 25 ID 1 17 2 5 9 3 1 II io 44 32 71 20 124 239 156 "5 -^7 186 87 213 SH i77 19s 76 no 41 305 117 14 24 29 3 51 13 94 40 PHYSICAL ENVIRONMENT 67 numbers in the open sea. In spring too this group, consisting of FurciHa 6 and early adolescents, and the later adolescents of the 16-20 mm. class are exceedingly abundant at the ice-edge, but again there is no evidence of this being the exclusive habitat of these stages, judging from the vast numbers of both groups, particularly of the latter in November, we find in ice-free water. As for the older stages, the scarcity of adults at the ice-edge upon which John and Fraser remark is seen to be confined exclusively to the winter months, June, July and August, when there is a corresponding, although less pro- nounced, scarcity of these stages in the open sea. The winter scarcity of older animals, however, both at the ice-edge and elsewhere, is perhaps hardly surprising when one considers the magnitude of the depredation that must be wrought by the multitudes of whales and other creatures that devour them during the spring, summer and autumn months. We have every reason too to believe (p. 399, Fig. 137 and p. 404, Fig. 139) that the dying-off of the adults that have paired and spawned contributes further to the winter decline. A comparison of the average catches at the ice-edge and in open water over the whole year shows that in both regions of abundance there is little significant difference in the average catch-figures for the three size groups wherever they are encountered in Antarctic waters. In order to obtain strictly comparable annual catch-figures the averages shown at the bottom of Table i are based exclusively on the months during which the three size groups, both at the ice-edge and in open water, severally have their optimum range and, as Table i shows, are at their maximum abundance. Thus, taking first the average annual catches at the ice-edge, the months principally affecting the under 16 mm. group are June to November, the 16-20 mm. group September to November and the over 20 mm. group September to May. The corresponding months for catches in the ice-free zone are March to November for the under 16 mm. group, and for the 16-20 mm. and over 20 mm. groups as for the ice-edge. Taking now a regional view of the matter it will be seen from Table i that the enormous numbers of larvae and adolescents we find at the ice-edge at certain times of the year is a phenomenon strictly characteristic of the East Wind-Weddell surface stream, and not of the vast extent of the circumpolar ice-edge wherever elsewhere it may be encountered. The ice-edge eastwards of 30° E, which, as Mackintosh and Herdman (1940) show, reaches far northward into the West Wind drift from May to December, is clearly, certain special parts of it excepted (see below), not involved, for whereas the average annual ice-edge catches of the under 16 mm. and 16-20 mm. groups in the East Wind- Weddell stream are 3145 and 1201 respectively, the corresponding figures for the ice-edge of the West Wind drift are a mere 18 and 21. The abundance of larvae and early adolescents at the ice-edge reported by John (1936) was recorded at one station in the East Wind drift, where 2046 larvae were encountered, and at thirteen stations located in two of the several cold tongues of this coastwise current that (p. 58, Fig. 4) get deflected into the West Wind zone (the 'special parts' referred to above) under the influence of submarine ridges. Even in these special parts, however, judged by East Wind-Weddell standards, John's ice-edge abundance was not a very marked one, the average catch- figure for the thirteen net hauls upon which it rested being only 123. The situation in the East Wind zone proper, disregarding that is the several tongues of it that encroach upon the West Wind drift, does apparently reveal (Table 2) a decided preponderance of larvae at the ice-edge in April and May. The preponderance, however, is unreal simply because in these high latitudes, as the distributional charts show (p. 322, Fig. 84; p. 348, Fig. 102), the main mass of the larvae is confined to a relatively narrow coastal belt which at this time of year is already almost completely frozen over so that in effect there is virtually no open East Wind water any longer accessible, such as there is being well outside the principal region of larval abundance there. 5-a 68 DISCOVERY REPORTS Table 2. Relative abundance of E. superba at the ice-edge and in the open sea in the southern or East Wind zone. Numbers of net hauls upon which annual averages are based are shown in parentheses {For corrections, see pp. 282-3) Number of Under 16 mm. 16-20 mm. Over 20 mm. net hauls Month Ice-edge Open sea Ice-edge Open sea Ice-edge Open sea Ice-edge Open sea Summer Dec. 17 — II 12 26 257 2 6 Jan. 9 10 33 72 280 236 10 72 Feb. 1 — 22 I 710 863 12 73 Seasonal average 6 5 26 35 474 540 24 151 Autumn March — 2 125 31 1147 354 10 56 April-May 378 41 14 — 717 225 6 21 Seasonal average 142 12 84 22 986 318 16 77 Annual averages Under 16 mm. 16-20 mm. Over 20 mm. Ice-edge Open sea Ice-edge Open sea Ice-edge Open sea 378 41 58 34 679 465 (6) (21) (32) (201) (40) (228) Observations in the eastern part of the Weddell drift, repeated in practically the same positions at different times of the year, throw much new light on the real extent of the association of the krill with the Antarctic pack. At Station 2316, for example, in 57° 15-5' S, 01° 13-4' E, 58,488 larval krill, the vast majority Second Calyptopes, were taken in the surface net on 14 April 1938, the ice-edge at this time lying in 65° S, well inside the northern boundary of the East Wind drift, and more than 460 miles due south. The larvae continued to be taken in enormous numbers in the ice-free water southward of Station 2316, their abundance ceasing abruptly south of Station 2320 in 61° io-6' S, 00° 437' E, the combined catches at Stations 2321, 2322 and 2323, the last at the ice-edge itself, amounting to only twenty-eight individuals in all. The larvae are thus seen to have been concentrated in a belt of open water some 250 miles wide, a fact which incidentally provides a useful indication of the probable minimum width of the Weddell stream in this particular meridian. Approximately a month and a half later, although not in the same year, another station was made very close to the position of Station 2316 — Station 1781 on 2 June 1936, 57° 41-8' S, 00° 19-8' W — where over 10,000 larvae, Furcilias 1-4, were taken in the surface net. As the winter was now well advanced the whole of the richly populated area traversed in April 1938 had become frozen over, and Station 1781, instead of being in open water as it would have been had it occurred earlier in the year, lay among the scattered young floes at the edge of an ice-field that stretched unbroken to the south. Later in 1938, on 18 August, a second station. Station 2394, was made at the ice-edge in 57° 18-5' S, 00° 52-2' E, a position little more than ten miles away from that of Station 2316, 14,976 young krill, Furcilia 6 and very young adolescents, being taken, as before in the surface net. Repeated observations in more easterly parts of the drift reveal a similar situation. At Station 2344 in 56° 18-4' S, 19° 32-6' E, 35,520 larvae, principally Furcilias 1-3, were taken on 26 April 1938 with the ice-edge lying in 67° S, again deep inside the East Wind zone and more than 640 miles to the south. Very few larvae were taken at ten stations made southward of Station 2344, and only two at the ice-edge itself. The mass of the larvae in this meridian were confined to a belt of open water PHYSICAL ENVIRONMENT 69 approximately 160 miles wide between Stations 2344 and 2346 in 53° 35-5' S, 19° 29-3' E, a consider- ably narrower belt than that encountered less than fortnight before along the meridian of Greenwich, but not inconsistent with the assumption that it is somewhere about here, in about 20° E, that the Weddell current is beginning to lose its momentum and is probably dying away. On 23 August of the same year, with the pack now practically up to the position of Station 2344, Station 241 1 was made on its northern edge in 56° 25' S, 19° 54-7' E, 11,177 larvae, predominantly Furcilia 6, being taken along with a few adolescents. Other instances of both summer and autumnal ice-free abundance with corresponding abundance at the ice-edge in winter could also be given. The above, however, are enough to show that it is simply the advent of the polar winter and the resultant freezing over of the sea that leads to the apparent concentration of the larvae at the ice-edge upon which John and Eraser remark. At the same time it is easy to see how these authorities, both former colleagues, were left with this impression, for it was not until after their work was pubUshed that these repeated Weddell observations, which have contri- buted so much else to our understanding of the whale food distribution as a whole, were for the most part made. It must be admitted that the existence of young or older stages of krill, in places remote from the nearest ice at the time of sampling, is not in itself enough to demonstrate that their distribution is entirely independent of that of the pack-ice. The figures for ' the open sea ' in Tables i and 2 represent our average gatherings from many stations, some of them far, others not so very far, from the ice. Many in fact are no more than 10-20 miles from it, although equally many are much farther away, especially in the Weddell drift and round South Georgia where they are separated by some hundreds of miles from the nearest pack. Now since the presence or absence of pack-ice in one region or another is seasonal it could be argued that the krill at some early stage, notably for instance when as a First Calyptopis (p. 97) it reaches the surface from deep water, is in fact in some way dependent on the presence or proximity of sea ice, and that when the pack melts (or retreats) it can as it were fend for itself when left behind 'in the open sea'. Then indeed its distribution would at least be affected by the ice. It is fair too to add that in the Weddell drift a great belt of ice persists in a low latitude across the Atlantic sector well into summer, being separated then by open water from a more southerly ice- belt girdling the Antarctic continent. This northern ice appears to last through December, but breaks up and melts away in early January so that by the end of the month it has vanished except in the more westerly parts of the current near the South Sandwich and South Orkney Islands. Populations of krill left behind in the Weddell stream would then be hundreds of miles from any existing pack-ice, but it could be said that some of them at least had been brought there along with the ice drifting eastwards from the northern part of the Weddell Sea, and that their distribution was not, therefore, entirely independent of the pack. Obviously, however, the massive accumulation of young surface forms, mainly Calyptopes and early Furcilias, that seems to fill the ice-free waters of the Weddell stream in autumn (p. 363, Eigs. 112 and 113), cannot be said to have been 'left behind ' by the retreat of the pack. On the contrary the great bulk of this teeming population could not yet it seems have been ' overtaken ' by the advance of the pack, for we find it there, on the whole an obviously not long born community, long after the northern ice has gone and long before the winter ice in this sector, creeping up from the south, has begun to approach even remotely near the low latitudes in which the Weddell stream is flowing. In fact the idea that the krill in its very early stages is in some way de- pendent on the pack, and is perhaps left behind in open water when the ice retreats, can only be true if it should prove that the advent of the larvae at the surface (see again p. 97) is exclusively or very largely an ice-edge phenomenon, and not, as we so repeatedly find (p. 301, Eigs. 70 and 71), a by no means uncommon occurrence in the open sea. It is distinctly possible of course, indeed I think likely, ^o DISCOVERY REPORTS that many of the earUest larval 'risings', as I have called them (p. 204), take place well inside the ice- belt, especially (Fig. 4 and p. loi. Table 17) in Weddell West, where no research ship in fact, at the right time of year, has yet penetrated, and in these cases the krill, very early in its developmental history, would clearly be dependent on the pack, or if not dependent, at any rate associated with it. In the more easterly parts of the Weddell drift (p. 311, Fig. 75 and p. 325, Fig. 86) we find as I have said an abundance of young surface forms long after the ice there has gone. In Weddell East for instance (Fig. 4), as late as mid-April, we strike an enormous concentration of Calyptopes (Station 2316) which clearly must have sprung (p. 313) from a rising that originally took place in open water farther west long after the last of the ice had vanished from this region. Still later, in May and June (p. 317, Fig. 79 and p. 328, Fig. 88), we find Weddell East, still largely ice-free, teeming with Calyptopes and early Furcilias, and these again could hardly have been carried along with ice that must have melted away some four or five months earlier. Finally, there is evidence from the well-sampled waters round South Georgia which are never^ invaded by pack-ice at any time of year. In these island waters it appears (p. 190) there is little if any successful spawning, the annual recruit- ment of the population being largely maintained it seems by surface-borne incursions of larvae, principally, our data (p. 327) show, furcilias, having their origin in the Weddell Sea. Such larvae must spring from Calyptopis risings a long way to the south or south-west, some of them no doubt in the ice itself, but since the first large scale incursions, at their earliest, do not appear to take place until April (p. 422) it seems clear enough, from the mean position of the summer ice-edge (p. 293, Fig. 65), that the time elapsed between the earUest major appearances of the Weddell larvae at the surface in January (p. 310, Fig. 74) and their advent at South Georgia must be spent by the young drifting forms largely in a packless sea. It is clear too, from the virtually year-long ice-free condition of these island waters, that once the larvae get there and subsequently grow to the adult state, they must go through this lengthy developmental phase entirely independently of the pack. So far we have been considering only the fringe of the circumpolar pack and its relation to the distribution of this species. Let us turn now to another matter, also of controversial opinion, the question of the abundance or otherwise of the krill deep inside the Antarctic ice-fields. The sea ice of Antarctica may broadly be divided into (i) non-permanent ice which, as Mackintosh and Herdman's ice-charts show, at certain times of the year, notably in winter and spring, but not at others, spreads over large areas of the Weddell and West Wind drifts and over the whole extent of the East Wind drift, and (2) permanent, or at any rate more or less permanent ice-fields of variable extent such as are found in the Weddell Sea west of the 30th meridian, in the Pacific sector south of 70° S, and along the northern fringe of the Ross Sea between 65° and 75° S. It is doubtful, of course, if any of the sea ice round Antarctica, with the possible exception of the high latitude ice-fields on the Pacific side and in the heart of the Weddell Sea (see again p. 49), could properly be described as permanent. The term, suitably qualified, is used here as a convenient contrast to (i), the non- permanent ice, which is known to clear away from the areas it covers more or less regularly year after year. The so-called permanent ice-fields I have in mind, however, are known to be very compact and to be composed of floes of great extent and thickness, and, relative to that of the non-permanent ice, their existence is prolonged and their dispersal highly irregular and unpredictable. Let us consider first the non-permanent ice. Here it must be admitted we have no direct observa- tions, the risk of freezing-in in winter when it is so widespread rendering it imprudent to venture a vessel more than a mile or so inside it. Even so there can be little doubt that the whale food, whether larval, adolescent or adult, in or under the non-permanent ice of the circumpolar West Wind drift, 1 Never it is true, in the sense that South Georgia is completely enveloped by the pack. The ice-edge (p. 417, Fig. 148) is in fact occasionally found not very far away from the south-eastern tip of the island in winter and especially in sprmg. PHYSICAL ENVIRONMENT 7, that is to say in such parts of it as are covered every winter, must be just as scarce as it has been shown to be (compare again Figs. 5 a and 5 h) when the whole of this area lies open to the south. On equally good grounds we may assume that temporary ice-fields in richly populated regions such as the Weddell and East Wind drifts must mask, at the right season of the year, populations of larvae, adolescents or adults no less rich than those that are known to exist in both great currents when they are not so encumbered with ice. Racovitza (i900^») of the 'Belgica', the only vessel that has ever wintered at sea in the pack of the East Wind drift, refers to the immense shoals of Euphausia that exist there below the winter ice, while Wilton (1908) records that Euphausia were commonly seen and captured in large numbers at the surface through holes cut in the sea ice near the ' Scotia's ' wintering station at Laurie Island. An isolated but highly interesting observation on the occurrence of krill below the winter ice of high East Wind lattitudes has just been published by Fuchs (1958). It was obtained while the Commonwealth Trans-Antarctic Expedition was wintering at Shackleton Base in Vahsel Bay when in May 1957 a large fish trap was being used through a hole cut in the ice over a depth of 500 fathoms. He writes : The hole through which the trap had been let down was frozen over and when we broke through, countless clusters of ice crystals an inch or more in diameter floated to the surface. At last a patch of clear water was obtained and we could see by the light of our torches numerous pink shrimp-like animals. These were Euphausia or 'krill', the main food of many species of whale. The trap proved to contain nothing but ' krill ' and ice crystals, but not intending to return empty handed we collected as many as we could of these pink Crustacea, thinking they would make a surprise dish for David Stratton's birthday next day.^ The pink colour and surface occurrence of these animals, which seem to have been large enough for the cooking-pot, suggest strongly they were adult E. superba, although in view of their coastal location (p. 124) the possibility that they could have been the neritic E. crystallorophias must also be kept in mind. As far as the permanent, or more or less permanent, ice-fields are concerned we have a number of direct observations, mostly from the Weddell Sea, which, although not very numerous, lead one to suspect that deep inside such regions the numbers of the krill may be much reduced. The more important of these were obtained in January 1932 when 'Discovery II' penetrated the Weddell Sea pack for a distance of approximately 780 miles from 59° S, in 21° W, to 70° S, in 24° W, returning through the ice for approximately the same distance on a north-north-westerly course to South Georgia. As she made her way through the pack shallow (loo-o m.) obUque nets were fished in the pools and lanes of open water that occur in the summer ice, some of the openings being so small that we had to tow on a circular course. The gatherings of the oblique nets examined on this cruise are shown in Table 3. Apart from the moderate catch at Station 818, which was made well south in the East Wind zone, they are very small. Indeed, compared with the enormous catches recorded elsewhere in the East Wind-Weddell surface stream,^ they are meagre to a degree. The apparent poverty of the whale food these figures suggest may be directly associated with a corresponding poverty of the diatoms upon which these euphausians feed, for even in summer the great floes characteristic of such ice-fields, some of them as much as a mile or two across, must cut oflF a great deal of light and so greatly reduce the photosynthetic processes essential to a flourishing flora. An alternative explanation may be that in the hub or centre of a clockwise circulation such as exists in this great embayment there is perhaps an eddy or area of slack water where the whale food 1 Martinez (1951) records that E. superba tastes like shrimp. - See Appendix where the average gatherings of the surface population, month by month and sector by sector, are shown for the whole circumpolar sea. 72 DISCOVERY REPORTS is scarce simply because the locality is not a very active part of the East Wind-Weddell surface stream in which the krill are otherwise so abundant. Beset in the heart of this vortex in the winter of 1915 the 'Endurance' expedition does in fact seem to have found a great scarcity of planktonic life. Shackleton (19 19) writes: By the middle of September we were running short of fresh meat for the dogs. The seals and penguins seemed to have abandoned our neighbourhood altogether. Nearly five months had passed since we killed a seal, and penguins had been seen seldom. Clark, who was using his trawl as often as possible, reported that there was a marked absence of plankton^ in the sea, and we assumed that the seals and penguins had gone in search of their accustomed food. Clark (191 9) himself, who used his townet almost daily throughout the drift of the 'Endurance', remarks on the scarcity of surface life deep in the heart of the Weddell Sea, adding that the ' lower water strata, down to about 100 fathoms, were only a little more productive, and Euphausiae were taken in the hauls — though sparingly '. Table 3. The catches of E. superba deep inside the Weddell Sea ice-field in January ig32 Station Under 16 mm. 16-20 mm. Over 20 806 — — — 808 — — 16 810 — — — 811 — — — 813 7 72 — 81S 16 55 — 816 3 — — 817 I I — 818 — 7 653 819 I — — 820 ■ — — "5 822 — — Total 28 i^ 784 Average 2 II 65 For most of the year I think conditions in the inaccessible core of the Weddell Sea will probably prove to be much the same as Nansen (1928) found deep in the Arctic basin. He writes: Misled by the abundance of vegetable as well as animal plankton they have found in the North Polar Sea near the outskirts of its ice masses, some travellers have assumed that similarly there is much plankton in the water in the interior parts of that sea. This is, however, a mistake. The North Polar Sea, covere4 in its interior by an almost continuous layer of thick ice, is extremely poor in plant as well as animal life. The sunlight is absorbed by the ice, and hardly any of those rays necessary for plant life are able to penetrate the thick floes and into the cold water beneath them. Extremely little plant life can therefore be developed in this sea — there is only just a little, chiefly in the water lanes between the floes in the short summer; and without plant life there can be no animal life. The heart of the Arctic basin, he continues, may therefore be considered as 'a desert in the ocean', adding that throughout the drift of the ' Fram ' the fauna encountered ' was so extremely poor in number of specimens that our townets might hang out /or several days- and, although we might drift along at a good speed, there was extremely little in them when they were hauled up '. He notes, however, that in view of the absence or extreme scarcity of the phytoplankton in this area the supply of nutrients, already high as in other parts of the polar sea, will build up enormously and so eventually contribute vastly to the immense wealth of plant and animal life on the marginal parts of the ice-field. 1 Except for ' plankton ' the italics are mine. ^ The italics are again mine and I have used them to accentuate the extreme poverty of the plankton this great explorer must have encountered. Compare, for instance, these immensely long periods of towing with the situation in certain ice-free parts of the Antarctic where in a matter of seconds (p. 152) over 100,000 krill may be captured. PHYSICAL ENVIRONMENT 73 The Weddell Sea, or at any rate such parts of it as are permanently covered, if conditions there are as Nansen describes, may also perhaps be conceived as a vast storehouse of nutrient salts. It may be remarked too that The Friendly Arctic of Stephansson (1921) refers essentially to the fringes of the polar sea, where life is indeed rich,i and that while he is strongly disinclined to believe the existence of Nansen's desert he produces no evidence (Stephansson, 1928) to suggest it could be otherwise. Dunbar (1955 a), following Stephansson, also questions the reality of Nansen's desert, noting that the results of the Papanin expedition of 1937, working deep in the Arctic pack, did not agree with Nansen's conclusions. This expedition, however, hauling its nets through holes in the ice from depths as great as 3000 m., must have been sampling in the main a bathypelagic population,^ the vertical horizon examined not being confined as Nansen's was to an impoverished euphotic zone. Even so, as Dunbar himself observes, the quantity of zooplankton encountered was 'not great', not 'in any remarkable bulk ' when compared with the wealth of life encountered on the fringes of the polar basin. He adds moreover, 'Whatever the production of life under the more or less permanent ice cover of the Arctic Sea, it is certainly less than in arctic water that is free of ice during the summer, and the latter production is, in turn, less than the production of life in the subarctic waters '. During the recent voyage of the U.S. nuclear submarine 'Skate' below the Arctic pack plankton populations were found to be 'extremely sparse' and 'no acoustic scattering layers, as commonly observed in other oceans and attributed to marine organisms, were found on the echo sounder trace ' (LaFond, i960). Bogorov (1938) calls attention to the work of Shirshov at the Russian drifting ice station 'North Pole', noting that Nansen's desert has now been found to be 'characteristic for the "winter" state only, whereas during the biological "spring", as noted by Shirshov, there develops a tremendous amount of planktonic algae and of diverse animals '. It is possible, however, that a more or less permanent winter state exists in the impenetrable core of the Weddell Sea. Fraser (1936) suggests that as a result of the reduction of their food supply through the cutting off of the sun's rays by ice, there will be a movement of the krill northwards to areas of richer grazing at the periphery of the ice-field. That this happens [he says], is proved by observations on the circumpolar cruise [of April to November 1932] and the voyage which the 'William Scoresby' made into the Weddell Sea [in January and February 1931]. The latter instance is most striking because the ship penetrated far into an area which is normally ice-covered for the greater part of the year, and the stations where krill was taken are all concentrated on the northern part of the line between South Georgia and the point east of the South Sandwich Group where the ship turned southwards into the water of the eastern Weddell Sea. No krill were taken in the part of the Weddell Sea which is normally ice- covered, although the observations included a station at the ice-edge with hauls throughout a twentj^-four hour period (St. WS 552). It cannot be doubted that krill does not penetrate far beyond the edge of the ice-field, and that to some extent at any rate the concentrations met with in the vicinity of the pack are caused by the movement, towards the periphery of the pack, of krill which was more uniformly distributed before the formation of the ice-field. This passage was written long before the publication of Mackintosh and Herdman's monthly ice-charts, and its author was therefore unaware that the eastern part of the Weddell Sea, far from being normally ice-covered, is open annually, as these charts show, from January right through to May, and accordingly should more properly be described as a region subject to periodic covering by non-permanent or temporary ice. As the route of the vessel lay across the Weddell drift from north to south, across a region (Fig. 5 b) now known to be rich in whale food, I was surprised, in spite of ^ Paulsen (1906), for instance, calling attention to the wealth of plankton north of Iceland, describes the water as 'literally living with herring'. ^ Dunbar notes that the variety of plankton animals captured included typical Atlantic forms and that these moreover were taken in 'the Atlantic layer of water beneath the upper arctic water'. 74 DISCOVERY REPORTS Eraser's findings, that no krill should have been encountered except at Station WS 542, at the northern end of the line. On looking into the matter, however, I have since found from the records of krill patches kept on board, and the examination of certain other nets fished on this cruise which Fraser could not have looked at, that there was in fact krill in considerable abundance along a substantial part of the route from Station WS 542 southwards to Station WS 547 in 62° 40' S, 17° 02' W, the gatherings including one enormous haul of over 10,000 individuals at Station WS 546 in 62° 09' S, 17° 12' W. Moreover, confirmed krill patches were seen at Station WS 546 and in great profusion between Stations WS 555 in 60° 27' S, 19° 36' W and WS 559 in 57° 19' S, 24° 50' W, pointing to the pronounced abundance of the whale food along the route traversed. However, even if there had been no krill in this supposedly normally ice-covered sea, the suggestion that their absence could be due to a deliberate migration in quest of better grazing seems rather speculative. In the other regions of heavy, more or less permanent, pack we have only a single observation from the Ross Sea in December 1928, Station RS 9 in 70° 02' S, 180° 10' W, located deep inside an ice- belt some 270 miles wide which stretched in this meridian from 68° to 70° 30' S. A meagre catch, consisting of forty-six Furcilia 6 and adolescents up to 20 mm. long, and eleven older adolescents up to 30 mm., was obtained in the horizontal surface net. As this station was made in broad daylight, however, and the towing on the surface, and because in broad daylight the older krill so readily avoid the surface net, the catch of the smaller animals only, those up to 20 mm. long, can be taken as in any way indicative of the poverty of the whale food we seem to have recorded there. While it would be unwise, therefore, to draw any definite conclusion from this one observation, it is perhaps worth while mentioning that the total yield of eleven towings made by the ' Terra Nova ' in this ice-belt in December 1910, along approximately the same meridian, and between latitudes 66° and 72° S, was only two individuals (Harmer and Lillie, 1914; Tattersall, 1924). There are no observations deep inside the great ice-belt in the Pacific sector. As Mackintosh and Herdman (1940) remark, the ice there seems 'stagnant and no ship has penetrated it to the land, although several have attempted to do so '.^ While there may be some grounds for believing that an eddy, or area of slack or ' dead ' water, deep inside the permanent ice of the Weddell Sea may partly be responsible for the apparent scarcity of the whale food there, it is still possible that the scarcity may not after all be a real one, for in the still, almost crystal-clear water of the pack and the continuous summer daylight of these high latitudes the krill may see and avoid the surface nets, both horizontal and oblique, far more readily than they can in the more turbid, diatom-rich and boisterous conditions of the open sea. Griffith Taylor (1916) for instance calls attention to the wealth of Euphausia in the open water between the ice-floes at the approaches to the Ross Sea, where, he notes, owing to their large size, they can readily be seen from the decks of vessels. In small pools deep inside the pack-ice off the Princess Astrid Coast Wild (1923) reports a similar abundance. Finally, it may be noted, the underlying implication of the remarks of Ruud and Fraser quoted at the beginning of this section is that it is in the pack or at its edge that the krill, especially the late larval and early adolescent krill, find the most suitable conditions for their growth and development. It will be shown conclusively later however (p. 355, Fig. 107) that the ice-edge and pack-ice generally, far from being the most suitable environment for this species, particularly when in its larval or adolescent state, is in fact a distinctly unfavourable one, slowing up the growth rate and retarding the develop- ment to a quite unmistakable degree. It may be noted too that the adult ice-edge scarcity reported by John (p. 65) during the April to November circumpolar cruise, a scarcity Tables i and 2 most emphatically contradict, can now be readily explained. For with a single exception all the ice-edge • Though see p. 51, note i. PHYSICAL ENVIRONMENT 75 Stations of this winter voyage lay well north in the West Wind drift where the krill, the adults especially, are now known to be very sparsely distributed. To sum up, of the two environments, the open sea and the pack, the krill are to be regarded neither primarily as creatures of the one nor primarily as creatures of the other, but essentially as creatures of both, their presence or absence, their abundance or scarcity, in the one or the other, being entirely dependent on the time of year or locality in which one happens to strike them. For clearly their abundance near pack or among icebergs, or at the edge of a freezing sea, is purely a matter of the chance or natural obtrusion of these encumbrances upon the open water of regions such as the Weddell drift, where in any case, in one condition or another, E. superba is extremely abundant all the year round. In winter, perhaps more so than at any other time of the year, the whole of this great surface stream from west to east (p. 366, Fig. 115) seems literally to be teeming with larval whale food, carrying vast numbers not only along the northern periphery of the pack, but also in wide stretches of the unfrozen drift to the north, the dense concentrations at the ice-edge being simply attributable to the fact that it is the southern part of the current that in the natural order of things periodically becomes frozen over in winter. In fact it is just as misleading to say that E. superba is a creature of the Antarctic pack as to say that certain fishes are creatures of the freshwater ice merely because of the annual freezing over of the lakes they live in. From the earliest times, it is true, an abundance of krill has been reported in the pack and we are repeatedly left with the impression that it is there rather than in the open sea that the whale food occurs in greatest profusion. But I am sure we get this idea simply because in high latitudes in the polar summer surface life of all kinds is so very easy to observe. Vessels working through heavy ice are often stopped in water that as a rule is flat calm and crystal clear, and observers, even casual observers, can readily see what is going on, not only at the surface, but quite a few metres below. Surface temperature For all that has been published on the systematics and general biology of E. superba it is a little surprising to find that what may well be ^ne of the most important aspects of its physical environment, the thermal conditions of the surface waters in which for the greater part of its life it is known to live, have been largely neglected in the literature or at any rate given little more than cursory attention. For instance, Tattersall (1924) remarks that it is 'a purely Antarctic species, circumpolar in distribu- tion and not yet recorded north of the mean isotherm of 6° C, while Ruud (1932) states that it 'has only been caught with absolute certainty in temperatures between —1-50° (or below) and +4° C. Mackintosh (1934) places it among his 'cold water' group of plankton animals, a group including species that may occur in large numbers anywhere south of the 3° isotherm. John (1936), while in general agreement with both Ruud and Mackintosh as to the temperature range, is more incHned to the view that its normal habitat is in the coldest Antarctic water ranging from the coastal regions of Antarctica to the northern periphery of the sea ice and not very far beyond. Bargmann (1937) and Sheard (1953) repeat this view, the former stating it is 'found in the colder water of the Antarctic zone, and along the edge of and under the pack-ice '. From the mass of data bearing on this question that has now been examined it can be shown that the relationship of the krill to surface temperature is a matter of considerable complexity — that the temperature range as might be expected not only varies from month to month and from season to season, but also from place to place, and that as a result of the constantly changing physical environ- ment in which the krill are growing certain developmental phases, owing to the time of year in which they appear in the plankton, spend virtually the whole of their surface existence in distinctly colder conditions than those encountered by other phases. 6-2 ,6 DISCOVERY REPORTS The monthly surface temperature range of the grand total of euphausians captured in the surface (loo-o m.) layer is shown in Table 4.1 The figures given here, and in the regional presentation of the data (Tables 5-9) that follows, show the average monthly catch per every 1° C and are based through- out on the material obtained in the surface (0-5 m.) and oblique (loo-o m.) stramin nets with which night and day we sampled the surface population (see again, however, p. 65). In both general and regional presentations of the data the population as before has been divided into three broadly dif- ferentiated developmental phases: (i) krill under 16 mm. long (vast majority larvae), (2) krill 16-20 mm. long (very young adolescents), and (3) krill over 20 mm. long (older adolescents and adults). Anomalous occurrences of E. superba have been disregarded. There are two such occurrences in our records, both of them in February, at Station WS 139 (surface temperature 6-15° C) and at Station 829 (surface temperature 6-29° C), one specimen of the over 20 mm. class having been taken in each instance. Both stations lie in Subantarctic water immediately to the westward of the S-shaped bend in the convergence (p. 58, Fig. 4) which occurs in 50° W about half-way between the Falkland Islands and South Georgia. In the neighbourhood of this bend hydrological conditions are complex, somewhat unstable and, as John (1936) has remarked, characterised by particularly strong water movements both across and beneath the convergence which resuh in the carrying of certain species (for example Euphausia frigida), normally inhabitants of the Antarctic zone, north of their usual habitat. A strong tongue of cold water flows towards this region from the Weddell Sea and no doubt it is this movement that is responsible for the occasional occurrence of E. superba in this unusual position, in conditions which as Table 4 shows Ue well outside the normal limits of its absolute temperature range. In the concluding remarks of his paper on the distribution of Rhizosolenia curvata Zacharias, Hart (1937) also calls attention to the water move- ments in this area, suggesting that it is there apparently that the position of the Antarctic convergence is most variable and there too apparently that mbcing across the convergence most frequently takes place. Disregarding these anomahes it will be seen that the maximum or absolute range of temperature in which the surface population of the krill may occur lies between —2-00° and +3-99° C, the mini- mum surface temperature at which it has been recorded being -i-Sg°^ and the maximum +3-90° C, the total range thus extending over nearly 6°. Tolerance of the maximum range it will be seen is confined as might be expected to the warmer months of the year, namely, from November to May, and even then it is the older stages including the adults (represented by the over 20 mm. group) and the early surface larvae (represented by the under 16 mm. group) that alone are involved, the former from November to March, the latter from March to May. It will be seen too, again as might be expected, that it is only in February, the warmest summer month, that the older stages occur in any real measure of abundance from extreme to extreme of the maximum range. At the end of autumn, with the onset of winter conditions, the temperature range noticeably shortens, the vast bulk of the total surface population, as the average catch-figures show, existing from June to November in sub- zero conditions. After October it lengthens again and continues to do so until it reaches its February maximum. 1 In this and Tables 5-1 1 I use the temperature measured at the extreme surface, because at many stations where towed nets were used this is the only point at which the sea temperature was measured. It can differ from the temperature lower down within the loo-o m. horizon sampled by the oblique nets, but seldom by very much except when the extreme surface is occasionally warmed up in summer. In any case there is reason to believe that the vast majority of krill taken in the oblique nets were quite close to the surface. Thus there should be no appreciable error in Tables 4-9. Tables 10 and 11 it is true show the catches from 250 m. to the surface, but even here any variations from the surface temperature should not be enough to affect the conclusions given at the end of this section. 2 At Station 1781 at the ice-edge in Weddell East in June 1936. 2 a JO ca to PHYSICAL ENVIRONMENT 77 O « -Oi S o S. *>■ TO E* J-l '(SI CO S to ■ 00 I I •^ m o\ •'t r^ M ■^ O* O m o ■<*• o 0» 00 _ oo oo f*^ so in oo ■»j- »n o^ -^ (>. ^-^ v-' ^W (V, ^ 3 U I ^^^ in -^00 c^ -^ ^^ hH in rj- iH -^ ^O w N -^ O in -^h ^ In O^ t*^ tH ^^^^ J " ^J 00 N O ^ 0^ Ov •-' ^ N vO m fO _^ •H M \0 fO T " """ "-I r; M o I I ii-)sO 00 ■^ ooo t^ r^ i-i oo o n ^— ' N N ro r*" T^ I 00 -"J- M N 00 * 00 "-1 I ■^ On Ov -^ O « 00 r^ "-I o •+ ro to vO sO On N \0 vO On \0 PO lO r>. Ov On 0\ ro M O O « ^ ^ (*) »nso \o o* ^^ n vo so OnnO ro ^ r^ r>> iM r^ \o VI ^^ ' N M m u-> f*^ i^ so 0"0 t^ r^ N o t^ o r^ On c^ I-. lo N 00 >-■ 00 1^ r^ n M 00 lO On -^ ■*- 1 00 lO u^ IT) O N ^D ITJOO I-" 0\ »n ^ ■^ fo 00 ^ \0 o* In \0 ^^Minint^infn n 'J- CO O Th m 00 o « I ) i-t in t^ o\ o^ c> ' M f^ ^ O O M *o" o > O CO fn -^ O ■^ u-jsO N O N O lo N On 00 »n ^^ N N O^ 00 O^ N Th M O lo m 00 O m 00 o O (J »o Thoo N I M in \o r* n IM U-) O IT) ro fO O^ 00 O -^ M fO -^ N u^ r<) On CO ^ T3S"o '1^1^'? 1 1 M 1 "S 1 r^ 1 00 00 CO '-' o O* U-)00 N M NO o N NO^ PI S- - N '!n '5^ So 'ri "«■ ^-'^-^^ invo t^ en in ■-• 'n "o"^ C" '^S '^ ^-^^-^ iM rosO SO N II 1 „ jn 2- o 00 On J M N « 00 Tl-00 1 -t N N "^ 00 in oo - JO "o"? C" 'P; "5 ^-^-' •- « in f) o> 1^ ■"I \0 00 i^'co'n" 0 ^-'^-'^-' ■-< •'J- M 00 ON 1 1 M in 00 in CO I^ o^ 0 On III 5-^ t^ 00 ^" 1 N -^NO "m^Co <> in *" O Bo a I On O ■> On On On On On O On On On On On O ^ CO N H, o *H I p t B POOOOOO'^-^ o o O O O O M .H p p p p o o o ^ ro N w 6 6 ^ I I -H ,-J B 2 e2 1 1 <-i 0 t> o On M -t- 1 Ov Ov Oi o. o> o •s e Ov O- Ov o o o u 1)- t-1 N -. 0 1 N 1 « •§ O O O 0 n n n >. >> j: o O 0 o o HH ^ S c o o o ■^ n « M 0 o 1 T s 1 E 00 " m 0^ 00 m 0» c o 6 ^ 0> 0» Ov O^ 0> O^ O^ Ov O^ ^ fO N ^ 6 oopopoora T3 J3 O 2 000-.0"- c ppopppo o 78 DISCOVERY REPORTS Disregarding average catch-figures of less than lOO, what we may call the optimum seasonal temperature ranges of the broadly grouped developmental phases shown in Table 4 may be sum- marised as follows: Under 16 mm. group (vast majority larvae) Autumn: -2-00° to + 1-99° C Winter: June —2-00° to +0-99° C; July and August -2-00° to -o-oi° C Spring: —2-00° to -o-oi°C Summer: Larvae virtually only taken in the vertical nets (see Table 10) 16-20 mm. group (very young adolescents) Autumn: Except sparsely in the East Wind drift, group virtually does not exist in the plankton (see Table 7) Winter: Group does not exist in the plankton Spring: -2-00° to -o-oi° C Summer: Except sparsely in the East Wind drift, group virtually does not exist in the plankton (see Table 7) Over 20 mm. group (older adolescents and adults) Autumn: -2-00° to +2-99°C Winter: -2-00° to -o-oi° C Spring: -2-00° to -o-oi° C Summer: December and January -2-00° to + 1-99° C; February -2-00° to +3-99° C So far we have been considering in general terms the thermal environment of the krill throughout its entire circumpolar range. Let us now examine the matter from a regional point of view and con- sider in turn the conditions in the principal water masses, (i) of the South Georgia whaling grounds (Table 5), (2) of the Weddell and East Wind drifts (Tables 6 and 7), and (3) of the Bransfield Strait (Table 8), in which this species is known (p. 61, Fig. 56) to reach its maximum abundance. A glance at this presentation of the data reveals at once that throughout the whole vast extent of its circumpolar range E. siiperba is nowhere, not that is to say in any measure of abundance, to be found living in the warmest upper hmit of the absolute temperature range (namely, between 3-00° and 3-99° C) except, as might be expected from its relatively low latitude, in the neighbourhood of South Georgia, and that even there it is only encountered in these conditions in February and then only in its older or adult state. In the main east-flowing part of the Weddell stream by contrast (Table 6) and in the East Wind drift (Table 7) it will be seen that the entire surface population throughout its development is confined to water which except in February and March, and then only in the Weddell drift, never gets warmer than i -99° C, and that it is only in March in the Weddell drift that the early larvae, and they alone, are found in some minor measure of abundance in the range 2-00° to 2-99° C. In the East Wind zone the absolute upper limit of the temperature range even in summer is only 1-99° C, and it is clear from the average catch-figures in Table 7 that the krill live and develop there in the coldest physical environment to be found anywhere in the Antarctic seas, the vast majority of them in surface temperatures which for 10 months of the year, from March, when the young ice first begins to form, until December, when the first signs of the spring break-up appear, must be below zero C. The prolonged winter cold that exists in the East Wind drift is only to be expected, since in the high latitudes in which this coastal current flows the sea freezes earlier and remains frozen longer than it does in the more northerly regions of euphausian abundance. The observations in the Bransfield Strait (Table 8), although not so comprehensive as in the South Georgia or Weddell regions, are nevertheless enough perhaps to indicate that the krill there exist in sub-zero temperatures from April right through to November. They show too that the optimum temperature range of the older adolescents and adults in February lies between o-oo° and 2-99° C, a summer range that approaches more closely to the conditions on the South Georgia whaling grounds than to those in the East Wind-Weddell stream. PHYSICAL ENVIRONMENT 79 ^ ^ f ^ ^ ■X ^ "O 'J- N O N f^ o Os O- OS ^ Qi O 0> OS OS o» Cs o ■^ f^ N ^ 6 M I o o o o o o N 2 lU CO 2 I coo pop f 2 S ^ ^ o c c B on'-' -— -^ « e2 g OS OS OS OS OS o t OS OS OS OS o- o o r * CO N ■"• o 1 1 u M o O o 0 o n n O >s >! o O o o n ^ ^ J= -P o O o o o o o c c 'I- m N 0 o 1 " o E 2 e2 o E 2 e2 0 o E •= -= « 8o DISCOVERY REPORTS J" CO ^ -^ -5^ too ■fcj 4> VJ ^ g ^ ?P >« T ■Oi K CO lq N 00 c» N r^ CJ T 1 1 - vO J^vLvLw m -* o o ON s - N Ti- CO i T 1 1 po. m 00 M r^ in f^ o ro r^ o y— sy— V/— s i^ c^ rosO O CO N Th M O N O ''t; CO CO CO -^ CO O* iTiOO CO tn\0 CO r^-vo Tj- t^ t/^ l-> >-> M »n -"^ CO CO o» moo CO ^^ 1 1 1 1 5^1 •-I n O' N r^ M V— ^^ ■-.«"— %o •* o> t~ vO m ■* 1 M \0 t^ vO m vO C t^ 00_ N fn ^^ 00 N >« N ■* M ^-^ MvD 1- ov r> •* r~ r^ - ^ M O •^ r^ 00 M ^00 N irt ^-^ M - <-< "I CI « , M t~ « 00 ■+ 0 t^ ■+ 1 N >o 0 m O^ N ■* o M M Cn 0? i vO 00 00 en '-' r^ ov o »o ^- « « N w o COOO O CO r^ "«■ 'zr "7? "ji "o" O 1 1 N 00 « M vO t^ II 1- t^ o ■*- o « M -^^-^ „ 1^ 1 1 m 0^ 1 1 S^ o> 0 o M •o •* r, o 2 N o NO VI t- 00 f^ N sO v> in O 00 II -■ O VO o_ \0 m 11 00 1 1 -^ VO VO N 1 i **2 00 N ^ N 1 1 :rs>s ' ' so «^ N 00 CO CO 3332" 3^ ^ M O o o o o O M I-. o o o 6 6- I I 2 *-• c o ^. on >> 2 ■S c c S ^ O* ^ (> o^ o* N M 6 I" c o s e2S 11 Po 2 c o s bo 0) (J ^3 2 B O 2 eE PHYSICAL ENVIRONMENT 8z I '^ 1 ^ ^ ^ "O w t^ « s ^ ^ L Jg s •1 to -« 2 N IW c 1 1 1 CO P '^ ^ — ' » J t a .« 1 1 1 " 53 ■« ^ /-vr-S^-N 1 ^S O N CO lO ■^ lr> in O (*) 00 ro ^ C" tri N a ■9h I*- ^, N Tj- N ^O vo ^O C a t>3 "a 1 'J- t^'tj'?; 1 ^ -^ fO N s 5 o 1 *-< -^^D w >o sO (*5 f^ g M f^ \D o 1 w t^ M O 1 ^-^ N M -rl- 00 '-' M in t^ Tl- M tC^ D U-) M Tl- o ^ 1 -^ w O fO lO Jp^www *^ 0 ro tJ- r-i K vO -O 0 00 ^ f) r^ •^ 1 O ^ O 00 *^ 1 J-!,^ '^"-' u 3 S 1 § § ^ i s 0^ O^ 0^ Ov 0\ o « « 6 0 p 1 o p ft 1 i t O j3 o CO u 0> 0^ 0> Ov O -* o J2 C O a O o p N O 0 88 - 6 o o 6 1 s p 1 u >. X. % £ o _>. c O S 1 1 C3 e2 2 e2 e2 — -- « 83 DISCOVERY REPORTS Is* 03 2 CO S 3 i c o *;t| to N 1 1- ^"11 1 '^ 1 1 1 1 N 1 VJ ^ s g" ^ c ■0. ■s '^ 00 CO 1^ M O CO r. 3 s 0 0 X S JS j= 0 n 0 [^ 0 0 0 catch observ ly catc O* On 0» O^ 0 0 •s ^ 1 2 ON ON 0- o> " 6 p 1 0 N 1 n n 0 o> o* o^ o^ ri N M 6 9 1 0 N 1 On ON ON O- fo CI H^ 6 0 1 0 1 0 3 0 0 0 0 n n >, >> XI 0000 0 0 onthly onthly month 0000 0 0 >. >> •S B 0 E f .c d 0 e x; X! a §8 0 0 0 0 0 0 c c 0000 0000 0 0 8 8 8 8 0 0 c C 0 t f^ M M 0 0 ■^ s £ CI « H. 0 0 e E a, t»> N •- 0 0 B s ■g. — — ca c^. P3 PJ OJ ass a rt S) > 0 0 > 0 > h h < hH<: h h < PHYSICAL ENVIRONMENT 83 The monthly surface temperature range of the krill in the sparsely populated West Wind drift, in- cluding such circumscribed parts of it as are affected by local encroachment of cold water from the East Wind zone, is shown in Table 9. Small though the average West Wind catch-figures generally are they are, nevertheless, enough to indicate that the scattered population in this wide circumpolar belt exists throughout the year in conditions broadly comparable with those on the South Georgia whaling grounds. There are several instances it will be seen where the average catch-figures in Table 9 point to some, if only minor, measure of West Wind abundance. Such instances are provided by the figures for the larvae and early adolescents (the under 16 mm. and 16-20 mm. groups) in December and again for the larvae in March, June and July. Strictly, however, these figures are not typical of the West Wind zone, for they are based almost exclusively (i) on observations in the north-going outflows of cold East Wind water that occur in the neighbourhood of the Cape Adare — Balleny Islands and Kerguelen-Gaussberg submarine ridges (p. 58, Fig. 4), and (2) on observations so close to the only approximately known northern boundary of the Weddell drift that for all we can tell they may have been located in the more northerly reaches of this krill-rich stream. Temperature range figures for the exclusively larval gatherings from the vertical nets are given in Tables 10 and 11, Table 10 covering the principal region of larval abundance in the East Wind- Weddell surface stream. Table 1 1 the West Wind drift, the South Georgia whaling grounds and the Bransfield Strait, three regions where our vertical gatherings of larvae in the vast majority of instances have been of very small or virtually negligible size. In both tables the average monthly catch-figures are based on the material obtained in the combined hauls of the three uppermost nets traversing the 250-100 m., 100-50 m. and 50-0 m. layers, and as might be expected with nets of such small size hauled for such short distances the figures in general are small and, except for from January^ to June in the Weddell drift, do not carry the major significance provided, for instance, by the much more voluminous data from the stramin nets. Small, however, though they are they are nevertheless of sufficient significance, having regard to the size of the vertical net and the circumstances in which it was fished, to be used in certain instances to augment the data from the stramin nets, particularly where such data are scanty as, for example, in the Weddell drift in January and February (Table 6, under 16 mm. group) and in the East Wind drift from February to May (Table 7, under 16 mm. group), and it is accordingly as such that they have been used in the final summary of the temperature range data that follows on p. 87. The pronounced falling off of the vertical catch-figures from July to November it may be noted is primarily to be ascribed to the vertical distribution of the larvae during that period. For whereas from January to June the young epiplanktonic population ranges in great abundance from the surface down to 250 m. (p. 90, Tables 13-15), and in consequence can readily be captured by any or all of the three vertical hauls made between those levels, from July to November (Table 15 and pp. 331 and 335), it tends to be massed permanently very close to the surface and so can only be sampled momentarily by the uppermost 50-0 m. vertical net. The average monthly catch-figures for the surface larvae we have occasionally recorded in the vertical nets outside the main East Wind- Weddell stream (Table 11), although on the whole very smalP are, nevertheless, again perhaps enough to indicate broadly that, the Bransfield Strait apart (Table 8), such larvae as are encountered in these northerly latitudes must develop in distinctly warmer summer and autumn temperature ranges than anywhere else in the circumpolar sea. 1 The Discovery expeditions surprisingly recorded very few surface larvae in the Weddell drift in January. Norwegian observations, however (p. 88), not included in Table 10, reveal heavy January concentrations there, indicating at the same time that the absolute surface temperature range of the Weddell larvae then lies between — 2-oo° and + 1-99° C, the optimum between —2-00 and +0-99° C. 2 A single, and exceptional, catch of 707 Calyptopes was recorded on the eastern side of South Georgia in March 1937. Without this gathering the average March catch-figure for the 3-00° to 3-99° C range would be nought instead of loi. 7-2 84 DISCOVERY REPORTS ■^ in ON N ■^ m O N bo Co .s -^ s r^ g ■ ,^ CO s ^ to 1 1 1 t ^ ^ N 1 1 1 1 '^ IX 00 ^ ^ § 'n "^S 'Fi'm ■+ t^ r; ■•^ f^ ^ 1 00 ^ "^ ■^ HH ro o ^ 'W t^oo 00 U-1 CO N ■(1, i; ■ ' » I ^ •^ ON « ■•j-xo r^ tT) t^OO 00 lo CO N N Tj-o r^ fo t^OO 00 m ro N e ^ u^ in IT) i/i s "nCo "rT 0 irioo" ro ^^^ M •I fn ■tl U-) e 1 1 1 1 I^ t) M r>- N «n ro\0 •^. T "fj N "^ '^ Ov T 1 ON 1 T « O '(> 'fT r^ 1 M 1 m m 1 1 1 1 1 n 1 1 J^ N irj 1 1 1 m 0 ■* J^ O 00 ^ Tt- iri u^ in ir> 1 M On 1 1 1 1 ^^^ 1 1 1 n^ N NO CO (*i 5 1 1 1 1 1 IT) NO CO •^ ? 1 1 lO NO -«*• ■^ 1 N vO N 0 moo ro '- l-l S y— . .--^ ^ r( m ^ »H * Cl N N o 0| Pi >A 00 )-) o r^ 1 1 O 1 1 1 1 M 1 1 w 00 lO ■+ ro m 00 On Tf NO 00 on N On NP ^,_^ .„ _ , , ___^ ^._^ CO ^_^ _ CO ^ 1 N r^ N tri (*l \0 N 1 N m O N lO n ro 1 1 00 1 1 O CO 1 ■+ ro r>. Th 1 1 1 N ^ CO N 1 iri N ON r^ Nft 00 ^ ^ so N ^^ ,.-v ,^ ,-v 00 \0 CO N CO CO CO N CO CO --i- fO N M CO CO ^ CO N N CO N ■^ CO ON 1 1 1 00 yi 1 1 1 f^ -)- Tl- 0 CO A, in * CO ^ 1 1 CO -t rn iri CO N s n rt o o O- o> O n £ o c^ O O o X rt & Tf r^ N O N o ^ 1 1 J2 0 >. 0 0 o o o n O >. >, +■ c o o o n o o „ ^ -fi j: o o o o o 0 o c c 1 '*■ f^ N "' o o 1 1 o e i 1 e2 s 1 CO ^ CO m lo O co -^ CO ^ N On O* O* O* On O* O* On On On t B o o o o o o o o o o o o -■ >- p o o p o o o Tj- CO N -H O 6 ^ I I o _>. If c ^ fO N M o 2 c o e a 2 § h h < o o o o o J3 o 2 2 c o o o > h H -5 PHYSICAL ENVIRONMENT 8s a IP -5 .^^^^,^.^,^ o I £ 3 3 ■a, ^ Tj-vO O N CO ^n-l o o m « M o - =» * 00 -"t irj " W d to eg I a t LQ =s I «. I i£.£S g t "a .a I i/l ir> N \0 « I (U 03 O 3 ■+ ■+ >0 fO 0\ ^ Oi O^ o O^ O^ o* Ov o fO N N. 6 ^ I o o o o o o CO N N- O c o s e2 c o B 0) iio 2 < 5 •§ >- Xl o _>. c o I B B 3 s § ,0 o ? E-i H < 86 DISCOVERY REPORTS It will be seen then, that the temperature range of this species, while varying as might be expected from month to month and from season to season, varies also, and varies quite significantly, from one geographical region to another. It is neither a creature of the coldest Antarctic water nor of the warmest, but essentially of both, its existence in one range of temperature or another depending entirely upon the time of year and locality in which it may be found. Table 1 1 . Vertical net hauls. Monthly surface temperature range of the larvae in the West Wind drift, Bransfield Strait and on the South Georgia whaling grounds (See heading to Table 4) Autumn Temperature Spring Summer , ^ April- Winter 1 Ull^C r ^ ' > r ■^ (°C) Sept. Oct. Nov. Dec. Jan. Feb. March May June July Aug. 4-00 to 4-99 (-) (-) (-) (-) (I) 6 (4)- (7) - (3)- (2)- (2) - (0- 3-00 to 3-99 (-) (-) (3)- (0- (16) II (28) I (7) 101 (II)- (0- (I) - (3)- 2*00 to 2-99 (3)- (I)- (2)- (I)- (") 15 (32) 2 (39) 17 (18 2 (2)- (I) - (3)- I -00 to 199 (2)- (0- (7)- (13)- (31)- (23)- (13) 32 (12) I (4)- (I) - (3)- 0-00 to 0-99 (4)- (10)- (II)- (13)- (12)- (8)- (10) 30 (20) 54 (3)- (-) (6)- — o-oi to —I -00 (8)- (6)- (72)- (II) I (10)- (4)- (7) - (10) 22 (3)- (-) (6)- -I -01 to -2-00 (10) I (8)- (20)- (5)- (2)- (-) (-) (2)- (5)- (-) (6)- Total monthly 13 — 55 15 369 117 2129 1389 — — 7 catch Total monthly 27 26 "5 44 83 99 83 76 20 5 28 observations Average monthly — — — — 4 I 25 18 — catch A comprehensive view of the variation that may be encountered, in both time and place, is given in Table 12, which, summarising the data in Tables 5-8, 10 and 11, shows the monthly temperature ranges of the surface population in the four principal regions of its abundance, beginning the year in each instance with January when (p. 102) the larvae first begin to appear in substantial numbers in the surface waters of the Weddell stream. In this summary the figures for the optimum range are based for the most part on average catch-figures of not less than 100 in the stramin nets and of not less than 50 in the vertical nets. In instances, however, where no such catch-figures are available, owing to the periodic scarcity of certain developmental phases in the plankton, the optimum range is based on the two highest adjacent average catch-figures in the absolute temperature range. From these figures then, taking the optimum range as the most significant measure for regional comparison, and considering in the first instance the surface population as a whole, it is evident that E. superba could be described as a creature of the coldest Antarctic water — in the East Wind drift from March to December, in the Weddell drift from July to December, in the Bransfield Strait from April to November and on the South Georgia whaling grounds from at least August, but probably before, to October. It could equally well be described as a creature that exists in from very cold to moderately warm Antarctic water — in the East Wind drift from January to February and in the Weddell drift from January to June — or as one that lives in from warm to moderately or very warm water — in the Bransfield Strait from December to March and in South Georgian waters from about the middle of November to the end of May at least, if not later. Taking now the three broadly grouped developmental phases in turn, and taking our data as before from the optimum range, it will be seen that in the East Wind drift the larvae spend virtually their entire surface existence in sub-zero temperatures, the few recorded in the low positive (o-oo° to 0-99° C) range in January being in fact the last Sixth Furcilias surviving (p. 371) from the previous PHYSICAL ENVIRONMENT 87 Table 12. Monthly surface temperature ranges of the krill in the four principal regions of its abundance, the figures in italics indicating scarcity or great scarcity of certain developmental phases in the plankton, the dashes where certain developmental phases do not, or virtually do not, exist, the asterisks where the surface population is wintering among or under the polar pack Absolute temperature range Optimum temperature range Month Dec. Nov. Oct. Sept. Aug. July June April-May March Feb. Jan. Dec. Nov. Oct. Sept. Aug. July June April-May March Feb. Jan. Dec. Nov. Oct. Sept. Aug. July June April-May March Feb. Jan. Dec. Nov. Oct. Sept. Aug. July June April-May March Feb. Jan. Under 16 mm. {vast majority larvae) 0-00 to +i'99 — 2-00 to +i'99 — 2-00 to +0-99 — 2-00 to — o-oi — 2-00 to —O-OI 0-00 to +2-99 + 2-00 to +3 '99 + 2-00 to +3 '99 + 1-00 to +3-99 -2-00 to +0-99 -2-00 to —O-OI * * - I -GO to +0-99 0-00 to + 0-99 - 2-00 to - 2-00 to - 2-00 to - 2-00 to - 2-00 to - 2-00 to -2-00 to -2-00 to -2-00 to -2-00 to - 2-00 to -fO-99 + 0-99 + 0-99 — O-OI — O-OI — O-OI -fo-99 -fI-99 +2-99 +2-99 + 1-99 — 2-00 to — I -01 * * « * * -2-00 to —O-OI — 2-00 to —O-OI — 2-00 to -O-OI — 2-00 to + 0-99 16-20 mm. (very young adolescents) 0-00 to +1*99 — 2-00 to -fi-99 — 2-00 to + 0-99 — 2-00 to —O-OI Over 20 mm. [older adolescents and adults) South Georgia 0-00 to -f I -99 — 2-00 to + 2-99 — 2-00 to +0-99 — 2-00 to —O-OI — 2-00 to —O-OI Under 16 mm. (vast majority larvae) + 1-00 to +i-gg — 2-00 to —O-OI — 2-00 to — I -01 — 2-00 to — I -01 — 2-00 to — i-oi 0-00 to + 2-99 + 1-00 to +3-99 + 1-00 to +3-99 o-oo to +3-99 0-00 to +0-99 + 2-00 to +3-gg +2-00 to +3-gg +2-00 to +3-gg — 2-00 to + 0-99 — 2-00 to —O-OI « Bransfield Strait-\ -2-00 to +0-99 -2-00 to —O-OI * * -2-00 to +o-gg -2-00 to —O-OI « * * — —I -00 to —O-OI —I -00 to —O-OI — 0-00 to + 0-99 — 0-00 to + 2-99 Weddell drift -2-00 to +0-99 -2-ooto+o-99 — 1-00 to— 0-01 -2-OOtO+0-99 -2-00 to -fo-99 — 2-OOtO— I-OI -2-00 to +0-99 —2-00 to +0-99 —2-00 to -I-OI -2-00 to —O-OI —2-00 to —O-OI —2-00 to -I-OI — —2-00 to —O-OI —2-00 to —O-OI —2-00 to —O-OI —2-00 to — O-OI — —2-00 to —O-OI —2-00 to +0-99 — -2-00 to +0-99 -2-00 to +1-99 — —2-00 to -4-1-99 —I -00 to -f2-99 — —2-00 to -1-2-99 -I- I "00 to -f 1-99 •I -00 to -(-0-99 -2-00 to -f-I-99 -2-00 to -1-0-99 East Wind driftf -2-00 to « ■I-OI — 2-00 to — I-OI — 2-00 to -1-0-99 — 2-00 to + 1-99 — 2-00 to —O-OI * * — 2-00 to —O-OI -2-00 to -f 1-99 — 2-00 to -I- 1-99 — 2-00 to -I- 1-99 — 2-00 to — I-OI * • — 2-0O to — O-OI — 2-00 to —I-OI -2-00 to -I-OI -J -00 to -1-0-99 16-20 mm. (very young adolescents) 0-00 to +i-gg — I -00 to —O-OI — 2-00 to — I-OI -2-00 to — I-OI Over 20 mm. (older adolescents and adults) 0-00 to -f 1-99 -I -00 to -I- 1-99 — 2-00 to —O-OI — 2-00 to -O-OI — 2-00 to — I-OI 0-00 to + 0-99 -f 2-00 to -f 2-99 -f I -00 to -f3-99 + i-oo to + 1-99 0-00 to -f 0-99 — I-oo to —O-OI * * — I-oo to —O-OI 0-00 to -1-2-99 — I-OI to —O-OI —I-oo to — o-oi — 2-00 to — o-oi —I-oo to — o-oi — 2-00 to —I -01 — 2-OOtO —O-OI — 2-00 to —1-01 —2-00 to —1-01 —2-00 to —1-01 — —2-00 to —1-01 — —2-00 to —1-01 — —2-00 to +o-gg — — I-OO to -fo-99 — —2-00 to -f 1-99 — 1-00 to —0-01 -2-00 to -f 1-99 — 2-00 to +o-gg — I-oo to —O-OI * * — 2-00 to # •I-OI — 2-00 to — I-OI — 2-00 to —I-OI — I-oo to +0-99 — I-oo to —O-OI # * # * * — 2-00 to — I-OI — 2-00 to —O-OI — 2-00 to -f 1-99 — I -00 to -f I -99 t Although both the Bransfield Strait and the East Wind drift are closed to navigation, the former from June to October, the latter from June to about mid-December, it can be inferred from the observed situation at South Georgia and in the Weddell drift that the 16-20 mm. group must in fact be absent from the plankton of both strait and East Wind zone from June at least until August. 88 DISCOVERY REPORTS summer's spawning. In the Weddell drift the larvae exist in a negative-positive environment from January to June and in an all-negative environment for the rest of the year. Exclusively positive temperature ranges for the larvae are only in fact encountered in the Bransfield Strait and on the South Georgia whaling grounds, in the former locality in March alone, in the latter throughout the six months December to May. In the Bransfield Strait the young stages endure an all-negative temperature range from April to November and on the South Georgia whaling grounds from July (probably) until November, the conditions in June in the latter locality most likely being positive- negative as they are then in the Weddell drift. Of the three developmental phases the very young adolescents, in all four regions of euphausian abundance, emerge as the group that endures the coldest physical environment of all, the vast majority of them spending their entire existence in the plankton in water that never rises above zero C. As for the older adolescents and adults it will be seen that in the East Wind drift they exist in negative-positive conditions from January to February and in all- negative conditions for the rest of the year, and that in the Weddell drift their monthly ranges are the same as for the larvae. It is again only in the Bransfield Strait and on the South Georgia whaling grounds that the older stages of the whale food encounter exclusively positive surface conditions, in the former locality from January (probably) to March, in the latter, like the larvae, from December to May. In both localities the periods of sub-zero existence for the older stages are as described for the larvae. Finally, it may be observed that although the krill, both as larvae and adults, are manifestly tolerant of surface temperatures which in summer and autumn (Table 4) range between — 2-00° and +3*99° C, it need not be supposed that this absolute range is a major factor determining the distribution of this species in the circumpolar sea. It must always, it seems, be to the movements of the East Wind- Weddell surface stream, and the farthermost limits to which it penetrates, that we must turn to find the ultimate determining factors. It is in this current system that the krill both young and old are found par excellence in their maximum abundance and if both it and the tongue of water flowing from it towards South Georgia be regarded as one great surface stream, as in fact it is, it is clear that the temperature range of the euphausians it carries must itself ultimately be dependent upon whatever measure of warming up this current may receive as it moves from high to low latitudes. It is con- ceivable then that if the flow of this great polar stream were stronger and it were to find its way into still lower latitudes the temperature range of the krill might be appreciably wider than it has been shown to be. As Deacon (1957^) has said, it would be 'useful to be able to form a precise idea of how the movements of water may facilitate dispersion of eggs or larval stages from successful breeding grounds and the subsequent replenishment of the stock. There seems to be growing evidence that such considerations may be more important than the eff"ect of a simple temperature difference '. THE LARVAL STAGES Vertical distribution In the spring and summer of 1929-30 a series of fourteen vertical plankton stations was worked from the Norwegian floating factory ' Vikingen ', the observations being spread over the main east-flowing path of the Weddell drift from longitude 40° W to the meridian of Greenwich. Two standard hauls, from 450 to 200 m. and from 200 m. to the surface, were made at each station. In the latter part of January and in early February large numbers of larval E. superba, principally the Calyptopis stages but including very small numbers of First Nauplii and Metanauplii and a few adults, were collected by this expedition and in the summary of his report on this material Ruud (1932) writes, ' It {E. superba] is a surface species, and both the adults and larvae in our material were taken, almost without THE LARVAL STAGES 89 exception, above 200 metres. In other words, it belongs to the Antarctic surface layer'. The later work of Fraser (1936) based on the analyses of over 2000 vertical net hauls, the majority of them extending down to depths more than twice those of the ' Vikingen ' series, has shown that of the three Calyptopis stages the First ranges in substantial numbers from the surface down to 750 m. and, therefore, strictly speaking, 'belongs' to both warm deep and surface currents, while the Meta- nauplius, of which Fraser found several thousands, is revealed as spending its entire existence in the warm deep layer.^ Fraser's discovery of the deep habitat of the Metanauplius is of far-reaching importance since as will be shown presently it proves to be the key to much of what has since become known of the life-history and vertical and horizontal movements of the larvae which are now to be described. In the meantime it may be noted that Ruud too records the Metanauplius from the warm deep layer, the few specimens of this stage he determined all having been taken in the deeper of the two net hauls he examined. He regarded their presence at this level, however, as abnormal, believing 'that these very young larvae, like the eggs, should be looked for close up to the drifting ice '. He did not, therefore, suspect them for what they must have been — stragglers from a much larger population living at depths beyond the range of his deepest nets. The analysis of over 5000 vertical samples completed since Fraser's work was published not only establishes beyond all doubt the permanent deep habitat of the Metanauplius in oceanic water, but reveals no less certainly the still deeper habitat of the Second Nauplius of which Fraser found only seven specimens and Ruud, in his shallow nets, none at all. The principal results of these later analyses with which I have incorporated the principal results of Fraser's are set out in Tables 13-15, Tables 13 and 14 showing the vertical distribution of the oceanic larvae of the Weddell and East Wind zones at stations where deep-living Nauplii and Metanauplii were encountered. Table 15 the corresponding distribution in this dual current system and elsewhere at stations where such early deep-living stages were absent. The salient facts emerging from these analyses may be summarised as follows: (i) It is when the Metanauplius is encountered, and then only, that the immediate product of its moulting, the First Calyptopis, ranges from the Antarctic surface layer well down into the warm counter-flowing deep current, it being particularly abundant down to 750 m. and probably reaching the extreme downward limit of its bathymetric range somewhere between this deep level and 1000 m. (2) The analyses reveal and repeatedly emphasise the deep habitat of the Metanauplius, showing it to be a stage that apart from occasional rare stragglers^ has not been recorded by us above the 500 m. level. It appears to be most abundant between 1000 and 500 m., the observations we have below this level suggesting that the lowermost limit of its vertical range, as revealed for example at Station 2594* (Table 13) and Station 2603 (Table 14), occurs somewhere between 1500 and 1000 m. (3) They reveal for the first time substantial concentrations of the hitherto rarely seen Second Nauplius and give some indication of its probable position in the bathymetric scale. Although recorded twice in large numbers between 1000 and 750 m. (Table 13, Stations 823 and 2594) it has in fact very rarely been encountered there. For the period November-April in the East Wind- Weddell zone we have 360 stations with net hauls down to 1000 m., the Second Nauplius, apart from an odd straggler or two, being absent from the 1000-750 m. level except at the two stations 1 That is, in deep oceanic water beyond the Antarctic continental shelf. See, however, p. 205. ^ These have been disregarded. They always occurred when the catch in the 750-500 m. layer was particularly heavy and, therefore, it is suspected, may have got left over in the net hauled from the deeper level. 3 The deep (1500-1000 m.) net at this station failed to close and fished open to the surface (see p. 100). It produced, however, a far larger gathering of Nauplii and Metanauplii than the 1000-750 m. net, much of it there is little doubt from the 1 500-1000 m. level. The estimated catch at the deeper level has been obtained by subtracting the 1000-750 m. catch from the large total taken in the 1500-0 m. net. 90 DISCOVERY REPORTS is s o "<3 <. "^ Qj i f«4 "« ^ 1 s 2 <3 1 •*-<. ^ ►-S^ "So-^ to !<; •is 1 CO 1 1 to 1 .2 — ■J2 t1 w « , ■*' ,!^ CO to ^ ^ •«^ t<> ^^ *J "« o to -JC* ,^ »• ■K. tS '-'J ?^ h." -S^ -TS 1 an ^ 1 •Ki =2 f <4J * r*i 1^ .s Q -JC ,o ^ ^ .■Ki 3 •*5; <^ o R o p' ^ o !"^ ■^ ^ ■Ki « •♦".i 2 i 3 ^ g ^ C3 ■1) h O -5: O i5 o. cola; 1;^ "25 CO CO 1^ CO to ^ ii to CO ^ t3 55 •2 a 55 3 J3 a 3 u i i i i i i i i i i i ir)0»'^0>oOOOOO >- N ir. t^ O >oO O O i-i H- M ro ■+ oooooooooo i i i i i i i i i i i i u-iOinOi/^OOOOO w-jOu^O»'iOOOOOO H«Mt^l^0»O0000 •- '- N ro -t- THE LARVAL STAGES 91 tS 1;^ 5 §< ■a * o- 13 « H CO I I CO '-' O -1- O N n MP) 1:^ c3 If ■a is t? 2s. CO o 3 55 I I ■+ N f*l o •* o CO J3 CO i i i i i i i i i i i mO"^Oi^OOOOO >-i ■a, 1 M N t^ 1 1 1 1 = '^ l« C^ 1- 1 1 1 = ^ N 1 1 1 1 — N III 8 ^o §- •* ^ = to =^ I- ^ 1-1 M ^ 00 i-i 1 1 , = ^R M M = "-.i* o o e o •2 * ! = a = .^ o n O 1/1 M M 1 ' ^ li^ N CO Ov •-■ O [ ^ CO 4 N 1 = -1 ^ r^ Ti- ^ M = M 1 = = it M M ?"= : = M M « ?, = tri 1 = N 00 =z ■« LS — 8f 1 1 1 1 1 1 ' 1 1 1 1 1 1 1 1 1 1 1 1 ' ^ Q 1 1 1 1 1 1 "0^ = = — = .3 fn M M M 1 = 1 1 1 1 1 1 = = = 1 ■ ^1 o>c» ■*■ ■" 1 1 ' ^ 1 1 1 1 1 1 1 " ■* " 1 1 1 in f-- 1 1 M « I = ■s: 1 m II ' ^ 5i 1 1 ^ "^ 1 1 = Si 1 ■* <* 1 1 1 ' 1 T 1 ►-< m ro 1 1 = to f*^ 1 >n " 1 1 1 1 ^z — 1 p* : — Js IZZ 1" ■* 1 ^^ ° 1 1 1 1 ^^ 0^ 1 1 ^00 1 i = ^) i-i 1 u% - 1 1 1 ' 1 — CO 11^ II ^ a fe O ■2 ^= rz Kt 0 1 OO fO iTiOO rt- 1 — K 1 1 \0 00 N 1 — 55 1 M N " 1 = a 1 1 f*^ N 1 = ~ ^ =1 o ^ ' — — ^ «• M M 1 '^ ! : — 1 1 1 1 1 "^ — ii 1 1 1 1 1 1 1 1 1 1 1 M = = 11 1 1 1 ! 1 1 • = 1 1 1 1 1 1 = 1 1 1 t 1 1 S Q 1 1 1 1 1 1 "^^ = = . — — c^ 1 "M M i = M M M 1 = .3 1 u^ m N 1 1 1 = 1 1 1 1 1 1 ' = i, i> i> 1, i> i> i: ii< \ i i i i i i i i C !. i i i 9 < lO O »0 O w^ O O < 3 O Q Q o, 3 E J3 BO C ;■§ « ^ S2 o o g a 2 o- >. I ^ •« O rt M •g e r § S t^ o Si J) ui^ rS f^ ^ o ™ = §8 i Si J= O ^ s S >, M 3 ■a & s c &^ Ola £ M P S^ o _-—'-'■ << < THE LARVAL STAGES 93 K o I*** -is ■c^ ji « •« E^ ^ti. s. "« ro -^ w K .■i' ^~ t^ K ■^ « ■W t*J ■41 t«5 lu s* ^ .o •12 « « s to ?^ S* "-^ O ■ft. •C? ^ w ^ ^ OJ -^ ►« « ^ J^ (^ § k. !>. -2 -IS <»-l •o i« s to ■u <.! ^ o ^ •S ^ ^ ^ s i^j « 2 vj g^ ^ s ^ ■JC « -o ^ -« •Ki o L ^ «■ ■^3 <3 §1- .a 8 S^ t>3 SQ 8 * * = '^ CO c 03 = "=^ =0 CO = "^ I I ■a, 55 = "^ "a 0? ooooooooooo inOinOw^OOOOO •-Mmr^oinooo •- HH M ro Tj- ooooooooooo irjOmOviOOOOO HHNinr^OmoOO "-I M N en -^ a 3 M ^ H 93 hn ^ E _^ c; u >, a o B 0 3 U > *j n c o .5 .5 2 c 2 « S *j *" "3^ O 01 u 3 > o .5 8 a M<" 2 c ° a ■« tm^ ^ c « „ 'S u « ^ 4^ •OS" ^ ° s ;= o « -2 O i i i i i i i i i i i"o loOioOu^OOOOOO MNu^t^O lOOO O O "-• '-' N ro ■»!- lO 94 DISCOVERY REPORTS I;' O CO t? « Q ?3 CO Q CO ■a. IN CO Q l-H 12 ■Oh c3 ^ o a CO O •£> 00 u^ CO CO O in ^ ; = i i i i i i i i i i r tTi O ^^ O loO ooo o HH N irir^o li^o oo HI >- « fn Tf i i i i i i i i i i i u^Ou^OmOOOOO i-'Nior^o»oooo !-■ w N M -^ i i i i i i i i i i i >- >- N f^ ^ iTi o in o >oOO O _ MNii-ir>'0»i^oo -. <-< M fO THE LARVAL STAGES 95 ■o. Q Q Q 03 CO a ^ a O A. 00 o 00 Q Q c ^ &5 Co !>3 c3 00 ■a, •0, Q CO a CO CO 6 O H- N O O « I I I I L I III I 0000000000 iDOioO>00 0000 "-"Minr^oioooo « w N ro -^ I 1 I i I i I I I I oooooooooo lOOu^OiOOOOOO " Mu^t^O vnO O O I- "-< N c^ Tt- 0660000 10 O >0 O U-) O I I I I 0000 _ o o o o Nu^r-o inp o o III! 0000 li-i O lo o J^ i ' I I I 0000000 •^000000 t^ O i^ o o o o t-" "-■ M CO Tj- li-, 96 DISCOVERY REPORTS ^ I: a O CO K O M ■ci.-^ O "0 ■{ OS o tN. vO •H )? r> ^ ^ r ^ •o, t-i <3 (J ^ .2 r^ so a i>5 crj ^^ CO Q O I 55 ■a- c 2 CO = m 1 ^ I 1 = = CO 1 1 1 1 Q 1= "O 1 1 ^ 1 1 ^n = ■a 1 1 1 ^ CO 1^ O O en Q t= I I I I I I I I I I I ooooooooooo I00>^'0"^00000 -H H^ N rn •t- i i i i i i i i i i i w~*0 w^O u^O OO OO i-Nior^o »no OO M "- N f^ -^ ooooooooooo u-)0"^0"^00000 — Mior^OinOOO « -I rt ro -t i I • I ' i i i ' i i OOOOOOOOOOO ii^O"^0 lOOO O 00 >-■ Nw^r^o "j'-o o Ox t-> ^ M m ■+ w^' THE LARVAL STAGES 97 mentioned above. At both stations, however, hydrological conditions (p. 102) seem to have been unusual. It would appear, therefore, that its normal habitat is below 1000 m. It may well in fact be far below, for at forty-four stations with still deeper net hauls in the Weddell and East Wind zones we find it only twice in the 1500-1000 m. layer (Stations 2594 and 2603) and in neither instance it seems could we rightly claim to have struck its normal habitat in deep oceanic water, Station 2603 having been made on the continental slope (Table 14) over a depth of only 1450 m., Station 2594, although truly oceanic, in the abnormal hydrological conditions referred to above. It could well be, therefore, that its true locus of optimum abundance in the deep ocean is below 1500 m. Only in slope waters, as for instance at Station 2603, is it likely that it might be somewhat higher. (4) Although Ruud with his shallow nets was unable to visualise the relatively short deep-living larval phase our much deeper observations reveal, his statement that the larvae belong to the Antarctic surface layer remains essentially correct, for as Table 15 shows it is for all practical purposes at this high level that by far the greater part of the larval existence is spent and where the development from Calyptopis One to Furcilia Six takes place. The developmental ascent It is evident from these analyses that wherever the three stages, the Second Nauplius, the Metanauplius and the First Calyptopis, or wherever the two stages, the Metanauplius and the First Calyptopis, occur together in substantial numbers the pattern of their bathymetric range relative to one another is always precisely the same, the Second Nauplius at its so far observed optimum^ ranging from 1500 to 1000 m., the Metanauplius at its optimum from 1000 to 500 m. and the First Calyptopis at its optimum from 750 m. to the surface. Clearly this constantly repeated pattern represents what may be described as a ' developmental ascent ' involving first the moulting of the Second Nauplius, followed at a somewhat higher level by that of the Metanauplius, the product of this second ecdysis, the First Calyptopis, becoming the earliest developmental stage of the new-born krill to reach the Antarctic surface layer. From the relative positions of the three stages in the bathy- metric scale one can judge with considerable accuracy the several levels at which these successive moultings take place. For instance, it is clear that as the larvae rise the Second Nauplius must begin, and, except for very rare instances, complete, its moulting to become a Metanauplius below 1000 m. and that the latter in turn must begin its moulting to become a First Calyptopis between 1000 and 750 m. and complete the process in the 750-500 m. layer. The final moulting of the Meta- nauplius must in fact be accomplished long before it reaches the 500 m. level, for if it were otherwise far larger numbers of this stage would undoubtedly have been found in the 500-250 m. layer than the few rare stragglers (which in any case probably came from a deeper level) that have so far been recorded there. That the early larvae undertake this remarkable climb, involving them in a journey of a mile or more to the surface, is confirmed by the obvious signs of moulting and growth (see foot- notes to Tables 13 and 14) that have been recorded in the deep Metanauplius and Calyptopis stages. For example, at Stations 854 and 855 (Table 14) many of the Metanauplii taken at the 1000-750 and 750-500 m. levels were seen to be in active moult, a characteristic half-way condition in which the incipient First Calyptopis form is often seen emerging, or about to emerge, from the Metanaupliar husk, while the small and stunted appearance of the First Calyptopes taken in the deep 1000-750 and 750-500 m. nets at Station 855, compared with the robust and obviously well nourished condition of the higher level First Calyptopes recorded at the same station, clearly points to the former having been the very recent product of the actively moulting Metanauplii along with which they were 1 Further observation, however, may show that its real position in the bathymetric scale in oceanic water is below 1500 m. 98 DISCOVERY REPORTS captured. Other examples of deep Metanaupliar moulting are provided by the analyses for Stations 823, 1138, 1144 (Table 13) and 1545 (Table 14), while the unmistakable growth of this stage during its upward passage at Station 647 (Table 13) is revealed by the length frequencies given by Fraser (1936) which are shown below. Table 16. Growth of the ascending Metanaupliiis as revealed by the length frequencies at Station 64^} {after Fraser), the frequencies in Roman type, the range of size {in mm.) in italics Depth 0-84 0-87 o-8g o-go o-g2 o-g4 o-gs o-gj o-g8 i-oo V02 1-03 1-05 Average (m.) length 750-500 ___ 4 2 10 22 41 5 7 3 4 2 0-97 1000-750 2 5 16 21 25 17 7 52 — — — — o-g2 So far we have traced the development of the early larva from the Second Nauplius to the First Calyptopis stage and seen that during this phase in oceanic water it rises from great depths to the surface. All now that is required to complete the early developmental history is to fit into its probable position in the bathymetric scale the rarely seen First Nauplius, a stage so far not yet encountered in oceanic water, even in our deepest nets. It seems a matter for simple deduction, however, to conclude that the main mass of it, superimposed upon that of the hatching eggs, must lie well below the 1500 m. level, and that our failure so far to sample it is simply because in the vast majority of instances our nets are not going deep enough. The actual bathymetric range of the newly-hatched larva is still of course a matter for conjecture, but it is not difficult to see that as it climbs it must moult and become a Second Nauplius long before it can reach the 1500 m. level, otherwise at least straggling specimens of it would surely have been found by now in the 1 500-1 000 m. layer. The stage undoubtedly exists, for unmistakable specimens of it, swimming freely about, have been found, although so far only in negligible numbers, all, paradoxically,^ from shallow, or from near shallow, water. Only thirteen specimens have been identi- fied with certainty, eleven by Fraser and me from the shelf water of the Bransfield Strait and Ross Sea, and two by Ruud, both from the deeper (450-200 m.) net hauls made at ' Vikingen ' Stations 1 1 and 12, close to the ridge of the Scotia Arc in the western part of the Weddell Sea. The scarcity of First and Second Nauplii in his samples led Fraser to suggest that both stages may have a very fleeting existence, while Ruud attributed his scarcity to the hypothesis that they, along with the hatching eggs, would normally be found in the as yet inadequately explored water close to or right under the drifting ice. As Rustad (1934) points out, however, although the movement of pack-ice is chiefly dependent on that of the water on which it floats, it is not invariably so dependent. The eff'ect of wind may be to set up a motion relative to that of the surface drift and this motion may at times be rather rapid. It is, therefore, he says inconceivable, if the normal habitat of the youngest larvae should prove to be directly below the ice, that they would never lose contact with it, from which it may be argued that our persistent failure so far to capture them by surface nets either close to or actually inside the pack must be attributed to the fact that these very young stages do not, as Ruud had supposed, exist in this particular environment. As for Fraser's explanation it is now clear that whether the two Naupliar stages are passed through with great rapidity or not the existence of the Second Nauplius at least is not so ephemeral that it cannot be captured in very substantial numbers provided that in oceanic water one searches deep enough for it. Although his data were not copious enough to allow him to visualise the full extent and finer detail of the phenomenon of the climbing larvae Fraser in a measure foreshadowed it in the following passage. ' Is it possible that the development of eggs of E. superba takes place in water which is ^ But see p. 205. THE LARVAL STAGES 99 deeper than the lower Hmit of the vertical nets and that the eggs obtained [i.e. the few he found in his samples] are the scattered product of dispersal of a much greater mass situated in still deeper water ? ' And again, four pages later, '. . .the present observations on the vertical distribution of the Metanauplius furnish very favourable evidence for the hypothesis that the eggs are to be found in deep water rather than at the surface '. Rustad (1934) also supposed that in this species there would be a developmental ascent, but from a very shallow depth, although he had no observations to show how it could actually come about. Deep and shallow horizontal dispersal In the south-western part of the Weddell Sea, as Brennecke (1921) has shown, the Antarctic continent is bordered by a wide continental shelf where the depth is only a few hundred metres. Over this shelf the water is cooled right through by convection in winter, and its salinity, already high, is increased as fresh water is removed and salt left behind when sea-ice is formed. Owing to its great density this cold highly saline water sinks from the shelf and, flowing down the continental slope, makes its way northwards as a cold deep bottom current. The later work of Deacon (1937) seems to show that the main, perhaps the only, source of this Antarctic bottom water as it has been called lies in the southern part of the Weddell Sea where Brennecke originally found it, it being no longer possible he says to assume as other authors have done that it ' is formed by the sinking of shelf water all round the Antarctic continent, nor as Sverdrup (193 1, p. 102) supposes, that it is deflected to the left on account of the earth's rotation as it flows northwards, turning towards the west. It is on the contrary formed only in one region, the Weddell Sea, and its principal movement is towards the east '. Several reasons [Deacon writes] can be suggested to explain the very large formation of bottom water in the Weddell Sea. The surface water in the southern part of the sea belongs to the current which flows towards the west along the coast of the Antarctic Continent. Even in the Indian Ocean and the eastern part of the Atlantic Ocean this current must have a high salinity in winter, and owing to the continued separation of sea-ice from it as it flows towards the Weddell Sea its salinity must keep on increasing. Brennecke (1921) recorded surface salinities as high as 34-49%o. The deep [i.e. the warm deep] water in the south and western parts of the Weddell Sea also has a lower temperature and salinity than it has in the open sea in any other part of the Southern Ocean ; like the surface water it travels along the continental slope from the western part of the Indian Ocean, and owing to the continued mixing with the surface and bottom waters it forms a weaker barrier between the surface and bottom layers in the Weddell Sea than it does anywhere else along the edge of the continent. The efltect of the earth's rotation on the current towards the west in the southern part of the sea and towards the north along the east coast of Graham Land also tends to make the water sink in the coastal region. This effect is also likely to be more powerful in the Weddell Sea than in any other Antarctic sector, because of the exceptionally high latitude to which the sea penetrates and because the westward movement is not confined to the surface layer as it principally is in the other sectors, but extended to the deep and bottom layers. The temperature distribution suggests that the principal movement of the bottom water from the Weddell Sea is towards the east across the Atlantic Ocean ; but the low temperature of the bottom water farther north shows there is also a strong northward movement. In his most recent account of the formation of the bottom water Deacon (1959), referring to the work of Fofonoff (1956), has written: . . .he [Fofonoff] shows that water cooled to freezing point in the Antarctic must have a salinity of at least 34-5 i%o before it can form mixtures with the warm deep water that are heavier than either type of water alone. It follows that as soon as freezing water on the continental shelf round Antarctica reaches such a high salinity it can form mixtures that sink down the continental slope to feed the bottom current. If it attains a salinity as high as 34-63 %„ it is heavier than any mixture it can form with the deep water and must sink directly down the continental slope unless it is confined in a depression on the shelf. Water with a sahnity less than 34-5 i%o, even if cooled to freezing point, cannot form mixtures with the warm deep water that are heavy enough to sink below it. They must therefore float above it and feed the surface current. 9-2 100 DISCOVERY REPORTS ANTARCTIC BOTTOM WATER Conditions ideal for the formation of freezing water with a salinity greater than 34'5i%o occur during the winter months on the wide continental shelf south and west of the Weddell Sea. In summer the water in the coastal current is well above freezing point and diluted to less than 34 %o. But with the approach of winter the temperature falls and the salinity rises as drainage from the land ceases, ice separates out, snow lies unmelted above the sea ice, and mixing between the surface and underlying water is intensified by increasing convection. A month after midsummer the salinity of the coastal current passing westwards across the Greenwich meridian is rising rapidly, and on the wide continental shelf much farther south and west it feels the effect of winter to the full. Now it will be seen from Tables 13 and 14 that although there is often a rather wide vertical gap between the deepest sample and the line I take as marking the transition from warm deep to Antarctic bottom water, there are significant instances (notably at Stations 762, 823, 2594, 637, 638, 11 44, 1994, 1713 and 855) where the gap is much narrower. From this, and the probability that the normal habitat of the Second Nauplius is below 1500 m., there would seem to be a distinct possibility that the still deeper hatching eggs are in the bottom water itself and that it is directly from this cold deep stratum that the new-born krill are climbing towards the surface. The remarkable west to east progression of the larvae in and under the Weddell stream revealed in Table 17 could, therefore, it seems, be brought about in two ways, in the first instance through the spreading of the bottom water northwards and eastwards away from its place of origin, carrying with it the developing and hatching eggs, and in the second instance, the developmental ascent having been accomplished, through transport in the surface stream. 1 Somewhat speculative though this may at first appear, it appears less so when one takes into account the major features of the distribution of the bottom water itself. In his chart illustrating the potential temperature of this layer at depths exceeding 4000 m. Deacon (1937, PI. xliv) shows its colder parts (including the very coldest) flowing directly below the Weddell stream, and it seems, therefore, to be more than simply coincidence that, outside the East Wind zone, it is from directly over these colder parts, conspicuously (Table 13, Stations 823, 2594, 1 138, 1 142, 1 144) from above the very coldest, that the young krill (Table 13 and p. 200, Fig. 28) are rising, and that where such deep water has escaped into the relatively shallow basin of the Scotia Sea there too we find them climbing. We have, moreover, direct evidence, although it has only been obtained once, that it is in fact from the bottom water that the larvae are coming up, evidence pointing to the overriding importance of the role this deep current must be playing in the dynamics of the larval dispersal. At Station 2594 in 61° 51' S, 00° 11-7' E, a position not far from the southern boundary of the Weddell drift in this meridian, there was cold water extending from the bottom in 5400 m. up to the abnormally shallow depth of 1440 m. (Discovery Reports, Station List, 1937-9). The extreme upper boundary of the cold layer may even have been higher, there being no hydrological observations between 1440 and 1000 m. to show where it actually lay. In the 1500-0 m. vertical net,^ which clearly at this station (see Table 13) must have fished for some unknown period in the upper part of the cold stratum, there were 3444 eggs, 1000 Second Nauplii and 400 Metanauplii of E. superba. No eggs were taken in any of the six nets fished between 1000 m. and the surface, and this leads one strongly to suspect that those captured in the deep haul must have occurred well below the 1000 m. level, an unknown proportion of them, there is little doubt, in the bottom water itself. Of a total of 2880 eggs in which the develop- mental condition could readily be determined, 2130, or 75%, contained clearly recognisable developing 1 Below part at least of the Weddell stream the warm deep water (see p. 123) might also it seems be involved in the eastward movement (see Deacon, 1937, pp. 90, 92). ^ This net unfortunately failed to close at 1000 m. as intended and was fished open to the surface. In view of the absence of eggs at all levels traversed by the other nets between 1000 m. and the surface, however, it is assumed that those that were captured were in fact taken below 1000 m. Moreover, 75 % of the eggs recorded contained advanced or very advanced Nauplii and this (see p. 183, Figs. 21 and 22) points strongly to a deep and not to a shallow concentration. •Ki THE LARVAL STAGES t O lOI I; a O a to 15 l3 I; •* r^ oo o « o o~ n ■>)- p., W o O a T3 s T3 1:^ § s ^ s I '5, in s Co jj o o -o .3 W a o .s 2 "o J2 ^ ^1 S- S ClO c ■3 "^ O 00 o^ J^ O^ iri »J- ■* in N N w N SO >-< so in so SO I I I o o o in o >n N in r^ Z O m 0 0 0 o o o m o u-i s N "n r^ 1 0 1 O i i o in o »o o u^ N in r^ o M as 3 < 3 o o o o o o o 0 >v^ o lo O 1 T I I I I o o o o o o O >o O lo o o lo N IT) r^ o »o 4s o 0 0 in O in ^ 1* in r^ o O i in o in o M in t^ o u O. < o r ? o o I . o o o in in r^ o o 0 in in f^ 1 I o o Q o O 0 in o in N in r^ '- O O i 1 o 0 in o 0 o in N lo r^ o lo o o o o "^ O in o X is I ' I o o o o o m O lo O O O o o o o >A O in O ?N in r^ M o o o o 6 in O m o O M in r^ O in lojj DISCOVERY REPORTS First Nauplii, 1296 of them being so far advanced as to appear almost on the point of hatching. Many of the most advanced had lost their shells, large numbers of which were found in the catch, and while the loss may have occurred through jostling in the net or immersion in the formalin fixative, the fact that it did occur suggests how very close to hatching, if not actually on the point of hatching, these ripe eggs must have been. Ripe however though they were, it is considered unlikely that hatching in oceanic water would normally take place at this comparatively shallow level. The high incidence of negative observation in the 1 500-1000 m. layer, and our complete and highly significant failure to strike the First Nauplius there, points it would seem to its being a much deeper phenomenon. Accordingly, I would suggest, the presence of these ripe eggs so near to the surface is perhaps to be ascribed to the unusually powerful, and probably purely local, upwelling of the cold deep water that evidently must have occurred in this particular region, an upwelling that also brought with it the Second Nauplius (see Table 13) to a considerably higher level than usual. The large catch of Second NaupUi in the 1000-750 m. layer at Station 823 (Table 13) could also it seems be ascribed to local upwelling, influenced possibly in this instance by the proximity of the submarine ridge of the Scotia Arc which, with minimum soundings of just over 200 fathoms, lay less than 40 miles away to the north-west. It would appear then from the data presented in Table 17 that the deep movement of eggs and larvae towards the east probably has its origin in the western part of the Weddell drift, in the sector I have called Weddell West (see again Fig. 4), in December, although considerations based on the occurrence of eggs and Second Nauplii in the Bransfield Strait in November (p. 290) suggest it may even be earUer. By February, probably, it has spread to the sector I have called Weddell Middle, and by March possibly to the far eastern sector, Weddell East.^ It is probable that by the end of March no eggs or Nauplii are left in the deep east-going current, the occurrence in April of deep Meta- nauplii only suggesting that by that time the Nauplii have completed their natural span or at any rate have already deserted the cold deep stratum below. There can be no absolute certainty of this, however, because although nets have been fished in April below 1000 m., all of them with negative results, there were none, as Table 13 shows, worked below this level at any of the stations where Metanauplii actually occurred. If there had been, it is possible, they might have revealed the last surviving Second Nauplii rising from below. From April onwards, it seems evident from the vertical distribution, eastward transport must for all practical purposes be confined to the surface drift. Eastward transport in the surface stream it seems has its beginnings in Weddell West in December when the larvae first begin to reach the surface in small numbers. It is in full swing there in Januarj^^ and by February has already progressed as far as Weddell Middle. It is not, however, until April that its eifect is seen on any appreciable scale in the far eastern sector, Weddell East. A natural outcome of the repeated populating of the surface stream by very young larvae from below is that in regions such as the Weddell drift, where spawning (p. 177) is spread over a long period, there will always, as might be expected, while spawning and hatching continue, be two broadly distinguish- able larval stocks in the sea, an older that has already reached the surface and a younger rising from below. This phenomenon is illustrated in Table 18 in which the figures in roman type show the total larval catch at all stations where deep rising larvae occurred and the figures in italics the corresponding catch at stations where the deep risers were absent. Although manifestly characteristic of the Weddell drift the dual age of the larval stock does not, our observations seem to show, have its counterpart 1 See, however, p. 208. « Table 17 does not show this. The Norwegian ' Vikingen' expedition, however (Ruud, 1932, p. 35, Table 10), recorded enormous numbers of First Calyptopes on the surface in Weddell West in January. THE LARVAL STAGES 103 Table 18. The two broadly distinguishable age groups of the larvae in the Weddell drift, the older group in the surface, the younger rising from below, the figures for January and February including data from the South Georgia whaling grounds {For further explanation see text) Calyptopes Furctlias Depth Second Meta- , * — , ^ — * , Month (m.) Nauplii nauplii ^23123456 April 250-0 — — ig 1487 704 343 362 50 17 9 3 — — 216 58 33 3 4 2 — — — 500-250 — — —11—82282 — — 67 8 3 i_____ 750-500 — — — — — ______ — 221 43 — — — — — — — _ 1000-750 — — — — — — — — — — — — 945 10 — — — — — — — _ 1500-1000 — — — — — — — — — — — March 250-0 — — g2 147 2g68 2235 510 145 302 246 45 — — 7393 1275 239 207 90 3 _ _ _ 500-250 — — — 12 3 2 — — — — — — — 647 114 18 16 3 — — — — 750-500 ___________ — "99 330 2 — — — — — — — 1000-750 — — — — — — — — — — — — 938 52 5 _ _____ _ 1500-1000 — — — — — — — — — — — Feb. 250-0 — — 4g 86 244 272 158 ig 1 — — — — 570 181 2 — — — — — — 500-250 — — 4 4 11 3 ^____ 750-500 ___________ -38 52 -------- 1000-^50 — — — — — — — — — — — 434 2308 66 — — — — — — — — 1 500-1000 — — — — — — — — — — — 1000 400 — — — — — — — — — Jan. 250-0 — — 10 g2 187 36 8 2 — — — — — 43 10 7— 3____ 500-250 — — 1 6 5 6 2 — — — — 750-500 ___________ 4 22 4 — — — — — — — — 1000-750 — — — — — — — — — — — 334 59 3 _______ _ 1500-1000 — — — — — — — — — — — Dec. 250-0 — — 1 — — — — — — — — 500-250 ___________ 750-500 ___________ 3 — — — — — — — — — — 1000-750 — — — — — — — — — — — 10 9 — — — — — — — — — 1500-1000 — — — — — — — — — — — J04 DISCOVERY REPORTS Table 19. Virtual absence of broadly distinguishable age groups among the larvae of the East Wind drift Calyptopes Furcilias Depth Second Meta- , ' , < * ^ Month (w.) Nauplii nauplii i 2 3 i 2 3 4 5 6 April 250-0 — — 47 ^4 — — 241 69 9 — I — — — 500-250 — — — — — --48 -------- 750-500 — — — — — — _ — 379 74 — — — — — — — _ 1000-750 — — — — — — 322 41 — — — — — — — — 1500-1000 — — — — — March 250-0 — — n 2 __ 575 16 — — ^ Z Z Z Z 500-250 — — 23 — — — — 149 — — — — — — — — 750-500 — — — — — 509 29 — — — — — — — — 1000-750 — — — — 18 442 I — — — — 1 500-1000 — — — — 510 95 — — — — — — — — — Feb. 250-0 — — 1 — ^ -- 138 -------- 500-250 — — — 1 --38 -------- 750-500 — — — — — 16879 — — — — — — — — iooa-750 — — — — -138--------- 1500-1000 — — — — Jan. 250-0 — — — ^ — I — — — — — 500-250 — — — — 750-500 — — ~~~~~~~zz 2 '^"""""ZZ — — — 1000-750 — — — " 5 '"""^""■ZZZZZZ 1 500-1000 — — — — among the larval communities of the East Wind zone. For there, as Table 19 shows, in February, March and April, when in the Weddell zone the dual age is most apparent, the only clearly distinguishable stock is that of the younger rising from below. There can be little doubt that our repeated recording of the young risers in the East Wind drift to the virtual exclusion of an older population, while pointing on the one hand to a much curtailed spawning season, a short-term phenomenon suggestive of the 'spontaneous' outburst (p. 44) of Ruud, is associated too with the brief and sporadic opening up of the ice-fields which in these high latitudes is known to occur irregularly and at unpredictable times and places all round the Antarctic continent (see p. 177). THE LARVAL STAGES 105 While then it seems clear that eastward dispersal of the larvae in the Weddell region is brought about by both deep and shallow transport, it is equally clear, at least in the East Wind zone and (p. 123) in the eastern part of the Weddell drift, that there must be a period when they are moving towards the south. That is, when in the course of their long upward passage they traverse, developing from First Nauplii to First Calyptopes as they go, the warm mainly south-flowing deep current sandwiched between the bottom and surface streams. The full significance of this movement is discussed on pp. 118-23 i^ ^^^ section dealing with the developmental ascent and its influence on the distribution and conservation of the euphausian population. Diurnal vertical migration It has been shown that the vertical movements of the First and Second Nauplii and of the Meta- nauplius and its immediate successor, the First Calyptopis, must at the outset be upward. Let us now consider the nature and extent of such subsequent vertical movements the larvae may undergo once, as First Calyptopes, they arrive in the surface stream. In a diagram based on a series of vertical stations, widely separated in space and time but arranged in a time sequence covering most of the 24 hr. of night and day, Fraser (1936, p. 116, Fig. 33) demon- strates what at first sight appears to be a pronounced diurnal migration of the Calyptopis stages in the course of which they become concentrated in the dark hours between 2200 and 0400 in the surface ( 1 00-0 m.) layer, while for the rest of the time throughout the daylight hours they appear to be massed at depths mainly between 500 and 250 m. At each station of this series deep-living Metanauplii occurred in moderate to substantial numbers, the later stages, however, being represented for all practical purposes by the First Calyptopis alone. Now since the latter is the stage which in the course of the developmental ascent makes its earliest appearance between 1000 and 500 m., it would seem that Fraser in the interpretation of his results had overlooked the possibility that the apparent massing of the First Calyptopes in daj^time at depths below 250 m., instead of being as he supposed the result of a migration from above, could equally well have sprung from the recent deep moulting of the Metanauplii along with which they were captured. In other words, in a species such as this in which climbing larvae are constantly feeding the surface population from below the establishment of the reality of diurnal vertical migration, in so far at least as the First Calyptopis is concerned, presents a more complex problem than it would seem to do in species that do not have their early existence at such enormous depths as those at which the first young krill appear. In fact, as will be seen presently, if we fail to reckon with the developmental ascent, and all it implies, the diurnal movements of the First Calyptopes cannot be worked out without sometimes involving a major misunderstanding of what is often taking place at levels below 250 m. The basic aspects of the problem are illustrated in Fig. 6 which presents the vertical distribution of the First Calyptopis at oceanic stations where deep Metanauplii were moulting and the developmental ascent was in progress. Stations where only one Metanauplius occurred or where Metanauplii only occurred are not included. In its construction Fig. 6 follows the conventional lines that Fraser used, the day being divided into six periods of 4 hr. each, the short-term gatherings of the 50-0 m., 100-50 m. and 250-100 m. nets multiplied by 5, 5 and 5/3 respectively to bring them into line with the deep gatherings from the 500-250 m., 750-500 m. and 1000-750 m. layers, the resultant catch-figures being finally expressed graphically as a percentage of the total catch between 1000 m. and the surface. A check on the significance of the resulting graphs is provided by tables showing the actual catch-figures upon which each is based. Figs. 8-1 1 , which follow presently, have been constructed in the same way. In his original figure portraying the vertical movements of the First Calyptopis Fraser (1936, p. 152, Fig. 72) shows it massed conspicuously in the 100-50 m. layer between 1800 and 2200 hr., io6 DISCOVERY REPORTS a period covering the daylight, dusk and darkness of the approaching end of the day. Clearly, however, as Fig. 6 shows, when co-existent with Metanauplii in the plankton, it is by no means always massed so conspicuously then at this particular level, but may in fact be encountered in equally conspicuous concentration at any level from the surface down to looo m. Fig. 6 in fact reveals a vertical distribution for this period so haphazard that I can only conclude that its motley pattern springs from the fact that sometimes our nets are striking deep concentrations of First Calyptopes that have but recently cast off their MetanaupUar husks and at other times shallower concentrations sprung from earlier moultings. In other words, wherever the deep Metanauplii are moulting and feeding the upper layers the expected level or levels at which the main body of the First Calyptopes will occur, at any time it seems of the day or night, will generally be unpredictable, since in many instances it must be dependent upon how far the larvae may have journeyed upwards towards the surface from the level of their first deep appearance. But let us look more closely at the graphs in Fig. 6. The massing of the First Calyp topis in the 50-0 m. layer at Station 647, for instance, does it is true suggest a pronounced diurnal movement up to the surface during the darkest hours of the night.^ It is also possible, however, that their appearance at the surface then coincided with the culmination of their long climb from deep water, and if this be true we are not confronted with an instance of diurnal movement here. But what of the deep and by no means insignificant occurrences of First Calyptopes at this and other stations that have been recorded both by night and by day? Are we to conclude that they too spring from diurnal movement? Surely the actively moulting Metanauplii^ with which they are so closely associated does not permit such a view. On the contrary, it can only be concluded that they are deep concentrations, recently sprung from their Metanaupliar parents, either at the outset of, or in some instances well embarked upon, their journey towards the surface, and that they have never in fact been near the surface at all. At Station 647 the developmental ascent is in full swing and large numbers of First Calyptopes, springing from the Metanauplii massed below 500 m., are climbing towards the surface. It is unlikely, however, that all of these climbers will actually reach the surface here, since the majority of them, being still in the warm deep current, must be moving away from Station 647 in a direction with a southerly trend. It is equally unlikely, I think, that the 361 Calyptopes at the surface sprang from the same batch of Metanauplii as gave rise to the climbers, for if they had, having regard to the move- ment of the warm layer in which earlier they must have spent some considerable time climbing, they would have appeared at the surface somewhere to the south of Station 647. They could only in fact have sprung from the same batch, or rather part of the same batch, if the Metanauplii in this locality had been spread horizontally over a considerable distance to the north. Some further evidence of the more or less unpredictable subsurface levels at which the First Calyptopis may be encountered when co-existent with Nauplii or Metanauplii in the plankton is provided by Fig. 7 presenting data drawn from (i) oceanic stations where only one Metanauplius occurred, (2) oceanic stations where Nauplii and Metanauplii alone occurred, and (3) stations in shelf or slope^ water where Nauplii or Metanauplii were also taken. Even with such small numbers as these the vertical distribution of the First Calyptopis from dawn to dawn presents, in the presence of Nauplii and Metanauplii, the same haphazard pattern as it does in Fig. 6. It is evident too, having regard to the exclusive occurrences of Nauplii and Metanauplii at Stations 2594, 1099, 762 and 764, that had our nets struck these organisms at the precise moment when they were moulting and not, as 1 This large catch was actually made at about 0130 hr. in 59° 29J' S. The major part of the vertical work at this station, however, took place between 0200 and 0600 hr. ^ See again p. 97 and footnotes to Tables 13 and 14. ' The seemingly anomalous near-surface occurrences of Nauplii and Metanauplii in shallow water, several instances of which appear in Fig. 7, are discussed on p. 205 and illustrated in Figs. 29 and 30. THE LARVAL STAGES 107 TIME STATION, 1400-1800 II04 25101 I800-2200 1965 855 637 1144 1138 1545 2346 1671 2600 1713 1994 2603 2200-0200 1492 620 518 854 I-O6O0 647 0600-I000 1WI97 638 823 ! I I 1400 636 B TIME STATION ■0 -SO 100 250 500 750 1000 50 100- 250- 500- 750- lOOO- CI M3 C- M3 10 2 14 13 90 155 2 I 36 59 142 49 ' 361 52 2 440 461 19 112 1 19 69 31 95 21 1215 4720 115 63 167 97 1 47 58 13 66 51 4S5 36 39 115 - 20 32 57 2 53 245 78 41 2 332 24 549 504 57 221 I 10 5 67 I 22 7 443 59 51 36 185 4 10 36 - 10 911 138 830 1 39 I 387 148 196 29 35 10 2 4 7 20 47 3 24 5 137 20 94 I50 282 I3702 7 577 36 28 2 2 2 3 18 115 25 31 50 100 ■250 •500 •750 •1000 Fig. 6. Four-hourly vertical distribution of the First Calyp topis at all stations in deep oceanic water where more than one Metanauplius occurred, the First Calyptopis in black, the Metanauplius in outline. The numbers of Calyptopes (C) and Metanauplii (M) recorded at each station are given in the catch-table below, the figures being actual catch-figures not corrected for length of haul. TIME STATION SO 100 250 SOO 750 1000 I400-1800 W383II04 1553 1988 1 15 m 1800 -2200 1100 1101 1140 1152 1214 1215 1546 1659 1662 2018 m I I m 2200- 0200 1098 2594 m 0200- 0600 1114 1554 0600- 1000 I099 1552 1000- 1400 762 754 -O TIME STATION SO 100 2S0 •SOO 7SO 1000 Fig. 7. Four-hourly vertical distribution of the First Calyptopis at other stations, including shelf stations, where Nauplii or Metanauplii occurred, the shaded portions in the lower part of the diagram indicating the approximate position of the bottom in relation to the deepest observations in shelf water. The Nauplii or Metanauplii are here distinguished from the Calyptopes by the figures enclosed in squares. ,o8 DISCOVERY REPORTS they did, before, concentrations of First Calyptopes confined exclusively to great depths will prove in the long run to be of just as common occurrence during the darkest as during the brightest hours of the day. It follows from the foregoing that any attempt to illustrate the diurnal vertical movements of the First Calyptopis, if it be based as in Fig. 8 (or as by Fraser in his Fig. 33) exclusively upon stations where deep Metanauplii are moulting, or any attempt whatsoever that fails to take account of the I400-I800 IBOO-2200 2200-0200 O2O0-O60O 0600-IOOO IOOO-I400 0-1 SO 100 250- 500 750 lOOO-J SOO 1000 0- C M - 298 286 361 ~ ^ C M 50- C M - 1305 373 20 _ _ C M 100- C M 12 6548 21 94 38 5 C M 250- C M 10 889 36 I50 32 31 C M C M 3 17 547 2318 23 69 282 3702 6 2! 1 17 C M 750- C M 15 146 2573 14 165 7 577 144 - C M Fig. 8. Pseudo-diurnal vertical migration. The four-hourly vertical distribution of the First Calyptopis based on the aggregate of the catch-figures shown in Fig.^6. disturbing influence of the deep moulting of the Metanauplius upon the results, must end up in a picture that in some respects is illusory. For such an illustration must inevitably include large numbers of larvae which, although to all appearances migrants into deep water from the surface, could never in fact have been anywhere near the surface at all. Fig. 8 in short provides a typical example of the many pitfalls oceanography falls heir to through the employment of nets that do not go deep enough. For if the lower limit of observation in this instance had been say 500 m. there would have been no reason to suppose that the orthodox pattern of rhythmic diurnal vertical migration it manifestly presents was anything other than real, or to doubt that the First Calyptopis made a daily descent into the warm south-flowing core of the deep current running counter to the surface stream. That the picture is not in fact real, but partly an illusion created by the presence of recently born, and still climbing First Calyptopes rising from below, can clearly be established if the dawn to dawn vertical distribution of the combined Calyptopis stages be based, as in Fig. 9, upon data drawn from stations where the First Calyptopis^ occurred free, that is, not in association with deep-mouhing Metanauplii, thereby excluding any possibility that deep, still climbing, Calyptopes could enter into and prejudice the resuh and making reasonably sure that the Calyptopes as a whole, the developmental ascent having been accomplished, had definitely reached the surface. Worked out in this way the vertical extent of the diurnal movement, if in fact it can be described as such, of the 1 But not the Second and Third, since neither, except in apparently one rare instance (p. 90, Table 13, Station 1138, footnote t), not included in the construction of Fig. 9, is involved in the developmental ascent. THE LARVAL STAGES 109 I400-I800 I800-2200 2200-0200 0200-0600 0500-I000 lOOO- 1400 100 250 SOO 750 1000 IT T^ A so 100 •250 •SOO •750 1000 SO 100 2 SO SOO 7 SO 1000 213 2767 351 17 119 824 12 1323 76 4 221 1793 55 382 21 7 IIS4 519 146 (lis) 27 Od 4 6(5) 41 (10) 229 (188) 0) 11(1) 2(1) — 13 (12) 6(5) 5(4) II 3(2) — 4(3-) 2(1) C48) (120) (37) (15) (38) (44) SO 100 2 SO •t-soo 7SO 1000 Fig. 9. Four-hourly vertical distribution of the combined Calyptopis stages showing absence of any pronounced diurnal movement. The distribution here is worked out from data obtained at oceanic stations where the First Calyptopis occurred free, the number of stations made during each 4-hr. period, where such fell within the life-span of the Calyptopes, being shown in brackets at the bottom of the catch-table. These include negative stations. The bracketed numbers in the catch- columns indicate the considerable part of the total catch (unbracketed numbers) which was obtained from vessels lying beam on to wind and sea in winds of force 4-5 or more. !400-l80O I8OO-22OO 2200-0200 O2OO-O6OOO6OO-IOOO I000-I400 o 50- 100 2SO SOO 750' 1000 SO 100- 2SO- SOO 750- lOOO- T T» o -so -100 2SO SOO -750 1000 412 32 9S 3S 246 162 20 32 112 470 23 (46) (Ml) (37) (15) (37) SO 100 2SO 500 ■7S0 1000 (44) Fig. 10. Four-hourly vertical distribution of the First Calyptopis based on data from oceanic stations where it occurred free. Note again the figures in brackets showing the number of stations (including negative stations) made during each 4-hr. period where such fell within the life-span of this stage in the plankton. free Calyptopes is revealed to be rather small. It certainly, for instance, does not seem to involve them in any very large-scale daily descent into the vv^arm current, and suggests too that such vertical movement as may be taking place, whether it be diurnal or not, is taking place in the main within the limits of the Antarctic surface layer. Owing to its being so consistently involved in the developmental ascent (p. 90, Tables 13 and 14) substantial numbers of First Calyptopes, as Table 15 shows, are seldom encountered free, even in the jio DISCOVERY REPORTS surface layers. In deep water below 250 m. Fraser records only ten free, Fry and I, in a very much larger series of analyses, finding only forty. In other words the First Calyptopis in deep water is nearly always associated with moulting Metanauplii. Even in such free occurrences as have been FURCILIAS 1-3 I400-I800 I800 -2200 2200-0200 0200-0600 O6O0-IO00 I000-I400 O- 50- lOO- 250- SCO- 750 1000 O- 50- 100- 250- 500- 750- 1000^ —I--T- 174 1155 268 36 10 276 18 24 299 4 1 (40) 692 (492) 36 (21) (3) 0) 12 (28) (68) (31) (10) (28) 598 1418 4B2 51(26) 270(209) 36(35) (25) 100 250 SOO 750 1000 -0 50 100 2SO 500 ■750 •1000 SO 100 250 SOO 750 1000 fURCILIAS 4-6 I400-I8OO 1800-2200 2200-0200 0200-0600 0600-I000 I000-I400 II -il T o so 100 250 500 750 1000 23 964 48 2 115 33 8 64 4 — 4 11 1 7 1 — 2 — 47 (46) ' 2 — — 23(2) (28) 0) — — — — (11) 0) — — — 3 -" (42) (129) (31) (17) (45) (35) O SO -100 -250 SOO -750 -1000 O SO 100 250 500 750 1000 Fig. 1 1 . Four-hourly vertical distribution of the Furcilia stages showing absence of rhythmic diurnal movement and more or less permanent massing of the main concentrations high up in the Antarctic surface layer. For further explanation see legend to Fig. 9. recorded, however (Fig. 10), there is little indication of any pronounced or rhythmic diurnal vertical movement and equally little of any mass descent into the warm core of the deep water during the dayUght hours. The Furcilias (Fig. 11) appear to undergo much the same rather restricted vertical movement, Furcilias 4-6, more so perhaps than the early Furcilias, showing a distinct tendency to be massed near the surface both by day and by night. THE LARVAL STAGES ii, Table 20. Diurnal vertical distribution of total larvae {Calyptopis 2 to Furcilia 4) recorded by Fraser at depths below 250 m. Depth 1400-1800 1800-2200 2200-0200 0200-0600 0600-1000 1000-1400 (m.) Second Calyptopis 500-250 39 2 2 7 14 45 750-500 I 2 — — _ _ 1000-750 21 I — J Third Calyptopis 500-250 93 18 — 2 9 203 750-500 — — I _ 3 J 1000-750 2 — 2 — First Furcilia 500-250 17 2 — — I 156 750-500 _____ J 1000-750 2 I — — Furcilias 2-4 500-250 49 3 — — 5a 132 750-500 _ _ _ _ 21 36 1000-750 I — — — Fraser (1936, p. 152, Figs. 72 and 73), in Calyptopes i, 2 and 3, in Furcilia i and in the combined Furcilia stages 2, 3 and 4, shows a very much more pronounced diurnal vertical movement than my Figs. 9-1 1 reveal, his results for Furcilias 5 and 6, however, being in general agreement with mine. Our conflicting findings, in so far at least as the First Calyptopis is concerned, can, however, partly be explained because the volume of material at my disposal was so large^ and partly because in presenting his data Fraser did so without regard to the possibility that the climbers could prejudice the result. In consequence his graphs for the period 0600-2200 hr., from dawn in fact throughout daylight to nightfall, are affected by the disturbing influence of the deep, still climbing, larvae (see again Fig. 6) that spring from the moulting Metanauplii at Stations 636, 637, 638 and WS 197. No such disturbing influence, of course, can have affected his graphs for Calyptopes 2 and 3, Furcilia I and Furcilias 2-4, for such larvae have long since accomplished the developmental ascent and their vertical movements can no longer be obscured by the presence in deep water of young climbers still reaching for the surface, as the vertical movements of the First Calyptopis so often are. Even so, the numbers of these stages he has recorded below 250 m., although distinctly suggesting (Table 20) a diurnal migration by some of the larvae well down into the warm layer by day, are not on the whole I think large enough to justify the conclusion that the mass of the surface population behaves in a similar way. Indeed the massive accumulation of larvae revealed by our two short-term 50 m. hauls above 100 m. (see again catch-tables for Figs. 9-1 1 which include Eraser's catch-figures as well as my own), throughout all hours of night and day, distinctly suggests it does not. If it did I should have expected equally large daylight gatherings from below 250 m., where our nets are hauled for five times as far, with daylight scarcity or even void at the surface. I therefore conclude that although both Calyptopes and Furcilias migrate away from the surface down into the warm deep layer, especially it seems (Figs. 9 and 11) between 0600 and 1 A conspicuous deep day massing of First Calyptopes, 70% of them in the 250-100 m. layer, the remainder in the 500-250 m. layer, appears between 1400 and 1800 hr. in Eraser's Fig. 72. This graph, however, is based on the occurrence of only seven specimens. 112 DISCOVERY REPORTS 1400 hr., this movement does not it seems involve a very significant part of the total surface population.' It is distinctly possible too that even the minor to moderate daylight gatherings we have recorded in the 500-250 m. layer and deeper may in many instances have come from a higher level, there being some doubt about some of the early vertical horizons recorded in R.R.S. 'Discovery' and R.R.S. 'William Scoresby'. Both vessels had to be laid beam on to wind and sea on station and in con- sequence were inclined to make so much leeway, their wires straying far from the vertical, that many of the nets they fished, more especially in the stronger winds, must in fact have been sampling higher levels than they are reputed to have done. The old 'Discovery' (Chaplin, 1932), with her massive hull and clutter of rigging, was particularly liable to drift, ^ and ' Scoresby ' too on station sometimes made leeway at quite astonishing speed. I recall one occasion in the Bransfield Strait particularly vividly when she must have been moving bodily to leeward at a rate of at least 2 knots. Here, at Station WS 391, we were working vertical nets in a 30 knot (Force 7) wind and our wires were straying far out to weather at an angle that seemed to be over 45°. Our leeway in fact was so great that I decided to tow the oblique (loo-o m.) nets without the engines as we drove bodily before the gale. The usual 200 m. of warp was paid away with a 56 lb. streamlined lead and bar, one stramin net on its heavy ring and one 70-cm. diameter silk net attached by their closing apparatus, at the end. Yet with all this massive clutter of gear on its heavy warp the nets at their deepest scarcely touched 120 m. I therefore conclude that the Calyptopes taken in the much less weighty 500-250 m. vertical net between 0200 and 0600 hr. on this occasion must in fact have come from a higher level. I think, too, that the 480 early Furcilias recorded by Eraser in the 250-100 m. net fished between 0600 and 1000 hr. at Station WS 527 (Fig. 11), with the ship drifting bodily to leeward before a 20-knot breeze, could also have come from nearer the surface. On the same grounds the accuracy of some of the early vertical horizons recorded in R.R.S. 'Discovery II ', before her then Commander, the late W. M. Carey, R.N. (Retd.), successfully laid her head on to wind on station, is also open to doubt.^ I feel, therefore, it must be not a little significant that so many of the larvae we have recorded as having apparently migrated from the surface down into the warm deep layer below 250 m. (see bracketed numbers in catch-tables for Figs. 9 and 11) should have been taken from drifting vessels moving steadily along before winds of force 4-5 (15-21 knots) or over. It follows from the foregoing that of the two broadly distinguishable larval communities in the sea, the climbers and those with the ascent accomplished, the latter, as Professor Ruud (p. 44) recognised so long ago, are essentially creatures of the Antarctic surface layer, and that there as they grow, especially as Furcilias, their vertical movements become confined to an increasingly narrow stratum so that they tend more and more readily to get carried along in the surface stream. The pronounced and steady decline in total numbers of larvae* taken below 250 m. as development proceeds (Table 21) provides further evidence of the lessening range of their vertical movements as they grow. Mortality it is true may also be contributing to the decline, but I do not believe it to be the overriding factor. In terms of average numbers taken per 250 m. haul, the vertical distribution of the total free popula- tion of grouped Calyptopes and Furcilias works out for daylight and darkness as in Table 22 where {a) shows the average gatherings based on all stations, both positive and negative, made during the * Reproductions of Eraser's Fig. 72, showing the pronounced diurnal vertical movement of his Calyptopes, appear in Gushing (1951), Zimmer (1956) and Nicol (i960). * Striking illustrations of how her plankton, hydrological and sounding wires used to drag out to windward, even in moderate weather, are given by Gunther (1928). ^ This, however, applies only to the first twenty of the grand total of 1320 odd vertical net stations she made during her long career in southern waters. * Excluding of course the still climbing First Calyptopes. THE LARVAL STAGES 113 life-span of each developmental phase in the plankton, (b) the corresponding gatherings based on positive records only, the stations involved in the construction of both (a) and (b) being shown by the bracketed numbers. Both presentations of the data reveal (i) a heavy massing of the total free population in the top 100 m., (2) that such vertical migration as is going on does not by any means seem to involve the total surface population, (3) that it is taking place, or seems to be taking place, principally in the 1 00-0 m. layer and does not involve either the night time or day time population in any mass descent into the warm core of the deep current flowing counter to the surface stream. Table 21. Decline in total numbers of larvae taken below 250 m. as development proceeds, the figures in brackets showing the numbers of stations worked during the life-span of each grouped developmental phase lyptopes Furcilia Furcilia Furcilia Furcilia Furcilia Furcilia 1-3 I 2 3 4 5 6 5" (302) 187 133 (190) 151 63 40 (299) 13 Table 22. Vertical distribution of total free population of grouped Calyptopes and Furcilias expressed as average numbers taken per 250 m. vertical haul, the asterisks indicating average gatherhigs of less than one («) Daylight Darkness Depth t Calyptopes Furcilias Furcilias Calyptopes Furcilias Furcilias (m.) 1-3 1-3 4-6 1-3 1-3 4-6 50-0 38 30 5 89 77 34 100-50 67 60 * 77 17 2 250-100 18 13 # 3 5 * 500-250 3 3 # * # * 750-500 * # * # # — 1000-750 # * # * # * (ISO) (107) (179) (A) (152) (83) (120) 50-0 96 68 32 235 119 93 100-50 168 134 3 III 26 6 250-100 46 29 * 9 8 # 500-250 7 7 3 # * * 750-500 # I # * * — 1000-750 # * * # # # (60) (48) (28) (58) (54) (44) The vertical movement that seems to be taking place seems to be confined principally to the Calyp- topis and early Furcilia stages, but not to Furcilias 4-6. They (the late Furcilias), it seems, tend to be gathered night and day in the top 50 m., where in view of the relatively small numbers we have recorded, it is distinctly possible they are assembled in shallow 'rafts' (pp. 234-6) through which the vertical nets fish only momentarily. Finally, to conclude this survey, the question might well be asked, are the conventional methods employed in portraying diurnal vertical migration when the net hauls involved are of variable duration in every instance justified? In other words are we always on safe ground when resolving vertical hauls of 50 sec, 150 sec. and 250 sec.,^ into hauls of a uniform 250 sec. duration? This practice is ^ Our vertical nets were hauled at the rate of a metre a second. 114 DISCOVERY REPORTS based on the assumption that the organisms whose diurnal movements one is attempting to portray are evenly (or randomly) distributed throughout each depth interval traversed by the vertical nets. But are the larvae of £■. superba in fact so disposed? It will be shown later (pp. 234-6) that there are considerable grounds for supposing they are not, but, like the older stages (p. 149), are disposed in shallow ' rafts ', no more than a yard or two thick, through which the nets in each instance pass perhaps for only a second or two. Assuming such a disposition, and assuming too that only one such raft existed in each vertical horizon, then the time correction applied in the case of the shorter-term TIME STATION O- SO- 100- 250 500- 750 1000- N5 OF METANAUPLII I400- -1800 1104 2610 O- 50- 100- 250- 500 750- 1000- I800 - 2200 1965 655 637 1144 1138 1545 2345 1671 2600 1713 1994 26031 63 517 34 557 1515 195 I05I 10 106 5 8 830 22O0 - 0200 1492 520 618 854 I0200- 0600] 647 24 6 20 184 0600- lOOO |WI97 638 823 !?: 115 7 43 looo- -woo 636 + TIME STATION O -SO -lOO -250 -500 -750 -lOOO N5 OF METANAUPLII 12 10 2 II 52 95 13 67 36 - 13 9 3 2 440 461 19 21 1215 4720 116 66 51 485 36 2 53 245 78 4 - 35 - 5 112 63 39 41 lO 167 115 5 I 90 119 97 4 165 69 2 31 47 20 22 59 148 II 10 6 142 49 196 29 5 5 - 35 4 7 5 351 - 20 - 94 36 ISO 28 282 2 7 2 5 - 31 2 I O SO lOO -250 500 7SO lOOO Fig. 12. Four-hourly vertical distribution of the First Calyptopis based on the catch-figures shown in Fig. 6 uncorrected (histogrammatically) for duration of haul. upper-level net hauls would be unnecessary since the nets in every instance of striking a raft would strike it for only a second or two and so produce samples which, whether the haul were long or short, would for all practical purposes be comparable. The four-hourly vertical distribution of the First Calyptopis, when co-existent with Metanauplii in the plankton, might for all we can tell then be as represented in Fig. 12 in which the catch-figures in Fig. 6 have been expressed graphically as they stand without any conventional correction having been applied to the upper-level hauls as hitherto. In this presentation of the data, which incidentally reveals even more emphatically than Fig. 6 does the haphazard and unpredictable subsurface levels at which, regardless of hour, the non-free First Calyptopis may be encountered, it will be seen that where significant numbers are involved, as for instance at Stations 171 3, 1994, 620 and 618, the concentration of the larvae near or moderately near the surface has not been recorded except in instances where the Metanauplii were scarce, leading one to suspect that the moulting of the latter at these particular stations had taken place some considerable time back in the past. Dr Fraser with whom I have closely discussed the whole question of the diurnal movements of the larvae agrees that it is a complex one and that in so far as the First Calyptopis is concerned there can be no satisfactory answer without there having first been separation of the deep climbers from those that have reached the surface. He makes the following comments which I present with some slight alteration in wording but not in sense. (i) Both Marr's and Eraser's method of presenting the vertical movements by histograms is inadequate to the extent that it does not take into comparative account the very different order of numbers at different 4-hourly periods. THE LARVAL STAGES 115 (2) Marr has included his actual results and this is a compensation for the percentage per level presentation in both his and Eraser's text-figures. (3) Marr has concentrated on his positive results but has not referred to his or Eraser's negative ones. (4) At stations where deep Metanauplii occurred neither Marr nor Fraser record any First Calyptopes in the top 100 m. between 0600 and 1800 hr. These are the main hours of dayUght and from Marr's and Fraser's combined observations it is reasonable to predict that very few or no First Calyptopes could be expected at this high level during this period of the day. (5) Marr and Fraser both show a relative increase of First Calyptopes at depth in daylight and a decrease at night. (6) Marr's Text-fig. 8 suffers from the disadvantage already referred to in that the percentages mask the actual results. Marr does not prove from this figure that ' large numbers of larvae . . . could never in fact have been anywhere near the surface'. (7) Marr's Text-fig. 9 combines all three Calyptopis stages. Fraser doubts the validity of this. (8) Marr's Text-fig. 10 (free First Calyptopis) shows there is only one specimen between 100 m. and the surface from 46 stations against 55 specimens between 500 and 100 m. at 1400-1800 hr., and nearly five times as many at loo-o m. than at 500-100 m. between 1800 and 2200 hr. At 0600-1000 hr. the concentration is 60 times as great below 50 m. as above it. The results for 1000-1400 hr. appear to support Marr's point of view but not if the numbers (not the histogram figure) relative to those of other times are considered. (9) Fraser agrees that the First Calyptopis population is composed of newly developed larvae climbing to the surface in the post-metamorphosis phase and those that have already arrived in the shallow water. What neither Marr nor Fraser knows is the rate at which the 'climbers' ascend and the duration of the First Calyptopis stage. Therefore there can be no assessment of the effect that the additions of First Calyptopes from below are having on the migration pattern of the larvae at higher levels. On the face of it and taking into account the orthodoxy of the pattern in both Marr's* and Fraser's figures it (the cUmbing factor) does not blur the impression that the diurnal migration is indeed happening. It has to be borne in mind also that there is no evidence to support the assumption that once the 'climbing' larvae come into the zone where the factors controlling diurnal migration are operating they (the climbers) will react differently from the 'resident' population. (10) Marr's page 113. A minor point. As an alternative to the rafting explanation for the diminished density of Furcilias 4-6, would not the expectation be that just as there are less Furcihas 1-3 than Calyptopes 1-3, so on ordinary mortality factors in a population, it is to be expected that there would be fewer Furcihas 4-6 still, assuming, naturally, the sampling of all stages to be adequate? Taking these points in turn, I reply, ( 1 ) My presentation of the data, as Dr Fraser acknowledges in (2), does at least carry with it the actual numbers on which each percentage per level histogram is based and it will be seen that in general (the period 0200-0600 excepted when our data are indeed very scanty) the numbers are adequate enough to be convincing. In any case they are all the data we have got and I am at a loss to think what more could be done with them or how otherwise they could be presented. I would point out too that in Figs. 9, 10 and 11 I show the numbers of stations, both positive and negative, made every 4-hr. period during the life-span of each developmental phase in the plankton, such stations being strictly confined to times and waters in which, our distributional charts show, the several developmental phases would be expected to occur. (2) See (i). (3) Dr Fraser also concentrates on his positive results and I see no objection to this approach to the problem. The positive results do at least show the actual disposition of the larvae at all times when we strike them. Negative results at all levels, and we have many such both for day and for night, merely show that the larval distribution, like that of the older stages (pp. 148-54), is very patchy (see also p. 219, Tables 45-7). Consistently negative daytime results at near surface levels, with consistently positive results below, would, I agree, be highly significant and would suggest strongly we were dealing with rhythmic diurnal vertical migrators. However, it will be seen (Table 22) that in so far as the ^ This refers to my Fig. 8 which I believe to be conveying a false impression. „6 DISCOVERY REPORTS free Calyptopes are concerned the average daytime gatherings in the top lOO metres, though some- what smaller, are not significantly smaller, than the corresponding gatherings at this level at night,i and that in so far as negative observations go, there are just as many, in fact rather more, at this level at night as by day. The same can be said for the early Furcilias. There are it is true considerably more negative daytime observations for Furcihas 4-6 than night-time ones, but clearly such positive results as we have point to a pronounced concentration of these stages, both by night and by day, in the top 50 m. In so far, however, as the Furcilias are concerned, especially the late Furcilias, it might it seems be expected that there would be more negative observations (and on the average smaller gatherings)^ in daylight than in darkness, for as Southern and Gardiner (p. 268) have shown for the much smaller freshwater Crustacea of Lough Derg, avoidance of the nets in the ' upper illumi- nated water layer ' in daylight is a factor that cannot be ignored. The average length of Furcilia 4 is 8-OI mm., of Furcilia 5, 9-52 mm. and of Furcilia 6, 11-34 J^"^- Towards the end of their life-span the majority of Sixth Furcilias in fact grow to between 1 1 and 16 mm. long (Fraser, 1936, Appendix I). And so we are dealing here not with such small and feeble organisms as copepods, ostracods and water fleas, but with distinctly large and active animals. (4) It is true that no First Calyptopes (with Metanauplii co-existent in the plankton) are recorded in daylight from 0600 to 1800 hr. in the top 100 m., but it is equally true (Fig. 6) that none or very few are recorded at this level at Stations 1965, 637, 1671 and 2600 at 1800-2200 hr., and only negligible numbers at Stations 855, 1144, 1545, 2346 and 2603 in the top 50 m. during the same period. These stations, moreover, almost without exception were made in darkness, it being in the latter part of the 1800-2200 hr. period that most, though not by any means all, of our night work was done. The fact that we do not record the First Calyptopes at these high levels in these particular instances is simply I think because, having only recently sprung from their deep Metanaupliar parents, they are still reaching for the surface. In any case we do in fact find (Figs. 9 and 10) large numbers of Calyptopes 1-3 crowding the surface zone, especially the top 50 m., at all hours of the day and night, when the disturbing influence of the deep moulting Metanauplii has been removed. (5) This again is true, as it is indeed for the Second and Third Calyptopes, but the numbers involved (Figs. 9 and 10), relative to those above 100 m., are in general small and suggest as I have already noted (p. iii) that such Calyptopes as do go down by day represent only a minor part of the total surface population. (6) Again perhaps I should ask how else are we to express the actual results? We can seldom ' prove ' anything in marine biology and if we do, sooner or later something or someone turns up to disprove it, which is perhaps all to the good. In general, or at best, we can only point to what seems to be taking place ; and it seems to me obvious that many of the deep First Calyptopes in Fig. 8 are in fact larvae that have never been at the surface, having only recently escaped from their Metanaupliar husks. Fig. 8 is based on Fig. 6 and this in turn on the basic data presented in Tables 13 and 14 (p. 90) and these, with their explanatory footnotes, provide it seems incontrovertible evidence that the deep Calyptopes captured along with Metanauplii in the same nets were in fact in the process of emerging from their Metanaupliar parents, or had only recently sprung from them. There is every reason too to suppose that the Calyptopes taken immediately above the zone of Metanaupliar abundance (i.e. immediately above 500m.), as for instance (Fig.6) at Stations 1965, 637, 1138, 1545, 1671, 2600, 2603, 647, WS 197, 638 and 636, had also only recently sprung from their Metanaupliar forbears, had never in fact been near the surface, but were still reaching for it both by night and by day. ^ Our average daytime gatherings for Furcilias 1-3 between 100 m. and the surface are actually larger than our night ones. * This seems to apply with particular emphasis to the relatively large Furcilias 4-6. THE LARVAL STAGES 117 (7) I was compelled to combine the Calyptopis stages in Fig. 9, just as I was for the Furcilia stages, because all earlier attempts to present the vertical movements by single stages broke down for lack of numbers.^ (8) This is true, but I question the significance of fifty-five rather deep specimens, the majority, however, in the Antarctic surface layer, at 1 400-1 800 hr. when one thinks of the fantastic numbers that must go to make up the Calyptopis (even the free First Calyptopis) population, and remembers that these fifty-five do not come from a single station but are in fact our total harvest from forty-six. The concentration recorded from 250 m. to the surface at 0600-1000 hr. is densest it will be seen between 100 and 50 m. and it does not upset my views that such vertical movements as the larvae undertake (i) do not involve the whole of the surface population, (2) are confined for all practical purposes to the Antarctic surface layer, and (3) do not involve them in a mass descent into the counter-flowing deep current during the daylight hours. (9) This is the crux of the whole matter and to avoid repetition I would refer the reader to begin with to my remarks in (6). Clearly what Dr Fraser has in mind here is that at many, although obviously not all, of the stations in Fig. 6 there could in fact be two distinct groups of First Calyptopes, the deeper between 1000 and 500 m. obviously still climbing, the shallower, between say 500 and 100 m., not climbers, but larvae that, having long since perhaps reached the surface, had returned to deeper daytime [and it could also be argued night-time] levels. This I agree is a distinct possibility, but I would point out that if it be true it would often seem to be involving the First Calyptopis in a massive night-time migration into deep, sometimes very deep, water, in other words in a paradoxical movement distinctly at variance with the ' orthodox '^ pattern we have come to recognise as typical of rhythmic diurnal migration. In view of this I think there can be at least some assessment of the effect the climbers are having on the migration pattern of the larvae at higher levels, and I therefore conclude that on the whole it is safer, especially in view of the migrational pattern in Fig. 9 and of my remarks in (6), to regard all First Calyptopes recorded at stations where Metanauplii were also taken, as travellers from deep water, the majority of them (Fig. 6) still reaching for the surface, their ultimate goal, a minority (as for instance at Stations 1944, 1492, 854 and 647) having actually accomplished, or almost accomplished, perhaps only recently,^ their long journey from the depths where they were born. I do not, it is true, know the rate of ascent of the climbing Calyptopis, but I have suggested (p. 120) on physiological grounds that it might be very slow. And so it could well be (see again Fig. 6) that those taken in the 500-250 m. layer (or even still higher) are still on their way to the surface and have not yet, as Dr Fraser puts it, been ' resident ' there. Nor do I know the duration of the First Calyptopis stage, but again I have suggested (p. 314), from what little we know of the northern euphausians, that it might last 7 days or even longer.* The occurrence, rare though it be, of pure, or almost pure, cultures of First Calyptopes at the surface, as for instance (Table 1 5) at Station 2280, without any trace of a climbing population feeding them from below, is also suggestive of a rather pro- longed existence in the First Calyptopis stage. For such larvae would already have climbed about 1000 m. and for all we know may have been in the surface zone for some considerable time before our searching nets came across them. Moreover, Station 2280 lies far out in the West Wind drift, a region in which we have never, except for one exceptional instance (Table 13, Station 647), struck the larvae rising from below. The Calyptopes encountered there it seems (p. 3 1 5) must have surfaced much farther ^ In his presentation of the vertical movements of the Furcilias Dr Fraser, with far smaller numbers than I had, combines Furcilias 2-4 and Furcilias 5 and 6. ^ I use the word guardedly, for in the science of the sea what is orthodox today may well prove to be unorthodox tomorrow. ^ In the sense that our nets were perhaps striking these concentrations soon after their arrival in the surface zone. * There is evidence (p. 314) from the earliest major appearances of the First and Third Calytopes at the surface in Weddell West that it might well last this long. 1,8 DISCOVERY REPORTS to the south and, therefore, could have been carried, as First Calyptopes, for many days in the surface stream. I would point out too that I have nowhere suggested that the 'climbers', once they have reached the surface, behave in any way differently from the 'resident' population. What I have in fact been suggesting all along is that both climbers, their journey accomplished, and residents, behave in the same way, undergoing the somewhat restricted vertical movements to which I have already called attention in (8). (lo) I agree of course that in view of the mortality factor we should expect to take fewer late Furcilias than Calyptopes. Yet rafting I think cannot be ignored. It is so strongly suggested by the enormous gatherings of Furcilias 4-6 we make with the horizontal (0-5 m.) stramin nets fished in both daylight and darkness at the same stations as the vertical nets are used. Such gatherings at times run to tens and even hundreds of thousands. At Station 2300, for example, there were only 261 late Furcilias in the 50-0 m. vertical net. In the horizontal net there were over 600,000. Many other instances could be given where the horizontal net took thousands or tens of thousands of speci- mens whereas the vertical net took only two or three or none at all, and the general impression we get is that the older Furcilias, especially Furcilia 6, are massed at the surface, both night and day, in an extremely narrow stratum, no more perhaps than a metre or two thick. This in fact is what might be expected from what we know (p. 149) of the shallow draught of the older swarms we so often see on the surface. Finally, if the Calyptopes and Furcilias did in fact have a daily rhythm, involving mass descent of the total surface population into the warm layer, then, assuming both surface stream and deep current to be travelling at approximately the same speed, with compensating northerly and southerly com- ponents, the larvae would not, as they manifestly are, get carried readily to lower latitudes in the surface drift but would tend to remain where they were ; and so the populating and annual recruitment of distant outlying districts of the krills' geographical range, such as for instance the South Georgia whaling grounds, where spawning (p. 190) is not a successful event, would become impossible to explain. Some biological implications of the developmental ascent and its influence ON THE distribution AND CONSERVATION OF THE EUPHAUSIAN POPULATION It follows from the restricted vertical movement of the larvae just described that once the larva, as the First Calyptopis, reaches the Antarctic surface layer it goes through the whole of its subsequent development there without ever, at any rate on any massive scale, becoming involved again in the warm, mainly south-flowing, deep water. In other words it is only during that period of its existence when, having left the cold bottom water, it changes from the Second Nauplius to the First Calyptopis form, that the larval population as a whole can be said to be moving towards the south. And this being so it is obviously a matter of major importance to determine, if it can be determined, how long the larvae actually spend in their warm deep environment and to what extent their sojourn there may affect or control the distribution of the whale food in the Antarctic seas. The role of the warm current as an agent of distributional control was originally postulated by Fraser (1936, p. 167) who in the concluding paragraph of the summary of his results writes, ' ... the continual abundance of E. superba in Antarctic waters and the replenishment of the stock of adolescents at the ice-edge is brought about by the rotary movement resulting from the assemblage of the earlier developmental stages chiefly in the southward flowing warm deep water and that of the later stages in the northward flowing Antarctic surface water '} 1 Mackintosh (1950), in a paper referring to some of my early findings, speaks in similar terms, noting that the adults live very close to the surface while the ascending larvae are carried southwards in the warm intermediate layer, to the benefit of the stocks in higher latitudes. THE LARVAL STAGES 119 Although we do not know the rate of development of the larval stages of the southern euphausians it is known from northern waters (Lebour, 1924) that laboratory reared specimens of Nyctiphanes couchii (Bell) take roughly 8 or 9 days to grow from the Nauplius to the First Calyptopis stage, while the plankton investigations of Ruud (i 927) and Einarsson (i 945) seem to indicate that the corresponding development in Thysanoessa inermis (Kroyer) and Thysanopoda acutifrons (Holt and Tattersall) takes about the same time. In laboratory reared specimens of Meganyctiphaties norvegica the same developmental phase, Nauplius to First Calyptopis, has also been found to take about 9 days (Mauchline, 1959). While it would be unwise to assume forthrightly a similar rate of development in Eiiphausia superba there seems to be some justification for supposing that it might not be of a widely dissimilar order. Fraser (1936, p. 113), referring to this species, says, 'Conclusions based on the presence of eggs in the plankton involve the assumption that the time between the laying of the eggs and the development into Naupliar and Calyptopis forms is very brief. The occurrence of all stages of development up to the clearly distinguishable First Nauplius in eggs in one catch, and the fewness of First and Second Nauplii in any of the catches, suggest that this is possibly the case '. While the existence of the Second Nauplius, however, is manifestly less ephemeral (p. 98) than Fraser imagined, the life-span of the First Nauplius on the other hand might well it seems be very brief, for in other euphausians where the development is known (Lebour, 1926) this stage is said to slough its skin immediately on emerging from the egg and become a Second Nauplius. If the development from First Nauplius to First Calyptopis does in fact take 8 days in E. superba and the eggs as we may well suppose are hatching somewhere between 2000 and 3000 m., say at 2500 (see p. 100 and Tables 13 and 14), then since the Metanauplius (p. 97) does not moult to become the First Calyptopis until it reaches the 1000-750 m. layer, it would follow that the ascending larva would be taking 8 days to travel a distance of at least 1500 m., and presumably another 5 to reach the surface. Such a rate of ascent, approximately 2-2 mm. per second, would mean that in its upward passage the larva, assuming it to have left the cold bottom water at say 2000 m. and to have entered the Antarctic suface layer at say 200 m., would, always provided the rate of climb were sustained, spend from 9 to ID days in the warm deep current moving towards the south (but see p. 123). Even for such relatively minute and feeble organisms as the Nauplii, Metanauplii and First Calyptopes of this species the speed of ascent, approximately 26 ft. in an hour, even if it could only be maintained for short bursts, is at least within the bounds of possibility. It seems to come within reasonable range for instance of the known swimming powers of the Calanoid, C alarms finmarchicus, which as Hardy and Bainbridge (195 16) have demonstrated experimentally, is capable of climbing at approximately double this speed, a higher rate no doubt associated with its greater size and more powerful swimming appendages. At any rate the speed could not be enormously less nor could it be enormously greater. For if it were exceedingly slow and the rate of development were correspondingly slow it is con- ceivable that all the larvae might perish from starvation long before they could attain the Calyptopis form when (Taube, 1915; Gauld, 1959) first they begin to feed. On the other hand if it were exceedingly fast, if for instance they took only 12 hr. or a day to reach the Antarctic surface layer, unless we assume a rapidity of development in larval euphausians unknown so far as I am aware in nature, we should expect to find the oceanic Nauplii and Metanauplii much nearer the surface than they have ever been found to be. It is possible, however, that the ascending larva, without unduly over-taxing the non-feeding period of its Naupliar and Metanaupliar existence, might travel con- siderably more slowly and spend a correspondingly longer time in the warm deep layer than we have supposed it to do. For as we have seen it does not have to traverse the whole extent of this layer before reaching the feeding stage but only about half, the First Calyptopis emerging from the non- feeding Metanauplius between 1000 and 750 m. where presumably it could obtain some sustenance ,2o DISCOVERY REPORTS from the rain of dead or dying phytoplankton sinking from above. Moreover, as Bargmann (1945) points out, the external food supply is not of immediate concern to the larva when first it leaves the egg, since the euphausian egg contains a large quantity of yolk, sufficient it seems to feed the young larva for some little time after hatching. Groom and Loeb (1890) give the average speed of the vertically migrating nauplii of the acorn barnacle, Balanus perforatiis, as about i mm. per sec, while Parker (1902) found that the copepod Labidocera aestiva could climb 1-83 m. in 18 min. and descend the same distance in six. Assuming these rates could be sustained upwards over long periods, in terms of time spent climbing up through the warm deep current in our southern waters, they would work out at a minimum of 21 days for the Balanus nauplii and at a minimum of 12 days for the Labidocera. Hardy (1953, 1956) finds that in I hr. the nauplii of Balanus sp. can climb nearly 50 ft., approximately quadruple the rate of Groom and Loeb's nauplii, but it seems doubtful if such a high speed could be sustained over the immense distance involved in the developmental ascent. So far I have been assuming the newly hatched krill rise to the surface through their own efforts. It may not be so, however, since for all we can tell they may be so constructed that they are able to ' balloon ' upwards although more likely perhaps they are impelled by some tactic sense about which we know nothing. However it comes about the long climb manifestly must demand of the rising larvae a large, indeed astonishing, measure of adaptability to environmental change, change in fact so gross, in so far for instance as pressure is concerned, ^ that one can scarcely imagine them surviving it except there be an extremely slow and gradual upward trend. The experiments of Hardy and Bainbridge (1951a) and of Knight-Jones and Qasim (1955) show that, starting from one atmosphere, certain plankton animals are responsive to small changes of from one half to two atmospheres in pressure (= from 5 to 20 m. in depth), in general becoming more active and swimming upwards in response to increases, and becoming less active, or totally inactive, and allowing themselves to sink, when the pressure is lowered. This behaviour, or high ' barokinesis ' as Knight-Jones and Qasim have called it, is usually they observe combined with negative geotaxis. It might be supposed from these experiments that the animals, if originally collected on the surface, were finding the slight pressure increases uncongenial and that they were rising in their experimental tubes'- towards the levels of lower pressure they would naturally find in the sea. The developmental ascent, therefore, could be described as an extreme instance of barokinesis, or an extreme instance of negative geotaxis, or a combination of both. We can only speculate, however, the very living existence of the larvae at the enormous pressures at which they hatch presenting a physiological phenomenon that perhaps for long will remain unexplained. Much obviously has yet to be done, above all now in the laboratory, before we can even begin to understand the * forces ' compeUing the larvae to rise or the mechanism controlling the physiological adjustments necessary to bring their long climb to its successful end. As Carthy (1957) has remarked of the smaller scale vertical movements involved in diurnal migration, ' if pressure is to be the controlling stimulus, either there is a daily cycle of diflferent preferences innate in the animal, or the effect of another changing stimulus is to alter the animal's pressure preferendum'. In the young rising krill the pressure preferendum would manifestly be subject to continuous, uni-directional, and in the long run, enormous, change. It will be interesting to discover what makes the newly hatched krill behave in this way, why they should go up instead of down or why indeed they should not remain just where they are. Referring ^ As Bruun (19576) has said, neither absence of sunhght nor low temperature is pecuHar to the deep-sea environment. Both are the common heritage of many other pelagic animals living at intermediate depths or at the surface near the poles. 'The unique dominating factor is pressure'. '^ The Hardy and Bainbridge tube was 20 in. long. THE LARVAL STAGES 121 to Other plankton animals and their responses to pressure Hardy (1956) asks, 'Have they some small pressure gauge — perhaps a tiny gas-filled vesicle expanding and contracting against a nerve ending? Or is it just the varying effect of pressure upon some vital chemical reaction within the body ? ' As Bruun (1957^) has said, ' It is definitely known that the hydrostatic pressure affects the physiological process, but this is known so far only from experiments with surface animals or deep-sea bacteria. It seems, however, safe to conclude that more knowledge about this pressure factor in relation to the bathyal and abyssal species may clear up many controversies. Some species have a greater tolerance towards varia- tions in hydrostatic pressure than others.' The tolerance of the climbing larvae, it seems, must indeed be enormous, and perhaps it is specific to each individual stage during this early developmental phase. A parallel phenomenon operating in the reverse direction is provided by the migration, or more correctly perhaps sinking, of full grown Sagitta gazelle (David, 1955) to enormous depths in order to shed their eggs, and both movements perhaps have their roots far back in phylogenetic history. As Young (1954) has recently said, 'We can say that the organism is instructed by the information it receives from heredity, from its environment and from whatever memory store it has built up during its Ufetime '. Huxley (1912) also remarks on the many structures and habits of animals 'that can only be made fully intelligible through their history '. In Eraser's original experiments (p. 55) gravid females taken on the surface, or from near the surface, were kept on deck in small aquaria in the Antarctic summer air. There, at atmospheric pressure, they shed their eggs. 'Attempts to induce segmentation ', however, were unsuccessful. In the same condi- tions Fry also got gravid females from the surface to lay their eggs but he too found they would not divide. Other laboratory workers have had more success. Lebour, for instance (p. 119), working with Nyctiphanes couchii, was able to rear this species from the egg right through to the Third Calyptopis stage. Again this was done at atmospheric pressure, in this instance, however, with a neritic species in which eggs, larvae and adults naturally occur near the surface or at their deepest never very far away from it. These findings distinctly suggest that in the Fraser and Fry experiments it was the vital element of pressure that was missing and that in Eiiphmisia superba the liberated eggs might be expected to divide only if gradually subjected to the increasing pressures, and ultimately to the enormous pressure maxima, they would naturally encounter if, as seems possible (p. 184), they sink in the sea. The resultant larvae might then be induced to develop by reversing the pressure trend. As ZoBell (1954) and ZoBell and Morita (1956) have recently shown with bacteria brought up from the deep-sea floor, reproduction is more rapid at high than at low pressures, luminescent forms only beginning to show their light when re-subjected to the immense pressures of their natural environment. The vital role of pressure, we may well indeed say enormous pressure, in the early developmental history of this species seems to be strongly emphasised by our widespread failure to find any evidence of successful hatching other than at great depths (p. 201) in oceanic water. Hatching has not so far been found to be a successful event in shallow shelf seas. Finally, it may be noted, it must happen occasionally that local upwelling of the bottom water, as for instance (p. 102) at Stations 823 and 2594, is assisting both eggs and larvae to rise. In so far as the laboratory will be concerned, perhaps alone concerned, in explaining the pheno- menon of the climbing krill, and other equally little understood phenomena of the open sea, Rae (1958) has recently called attention to our pressing need for establishing 'a reasonable variety of healthy plankton species in captivity ', adding that if we succeed in so doing 'we open up entirely new facilities for a wide range of studies that have not yet been applied to oceanic animals '. It seems likely then, that in their ascent towards the surface the larvae are carried in a current with a southerly component^ for a period which at its shortest might be put at 10 days and at its longest at ^ Except perhaps (p. 123) in part of the Atlantic sector. 122 DISCOVERY REPORTS possibly 20 or even more, in other words long enough for them to undergo a latitudinal displacement the extent of which would depend on the speed with which the warm deep water was moving, wherever that is it should happen to be travelling towards the south. Although little is known, experiments with drift frames,^ undertaken by Dr G. E. R. Deacon off South Georgia in October and December 1936, indicate that there it may be moving in a south-easterly direction at the rate of about 3 miles a day. The drift frames, or drogues, employed, each consisting of two 6-ft. square brass frames covered with cotton drill and arranged in two planes fixed to intersect each other at right angles, were suspended from light floating pellets 20 in. in diameter on ordinary Kelvin sounding wire, the distance and direction in which they travelled being recorded in relation to an anchored buoy. The results obtained at four diff'erent levels in the warm deep layer are shown in Table 23. Table 23. Rate and direction of travel of drogues in the warm deep current near South Georgia Depth Miles Bearing from (m.) per day anchored buoy 250 3-8 143F 500 2-S 145F 750 2-6 137° 1000 3"o 144° These figures no doubt must be accepted with some reserve since they may have only a local significance. Nevertheless, we might reasonably I think regard them as providing at least some indication of the possible order of the speeds that might be expected in the other parts of the Antarctic such as, for example, the Weddell and East Wind zones, where the deep intermediate current is known to be carrying the climbing stages. It is possible indeed, in so far at least as the South Georgia area is concerned, that the speeds recorded there were too low, for in the final reckoning no allowance was made for the frictional drag on the suspending wires, nor was any made for the considerable counter drag that must have been exerted by the Antarctic surface layer which at one point was found to be travelling in exactly the opposite direction to that of the deep current at a rate of 7-3 miles a day. Over a century ago Maury (1855) called attention to the sometimes quite remarkable speeds that subsurface currents may attain. In the following passage he refers to the work of Lieutenants J. C. Walsh and S. P. Lee of the United States Navy. They made some interesting experiments upon the subject. A block of wood was loaded to sinking, and, by means of a fishing-line or a bit of twine, let down to the depth of one hundred or five hundred fathoms (six hundred or three thousand feet). A small float, just sufficient to keep the block from sinking farther, was then tied to the line, and the whole let go from the boat. To use their own expressions, ' It was wonderful, indeed, to see this barrega move off, against wind, and sea, and surface current, at the rate of over one knot an hour, as was generally the case, and on one occasion as much as if knots [i.e. 42 nautical miles per day]. The men in the boat could not repress exclamations of surprise, for it really appeared as if some monster of the deep had hold of the weight below, and was walking off with it'. Both officers and men were amazed at the sight. Although with the above exceptions there are no direct measurements to prove it, it is generally thought that the southward movement of the warm deep water is exceedingly slow, being probably of the order of little more than a mile a day. Such a slow rate of transport, if true, clearly could not bring about any major southerly displacement of the larvae unless their ascent from deep water is a very much more protracted phenomenon than even at its longest I have supposed it to be. It would mean at the most that they would travel no more than 20 or 30 miles to the southward of their deep point * Deacon gives a full account of this work in the 22nd Scientific Report to the Discovery Committee on the work of 'Discovery II'. THE LARVAL STAGES 123 of hatching before they appeared in the Antarctic surface layer. If, however, the developmental ascent does take very much longer than has been suggested,^ let us say it took not 8 + 5 days but 24 + 1 5 days, and if the higher deep current speed indicated by the South Georgia experiments be accepted, then a major southerly displacement of the larvae, involving distances of 100 miles or more, can well be imagined. A displacement on this scale would mean that the deep larvae in the East Wind zone, never, it will be shown (p. 200, Fig. 28), very far away from the land, would constantly be carried deeper (in the latitudinal sense) into this coastwise current, while those hatched deep down in the far southern reaches of the Weddell drift, such as, for example, the larvae at Station 2594 (p. 90, Table 13), would, when they appeared at the surface, no longer it seems find themselves in the east-flowing current under which they were born but in the west-flowing East Wind stream some considerable distance to the south. In other words in high latitudes all round Antarctica, wherever in the East Wind zone there should be an adult breeding stock (p. 194, Figs. 24 and 25) its progeny, whether hatched near, or some distance from the land, would tend automatically to become compressed within a relatively narrow coastal belt in which, first as larvae, and later as adults, they would move westward in the surface stream until eventually some of them would find their way into the Weddell Sea where they would become augmented by fresh influxes of larvae deep-hatched in the far southern reaches of the Weddell drift. On the Atlantic side of Antarctica between 30° W and 30° E the warm deep water appears to be traveUing mainly towards the east (Deacon, 1937, p. 86, Fig. 22, and pp. 90 and 92). There is, however, between 10° and 30° E a strong indication of south-westerly movement reaching far south into the East Wind zone where it penetrates well west of 0°. Model (1958) shows the same movement even more con- spicuously. As Deacon (1937, p. 81) points out, since there are general movements towards the north in both surface and bottom currents all round the continent, these currents ' can only exist if there is a compensating movement towards the south in the intermediate deep water '.^ Between 30° W and 30° E the deep water is generally warmer below the westerly current near the continent than it is below the cold more northerly surface current flowing eastwards from the Weddell Sea. This gives a clear indication that the main southward movement in the warm layer occurs in a sweep round the end of the Weddell drift, but the fact that every deep observation in the Weddell Sea shows a warm layer sandwiched between two very cold ones indicates everywhere a very active supply of warm deep water. There can in fact be little doubt that here as in the Ross Sea (p. 124) its influence as a major instrument of southerly dispersal does not cease until it comes up against the continental slope, as for instance in 0° (Discovery Reports, Station List, 1937-9), where we do in fact find it penetrating practically up to the coast. But perhaps the overriding consideration is this. Unless there be somewhere, between 30° W and 30° E, active penetration of the East Wind zone by warm deep water from the north the continued existence of the Atlantic krill population it will be seen (p. 432) becomes hydrologically impossible. The developmental ascent in relation to the local distribution OF krill in the Ross Sea At the head of the Ross Sea there is a vast expanse of shallow water from which the warm deep current (Marshall, 1930; Deacon, 1939; Discovery Reports, Station Lists, 1927-9 and 1935-7) ^^ virtually excluded.^ The distribution of adult and adolescent whale food in and on the outskirts of 1 There is some evidence (see p. 312 in the main distributional part) that it might in fact occupy the better part of a month, if not longer. ^ See also Deacon (1957a). ^ Wright and Priestley (1922), referring to the annual calving of bergs from the Ross Barrier face, state that the prime cause is melting by the warm water lying beneath the barrier, which is most effective close to the edge. 'This action', they continue, 'suggests the presence of a warm current from the north in the lower layers of the Ross Sea'. There is nothing, however, in our hydrological data to suggest (see below) that such a current exists. 144 DISCOVERY REPORTS this region, based on the plankton gatherings of the ' Discovery ' (1901-4), the 'Terra Nova ' (1910-13), the floating factory 'C. A. Larsen' carrying a Discovery Investigations observer (1928-9) and the 'Discovery II' (1936), is shown in Fig. 13, the combined observations of these vessels, so far as I know, providing the only data available from net-hauls relating to the euphausian population of this region of which, whether published or not, there is any record. Not one of these expeditions it will be seen records the krill in any measure of abundance southward of a line in approximately 73° 30' S, which, our hydrological observations between 175° E and 175° W show, appears to be the limit of major southerly penetration of the shelf by warm deep water in this particular locality, and none record it at all southward of a line in 74° 45' S between the same meridians where the last most southerly traces of a warm deep penetration can be detected. All to the southward, between 175° W and the Victoria Land coast and up to the Great Ross Barrier, the water is pure shelf water, cold to the bottom, the krill that have been recorded in it being represented exclusively by that other large southern euphausian, E. crystallorophias, which in January 1936 was seen from 'Discovery II' swarming in great profusion near the barrier face, particularly in the neighbourhood of the Bay of Whales. No doubt it is this species that is the food of the minke whales that frequent the barrier region and doubtless too it was this that Lindsey (1937) found in the stomachs of the Weddell pups he examined near Little America and not as he records (p. 134) E. superba. In a preliminary note on the biological collections of the 'Discovery' (1901-4) Hodgson (1905) refers to the 'profusion' in which the large E. australis (= E. superba) was encountered between latitudes 66° S and 72° S, but he does not mention it farther south. The combined observations of four well-equipped expeditions operating at intervals over a period of 35 years point fairly conclusively it would seem to the absence of E. superba from the greater part of the shelf water of the Ross Sea region, and it seems likely that the factors principally contri- buting to this interesting distributional phenomenon are (i) the failure of the warm deep current, carrying the ascending larvae, to penetrate more than a short distance on to the shelf,^ and (2) the absence of any influx of krill, whether larval, adolescent or adult, west-borne upon the surface stream that enters this region from the east and flows clockwise round it. Between 175° E and 175° W the warm deep current overflows on to the shelf as a narrow stratum averaging less than 150 m. thick, penetrating as such as far as 73° 30' S, where, as a major carrier of the larvae, its influence must virtually cease. It continues, however, to the southward as a barely detectable trace to about 74° 45' S where, as Fig. 13 shows, the last, virtually negligible^ and most southerly occurrences of the krill have been recorded. It seems likely too that the majority of such south-borne larvae as do get carried on to the shelf and subsequently rise to the surface there must quickly be carried off it again by the cyclonic movement of the surface stream. The non-recruitment of the high southern reaches of the shelf by surface-borne influxes from the east is probably associated with the impenetrable and more or less permanent nature of the pack-ice that for long has been known to exist in that direction. Although so far only its northern outskirts have been explored it seems that the high latitude Pacific ice-cover is very compact and thick, probably continuous over vast stretches of ocean and merging imperceptibly with the continental ice-sheet, presenting in all it would appear such an effective barrier to the direct passage of sunlight so essential to the existence of a diatom flora that no euphausian population it seems could exist. Mawson (1928) has already called attention to the probable conditions below the shelf-ice bordering 1 Riedel (1958) has just called attention to the possibility that the absence of certain cosmopolitan Radiolaria from Antarctic shelf stations south of AustraUa may also be due to the failure of the warm deep water to penetrate on to the shelf. * The total yield from thirteen net hauls in these, the highest known latitudes of the krill's occurrence, was only fifteen individuals. THE LARVAL STAGES 125 the continental land, noting that ' a general absence of living plant life is to be expected ' and that this in turn 'must greatly limit the possibilities for animal life'. It seems probable, therefore, that the adolescent and adult krill population (p. 394, Fig. 135) known to frequent the lower ice-free lati- tudes of the Pacific sector is carried in the East Wind stream north-westwards diagonally across the mouth of the Ross embayment, by-passing the higher latitudes of the shelf. Although more hydro- logical evidence is needed to substantiate such a movement it is interesting to record that Taljaard EAST IBO WEST I70" c:- ^ A DEPTH KEY Ni»J A \ 's^ SHELF WATER "XT \ ♦ LESS THAN 500M SHELF WATER MAINLY 500-750M E.SUPERBA O 100 100-1,000 1,000-10.000 Q, 2 < O Y- - O Fig. 13. Distribution of whale food in the Ross Sea area based on the gatherings of our i-m. stramin nets with additional data from the 'Discovery' and 'Terra Nova'. (1957) States that in the neighbourhood of Marie Byrd Land the East Wind current shows evidence of dividing into two streams, the southern and swifter branch following the edge of the Ross Barrier and Victoria Land coast, the northern branch flowing west-north-west towards Cape Adare and the Balleny Islands where it is joined again by the southern stream. Debenham (1923, end map i) and Hansen (1934, Pis. iv-vii) both show west-north-west diagonal movements across the mouth of the Ross Sea in addition to the current that follows the barrier face. Although no adolescent or adult krill have so far been recorded in the purely shelf water of the Ross Sea area, water, that is, unadulterated by any trace of the warm deep layer, a very few early larvae, undoubtedly belonging to this species, have in fact (p. 301, Figs. 70 and 74) been encountered there, six of the ten stations made by 'Discovery II ' in the coldest water of the shelf in January 1936 producing an aggregate of four Second Nauplii, five Metanauplii and six First Calyptopes. The occurrence of these larvae, very small though their numbers be, presents an interesting problem in local distributional dynamics, for in view of the absence of a breeding stock it raises the obvious 126 DISCOVERY REPORTS question as to how they got there. Since clearly they could not have been produced in situ and it is unlikely as mentioned above that they could have entered as surface-borne influxes from the east (the Nauplii and Metanauplii in any case being essentially deep living forms), it appears probable that they represent isolated stragglers from a larger population that perhaps had its origin on the continental slope! to the north or east, stragglers that got carried into the shallow water, originally perhaps as eggs, by the warm deep current as it came up against the shelf on which, after a short and weak penetration, it presently died away. The occasional transference of eggs or larvae from the oceanic water to the more southerly reaches of the shelf may well be more effective in the south-eastern part of the Ross Sea than elsewhere, for as the bottom contours in Fig. 13 show it is there, between 170° and 180° W, and in about 75° S, that the deep oceanic water, and with it presumably the warm deep layer (although we have no hydrological observations to show it), extends much farther south than it does for instance in the north-west in the neighbourhood of Cape Adare. It is interesting to record that when C. A. Larsen originally began hunting the whales in these high latitudes in December 1923 they were wild and difficult to approach, Villiers (1925) remarking that they ' did not come into the Ross Sea in any numbers until after the New Year, and when they did come they found their food very scarce, and so were restless and hard to catch '. It seems, therefore, that these shy whales were either hungry or else resentful of the intrusion of the catchers and new to the experience of the hunt. It is more likely in fact they were hungry, for whales are well known to be more approachable when feeding than when not, Gunther (1949), to quote a particularly reliable observer, noting repeatedly while whale-marking at South Georgia, that when engrossed in a meal they would show neither concern for the approaching vessel nor for the din of the marking guns. He describes the ' supreme indifference ' with which they would accept the marking and the presence of the marking vessel, adding that they 'blew leisurely, sometimes swimming towards us, beneath us and by our side ', apparently preoccupied he says with the question of their meal. Bennett (1931), too, remarks on their lack of shyness when feeding adding, moreover, that it is ' usual for whales to be stupidly "tame" when first a district is exploited'. In similar terms Andrews (1916) refers to the feeding of Balaenoptera borealis, mentioning one occasion when he watched a sei pursuing a school of sardines on the surface, its high dorsal fin exposed and so absorbed in its meal that it ' allowed the ship to approach at once and was killed without difficulty '. Larsen, it may be observed, began hunting well out in the middle part of the Ross Sea shelf, about 1 50 miles north of the Barrier face. He began, therefore, in a locality where E. siiperba has not been recorded and where in fact the majority of Villiers's restless (or hungry) whales seem to have been seen. Later, when 'Sir James Clark Ross', Larsen's factory, moved to Discovery Inlet, well-fed whales, principally blue but with some fin, began to be taken close to the sheltered mothership and there can be little doubt that both species must have been feeding on the swarms of E. crystaUorophias that occur in such profusion along the ice cliffs there. From a photograph by Kohl (1926), of Crustacea captured in Discovery Inlet, showing ' rechts die Gattung Euphausia die die Hauptnahrung der Blauwale darstellt ', it seems clear enough in fact that they were. THE OLDER STAGES Economic and ecological importance Euphausia superba, the staple food of the large southern baleen whales, stands perhaps alone among plankton animals in the economy and ecology of the sea. Writing of the South Georgia whaling grounds Hardy and Gunther ( 1 93 5 , p • 208) say, ' Numerically this species is by far the most important member of 1 At the single slope station we made in this locality (p. 93, Table 14, Station 1662) five Second NaupHi and ten Meta- nauplii were taken in the 1000-750 m. net, the sounding being 1180 m. THE OLDER STAGES 127 the order in this area; indeed one is tempted to beHeve that ecologically it is the most important zooplankton organism of the Antarctic'. Tattersall (1913) refers to it as the euphausian /)ar excellence of the Antarctic seas. Because of its vast abundance there exists an industry into which in the last fifty odd years huge capital sums have been sunk and into which millions more are being poured today, an industry which for the most part has yielded great profit to those who engage in it, which has provided the world with an essential raw material of prime importance both in peace and war and above all one which, through the wisdom and foresight of the existing international control, might yet, as in the past, continue to benefit mankind for many years to come. It is not, however, the great rorquals of commerce only, the blue and fin whales, that depend on the krill for their existence. The smaller humpback and sei,^ the formerly abundant but now scarce and rigidly protected southern right,^ the minke or little piked whale,^ the millions of Crabeater seals that haunt the pack-ice,* and the hordes of Adelie, Ringed and Gentoo^ penguins in their coastal rookeries all feed, the majority of them for all practical purposes exclusively,^ upon this little pelagic prawn.^ Of the large, or moderately large, Balaenopterids resident and hunted in southern waters, Bryde's whale (Balaenoptera edeni Anderson) seems to be the sole representative that does not prey upon the Antarctic krill. For long after Olsen (1913) originally described it from Saldanha Bay it was thought to be rare and to occur only locally. Recent reports, however (Best, i960), have revealed it to be widely spread and far more abundant than was formerly believed. It feeds principally on small sub- tropical Clupeoids, pilchard or anchovy, and so far, as Best remarks, has not been recorded from the Antarctic. Probably because of its feeding habits he adds it tends to remain in the same locality throughout the year and has no need to make a long migration. Vivid accounts of the magnitude of the toll exacted by a single penguin rookery are given by Wilson (19076) and Levick (19150) in their remarks on the Adelies of Cape Adare, where, it is estimated, some three quarters of a million birds congregate annually to breed. The quantity of euphausians, says Wilson, brought ashore there daily must be immense, adding that ' parents hurry in ashore from one end of the twenty-four hours to the other without cessation, their stomachs loaded with a mess of shrimps. Chickens by the hundred stand in little groups of twelve or twenty, now in a state of bulging 1 At least during its annual wanderings (Matthews, 19386) into Antarctic waters. Since the sei does not range into such high and cold latitudes as do the blue, fin and humpback whales (Mackintosh, 19426; Clarke, 1957) it is distinctly probable that it is on the northerly, relatively warm South Georgia whaling grounds that it exacts its most significant toll of the krill. ^ Brown (1958) refers to the increasing abundance of this species during the last few years at Tristan da Cunha, noting that one observer (Elliott, 1953) states that it is 'becoming increasingly numerous and almost a pest'. It may be on the increase too in the Patagonian region, Ruud and Oynes (1959) reporting a school of twenty there in 38° 4' S, 56° 59' W in November 1956. It seems possible, therefore, that the southern right is now in fact beginning to recover from the ravages of last century. ^ A minke harpooned from the whaleship 'Antarctic' near the Balleny Islands in January 1895 is stated by Bull (1896) to have had the stomach filled with small red ' shrimps'. Mr Gordon Williamson, biologist and whaling inspector in the floating factory 'Balaena' in 1957-8, recently gave me two samples of E. superba he collected from two small piked whales which appear to be a colour variety of the well known minke of Lacepede, but may prove (Williamson, 1959) to be a separate species. Mr Williamson has just called attention to the exceedingly large numbers of minkes that frequent the Antarctic seas 4 Crabeater seals, sometimes in enormous numbers, have been reported by every major expedition that has passed through the pack in the southern summer or, as the Belgian expedition did in 1898, wintered in it. In view of the vast extent of the circumpolar pack this species may prove (Bertram, 1940; Laws, 1958; Bertram, 1958) to be the world's most abundant seal. SchelTer (1958) puts its numbers at between two and five million. ^ According to Murphy (1936) the stomachs of Macaroni penguins have variously been reported to contain Euphausiid crustaceans, cephalopod beaks, or both. ^ The southern right, however (Matthews, 1932), like the sei, only in the higher latitudes of its range. ' The largest specimen I have seen, a female, measured 65 mm., or a little over zh in. J28 DISCOVERY REPORTS repletion, now in a ravenous hurry chasing some unfortunate aduh but lately arrived with spoil from the open sea'. Or to quote Levick: Hitherto they had been merely remarkable for their spotless plumage, in contrast to their former dirty state, but now their shape too was greatly altered, for their meals, in place of merely satisfying their own individual wants, had now to provide for the offspring as well, and they were in consequence so distended with their heavy load of Euphaiisia that they were obhged to lean back to counterbalance the weight of their bellies that bulged before them as they walked. Frequently they would find to their cost that they had attempted too much, and overcome by the labour of their journey over the rough ground, they would be sick, depositing the whole load on the ground, and having perforce to return to the sea for more. Manifestly the quantities of krill consumed, and the local effect of this on the euphausian population near a large rookery, must be enormous, the foraging parents, with two ravenous chicks to rear, each having to eat for three. In the but moderately full stomach of an adult Gentoo Bagshawe (1938) counted the remains of 960 E. superba^ from which it seems distinctly likely that the great rookery at Cape Adare alone, with its nestlings, assuming half the aduh population are fishing and the other half nursing, is taking the staggering figure of over 370 million euphausians from the plankton per day. The insatiable appetites of the Adelie chicks, and the astonishing rate at which they grow, has been well described by Levick (191 5 6) who gives figures showing that in the first 12 days after hatching they increase their birth-weight 14-fold, from 3 to 42 ounces. All this indeed is impressive enough. It does not, however, by any means cover the host of animals that harry and devour these unfortunate shrimps. Among the vast flocks of petrels and other oceanic birds that frequent the whaUng grounds there are a number of species which, although not exclusively krill-eaters, habitually include E. superba, sometimes in very substantial quantities, in their otherwise catholic diet. A Ust of such birds, compiled mainly from the comprehensive records of Murphy (1936) with additional records by Pirie (1905 a), Wilson (19076), Wilton (1908), Falla (1937), Roberts (i940«) and Bierman and Voous (1950) is given below. While it probably includes the chief krill-eaters there may well be others, since a number of birds vaguely recorded as feeding on 'Euphausians', 'pelagic shrimps' and so on, without specific reference to E. superba, have not been included. Diomedea exulans Linnaeus Wandering Albatross^ Diomedea melanophris Temminck Black-browed Albatross Phoebetria palpebrata (Forster) Light-mantled Sooty Albatross Priocella glacialoides (Smith) Silver Grey Petrel Thalassoica antarctica (Gmelin) Antarctic Petrel Pagodroma nivea (Forster) Snow Petrel^ Oceanites oceanicus (Kuhl) Wilson's Petrel* Pachyptela desolata Gmelin Antarctic Whale-bird Sterna vittata Gmelin Antarctic Tern Sterna macrura Naumann Arctic Tern Sterna vittata georgiae Reichenow Wreathed Tern Daption capensis Linnaeus Cape Pigeon* Halobaena caerulea Gmelin Blue Petrel 1 A bulging, completely filled, Gentoo or Adelie stomach could well I believe hold considerably more than 960 krill, Hartley (1934)' finding that in the much smaller Kittiwake gull 350 Thysanoessa inermis, which grows to about 30 mm., is merely an average single meal. 2 This species has not been seen taking fresh krill from the plankton. It is known, however (Matthews, 19296), to be partial to krill vomited up by harpooned whales. It seems reasonable, therefore, to include it among the krill-eaters. 3 Wilson states that the food of this abundant species consists almost entirely of E. superba. * Fisher (1954) believes this to be without question the world's most abundant bird. * Discovery Investigations record. THE OLDER STAGES 129 As these thirteen species alone represent a very substantial part of the huge oceanic bird population of Antarctica it is probable that the quantity of krill they devour is second only in magnitude to that consumed by the whales, seals and penguins themselves. Their diet as already said is a catholic one and during the whaling season they are often to be seen in vast numbers gorging on the scraps and garbage arising from the slaughter, at times it would seem to the virtual exclusion of their natural diet. Bennett (i 931) who spent a number of years with the early whalers and saw the rise of the southern industry to its present dimensions, records the following passage : When the day comes which sounds the death-knell of Antarctic whaling, it will also see the doom of millions of birds that at present prey on the waste produced by this industry. For whaHng artificially produces an abundance of food that has enabled all sorts of birds to increase enormously in numbers. When an end comes to it, the natural food supply will only be able to support a tithe of the present bird population At present the birds, thanks to the whaling trade, get plenty of food, and get it easily. But the withdrawal of this source of nourishment can only spell disaster on a gigantic scale. 'No one', he continues, 'can doubt this who has seen the myriads of petrels of all kinds '^ that gather at the whaling stations and round the floating factories to gorge upon the waste. Whether in fact such a holocaust would ensue seems doubtful, for the chief predators of the krill, the whales themselves, now supposedly so reduced in numbers as to make whaling uneconomic, would be taking less from the plankton and leaving more for the birds. Even so there can I think be little doubt that as a result of whaling the numbers of Antarctic oceanic birds have increased, and possibly increased enormously, and so it could well be that the mortality among the krill for which they are responsible is even heavier today than in the years before the advent of the modern whalers upon the Antarctic field. Mawson (19406) refers to the impact of sealing on the bird population of Macquarie Island, stating that round about 1911-14 carrion birds such as skua gulls and giant petrels had 'increased greatly in numbers, owing to the abundant food supply available in the flensed carcases of the seals left on the beaches '. A similar increase, as a result of whaling, seems to have occurred too in the Brown Skua population of the South Shetlands and Graham Land (Hamilton, 19346). Falla (1952) also calls attention to the steady increase in numbers of scavenging or turned-scavenging birds that has accompanied the growth of the summer whale fishery, adding that since the luinter food supply is also of concern to this artificially swollen and ravening population some species in order to survive have evidently been compelled to increase their winter range, the Cape Pigeon and Giant Petrel even as far north as Tory Channel in New Zealand. The extensive circumpolar journeys of young first year Giant Petrels, recently demonstrated by Antarctic banding (Carrick, 1959), could also it is suggested spring from a growing necessity in this species to increase its winter range. Other bird populations are known to have increased enormously both in numbers and range as a result of human activity, the fulmar providing a striking example. Originally a high Arctic bird and a plankton feeder, as a result of the easy food supply provided in former times by whaling and more recently by trawling, it has multiplied and gradually spread southwards far into temperate waters (Crisp, 1959). Quantities of krill have been reported in the stomachs of the Giant Petrel or Stinker, Macronectes giganteus (Gmelin), but in this case it comes as a rule to the Stinker third-hand, being derived not directly from the plankton but from the stomachs of the nestling penguins it is known to attack and devour (Matthews, 1929; Murphy, 1936; Ardley, 1936). It seems possible, however, that even this 1 A remarkable photograph of Bennett's myriads, 'Birds preying on refuse', is given in his book facing page 189. It is the most striking of the many such congregations I have seen. 13 DM jjo DISCOVERY REPORTS large carrion-eater sometimes takes its krill directly from the sea, Wilton (1908) recording that the stomach contents of a speciman shot on the pack hundreds of miles from land 'consisted mostly of Euphausia '. The Brown Skua of South Georgia, Catharacta skua lonnbergi (Mathews), is also, it seems, a cathoUc feeder, much prone to carrion, that comes by its krill third-hand, Stonehouse (1956) recording that during the nesting season it commonly attacks Gentoo chicks, tearing them open, removing the viscera and crop contents and regurgitating them for the benefit of its young. ' On more than one occasion, ' he continues, 'a bird was seen to take off from its roost, circle high over its territory, and fly off out of sight along the coast, returning one or two hours later to deposit a cropful of krill (no doubt removed from the crop of a Gentoo chick) at the feet of its chick The possibility that the krill was collected directly from the sea^ cannot altogether be excluded, but it seems extremely unlikely in view of the general nature and feeding habits of the birds'. Routh (1949) reports the occurrence of vast flocks of the Short-tailed Shearwater, Puffinus tenui- rostrts (Temminck), the famous 'Mutton Bird' of the Bass Strait, which he says would spend much time resting on icebergs near the patches of krill on which they fed ' in a seething mass '. There can be little doubt, however, that he must have mistaken the closely allied Sooty Shearwater, Puffinus griseus (Gmelin), for this species with which, as Murphy (1936) points out, at a distance it can readily be confused.2 palla (1937) records P. griseus as breeding on Macquarie Island, stating that it was seen in large flocks by the Australian expedition as far south as 64° S in 131° E in February 191 3, while Mawson (1942), in the same month and year, refers to flocks of 'dark petrels (Mutton Birds?)' in 65° S, 118° E. Dr T. J. Hart (personal communication) has also seen it in great abundance still farther south near the Balleny Islands, apparently he thought feeding on the krill. Reports of 'Mutton Birds' in this locality by Balleny (1839), McNab (1839), Borchgrevink (1901) and Hanson (1902) no doubt also refer to this species. Balleny twice records 'numerous flocks' and from all accounts it would appear to be extremely plentiful in East Wind waters south of Macquarie Island throughout the southern summer. So far, however, it has not been reported, except in negligible numbers, far south in the Pacific sector (Holgersen, 1957). From an examination of ninety-five stomachs, representing fifteen species, Bierman and Voous (1950) conclude that the 'Opossum-shrimp {Euphausia spec.) is the basic diet of all Antarctic sea- birds', adding that cephalopods come next in importance^ and that fish and pteropods are seldom eaten. Put in this way the Opossum-shrimp could, of course, refer to Antarctic euphausians in general. However, since the authors refer to it unequivocally as E. superba earlier in this paper and make repeated references to the presence of 'orange' or 'bright orange remains of Euphausia' in the stomach contents they examined, there can be little doubt that it is to E. superba, the dominating shrimp of these southern waters, that they allude. The ravages of fish, while possibly little less severe than the toll exacted by oceanic birds, are not so easy to gauge, for few Antarctic ichthyologists seem to have bothered very much about the food of the specimens they handled. Mr N. B. Marshall, however, has told me that among the collections preserved in the Natural History Museum he has found a number of substantial krill-eaters. The majority are well known oceanic or coastal forms, for example, Electrona antarctica (Gunther) and 1 The italics are mine. 2 Kuroda (i960) also calls attention to how easily these two can be confused except at very short range. 3 Perhaps the largest and best documented collection of southern cephalopod remains ever to be made has recently been described by Dell (1959) from stomach material brought back by the B.A.N.Z. Antarctic Research Expedition. He calls attention to the enormously important place cephalopods fill in the food chains of Antarctic vertebrates, especially birds, and notes that although such remains can contribute little to our knowledge of the Antarctic cephalopod fauna, they do show that squid are far commoner in high southern latitudes than conventional collecting gear lets us think. THE OLDER STAGES 131 Lampanyctus braueri (Lonnberg), the most abundant and widely distributed of the Antarctic Mycto- phids; Notothenia rossii Richardson, ^ a coastal form so numerous in the neighbourhood of South Georgia (p. 132) that recently (Olsen, 1954) two unsuccessful attempts have been made to exploit it commercially; Notolepis coatsi DoUo, a widely distributed^ circumpolar oceanic species, and the more recently described Dissostickus mawsoni Norman which grows to a length of four or five feet. As Electrona, Lampanyctus and Notolepis have repeatedly been recorded in small numbers^ from blue and fin whales' stomachs all three it seems are swallowed fortuitously (Clarke, 1950; Marshall, 1954) while feeding on the dense concentrations of euphausians on which the whales are known to browse (see PI. III). A new genus of Chaenichthyidae recently described by Nybelin (1947) to which he gives the name Neopagetopsis ionah also appears to be a krill-eater. The specimen he describes was collected from the stomach of a baleen whale somewhere near the Balleny Islands in February 1939 and in his remarks on its probable bathymetric distribution Nybelin says, ' Its stomach was filled with Euphausiids and it was no doubt catching food in a patch of "krill" when together with these, it was swallowed by the whale'. Mr Marshall tells me there are two other specimens of this genus in the Natural History Museum, both from whales' stomachs, Tattersall (191 8) records it from the stomachs of Trematomus lonnbergi Regan and Prionodraco sp., and Waite (191 6) from the stomachs of Prionodraco evansii, and it seems likely too that Champso- cephalus gunnari Lonnberg from the South Georgia area is a krill-eater, Lonnberg (1906) recording that the stomachs of the specimens collected by Mr Erik Sorling in 1905 'were filled with the remains of shrimp-like crustaceans (perhaps large Euphausiids)'. C. gunnari may well in fact be a major predator in these island waters, Olsen (1955) quoting reports from whalers that it has been seen off the coast at the surface feeding on the krill in shoals, and mentioning at the same time another schooling species, Pseudochaenichthys georgianus Norman, hooked at the surface near Cape Buller, that was evidently feeding on a swarm. C. gunnari has recently been found to be of considerable local abundance. Next to Notothenia rossii in fact it might be the most abundant of the South Georgia fishes, Olsen (1955), trawling in West Cumberland Bay in the autumn of 1951, once having taken 4000 in a single haul. The partiality of Notothenia rossii to E. superba at South Georgia is well known to the local whalers, was noted by Lonnberg (1906) and has been recorded on more than one occasion by members of the former Discovery Committee's staff. The late Dr E. R. Gunther, to quote one of them, describes how he saw, lurking beneath a swarm, shoals of large A^. rossii swimming in packs of from twenty to thirty fish in close formation and harrying and devouring the krill above. A fish 70 cm. long taken on a handline was gorged to capacity with the freshly eaten quarry.* In other notes made while watching the krill which have been seen swarming round the Government Jetty at Grytviken, Gunther records the capturing of small adolescent krill by young Nototheniid fish, some 4-7 in. long, suspected of being A^. rossii, and describes how 'Crocodile Fish', Parachaenichthys georgianus (Fischer), lurking near the bottom, would suddenly leap up and dart at and eat the krill overhead. 1 Nybelin (1947, 1951) now regards this Nototheniid as probably belonging to a distinct subspecies, A'', rossii marmorata Fischer. ^ Its wide distribution is inferred from the frequency with which its larvae and young stages have been taken in our circumpolar plankton hauls. Full grown adults have never been captured in a townet, the relatively few specimens that have been described having been collected principally from the stomachs of whales. The adults, therefore, are evidently very fast moving animals that readily avoid capture by ordinary slow-moving oceanographical apparatus. * Generally from one to a dozen have been found at a time. Mackintosh (1942 a), however, records a notable instance where a blue whale had swallowed no less than fifty Notolepis coatsi. * This and other information for which I am indebted to the late Dr Gunther has been extracted from the original logbooks of the Discovery Investigations in which he left many valuable field notes on krill and its behaviour. 13-2 132 DISCOVERY REPORTS Matthews (1951) also refers to the local abundance of A^. rossii and the massive impact of its assault on the krill in these island waters : ' In about twenty minutes we came to a place near a small iceberg where the birds were gathered in their tens of thousands, and the krill was so dense near the surface that it made the water quite soupy. And there were the fish, about two fathoms down, thicker than salmon under Galway bridge, so closely packed that it looked like a solid bottom of huge pebbles just under our keel '. He adds that the fish were two to three feet long, ' splendid great bull-nosed fellows', and that they were gobbling up the krill as hard as they could, their stomachs all being crammed with it. Olsen (1954) states that the adults of this species feed almost exclusively on E. superba and calls attention to the immense numbers of them that frequent the continental shelf on the east side of the island from Gierke Rocks in the south to Willis Island in the north. ' As soon as it is light enough in the morning', he writes, 'the "sea-fish"^ begins to hunt the "krill" at the surface, where it is found gambolling about. In good weather it was possible therefore on light days to see shoals of fish as far as the eye could reach, but when the sun went down, all was still again '. While all the fishes I have mentioned are either shallow-water coastal forms or, if oceanic, inhabitants of the Antarctic surface layer, ^ there is evidence that much deeper living species, even those commonly regarded as bathypelagic, may occasionally come to the surface to prey upon the krill. In his report to the Discovery Committee on the work of R.R.S. 'William Scoresby' in January and February 193 1 Dr N. A. Mackintosh records at Station WS 540 'an eel-shaped Stomiatoid fish, with brilliant luminous organs and silvery scales ' which was feeding on krill close to the surface and could plainly be seen by the light of a cargo cluster. In the Note and Sketch log for that period Gunther gives a vivid description of this incident which I quote (see also Clarke, 1950; Hardy, 1956) exactly as that brilliant natural observer recorded it: In the midst of a swarm of krill which had drifted round the ship, an eel-shaped silvery fish 9-12 inches in length was observed by the light of a cargo cluster. From a pair of luminous organs in the orbital region, the fish emitted a beam of varying intensity of strong blue light which shone directly forwards for a distance of about two feet. The fish had the habit of lurking at a depth of 2-6 feet below the surface, poised at an angle of about 3 5 "-40° from the horizontal, head up — this gave the beam an upward tilt : occasionally the fish swam round and with quick action snapped at the cloud of krill above it. In its manner of lurking and of snapping prey it was semblative of the freshwater Pike. From the anal region was seen to trail a length of brown substance which it was supposed might have been either genital or faecal product. Efforts to catch the fish with a hand net failed owing to the drifting away from the ship of the krill swarm. The supple silver of the fish, the parallel beams of blue which emanated from its forehead, and by the Hght of the cluster other animals seen swarming at the surface — particularly Ctenophores with iridescent combplates which in waving bands gleamed the spectrum — combined to give an effect of consummate wonder and beauty. This record is of special interest as it appears to be the first notice of the existence of Stomiatoid fishes in Antarctic waters. Although the fish could not be captured there is little doubt that those who saw it assigned it to the correct group, for Mr Marshall tells me that among the undescribed collections of Discovery fishes in the Natural History Museum there is a specimen of the genus Melanostomias taken at a depth of 450-270 m. at Station 1356 in 60° 12-8' S, 19° 37-5' E. Since the krill themselves (p. 168) are not bathypelagic, but tend to cleave, especially at night, very much to the surface zone, Notolepis coatsi would appear to be another bathypalegic form, if it is in fact a bathypelagic form,^ that comes up to the surface to feed. 1 That is, as distinct from the young kelp-living N. rossii of the South Georgian fjords. * Notolepis coatsi perhaps excepted (but see p. 155, note i). ^ But it may not be. Referring to its frequent occurrence in whales' stomachs Nemoto (1959) states that the fish has been seen chasing E. superba on the surface and since this evidently was a daylight observation it suggests that it need not necessarily be a deep dweller that comes up to the surface at night. THE OLDER STAGES 133 There is little doubt then that many fishes do feed more or less heavily on the krill. Some of them (e.g. Notothevia rossii) are known to be exceedingly abundant in certain places, but there is no knowing what quantities of krill-eating pelagic fish may exist in the vast area of the Antarctic seas in which this euphausian is known to range, for as yet no suitable method has been devised for quantitative sampling of the larger and more active pelagic animals. For all we know there may be vast numbers of oceanic fishes subsisting on the krill, devouring it on a scale far greater than our existing knowledge permits us to believe. In his Living Resources of the Sea Walford (1958) surprisingly seems to regard the Antarctic fish population as playing only a minor part among the predators of the krill as a whole. He writes : One of the most striking features in the economy of life in the antarctic seas is that few fishes have the form and swimming powers to utilize krill (planktonic crustaceans) as a source of food. The evolution of pelagic types of nototheniid fishes has been limited, and it is the warm-blooded animals, the whalebone whales, crab-eater seals, and penguins, which make most use of the krill. In oceanic circumpolar waters, a number of small bathypelagic fishes feed particularly on krill and are often to be found with the latter in the stomachs of whalebone whales.^ These fishes, however, would hardly repay commercial exploitation. Obviously, however, the coastal and exceedingly numerous Notothenia rossii has both the form and speed to capture and devour these large euphausians on an enormous scale and there seems little reason to doubt that other coastal or inshore forms are equally well equipped. As for the other krill- eaters, both coastal and pelagic, time has yet to show that they are really so scarce as our sluggish bottom apparatus and plankton nets reveal them to be. As Marshall (195 1) has said of the pelagic forms, ' it is highly likely that they are much more numerous than is indicated by collecting gear '. How abundant, and evidently closely packed, they might well in fact be, is suggested in the following passage from Worsley (1959) in his vivid description of the escape of Shackleton's boats from the ice of the Weddell Sea: 'Lying about in the slush and on the "pancakes" were countless thousands of dead fish, some of which were eight inches long. They had been caught and frozen by the sudden freezing of the sea. They looked like splashes and bars of silver glistening in the sun. The petrels and Cape pigeons were enjoying an unusual feast. Like the birds, we would have relished a splendid meal of fish, but we dared not waste time gathering them '. Nordenskjold (1905) also reports a concentration of dead fish floating at the surface in this locality, and in January 191 2 (Filchner, 1922) the ' Deutschland ' encountered fish in great abundance among the open floes of the East Wind drift in 67° 27' S, 26° 16' W. The silvery colour and bar-like appearance of Worsley's fish recall Dollo's (1908) original descrip- tion of Notolepis coatsi, itself (Marshall, 1955) a voracious krill-eater, and if it was in fact this species he saw then one can well imagine it, growing as it does (Marshall, 1954) to a length of at least 2 ft., as a major predator of the krill, especially if, as this instance of mass mortality suggests, it congregates densely in shoals. Pelagic fish larvae too are sometimes encountered in great concentration. I saw them, in clouds, swarming near the Ross Ice Barrier in January 1936. Hodgson (1905) refers to the wealth of fishes in McMurdo Sound, noting that they were abundant all the year round and that the hundreds of Weddell seals in this locality lived almost exclusively on them. He mentions one very large Nototheniid weighing 39 lb., 3 ft. 10 in. long, with a girth, just behind the pectoral fins, of just over 2 ft. The foregoing survey probably includes most of the larger animals that feed exclusively or very sub- stantially on the krill. But there may be others, perhaps many others, especially among the fishes and (p. 134, note 2) squid. It would not be complete, however, without mentioning the Weddell seal, the Leopard seal, the Ross seal, the Emperor penguin and the Antarctic shag, all of which have been 1 Suggesting strongly, however (p. 168), that they were at least not bathypelagic when swallowed. 134 DISCOVERY REPORTS reported^ from time to time with krill in their stomachs. Apart from their swallowing it fortuitously, however, or as an occasional morsel, it is doubtful if these animals take it in any great quantity directly from the sea. More likely they obtain it indirectly, as the Giant Petrel does, from the stomachs of the larger quarry upon which they habitually prey. As adults they feed principally on fish or squid,^ the Leopard seal, besides being partial to fish, being a voracious hunter of penguins as well. In view of its partiality to krill-eating penguins and fish it would be surprising indeed if E. superba were not found occasionally in the stomachs of this seal. Rustad (1930), however, does mention one instance where the krill in the stomach of a Leopard were so fresh that they obviously must have been taken directly from the sea 'and not merely passed in as stomach contents of the fishes', while more recently Mr L. Catherall, who wintered at Hope Bay, North Graham Land with a Falkland Islands Dependencies Survey party in 1956, reports^ a young Leopard with the stomach fully distended by 'completely undigested ' krill. Other evidence that the young Leopard may be partial to, or even perhaps a sub- stantial consumer of, E. superba comes from the private diary of Nicolai Hanson (1902) who died at Cape Adare in 1899. Referring to a specimen shot in the pack earlier that year he remarks, 'It was a fine young Leopard-Seal. Like the one Borchgrevink killed, his stomach was full of small shrimps *. In general, however, as Bertram (1940) points out in his remarks on the diet of the adult Weddell, it is obvious that the Crustacea found in the stomachs have ' been swallowed along with the fish that had already devoured them'. The Weddell pups on the other hand do it seems feed for some time directly upon the krill, Lindsey (1937), in his study of the Weddells of the Bay of Whales, having shown there is a short transition period after weaning when the young seals subsist largely on Crustacea before taking to the adult diet. ' The change from a milk diet ', he writes, ' is gradual com- bining amphipod and isopod crustaceans with the milk for a time. The stomach of a 5 5 -day seal contained milk and a few Euphausia superba and other crustaceans. Another pup. . .2 months old, was filled with Euphausia'. Lindsey's pups, however, could not in fact have been eating E. superba since this species it is now known (p. 124) does not extend, except for a very short distance, into the shallow shelf water of the Ross Sea. In all probability they were feeding on E. crystallorophias, that other large euphausian superficially resembling the krill, which swarms in great profusion in this locality. The Crabeaters examined here by Lindsey (1938) and Perkins (1945) also it seems must have been feeding on this species. Nevertheless, it may readily be inferred from Lindsey's observations that in other parts of Antarctica where E. superba was the dominant swarming euphausian it would be to this species among others that the Weddell pups would turn during their brief period of crustacean feeding. It would appear too that the occasional adult Weddells that wander off on the pack also take krill in some quantity from the sea, Barrett-Hamilton (1901a) in his account of the seals of the ' Belgica ' expedition stating that the Weddells collected on the pack by Racovitza were feeding on ' the Schizopod crustacean Euphausia '. He repeats this in the Antarctic Manual (Barrett-Hamilton, 1901^) and again (Barrett-Hamilton, 1902) in his account of the seals collected by Hanson in the 'Southern Cross', remarking that the Weddell, like the Crabeater, 'feeds when on the pack-ice, according to Dr Racovitza, on Euphausia '. The toll of the krill exacted by the straying Weddell, however, must be very small, negligible in fact when compared with the major ravages of whales, 1 Wilton (1908); Pirie (1908); Bruce (1913); Tattersall (1918); Rustad (1930); Murphy (1936); Hamilton (1939); Allen (1942). 2 Wilson (1907a); Levick(i9i56); Matthews (1929a); Lindsey (1937); Hamilton (1939); Bertram (1940); Cendron (1952); Stonehouse (1953) et al. Although krill-eating fish are probably mainly responsible for the presence oiE. superba in the stomachs of these essentially carnivorous (or non-plankton) feeders, it may be noted that some species of squid (Brachi, 1953), many more perhaps than we are aware of, also feed upon the krill. Bierman and Voous (1950) counted 100 Euphausia eyes (probably E. superba) in the stomach of an Emperor penguin and in the stomach of another up to 124 cephalopod beaks. ^ In a letter to Dr R. M. Laws formerly of the National Institute of Oceanography. I THE OLDER STAGES 135 Crabeaters, birds and fish. It is rarely encountered on the pack (Wilson, 1907a; Bruce, 1913 ; Brown, 1915a ei al.) far away from land. As Wilson (1902) remarks it is essentially a 'shore seal ', not a single specimen he adds having been seen from the 'Southern Cross' (Bernacchi, 1901) throughout her voyage in the Ross Sea ice. There can it seems, therefore, be little justification for the statement by Ekman (1953) that E. superba forms 'a considerable part' of the food of this exceedingly abundant coastal species. Pouting (1923) vividly describes the feeding of the Adelie penguins at Cape Royds, the parent birds in a continuous flow streaming landwards over the fast ice, their stomachs ' swollen with the load ' of Euphausia they had gathered for their chicks. Here again, however, as with Lindsey's Weddells and Crabeaters, it must have been E. crystallorophias they bore. And it must be to this species too that Grifiith Taylor (1930) refers in his note on the teeming wealth of Euphausia off Cape Crozier among which hordes of Adelies hunt and feed, and this the ' Euphausia ' so often reported^ to be supplying the multitudes in McMurdo Sound. Dell (1952) states that the Emperor penguin, at least for part of the year, feeds almost 'entirely' on E. superba. There is nothing, however, in Antarctic ornithology or narrative to suggest that the krill ever feature thus largely in the diet of this predominantly squid-eating bird. It may it is true be responsible for some minor measure of depredation, particularly it would seem during its periodic wanderings off on to the pack. Falla (1937) for instance records that three birds collected on the ice north of Queen Mary Land in February 1931 'had been in the water fishing for "kril " {Euphausia), of which they contained fair quantities '.^ Finally, turning to the plankton, we find indications that even from this the krill may be far from secure. Both Hart (1942) and I have seen single small individuals in the gut of the large Sagitta gazellae, while the ' Euphausiid larvae ' recorded by David (1955) in the gut of this arrow worm would no doubt include some larval krill. Hart, in his Antarctic food chain (p. 45), shows that young E. superba are also included in the diet of the pelagic amphipod, Parathemisto gaudichaudii, which swarms in enormous numbers in parts of the circumpolar sea. Other voracious plankton animals, notably coelenterates, are doubtless also involved, for, as Bigelow(i926) emphasises, the various species of medusae, large and small, all ' belong to the piratical category ', the total destruction they wreak on euphausians, copepods and other animals being ' beyond any estimation '. E. superba then manifestly occupies a key position in the economy and ecology of the Antarctic seas. It supports life on a vast scale, a complex variety of life ranging in size from that of the small organisms of the plankton to that of the immense bulk of the blue whale, the most gigantic animal that has ever existed. It must exist in enormous numbers, in an astronomical abundance far exceeding that of any other pelagic euphausian, and while any estimate of its bio-mass (p. 170), in the present state of our knowledge, must involve much speculation there are perhaps some grounds for supposing, as Mackintosh (1934) has written, that in sheer mass of living matter it might be 'equal to all the rest of the Antarctic macroplankton combined '.^ 1 Murray (1909); Cherry-Garrard (1922); Joyce (1929); Debenham (1930a) et al. ^ Little is known about the feeding of the King Penguin. Stonehouse (i960) states that it is clear that cephalopods form the bulk of the diet, although it has not yet been possible to determine the species. He notes, however, that Notothenia rossii may be included in the feed, and since this fish is a voracious krill-eater it would not be at all surprising if E. superba were to turn up occasionally in the stomachs of these large birds. ^ This was written long before the full facts of the distribution of the krill became known and could now only be said to apply, if it does in fact apply, to the macroplankton of the East Wind-Weddell surface stream to which the vast bulk of the euphausian population (p. 61, Figs. 56 and 5 c) is now known to be confined. The bio-mass of the combined macroplankton of the circumpolar West Wind drift manifestly it will be seen vastly exceeds that of the scattered and meagre krill population there. 136 DISCOVERY REPORTS Perhaps the only animals playing a comparable ecological role in these southern waters are the swarming anomurans Munida siibnigosa (White) and M. gregaria (Fabricius), with its pelagic post- larval Grimothea stage, upon which, it has so often been recorded,i such a multitude of predators, whales, seals, birds and fish, gorge and fatten in the lower latitudes of the Patagonian shelf and along the coasts of New Zealand. Clearly, however, Rayner (1935) under-emphasises the importance of the krill when he states that these other swarmers ' fulfil in the economy of the Falkland Islands region a role similar to, but wider than, that of krill, Euphaiisia superba, in south Georgian waters '. For surely it is the krill, in every sense, and above all geographically, that fulfil the wider role. Rayner's work on the growth of the Falkland species of Munida, as Hart (1946) observes, shows that, as compared with E. superba, they are comparatively long-lived (5 years or more of post-larval life in M. subrugosa) and are sexually mature from the end of the first year of post-larval life. It may be, therefore, as Hart concludes, that the ' distinctive differences ' in pelagic life of the Patagonian and Antarctic regions are ' affected by this difference in the life history of their respective key-industry animals '. It has sometimes been suggested by commercial interests that these teeming and seemingly inexhaustible fields of southern shrimps might themselves be harvested, with advantage to the whales that for so long have suffered at the hands of the industry. For in addition to the oil and protein upon which the whales build up their huge bodies the krill, especially in their eyes, provide a rich source of preformed vitamin A^ which the whales concentrate and store in their livers. If, therefore, effective measures could be devised for the capture and processing of the euphausians themselves a major reduction in the wholesale slaughter might be expected to follow and the southern whale population as a whole preserved safe from extinction perhaps for many years to come. As Fisher (1958) remarks, however, so long as the whale stocks are allowed to survive through annual control of the catch, they will continue to concentrate fat and protein in their bodies and vitamin A in their livers far more cheaply and efficiently than any man-made krill-catching factory could do. ' Meanwhile ', he adds, 'the innumerable krill remain as a huge reservoir of concentrated vitamin A, on which all larger marine animals, and ultimately man, may draw'. As Davis (1955) has said of the plankton as a whole, its main value to man 'will undoubtedly remain an indirect value for a long time to come, in its role as the primary food for larger animals '. No doubt it will. But no doubt too when the time comes it will be to the immeasurable^^ wealth of plankton animals in the polar seas, and above all perhaps to the southern krill, that one day he will turn and as Hardy (1930) wrote over 30 years ago, 'take both food and power direct from these vast planktonic stores of solar energy '. Bogorov (1958) calls attention to the annual toll of organisms now being harvested from the sea, observing that it has already reached the staggering figure of thirty million tons, adding that progress in mechanisation and advances in fishery technique generally will soon set a limit to our exploitation of the known, mainly onshore, fishing grounds, making it necessary in the near future, if overfishing 'with all its catastrophic consequences' is to be avoided, to exploit the riches of the world ocean in distant waters. Again, perhaps, it might be to the Antarctic that the fisherman will turn, for no one really knows what quantities of fish it may hold, or at any rate could hold, by virtue of the super- abundance of its krill. It may indeed take a long time before such visions become fact, for as yet, as Jackson (1954) points out, there seems little prospect of our developing a plankton harvesting industry that could be run as an economic concern. Even so, as Baalsrud (1955) has emphasised, that is no reason why further 1 Chilton (1909); Matthews (1932); Ommanney (1933); Hamilton (1934a); Rayner (1935); Hart (1946) et al. 2 Thompson, Ganguly and Kon (1949); Batham, Fisher, Henry, Kon and Thompson (1951); Fisher and Kon (1959). ' Perhaps some day, a long way ofF, it will be measurable. THE OLDER STAGES 137 investigation of the many and complex problems involved should not go ahead, for the latent resources of the sea, as Schaefer and Revelle (1959) have called them, 'will almost certainly become important food resources in the future ', as technology improves. The krill population may be regarded as an open biological system in which the incoming energy represented by recruitment and growth is balanced by the outflowing energy represented at present by its losses from natural causes (Bertalanffy, 1950). When the ultimate assault of man is included in this equation it will have to be directed wisely, the mathematical models of exploited fish populations of Beverton and Holt (1956) having shown that a truly rational exploitation of a fish resource requires in the end ' a high degree of co-operation between fishing units and the adoption of a conscious and balanced fishing attack ; in other words man's predation must become social rather than individual ', and he must be prepared, they add, to seek a full understanding not only of the characteristics of the populations as prey 'but of his own behaviour as predator'. Zenkevich (1957) has uttered the same warning note. 'You know', he said at a recent UNESCO symposium, 'that today we mainly make use of fish, but tomorrow we shall utilize other plant and animal raw stuff of the ocean. We must be prepared for this, but prepared like rational owners. Sea fishery must, as far as possible, be rationally conducted. . .'. The ecological phenomenon presented by the southern krill is perhaps without parallel among euphausians, if not indeed among plankton animals as a whole. It is clear, however, that other members of the order, although less widely distributed, in their own way play equally striking, if more local, roles in the economy of other seas. The swarming Nyctiphanes australis, for instance, of the Australian coasts (Dakin, 1953) at once springs to mind, while in northern waters the shoals of Thysanoessa inermis and Meganyctiphanes norvegica} seem to support life on a scale only surpassed perhaps by the Antarctic krill. Bigelow (1926) is particularly interesting on the euphausians of the Gulf of Maine, recording their swarming at the surface there and their importance as the food of whales and of many fish and birds. Euphausiids are often extremely plentiful near the surface in the Eastport-St. Andrews region at the mouth of the Bay of Fundy, where the smaller-sized herring can be seen chasing them to and fro right up to the docks, and they are so conspicuous when schooling that they must have been seen and commented upon by local fishermen from the first settlement of that coast. The earliest published reference to their local abundance there, or in any part of the gulf, for that matter, seems to have been in 1879, when S. I. Smith (1879, p. 90) described Meganyctiphanes norvegica as occurring at the surface in the Eastport region in 'swarms, filling the water for miles', and as 'usually accompanied by schools of mackerel, young pollock, and other fish, and in autumn by immense flocks of gulls, the fish and smaller gulls appearing to feed almost exclusively on Thysanopoda at such times'. Such occasions he recorded for April, August, September and October, adding that Verrill found these shrimp swarming in myriads in the ripplings in the center of the Bay of Fundy in 1869, and that they are often so abundant among the wharves at Eastport that they may be caught there by the quart. Moore also wrote (1898, p. 401) that 'during the summer and fall dense bodies of Thysanopoda are seen swimming about the wharves at Eastport and at other places in the vicinity, and they are also extremely abundant on the ripplings at Grand Manan, which has long been famous as a herring fishery. Excepting the eyes and the phosphorescent spots beneath, which are bright red, the bodies of these shrimps are almost transparent, yet such is the density of the shoals in which they congregate that a distinct reddish tinge is often imparted to the water. In the summer and early fall of 1895 they were especially abundant about the wharves at Eastport, and on one occasion, at least, they were left at low water several inches deep over a considerable area of one of the docks'. Moore believed that Thysanoessa inermis was the species chiefly concerned, but in the light of subsequent observations it is probable that then, as now, it was outnumbered there by Meganyctiphanes. Our own observations, with information communicated by Doctor Huntsman, show that the passage of time has seen no diminution in the abundance of the latter in the Eastport-St. Andrews region in summer and early autumn. 1 Sars (1875); Allen (1916); Bigelow (1926); Stott (1936); Ruud (1936); Einarsson (1945); Fisher and Lockley (1954) et al. 14 DM igg DISCOVERY REPORTS Elsewhere, Bigelow (1926) remarks on the vast quantities of euphausians, principally Thysanoessa inermis, eaten by blue, fin, humpback and North Atlantic right whales off the New England coast, referring particularly to CoUett's (1877) statement that 'sulphur-bottom stomachs frequently contain 300 to 400 liters of shrimps, and that occasionally one is taken with up to 1,200 liters of Thysanoessa '. Sizes and quantities eaten by baleen whales Before we can begin to consider E. superba in its primary role as the diet of the Antarctic Balaeno- pterids it is necessary first to get a clear idea of what size or sizes of this euphausian constitute the actual feed and what, being too small, escape through the baleen plates and so do not contribute to the diet. In the early days of the Discovery Committee's work at Grytviken, Mackintosh and Wheeler (1929) examined the stomachs of 519 krill-eating whales, the majority of them blue and fin. 451 of them, or 87%, contained E. superba and 68 were empty. The stomach contents were roughly classified into large krill from 55 to 65 mm. long, medium krill from 40 to 50 mm. and small krill of 40 mm. and less. As far as can be gathered from this rough classification the whales examined appear to have been feeding for the most part on euphausians over 40 mm. long. It was noted, however, that in November 1926 there was a marked predominance in the stomachs of the smallest krill over the other classes while from January to May 1927 there was a conspicuous mixing throughout of all three classes, the smallest being particularly abundant. Some two years later Wheeler again visited Grytviken where he measured to the nearest half centimetre a total of 39,081 E. superba from blue and fin whales killed on the South Georgia grounds from October to April 1929-30. These hitherto unpublished measurements, arranged in 10 mm. groups, are shown in Table 24. At first sight they give the impression that blue and fin whales, those at any rate taken in the neighbourhood of South Georgia, feed almost exclusively on euphausians over 30 mm. long, 88 % of the grand total measured being of that order of size. The figures for December, however, show that occasionally at least, much smaller krill of the 21-30 mm. group may constitute a by no means unimportant part of the diet, nearly 30 % of the 1 2,037 measured in that month falling within this range. We get much the same general picture from the observations of biologists and others who have worked in pelagic factories. Ruud (1932) for instance gives details of the measurement of 1 785 stomach krill collected from whales taken by the floating factory 'Vikingen' from October to March 1929-30. According to his figures whale food over 30 mm. long occurred commonly in the stomachs throughout the whole period the factory was operating, while in October, November and December the 21-30 mm. group and under formed an important contribution to the diet, occurring more frequently and in far greater abundance then than later in the season. From these records it would appear that the staple diet of the southern rorquals consists of euphau- sians at least over 20 mm. long. A good deal of evidence, however, has recently come to light which suggests that in spring and early summer, notably in October, November and December, great quanti- ties of much smaller and younger krill may also be consumed. While exploring the possibility of using whale meat on a large scale for human consumption on board the pelagic factory 'Terje Viken' in December 1939, I examined on the 22nd of the month a fin whale killed in the eastern part of the Weddell drift which had eaten, along with older krill, very large numbers of the last larval stage, the Sixth Furcilia, together with early adolescents up to 20 mm. long. The stomach of this animal was bulging with the meal it had eaten, the larval, adolescent and older krill being present in the following proportions: Sixth Furcilias and early adolescents up to 20 mm. 21-30 mm. Over 30 mm. (%) (%) (%) 21 46 33 THE OLDER STAGES 139 The sample examined consisted of 129 individuals measured to the nearest millimetre. For several days after this the stomachs of the blue and fin whales brought to the factory all contained a high proportion of Sixth Furcilias and early adolescent krill and in two notable instances whales were recorded as having been feeding almost exclusively on this very young brood. The smallest euphausian measured at this time was a Sixth Furcilia only 11 mm. long. Among his 'Vikingen' measurements Ruud also records some very small individuals, the smallest 12 mm. long and almost certainly a SLxth Furcilia. Table 24. Stomach krill measured at South Georgia during the season ig2g-jo with the percentage figures shown in brackets Month 11-20 (mm.) 21-30 (mm.) 31-40 (mm.) 41-50 (mm.) 51-60 {mm.) Total measured October 7 335 (5) 1672 (27) 3821 (61) 450 (7) 6285 November 405 (8) 528 (10) 2930 (55) 1458 (27) 5321 December 174 (2) 3538 (29) 2181 (18) 3724 (31) 2420 (20) 12037 January 220 (7) 2751 (84) 261 (8) 30 (I) 3262 February 168 (4) 3627 (76) 825 (17) 157 (3) 4777 March 50 (I) 3701 (75) 1 189 (24) 23 4963 April 98 (4) 1832 (75) 494 (20) 12 (0 2436 Seasonal total 181 4814 16292 13244 4550 39081 Seasonal % — (12) (42) (34) (12) That these instances of such small whale food are not merely isolated ones without any real significance is confirmed by the observations of other biologists who at various times during the last 20 years have worked in factory ships for the former Discovery Committee. Measurements of stomach krill by Mr A. H. Laurie,^ for instance, on board the 'Southern Princess' in 1932-3 (Table 25) and others by the late Mr Paul R. Crimp^ in the 'Hektoria' in 1939-40 (Table 26) lend strong support to the view that late larval and early adolescent krill, in other words krill approaching the end of their first year of growth, may contribute substantially to the diet of whales especially during the early part of their sojourn on the southern grounds. Peters (1955) also reports a marked consumption of young, approximately i-year old, krill during the early part of the season, especially in November and December, but Nemoto (1959) states that his stomach samples did 'not show such tendencies'. Nemoto's samples, however, came largely from the East Wind drift and since these high latitudes are not open to shipping until January it is hardly surprising that he did not find the same abundance of very small whale food then as is encountered (p. 370, Fig. 118 and p. 377, Fig. 123) in the more northerly plankton in spring. He states that throughout the season his samples consisted of i-year and 2-year swarms in about equal numbers and this our summer East Wind measurements seem to show (p. 397, Fig. 136) is what would be expected. He finds in fact that it is towards the end rather than at the beginning of the pelagic season that his East Wind whales were feeding most heavily on small (14-15 month old) krill, and there seems to be a simple enough explanation for this. With the 1 To the nearest half-centimetre. ^ To the nearest miUimetre. 14-2 140 DISCOVERY REPORTS Table 25. Measurements of stomach samples from fifteen whales taken by FjF 'Southern Princess' in the season 1932-3 with the percetitage figures shown in italics Total 51-60 mm. measured — 228 — 168 — 195 — 185 — 176 — 218 — 131 — 192 — 192 — 195 — 172 — 198 — 19s — 130 — 170 — 2745 advance of summer in these high latitudes the pack retreats and the scene of hunting moves south- wards with it. As the retreat progresses more and more krill that have spent the winter and spring below the ice are uncovered, the last to be freed being the most southerly and those that have ' wintered ' longest below the pack. As a result of their prolonged existence in such cold and dark conditions the growth and development of these high southern swarms is much retarded, the Sixth Furcilias that spring from the previous summer's spawnings surviving in some instances into March more than a year later, for as long in fact as 12-15 months after they were born.^ It seems clear enough, therefore, that many of Nemoto's end season East Wind whales must have been feeding largely on this stunted and retarded population (see p. 397, Fig. 136). 1 The effect of ice on the growth and development of the krill is discussed on p. 355 and illustrated in Figs. 107 and 108. Date 11-20 mm. 21-30 mm. 31-40 mm. 41-50 i7xi32 210 92 15 7 3 1 — 17x132 16 9 24 14 48 29 80 48 17x132 176 90 19 10 — — 22xi32 65 35 "5 62 S 3 — 22xi32 41 23 108 61 19 11 8 5 I xii 32 67 31 127 58 24 11 ^^ Sxii 32 27 21 93 71 II 8 ^^ 8 xii 32 7 4 150 78 35 18 8 xii 32 II 6 73 38 108 56 16 xii 32 85 44 35 18 25 12 50 26 20 xii 32 92 53 64 37 4 2 12 7 ioi33 — 65 33 97 49 36 18 21 ii 33 16 8 65 33 90 46 24 12 8iii33 — 10 100 20 — 8 77 15 i9iii33 — — 120 7-' 50 29 813 30 963 35 689 25 280 10 THE OLDER STAGES 141 Table 26. Measurements of stomach samples from seventeen whales taken by FjF 'Hektorta' in the season 1 gjg-40 with the percentage figures shown in italics Total in 51-60 mm. sample — 235 — 102 2 120 2 2 126 2 — 222 16 169 9 30 171 22 217 10 — 324 1 321 2 236 1 3 171 2 2 215 1 12 369 3 7 177 4 " 255 4 2 132 1 112 3562 3 Another observer, Mr D. A. Parry in the F/F 'Empire Venture', reports very small krill only 1 1 mm. long — obviously the Sixth Furcilia stage — in the stomachs of a number of whales killed near 61° 23' S, 47° 28' W during the first 4 days of April 1946. This is an interesting observation since it suggests that the late larvae may feature in the diet towards the end as well as at the beginning of the feeding migration. It is even more interesting from another point of view, for the position is located in the western part of the Weddell drift, a region since found (p. 332, Fig. 91) to be the only part of the circumpolar sea in which at this time of year the Sixth Furcilia may be encountered in any measure of abundance. Bale 11-20 mm. 2i-jo mm. 31-40 mm. 41-30 mm. 19 xii 39 167 71 48 20 7 3 13 6 28 xii 39 — 27 26 73 72 2 2 5140 4 3 52 43 57 48 5 4 10 i40 55 52 41 9 7 8 6 13140 6 2 70 32 135 61 II 5 22 i 40 — 4 2 21 12 128 76 31140 — I 1 5 3 135 79 I ii40 — 13 6 103 47 79 36 6 ii40 3 I 298 92 18 6 5 1 12 ii 40 — 276 86 42 13 2 i 14 ii 40 I 94 40 96 41 43 18 22 ii 40 S 2 102 60 44 26 17 24 ii 40 — II 5 98 46 104 28 ii 40 — 12 216 129 — 3 59 35 10 iii 40 — 13 7 89 50 68 55 12 iii 40 — S 2 117 46 122 48 16 iii 40 — 10 8 90 68 30 23 241 1088 1220 901 7 31 34 25 142 DISCOVERY REPORTS The consumption by whales of such small euphausians may well be of commoner occurrence than is generally supposed, for as Ruud and others have noted the younger krill are digested, and therefore become unrecognisable, far more rapidly than the more heavily chitinised adults, the latter persisting longer in a recognisable state in the stomachs. Even when fresh very small krill, especially those as small as the Sixth Furcilia, might readily indeed escape the notice of the casual observer since in a well filled stomach in which the young and old broods are both represented the smaller animals tend to be masked by the sheer bulk of the larger. It is evident then that numerically at least, if not in bulk, krill of hitherto unsuspected smallness may form an important constituent of the diet at certain times of the year. Nevertheless, in so far as sheer bulk of intake is concerned it is likely that, as Peters (1955) has also remarked, it is upon the older euphausians over 20 mm. long that the whales mainly depend for their annual fattening up. During the feeding process the baleen plates must behave selectively, retaining the large euphausians more readily than the small. At the same time it is difficult to avoid the conclusion that the supply of young krill, the staple whale food of the future, might become seriously depleted if the whales were to frequent the feeding-grounds in large numbers at a time when older krill was scarce (see charts on pp. 363, 366, 375, 380, 399 and 404) and there was little else for them to eat but the late larval and early adolescent brood. Fortunately, when such a situation arises, as it does during autumn and winter, the whales have largely deserted their southern haunts and dispersed to warmer waters to breed. Thus as they do not return to their feeding-grounds until spring, the rising generation of whale food is granted annually some three to four months' respite from the ravages of its largest predators. That the baleen plates of both blue and fin whales should be capable of dealing effectively with euphausians as small as say from 1 1 to 20 mm. long is not surprising when it is recalled that in northern waters, as CoUett (1886 and 191 2) and Milais (1906) have shown, the sei whale, although the fringes of its baleen plates are admittedly of a finer texture than those of the larger rorquals, is well known to feed quite commonly on much smaller organisms such as the copepods Calanus finmarchicus and Temora longicornis. In view of its abundance and pronounced tendency to occur in surface shoals (Hardy and Gunther, 1935, Fig. 133) it is surprising that the large pelagic amphipod, Parathemisto gaudichaiidii, is so seldom found in the stomachs of the large southern baleen whales. The seis that penetrate into the Pacific sector (Nemoto 1959), are so far the only recorded substantial consumers. At South Georgia the seis examined by Discovery observers were feeding on krill (Matthews, 19386). In the stomachs of the other well-known krill-eaters, blue, fin and humpback,^ this swarming amphipod has not been recorded except as an occasional and evidently fortuitously swallowed morsel (Mackintosh and Wheeler, 1929). As Mackintosh and Wheeler remark it is so abundant in the plankton that the whales ' can hardly help swallowing a certain quantity '. If blue, fin and humpback whales preyed deliberately on the Parathemisto swarms, as well as on E. superba, we should expect to find them not only in conspicuous masses in the stomachs but like the krill (p. 148) in discrete masses, and this has never been recorded. In northern waters Bigelow (1926) calls attention to the same phenomenon in the Gulf of Maine, noting that 'the large, easily recognized, pelagic amphipod Euthemisto, locally and temporarily so abundant, has never been recognized in the stomachs of any of the whalebone whales. Is it eaten? And if not, why not?' I would suggest that the most likely answers to these questions are that in southern waters the whales concerned are able to distinguish between the bright red of the krill swarms and the blackness of the Parathemisto swarms and that for some unknown 1 Little is known about the southern right. At South Georgia, from a single record, it would seem to be krill-feeder (Matthews, 1938 a). THE OLDER STAGES 143 reason they avoid the latter. It certainly cannot be that it is taken but rejected wholesale by the baleen plates, for growing as it does up to 30 mm. long it is much too big for that. It is bulky besides, being broader in the beam than the slender krill. We do not of course know if the baleen whales can in fact tell one colour from another. It has been suggested, however (Mann, 1946), that colour vision in these animals is a possibility. Z 36 OCTOBER NOVEMBER 19 23 25 30 4 7 9 IS 16 18 22 28 I 2 5 9 P 17 19 28 3 4 5 8 KD 12 I Scale percent ■ | I | i I 1 T °^°° I '1 iiiii 30 40 40 40 40 40 40 40 40 40 40 70 49 216 27 III 58 37 I20 97 40 40 40 20 40 40 18 37 42 55 99 79 702 61 9 40 lOO lOO I03 NO 42 3IO 40 40 40 30 216 100 lOO 100 100 lOO 52 280 27 189 48 379 61 113 40 116 40 116 588 325 522 DATE 68 64 60 55 \ 52 i 48 ' 44 I 40 ■» 36 z 24 a 20 I 16 U 12 "^ I DATE 68 64 to ^ 60 O 56 cr O 52 2 48 s '^ . 36 O 32 < 28 "^ 24 f 20 O 16 ::i 12 S o FEBRUARY 13 19 20 23 31 2 3 6 II 12 13 14 18 22 23 24 28 I ^MRCH I 7~\ m I 1 fmuuw O so 100 O iO lOO Scale per cent STOMACH SAMPLE NET SAMPLE 146 40 40 50 159 IOC 32 164 40 40 171 146 318 40 36 III 40 207 122 287 40 904 86 539 lOO 818 186 82 lOO lOO IOC 100 118 Fig. 14. Length frequencies of krill from the plankton and from whales' stomachs compared, both stomach samples and plankton samples having been collected on the same, or approximately the same dates. The number of older (O) and younger (Y) krill measured is given for each sample. The remarkable exclusiveness of the diet of the southern whales is paralleled among many far- eastern fishes, the Pacific cod for example, with over lOo abundant species to choose from, consistently selecting only four or five (Moiseev, 1957). As a result of the 2-year life-cycle of the krill the feeding-stuff available to the whales consists of two distinct broods, a younger, rising yearling or yearling class, and an older near-mature or 2-year old breeding stock that is reaching, or shortly about to reach, the end of its natural span. The whales feed indiscriminately on both. Sometimes the young brood is present in the stomachs, sometimes the old, or, as often happens, both broods may be present in the same stomach. In Fig. 14 I have plotted in 4-mm. groups the length frequencies of samples taken from whales' stomachs and compared them with samples taken directly from the plankton on the same, or approximately the same, dates (not J44 DISCOVERY REPORTS necessarily, however, in the same year) by our stramin nets, the stomach length frequencies being based on the 'Vikingen' and 'Hektoria' measurements of Ruudand Crimp. Although a high propor- tion of them come from the Weddell stream, net and stomach samples are not necessarily from identical regions, but for purposes of this rough comparison it is the approximate date that matters. It will be seen that as the season advances the whales tend to sample the krill in much the same way as our nets do, from October to March the young and old broods being represented in the same stomach with much the same frequency as they are encountered at the same time in the plankton. There is a distinct tendency, however, and perhaps a natural one, for the whales to be the more efficient samplers of the older krill, a tendency doubtless associated with the measure of avoidance the larger, and particularly the very large, euphausians undoubtedly (p. 262) take of our nets. Einarsson (1945) also remarks on this tendency in a reference to the various animals, notably whales, that ' are more mobile in their pursuit of the Euphausiids than our nets are'. At the same time it is clear that the spacing of the baleen plates puts a limit to their sampling of the young 10-12 month old broods encountered from October to December. Here it is the nets that are the more efficient samplers, the tendency being for the whales to sample such broods principally in the upper reaches of their length range. And so it may be supposed that while the adult krill suffer heavily through the depredations of their immense predators the young yearling swarms survive in at least some measure of immunity to contribute to the mass of the available feeding-stuff later in the season. Although some slight information is available from the north we know very little about the quantities of food consumed by the large southern rorquals, and as Thorson (1958) has said of predators in general facts about this are urgently needed. All we can be certain of is that depredation must occur on a stupendous scale. Bennett (1931) remarks, 'The amount consumed daily must be enormous, for whales taken in areas where "whale food" is plentiful have the stomach so distended as to be on the point of bursting when that organ is removed from the body. In the case of large whales, the stomach contents consist of many cart-loads'. Perhaps one cartload, however, or at most two (PI. Ill), would be nearer the truth. Upwards of two cartloads are clearly indicated in an impressive photograph by H. W. Symons^ published by Slijper (1958, Fig. 167), and in an equally striking illustration by Hjort and Ruud (1929) a bulging fin-whale stomach might it seems be carrying a similar quantity. Hardy (1959) gives much the same broad volumetric estimate, remarking that it is 'very impressive, when a whale-stomach is cut open on the flensing platform, to see a cartload or so of these little shrimps come gushing out'. I think, however, that Budker (1957) may be going too far when he suggests the possibility that a large baleen whale could swallow a ton of krill at a single mouthful. Hjort (1933) probably comes closer to the truth when he says 'they take a huge mouthful, say several barrels full '. Guldberg (1887), citing Collett (1877) on the quantities of Thysanopoda inermis (= Thysanoessa inermis) eaten by blue whales in northern waters, states that ' Collett fand gewohnlich 2-3 Tonnen (3-400 Liter) im Magen; aber die sehr grossen Individuen, wenn sie sich richtig voU gefressen hatten, batten bis 10 Tonnen'. Here, clearly, Guldberg is using 'Tonne' in its German meaning of tun, cask or barrel, and not, as it also means in German, as the metric ton or 1000 kg. Assuming his euphausians to have been adult T. inermis, which grows to about 30 mm., and that a litre of these would weigh about 2 lb. ,2 Guldberg's 300-400 litres would work out at between 6 and 8 cwt. and his 10 Tonnen (= say 1300 litres) at approximately i| English tons. If, however, his Tonne 1 An equally remarkable photograph by Professor Johan T. Ruud, showing what seems might be at least two cartloads (perhaps 2 tons), appears in Clarke's (1954) Elements of Ecology. It is said to be of the opened stomach of a southern right. 2 A litre of E. superba in the 21-30 mm. range, drained of its fixative through muslin so as to simulate the packed and damp condition in which it would come from a whale's stomach, does in fact weigh 2 lb. THE OLDER STAGES 145 refers, as I imagine it must, to the whaleman's barrel of 170 litres, 2-3 such barrels would work out at between 6 and 10 cwt. and 10 such at i-|- English tons. Ponomareva (1954) gives the average wet weight of T. inermis in the 15-26 mm. range as 0-57 g. from which it would appear that i|tons of this euphausian would amount to at least 2,600,000 individuals, a figure approximating closely to the number of E. superba it seems distinctly likely (see next paragraph) might be crammed into the full stomach of an average-sized southern baleen whale. Routh (1949) estimated that the full stomach of a blue whale taken by the floating factory ' Balaena ' in the season 1946-7 and said to measure between 75 and 80 ft., contained roughly 5,000,000 £". superba. He does not, however, give details of their size. Heyerdahl (1932) publishes figures showing that the average weight of the individuals in a krill sample ranging from 28 to 65 mm., a range that might well (p. 139, Tables 24-26) have been represented in this stomach, is o-8o g. from which it would appear that 5,000,000 of these euphausians might weigh as much as 4,000,000 g. or approximately four^ tons. This would not, of course, represent an average figure, since many whales are larger than 80 ft. and many are smaller and (Nishiwaki, 1950) have correspondingly larger or smaller stomachs. Allowing, however, the following broad assumptions (i) that the average population of the large Antarctic baleen whales, as Mackintosh and Brown (1956) have shown, was, at its seasonal maximum, about 210,000 over the period 1933-9, (2) that on an average (a very conservative average) each individual of this population spends say 90 days on the feeding-grounds, (3) that each individual fills its stomach once a day, and (4) that the average weight of contents of a full stomach might be 2 instead of 4 tons, it would appear that, at a very low estimate, between 1933 and 1939 the krill were being grazed down at the staggering rate of 210,000 x 90 x 2 = 37-8 million tons a year^ by large baleen whales alone, and in actual numbers at the rate of say 2,500,000 x 210,000 x 90, or over 47 million million. This last figure is almost certainly too low since it takes no account of the very small euphausians, the Sixth Furcilias and early adolescents up to 20 mm. long, upon which the whales feed more or less heavily during the early part of their sojourn in the south. It is distinctly possible in fact, that in sheer numbers the krill are being destroyed by the southern whales on a vastly greater scale than has been estimated here. At Station WS 540 Gunther found that in a sample from a patch of yearling E. superba ranging from 17 to 32 mm. (with a 30% mode at 24 mm.) there were 11,152 euphausians to the litre. An average-sized whale gorged to repletion on such a patch, and probably having eaten at least 1000 litres would, therefore, be carrying substantially over twice Routh's estimated 5,000,000. The whales, of course, do not feed exclusively on such patches, but (p. 143, Fig. 14) they often do, as well as on patches of much smaller individuals, and this must be borne in mind in any estimate of the total destruction they wreak. I would emphasise, too, that both Vangstein (1956) and Ruud (1956) have already expressed the opinion that Mackintosh and Brown's estimate for the Antarctic whale popula- tion may be too low, and if it is, then my estimate for the annual toll it exacts from the krill could be correspondingly low. Ruud writes: There is reason to stress that Mackintosh & Brown's results must be accepted with all possible reservation. It is possible that some of the best whaling fields have been poorly covered by the 'Discovery 11' cruises. The number of whales they sighted, 2602, along a course of nearly 47,000 nautical miles, seems small. . . . But with all possible reservations we are bound to assume that Mackintosh & Brown have reached a correct order of size in their estimate of the stocks. They amount to some hundreds of thousands. 1 Using Ponomareva's figure for the average wet weight of 15-26 mm. T. inermis, and assuming that to be the same as in the corresponding length range in the southern krill, Routh's 5,000,000 euphausians would weigh 2-8 English tons. ' Assuming a weight of say 50 tons for an average-sized whale this estimate in fact seems very low indeed when compared, for instance, with the 18 kg. of animal food required in the course of a year's feeding (Petersen, 1918) to produce I kg. of plaice. 13 DM ,46 DISCOVERY REPORTS The fantastic numbers in which the krill are consumed by whales alone may be compared with the impact of the northern herring upon Temora longicornis. In January 1872, Howell (1921) records, Kiel harbour was crammed with herrings for three weeks, ' and each herring was crammed with a compact pink bolus consisting of about 60,000 Temorae'. Using Guldberg's figures, which suggest on average full stomach content of about i ton,^ the annual toll of the krill exacted by the southern rorquals would work out at about 19 million tons, still an enormous figure. A ton in fact might not be an improbable estimate for an average full stomach content, Nishiwaki (1950) having given figures showing that in 61 whales, 32 blue and 29 fin, ranging fairly from 61 to 89 ft. in total length, the average weight of the stomach when empty is over | ton, the lightest, in a 61 -ft. fin, weighing 2 cwt., the heaviest, in an 83 -ft. blue, over \ a ton. In the much smaller northern sei whale Fujino (1955) finds the average weight of the empty stomach in twenty 40-45 ft. specimens to be over 2 cwt., a similar figure being found by Omura (1957, 1958) for the empty stomach of a 38-ft. North Pacific right. It is interesting to compare these recent Japanese figures with the earlier findings of Captain T. Sorlle, who, while manager of the Vestfold Whaling Company, South Georgia, was the first to dis- member baleen whales and weigh the viscera piecemeal. He obtained the following results. Stomach of a blue whale, 20-30 m. (66 ft.) long, just under | ton (Sorlle, 1924) ; stomach of a large blue whale, 27-18 m. (89 ft.) long, a little over 8 cwt. (Laurie, 1933; Rammner, 1956). Krogh (1934), using the growth curve for blue whales given by Mackintosh and Wheeler (1929) and the length-weight relationship in this species given by Laurie (1933), calculates that in 2 years the southern blue whale grows from a weight of 23,000 kg. at weaning to one of 79,000 kg. at sexual maturity, observing that this enormous increase takes place mainly, if not exclusively, during the two summer seasons (approximately 12 months) the young whales spend in the Antarctic. Walford (1958) estimates that such a growth-rate would require a regular intake of at least no quarts of plankton a day, adding that when the respiratory requirements are added the daily ration would come to 740 quarts. The higher figure would work out at a daily consumption of nearly | ton, for young whales alone. It should be noted, however, that more recent work,^ with improved methods of age determination, indicates that the period from birth to sexual maturity is normally substantially more than 2 years, so that f ton for young whales is likely to be an overestimate, although the observations of Betesheva and Nemoto given below suggest it may not be far out. In the stomach of a sei whale Betesheva (1954) reports 600 kg. (about 12 cwl:.) of squid, and a young 57-ft. fin is reported (Nemoto, 1959) to have eaten a stomachful of 759 kg. (f ton) of Alaska pollack. Surely the vast stomachs of the adult southern blue and fin whales would hold much greater quantities than these. If the fast-growing rat can eat up to 7% of its body weight per day (Donaldson, 1924), a daily ration of i ton does not after all seem an unduly excessive figure for an average-sized, say 60 or 70 ton, and also fast-growing, baleen whale, especially when one takes into consideration that it feeds only while in the Antarctic and there perhaps at the outside only for 6 months of the year. 1 Referring to the stomachs of the southern blue and fin whales Feltmann and Vervoort (1949) remark 'Hierin bevindt zich vaak meer dan 1000 kg. "krill"'. They give no details, however, as to how they arrive at this figure, nor does Douglas (1953) give any authority for his statement that blue and fin whales eat 2-3 tons at a single meal. On the other hand, the 1000 kg. mentioned by Schubert (1955) seems to be a very reasonable estimate. It is based on a photograph by Dr Nicolaus Peters of a bulging opened stomach from which the krill in their millions are gushing out. In The Emperor's Nezv Clothes, written on board a whale factory, Hjort (193 1) wrote, 'Thegreat whales we see on the flensing platform Hve on a single species of Crustacea, tons of which may be found in the whale's belly'. I wonder if he really meant tons or was speaking figuratively. Klumov (1961) states that average-sized fin and humpback whales probably eat from i to li tons of plankton per day. * See Ruud, Jonsgard and Ottestad (1950). THE OLDER STAGES 147 Kritzler (1952) gives Interesting figures for a y^-ft. Pilot whale which died after 9 months in a Florida aquarium having grown a foot in length, put on 200 lb. in weight and eaten 7 tons of squid at an average rate of | cwt. a day. The weakest link in this argument is perhaps the assumption that the whales fill their stomachs once a day, although the error involved might to a large extent be offset by the conservative figure put for the average time they spend on the feeding-grounds. Even this assumption, however, is not without some substance, Nishiwaki and Oye (p. 45) having noted a distinct tendency for afternoon- caught Antarctic whales to have the stomachs empty and concluding that both blue and fin whales are in the main early morning feeders and that (whether they fill their stomachs daily or not) they feed at least once a day. Nemoto (1957, 1959) also finds a tendency towards early morning feeding in the baleen whales of the northern Pacific, with decreased activity during the daylight hours and increased activity, not however always distinct, in the evening. It is perhaps of interest, therefore, to recall that it was maintained by the gunners of the old Arranmore Whaling Company, County Mayo, that the baleen whales they hunted were most plentiful on the surface at sunrise (Lillie, 1910). Obviously, however, the whole question calls for a comprehensive programme of weighing, counting and measuring of euphausians from Balaenopterid stomachs. There is much need too for extensive observations on the feeding habits and for data on the state of digestion of the krill that are swallowed. A stomachful of uniformly fresh euphausians, for instance, would indicate a recent single meal, and if this condition were observed to occur regularly at certain hours then we might be able to tell the approximate time or times when the feeding took place. In the hake, as Hickling (1927) has shown, weighing experiments indicate a fall in the mean weight of the stomach contents about midday and a tendency for the stomachs to be heaviest about midnight. From this and the fact that their prey was usually freshest in the morning and most ' cooked ' in hauls later in the day Hickling was able to deduce that hake are night feeders. As Clarke (1954) has said, 'When we can arrive at some reliable estimate of the size of the whale stocks, and have somehow found out the rate of feeding, this information could be combined with studies on growth and size distribution to calculate and compare the total annual consumption of food and the total annual increment of whale tissue '. Summarising, the large southern baleen whales feed on krill ranging from larvae at the Sixth Furcilia stage, no more than 1 1 mm. long, up to adults of the largest size, 60 mm. long or more. The bulk of the diet, however, or ' staple whale food ' as I shall call it, seems to consist of euphausians over 20 mm. long. The smaller (larval and early adolescent) forms which provide an important contribution to the feed in spring, the so-called 'blue whale krill ' (Ruud, 1932) of the South Georgia whalers, I shall refer to collectively as ' the small whale food '. Influence on the foetal growth-rate of whales Laws (1959 a ; 1959*) has recently shown that during the last 5 months of pregnancy the foetal weight of southern blue and fin whales increases with unparalleled rapidity, the weight of the blue foetus rising in that time from 20 to 2500 kg., a gain of 2-44 tons, of which over two are put on in the last 2 months. It is interesting, as Laws remarks, that the beginning of this phenomenal burst of grovv^th should coincide with the advent of the mothers upon the richly spread table of the Antarctic feeding-ground. On the slenderest of grounds Naaktgeboren, Slijper and Utrecht (i960) express the view that the 'remarkable deflection in the curve of Laws (1959, 1959*) is likely to be a result of the small number of foetuses involved '. Laws' very accurately measured material, 1 1 12 blue and 956 fin whale foetuses, could hardly in fact be called small, and, replying to his Dutch critics (Laws, 19606), he notes they seem to have missed the main point of his paper on foetal growth. ■" %■ 148 DISCOVERY REPORTS Patchiness Although known to occur over vast areas of the Antarctic seas it is clear from the evidence of plankton nets, from eye-witness accounts of its occurrence at the surface and from the many observations that have been made on the stomach contents of the whales themselves, that E. superba is far from evenly distributed throughout its range. In recent times^ indirect evidence of what appeared to be a tendency for it to occur in circumscribed shoals or swarms was first obtained by Mackintosh and Wheeler (1929) during the 1926-7 whaling season at South Georgia. Referring to the diet of the whales that season, they say, ' The krill differed from that of other seasons in the fact that there was in most cases a noticeable mixing of Euphausiids of different sizes. These were not ahvays mixed indiscriminately in the stojnach.^ Large or small individuals might be found together in different parts of the mass of stomach contents, or patches of large ones might occur in a mass of smaller forms suggesting that the whale had been feeding on separate shoals which differed in respect of the sizes of the individuals '. In the ' Vikingen' stomach samples Ruud (1932) also found a tendency for the small and large indivi- duals to 'keep apart in separate shoals '? Since 1927 continuous sampling of Antarctic surface waters for distances of up to and over 30 miles by repeated series of consecutive nets has shown conclusively that patchiness or unevenness in distribution, suggesting a tendency to congregate in dense swarms or patches, is one of the most remarkable features of the krill in the sea. From the results of one such net series worked off South Georgia in January 1927 Hardy and Gunther (1935, pp. 256-7 and p. 353) estimate that the patches, assuming them to be circular, are probably not larger than some 200 yards across and probably not more than about | mile apart. This concentration of the whale food into relatively small densely crowded areas they remark, 'helps one to understand how the large rorquals are able to collect sufficient to form an ample meal'. Indeed, when one thinks of the vast body of the whale and the relatively minute creatures on which it feeds it is difficult to conceive, were the habit of the krill otherwise, how the great southern whale population could even exist. Direct observation on the whale food at sea largely confirms the conclusions reached by Hardy and Gunther both as to the order of magnitude of the patches and the degree of their separation one from another. During a cruise undertaken by R.R.S. 'William Scoresby' into the Weddell Sea in January and February 1931 an area of profuse patchiness was encountered and observed in great detail by Gunther who was on board at the time. The patches were particularly abundant in the neighbourhood of Station WS 540 in 57° 55' S, 21° 21' W where they were seen continuously over an area estimated to be at least 150 square miles in extent. They were separated from each other by short gaps at most a third or a quarter of a mile wide and although they showed considerable variation in size none was exceptionally large, an average-sized biggish patch covering an area of about 2500 square yards, or say 60 by 40 yards. The biggest seen was a narrow belt 150 yards long by about 20 wide.* They varied greatly in shape (Fig. 15). Some were irregularly circular, some oval, some irregularly oblong, others occurring as ribbon-like behs which sometimes had narrow tributary out-streamers resembling the ^ See, however, early records, pp. 40-42. ^ The itahcs are mine. » The same phenomenon has been found in feeding sahnon, Aron (1958) recording an instance where the stomach con- tained amphipods and euphausians in almost 'pure cultures, with a sharp division between the two groups'. While allowing that the fish might have selected first individual euphausians and later only amphipods from among the plankton, he adds, ' But it is easier to believe that the stomach contents represent actual pure cultures as they occur in nature and that the salmon fed first on a swarm of euphausiids and then on a swarm of amphipods'. ^ Some patches, however, I suspect may be much larger than this. An entry by Gunther in one of our rough deck logs for 18 February 193 1 (time 1700) reads as follows: 'A large extent of krill (suspected) imparting a dull plum hue to the green water around. The patch extended towards South Georgia as far as the eye could see. Birds present. No net shot. Sea calm with long easy swell. ' THE OLDER STAGES i49 pseudopodia of an amoeba. Many were of such irregular outline that they can only be described as amorphous. In a number of instances they were seen to have gaps, resembling vacuoles, of clear water inside them, giving them the appearance of having been hollowed out. They occurred right on, or very close to, the surface, or from i to 4 m. below, each individual patch without exception seeming to consist of a narrow stratum or ' raft ' of densely crowded euphausians no more than a metre or two thick. The patches in general were of more or less uniform thickness Fig. 15. Rough sketches of krill patches made by the late Dr E. R. Gunther, Weddell Sea, January to February 193 1, the approximate dimensions of some of them being given in yards. The spacing of the patches has no relation to the natural spacing. although variations in thickness did sometimes occur within individual patches. In colour consider- able variation was noted, the degree of discoloration of the sea seeming to depend upon the depth of water through which a patch was viewed rather than upon its own density. If right on the surface a patch would present a brilliant red or vivid blood-red appearance while deeper down, at say 2 or 3 m., the colour would appear a dull rusty red or mahogany brown, fading away to a pale indeter- minate cloudiness deeper. Even surface or near-surface patches would not always produce the charac- teristic reddish discoloration. Sometimes it would be ochre-coloured or pale straw yellow, possibly owing to variation in the pigmentation of the euphausians themselves. Matthews (1951) also remarks on the shallow draught of the congregating krill and provides us with a vivid picture of the astronomically large numbers that must be concentrated in a single patch. He is speaking here of a big one he saw near South Georgia. 'The shoal of krill, the food of the whales, consisted of little shrimps about two inches long — they were only about half-grown^ — and reached from the surface down to about two fathoms, the fish swimming underneath and feeding off the ceiling. There must have been hundreds of tons of it, for the sea was thick with it like pea soup. ' 1 They must in fact have been full grown. ijo DISCOVERY REPORTS Towards the end of this cruise ' Scoresby ' ran into an area described by Gunther as being of 'exceedingly widespread patchiness' a mile or two to the north of Station WS 558 in 57° 41' S, 23° 12' W.i Here patches of krill of irregular outline were scattered about haphazard in all directions and were so numerous that in the course of a lo-mile run the observers were sometimes recording them every 10 or 30 sec. The majority were quite small ranging from some a yard or two across to others of from 50 to 100 square yards in area. Larger patches, one of 1800 square yards, were also seen. Elsewhere in a posthumous pubUcation Gunther (1949) gives an admirable description of patches he saw while whale-marking off South Georgia in the season 1936-7. On nth January whales in the act of feeding came under direct observation. A heavy concentration estimated to number 100-200 whales was centred over a patch of whale food in an area of 4-5 square miles. The krill {Euphausia superba) could be seen in a layer no more than 3 m. below the surface, and in places it affected the colour of the sea. An easterly gale the previous day had left a very heavy swell, but the wind had since moderated and waves were no longer breaking. The sky was grey and the bad visibility was reduced from time to time by patches of mist. The sea was grey too, but in places the krill imparted to it a barely noticeable tinge of ochre. In one place where the krill was unusually thick and might have been closer to the surface a patch of half an acre or so had a brick-red tinge. The krill was irregularly spread over the whole region with large gaps between the swarms; some measured a few feet across and had an indefinite contour like that of a gorse bush, and others extended in long wavy bands from one to several feet or even metres in width. The krill did not seem as dense as patches of it often are, and it looked as though it had been broken up by the recent gale and had been depleted by the depredations of the whales. Krill patches are by no means exclusively a feature of the open sea. They have been observed in great profusion in deep embayments of the land, such as Cumberland Bay on the east side of South Georgia, and have commonly been recorded right inshore, sometimes for instance near the Govern- ment Jetty at Grytviken at the head of Cumberland Bay. Farther south at the South Shetlands (p. 41) Webster records their stranding right inside Deception harbour and I have seen them too (Marr, 1937) cast up on the beach in Admiralty Bay. Still farther south in De Gerlache Strait Bagshawe (1938) records E. superba in millions a yard or two from the shore near Andvord Bay. Nor is patchiness confined to the adult population alone. It has been seen in half-grown individuals, in still earUer adolescent forms and, judging from the enormous variation in catch-figures (p. 219, Tables 45-7) revealed by the random sampling of our nets, would appear to be a feature of all developmental stages of this species from hatching onwards. Perhaps the strangest situation of all in which the krill has been reported was in a wave-cut cavern in the 120-ft. high Mertz Glacier Tongue, into which Mawson (1942) penetrated for 120 ft. by boat in December 191 3. 'Looking down through the clear blue waters', he writes, 'many shrimps (krill) about an inch in length could be observed swimming about near the surface'. When looking down on a krill patch through a metre or two of water one gets the distinct and some- times alarming impression^ that one is seeing the bottom and that the ship has blundered into shallow soundings, and it may well be that a number of the alleged shoals and ' muddy ' patches reported by early navigators in southern waters were in fact sightings of swarms of this, or other closely packmg animals such as (p. 136) Grimothea, not far below the surface. Close inspection of a small patch at the jetty at Grytviken revealed it to be in a continual state of flux, constantly changing shape, expanding and contracting, elongating this way or that, partially dividing only to re-form and become one whole, but never actually breaking up, in a manner distinctly amoeboid. An admirable description of this phenomenon is given by Hardy (1935, p. 210) and is quoted in full in the section which follows. 1 Possibly this was part of the same patchy region encountered earlier at Station WS 540. * As Captain Cook's watch-keeper (p. 40) evidently did. THE OLDER STAGES 151 In daylight the presence of patches is often betrayed by large, sometimes enormous, congregations of birds. In one region of patchiness encountered off the south-eastern tip of South Georgia in February 1931 Gunther records a 'vast gathering' estimated to total some 5000 birds which in flocks of from 200 to 300 was spread over an area of about one square mile. It consisted mainly of Whale Birds, ^ but included numbers of Wandering Albatrosses, Black-browed Albatrosses, Cape Pigeons and the South Georgian Diving Petrel, Pelecanoides georgicus Murphy and Harper. Among other such gatherings Gunther notes the Sooty Albatross, Wilson's Petrel, Silver Grey Petrel, Antarctic Petrel and Snow Petrel, all the species just mentioned it will be seen, with the exception of the unidentified Whale Bird (or Birds) and the Diving Petrel, being recorded in the list of krill-eaters given on p. 128. Bruce (191 5) vividly describes a triple association of krill, birds and whales, 'myriads of Cape Pigeons and thousands of finners ', he saw between the South Shetland and South Orkney Islands in December 1892. The sight of these whales and birds . . . will for ever remain one of the most vivid of my Antarctic recollections. Whales' backs and blasts were seen at close intervals quite near to the ship, and from horizon to horizon, while Cape pigeons were tumbling over each other after small pieces of fat thrown over the ship's side, just as do fulmar petrels after scraps of whale fat in the northern hemisphere. These Cape pigeons were captured with an ordinary hand landing-net over the side of the ship in such numbers that our crew of forty-seven hands were furnished with a very full supply of 'sconce'. The sea was swarming with Euphausia. Matthews (1951) also refers to the vast numbers of birds that congregate to feast on the surface patches : As far as we could see in every direction the sea was covered with birds, mostly White mollies, sitting close together and all facing the breeze. Here and there among them were smaller flocks of Blue mollies and Wanderers, and the spaces were filled with immense numbers of smaller birds. Cape Pigeons and Shoemakers, while over all skimmed great flocks of little dove-grey Whale-birds, probably hundreds of thousands, their wings flashing as all the birds in a flock turned together like sandpipers over an autumn mud-flat. Almost identical congregations are mentioned by Cook (1777) and Forster (1777) vA\o note par- ticularly the vast numbers of ' blue peterels ' (Whale Birds) they saw as their vessels were approaching South Georgia. Discoloration of the sea by patches leads one to suppose that the individuals of a patch are so densely crowded that they must practically be jostling one another, close-range observation distinctly suggesting they are. In some notes he made on a patch seen in particularly clear water near South Georgia in January 1931, Gunther describes the krill as being densely packed in close formation and all headed in the same direction parallel to one another. Each animal was distinctly separated from its neighbour by an average distance of about two-thirds of an inch, although heads and tails were often overlapping. At the beginning of his description quoted on p. 154 Hardy provides us with another instance of how densely crowded they must be. 'For a whole day', he writes, 'there was a dense swarm, like a red cloud, of closely packed euphausians {Euphausia superbd) against the jetty at our shore station. There must have been thousands and thousands in a close swarm some four feet across'. These observations, particularly Gunther's, suggest a density that might well be of the order of at least one euphausian to the cubic inch. It is interesting, therefore, to record that in a photograph recently taken from a bathyscaphe, 610 m. down, Peres (195 8) shows unidentified swarming euphausians, ' Nuages d' Euphaiisiacea au voisinage du fond ', that might well it seems be packed as tightly as this. Moore's observations on the swarming of a species that was evidently (p. 137) either Thysanoessa inermis or Meganyctiphanes norvegica distinctly suggest a density of the same high order. Similar concentrations have been reported for other swarming plankton animals, VanhoflFen (1902), 1 Most probably f Matthews, 19296) Banks' Whale Bird, Prion banksi Gould, or possibly Prion desolatus Gmel., or both. JJ2 DISCOVERY REPORTS for instance, recording thick clouds of salps at the surface so densely packed that the individuals were actually touching one another (see also p. 154 below). In January 193 1 an attempt was made in R.R.S. 'William Scoresby ' to calculate what the natural density of a typical patch might be, the observations being carried out at Station WS 540 where as already mentioned (p. 148) an area of widespread and prolific patchiness was encountered. As will be shown later (p. 262), when a ship towing a stern net ploughs through a surface patch in daylight the patch may sink bodily away or split asunder to right and left, in either event leaving an avenue of clear water astern in which the net catches nothing, or next to nothing. In view of this behaviour the sampling in this instance was effected by means of a net, an ordinary i-m. diameter stramin net, which instead of being towed from the stern in the usual way was rigged to fish from the head of a forward boom (Fig. 16) projecting 12 ft. out from the port rail, enough warp being run out to allow it to fish alongside the vessel, about midships, and well clear of the side. With this arrangement, which may be referred to as lateral or alongside towing, it appears that the sundering patch, although not of course the sinking one, splits away directly into the path of the approaching net with the result that enormous daylight gatherings can be made, involving large animals which in the ordinary way avoid, or (p. 264) are caused to avoid, our daytime surface nets. Moreover, with a net so rigged, by towing on the edge of a patch and by hauling or slacking away on the guide rope, it is possible to manoeuvre the net directly into the patch without the latter having suffered any major disturbance resulting from the passage of the vessel. Towing thus at Station WS 540, on the edge of a surface patch undisturbed by the moving hull, the lateral net was fished for 36 sec. before emerging from the discoloured area and passing into the clear water beyond. During this brief operation no less than 144,976 krill were captured, the vast majority of them large euphausians of the staple whale food class. As a net of this diameter towed at ii knots will fish through a column of water approximately 28 cubic yards in volume in 36 sec, assuming it filters the whole of the water in its path.i the figure 144,976 represents a density of approximately 5200 to the cubic yard, or say one to every 8 cubic inches which, although considerably less than the apparent density, is broadly in keeping with appearances. It would mean that the krill in this instance were spaced at about 2-inch intervals. This, however, must be regarded as a con- servative estimate since it does not allow for the fact that an unknown but probably high percentage of these large euphausians must have been taking avoiding action and (see p. 258) were successfully eluding the net. In other words, the actual density may well in fact have been of the order of about one to the cubic inch as eye-witness accounts suggest. Crowding of marine animals on a comparable scale is known from other ecological niches. It is estimated, for instance, that there are 3000 Thoracophelia mucronata, a Californian beachworm, in a cubic foot of sand (Fox, 1957), Beklemishev (1958c) estimating that the long and narrow belts of Thalia longicatidata, sometimes seen in the tropics, are crowded 2500 to the cubic metre. Further evidence of close surface packing by the Antarctic krill and other euphausians is provided by instances of the choking of pumps by these animals, of their being washed on board vessels in stormy weather, of their being involved in mass strandings or discharged from ships' hoses while washing down. In northern waters, for instance, Lachambre and Machuron (1898) record an occasion when the pump used by the Swedish explorer Andree in the preparation of hydrogen for his balloon 1 In practice it would not do so. There is loss of water through 'out-thrust' and in this particular instance there must have been further reduction in filtering power owing to the gradual blockage of the meshes of the net as more and more euphausians were crowded into it. The figure of 28 cubic yards is, therefore, too high and the resulting estimate of 5200 krill to the cubic yard in consequence too low. Igarashi (1957) analyses the filtering rate of a plankton net mathematically, showing that it decreases directly with lapse of time and inversely with the speed of towing. THE OLDER STAGES 153 suddenly stopped because it had taken up such a quantity of ' shrimps '^ from the sea that all the cocks were choked up.^ Gunther again, in 'William Scoresby', records an occasion when during a rough night oflF South Georgia, with the wind almost at gale force and the sea running very high, large numbers of E. superba kept coming on board for a period of 10 min. They were live krill for he adds that in the darkness 'specimens were easily collected by reason of their luminescence', an observation suggesting this large species is perfectly at home in a turbulent and breaking sea, neither seeking, as some plankton animals are said to do (France, 1894; Schouteden, 1902), the tranquillity of the deeper layers nor finding in such rough conditions what Gushing (1951) has described as a 'physically uninhabitable' zone. The mounting catching power of our surface nets (p. 260, Tables 51 and 52) that develops as the waves get higher provides further support for this view. BOOM fiUlOE ROPE Fig. 16. Lateral towing on the surface from a boom. A mass stranding of Nyctiphanes australis, the most abundant euphausian of the Australian coasts and one that like the krill may occur on the surface by day so densely crowded that the sea is coloured red, is described by Dakin and Colefax (1940) in the following illuminating passage: This species occurs in our waters at times in 'mass formation'. Great shoals occur so that a bucket will function for capture as well as a tow-net. At night the effect is remarkable. Perhaps one of the most extraordinary occurrences of this nature was the night of September 14th, 1938, when such a shoal actually appeared within the limits of Sydney Heads and so became washed up in myriads on certain beaches inside the harbour.^ Crowds of people noted the luminescence. An amusing and odd effect was produced, since either the squashed Crustacea, or their secretions, adhered and continued to produce light on the feet of dogs which were paddling about and then running over the sandy beach of Watson's Bay at the time. 1 Almost certainly Thysanoessa inermis which Stott (1936) describes swarming in great profusion off the Nordenskjold Glacier, Spitzbergen. 2 The filter in the evaporator intake of the diesel-electric ship ' Ob ' was similarly choked by E. superba while passing through a dense swarm off Enderby Land early in 1957 (Beklemishev, 1958 c). ^ In northern waters Moore (p. 137) records the stranding of Thysanoessa inermis in almost identical circumstances. See too Webster's account (p. 41) of the stranding of E. superba at Deception Island and Wilson (1905; 19076) who reports its stranding on the pack, thrown up by wavelets breaking on the floes. 16 DM IS4 DISCOVERY REPORTS In our own waters M*^Intosh (1887) reports a mass stranding of Meganyctiphanes norvegica on what seems to have been an equally grand scale, an occasion when this species was washed up in such multitudes on the east coast of Scotland that miles of sand in St Andrews Bay 'were strewed with their bodies which receding wavelets left in streaks and curves ', Mcintosh adding in a footnote that when dried 'they closely resembled chaff, for which, indeed, the uninitiated took them'. On another occasion (Robertson, 1892) the surface waters of Upper Loch Fyne appear to have become so choked with M. norvegica that the Duke of Argyll had over 24 lb. cooked and served for breakfast ! Fisher, Kon and Thompson (1953), referring briefly to the swarming of M. norvegica in the Mediterranean, cite Lo Bianco (1902, 1903) who reports daylight concentrations on the surface at Capri so dense that fishermen, casting for bait, were taking it sometimes by the hundredweight. In the same paper Kon records its spectacular swarming in the port of Monaco, reporting an occasion in February 1952 when it was thrown up on the beach in such astronomical numbers that municipal carts had to be employed to remove the bodies. The intense surface crowding of the Antarctic krill and of other euphausians in both southern and northern hemispheres is paralleled among the aggregate forms of certain Salps, Sars (1829) recording an historic occasion when these animals, evidently Salpa fusiformis, appeared in such fantastic numbers on the surface off Bergen that in certain places where they were thickest boats had difficulty in putting out to sea. Yet another parallel is presented by shoaling fish. There was an occasion in January 191 3 when the Pacific herring {Clupea pallasii) crowded into an Alaska bay in such enormous numbers that hundreds of thousands, perhaps millions, became stranded, their bodies left piled in a solid mass, in places several feet deep, when the tide went out (Brongersma-Sanders, 1957). Similar catastrophes, involving the stranding of ' thousands of tons ' of dead fish, are well known on the coast of south-west Africa near Walvis Bay (Brongersma-Sanders, 1948; Copenhagen, 1953; Hart and Currie, i960). So far then as can be seen from the decks of vessels in daylight the krill are manifestly concentrated in dense patches on or very close to the surface of the sea, and although in the rather turbid water of the Antarctic nothing as a rule can be seen beyond a few metres down, it will be shown later (pp. 157-70) that it is largely to the surface zone, or to not very far below it, that the main concentrations of the whale food are in fact confined. Habits and behaviour In view of the relatively large size to which it grows, its vivid colouring, shoaling habit and pre- dominantly surface habitat, the behaviour and movements of E. superba can be studied at closer range and observed perhaps with greater accuracy than those of most other organisms of the plankton. Continuing his description of the krill swarm from which I have already quoted, Hardy goes on to say, They were all swimming hard and going round and round, sometimes in a circular course, and sometimes in a 'figure 8', but never breaking away from the one mass. The cloud would sometimes change shape, elongate this way or that (There appeared to be some guiding ' principle ' — almost as if there was some leader in command of the whole !). At times they would form into two such moving parties and one would tend to separate from the other, so that the swarm became dumb-bell shaped ; but as soon as the connecting link became of a certain thinness the one part would turn back and flow into the other to form one big swarm again. It was drawn into the whole like the pseudopodium of an amoeba ; indeed the whole swarm appeared to behave as one large organism. It was for the most part at the surface, but at times the whole would sink down almost out of sight to rise again. This would happen apparently spontaneously, or again happen if some sudden disturbance occurred, the approach of a boat for instance. They were so close to the pier, at times even below it, that one could look straight down on to them and observe them with ease. I put in my walking stick and stirred the whole swarm up quickly so as to scatter them in all directions; but within half a minute they were all back again in their old formation. THE OLDER STAGES 'SS In an unpublished report on the field-work of R.R.S. 'Discovery 11', the late Dr Stanley Kemp records that while the ship was making her way through the pack not far from Bouvet Island m October 1930 great quantities of krill were almost always to be seen and he goes on to say, 'The young stranded themselves in numbers on floes that we momentarily submerged in our passage, while the adults with greater activity often jumped clear of the water and landed kicking on ice some 10 or 12 inches above the surface'. Jumping clear of the water, presumably when disturbed, may be of commoner occurrence in this species than is generally supposed. It has not been recorded very often, probably because it is a thmg that would tend to escape notice except under the conditions of absolute calm that are to be found in the polar pack. I saw a remarkable exhibition of it while ' Discovery II ' was steaming south through the Ross Sea pack-ice in January 1936. I quote from a note I made at the time: On the evening of January 7 as we were steaming across a large pool on the northern edge of the pack we saw an unusual sight. In most directions there appeared every now and then a sort of fluttering in the water as if fine shot had been sprinkled on it. As we looked more closely we could see that the disturbance was caused by krill breaking surface rapidly or jumping clean out of the water. The phenomenon seemed to be taking place at scattered points over a wide area— probably all over the surface of the pool which was several square miles in extent. The evening was one of exceptional beauty with brilliant sunshine, the air being quite still and the surface of the sea dead flat and glassy. We do not know if such behaviour has been observed before but we rather think it would pass unnoticed except under such perfect conditions as we had that night. We did not ascertain what was causing the krill to behave in this fashion but thought they may have been disturbed by the harrying of fishi as there were no penguins about, or perhaps that they were simply enjoying themselves! In regard to the last suggestion it is interesting to record that Olsen (p. 132) also refers to daytime congregations of krill on the surface 'gambolling about'. Gray (i960) has spoken in similar terms of the jumping of salmon in open water, noting that while it has been regarded from time to time 'as an attempt to rid itself from external parasites. May it not also be an indication of a general feeling of joie de vivre or even a process of limbering up before the big race ? ' Dr T. J. Hart (personal communica- tion) has seen the phenomenon of the jumping krill on several occasions. He describes it as like rain falling on the water and says it can sometimes be heard. Possibly the most revealing of all eye-witness accounts of the habits and behaviour of this euphausian is contained in some rough notes made by Gunther while watching a swarm of krill near the Govern- ment Jetty at Grytviken in February 1931. I am indebted to these notes for the following paragraph. Round about noon a swarm of half-grown krill appeared off the north side of the jetty swimming at the surface. The day was calm with the sun shining brightly out of a clear sky, and at the time of the observations the tide was ebbing at a rate of about a third of a knot (about 30 ft. per min.) southwards under the jetty. In spite of the current the swarm as a whole showed no tendency to drift towards or under the jetty, but, heading upstream for several hours, maintained its position where first seen in the full glare of the sun. Indeed the krill seemed to have a definite preference for the sunlight, for there were none directly under the jetty or in the shadow it cast upon the water. One or two that did chance to get into the shadow quickly swam out again, and making headway without difficulty against the current^ regained their original position in the sun. On the south side of the jetty there were a few isolated individuals stemming the current head on. Some swam as far as, but no farther than, the shadow, and it seemed as though that constituted a barrier the krill deliberately avoided. Later in the 1 See next paragraph but one below. Predatory fish are known to cause such behaviour among other swarming euphausians, Bigelow (1926), for instance, referring to an occasion in the Gulf of Maine when dense shoals of Thysanoessa raschii were seen just below the surface, the shrimps ' breaking water' to avoid the pollock that were after them. It will be recalled, too (p. 132 note 3), that Nemoto (1959) has seen the supposedly bathypelagic Notokpis coatsi chasing the krill on the surface. 2 The italics are mine. l6-3 is6 DISCOVERY REPORTS afternoon numbers of Crocodile Fish {Parachaenichthys georgiantis) lurking near the bottom were seen to be preying on the swarm. The fish would leap up and dart at the krill overhead and although many were captured and eaten, others with a single kick of the telson would spring backwards 4 or 5 in. to avoid the fish, often breaking surface as they did so or leaping clean out of the water altogether. The backward kick is accomplished with astonishing speed and it is probably with this that the krill most frequently avoid approaching danger or obstacles to their forward way. I quote from Ommanney (1938): 'The "krill" is a creature of delicate and feathery beauty, reddish brown and glassily transparent. It swims with that curiously intent purposefulness peculiar to shrimps, all its feelers alert for a touch, tremulously sensitive, its protruding black eyes set forward like lamps. It moves forward slowly, deliberately, with its feathery limbs working in rhythm and, at a touch of its feelers, shoots backwards with stupefying rapidity to begin its cautious forward progress once again'. In another note on a swarm of adult krill he watched in the open sea off South Georgia Gunther mentions that the euphausians persistently avoided the ship's side, the whole swarm keeping a good 5 or 6 ft. away and never coming within 4 ft. of it. In the same note he states that when a lead on a line was thrown into the swarm it parted suddenly to right and left leaving an avenue of clear water some 5-9 in. wide, tentatively concluding that similar avoiding action might well be taken by krill of this size when touched by, or confronted with a clear view of, the warp and bridles of an approaching net. Clearly then, by the time it is half-grown, if not before, E. superha far from remaining a passive drifter, has on the contrary become a creature of great agility, powers of locomotion, purposeful intent and not a little awareness. It has been shown, for instance, that it can stem and without difficulty make headway against a current of considerable force, can gather enough momentum to break surface and leap surprisingly high out of the water, is by no means indifferent to sudden disturbances and incursions in its vicinity and is prepared and able to take active measures to avoid them. Indeed, when one considers its reaction to the approach of a boat for instance, or to the quick thrust of a marauding fish, or when suddenly confronted with or touched by a foreign body such as a lead on a line, it is not surprising that the proper interpretation of the catch-figures revealed by the random sampling of our nets should present a problem (pp. 258-84) of considerable magnitude. The principal facts that emerge from these eye-witness accounts are (i) E. superba quite early on in its development, already that is when half-grown, ceases to be strictly planktonic and begins to behave, particularly in well illuminated conditions near the surface, with much of the vigour and awareness of a nektonic organism, (2) the patch or swarm is a unit, complete in itself, that does not break up except under the influence of sudden or violent external stimuli, a unit that even so will quickly re-form, and (3) the individuals of a swarm, all it seems heading in the same direction, can maintain themselves, regardless it seems of surface drift, over a given position for hours — Hardy for instance mentions a ' whole day ' — on end. It will be recalled that Gunther's swarm 'appeared' off the jetty. It was not there in the morning. In other words it must have come from seaward, it being unlikely that it came from up-harbour where the water is foul, much contaminated with factory refuse and red with the blood of dismembered whales. Moreover, it must have moved in against the tide which was ebbing when it first appeared. All this suggests that the swarms we see on the surface, in other words swarms in early or late ado- lescence, or in their adult state, do, as swarms, move bodily about, not necessarily however helplessly under the influence of tidal streams or other more far-reaching water movements, but evidently under their own very considerable locomotive power. As Marshall (1954) has said, ' . . .the well-muscled, shrimp-like euphausiids and mysids and the smaller deep-sea prawns are almost certainly active enough to keep station in the face of moderate currents. Perhaps it would be better to place these organisms THE OLDER STAGES '57 in the microneckton, a group intermediate between the thrusting neckton and the feebler swimming plankton '. And where indeed but high among such a near-nektonic community are we to place the full- and even half-grown southern krill? Uniformity of individual action may well be fundamental to the existence of the swarm, for it is difficult to see how it could remain as such unless every member of it were behaving in exactly the same way — heading in the same direction/ changing or reversing direction in unison as witnessed by Hardy, and moving at the same speed. Without some such pattern of behaviour it is conceivable that the strong swimming krill would scatter and disperse in all directions so that the swarms would cease to exist and the whales dependent on them would either go hungry or be hard put to it to obtain an adequate meal. Vertical distribution and migration While it has been established through direct observation, going back (p. 42) a long way in history, that dense concentrations of the older stages of this species, variously described as patches, shoals, ' rafts ' or swarms, occur at or very close to the surface of the sea, there is no means of telling, except through the medium of the data provided by the catches of townets covering a wide bathymetric range, whether this congregating of the krill is a phenomenon exclusive to the surface zone, or whether it be one that may extend into the deeper waters beyond the range of human vision. These are important questions, for if it can be shown that the main concentrations of the whale food are confined to a relatively narrow surface zone, then we should have reasonable grounds for supposing that the whales themselves must be largely, if not exclusively, surface feeders, or at any rate that there would be little point in their ranging far into deeper water in quest of an adequate meal. The material available for study is derived from 2960 net hauls of which 774 were horizontal hauls ranging from the surface down to 250 m. and 2186 oblique hauls covering a bathymetric range extending from the surface down to a maximum of 2000 m. The majority of the observations were taken with the lOO-cm. diameter stramin net. In exploring the deeper strata, however, more especially the depths below 250 m., we made extensive use of the large 200-cm. diameter TYF, an apparatus with four times the aperture of the N 100 and one besides that quite often was fished for as much as up to twice as long or more. To express the catch of the large net, therefore, approximately in terms of that of the smaller, the following equation has been used, c = cTm, where C is the corrected catch, c the catch of the TYF, t its fishing time and T the standard duration of haul of the Nioo which (p. 59) was 30 min. All the nets concerned, with the exception of the uppermost (loo-o m.) oblique nets and the horizontal nets fished in the immediate vicinity of the surface (0-5 m.), were closed at the end of their period of tow. In working out the vertical distribution of the staple catch I have discarded as before (p. 65) all data derived from horizontal nets towed on or very close to the surface in daylight, since obviously such data if used would have led to a gross underestimate of the surface density of the older stages of the euphausian population. Surface nets excluded, in the several presentations of the data that follow no attempt has been made to show the actual depth (accurately measured for all nets below 5 m. by Kelvin tube or depth gauge) at which each individual net was fished. There has instead been a grouping of depths of tow into selected broad horizons. Thus, taking the oblique hauls as an example, nets fished at say 400-200 m., 390-250 m., 520-300 m., 480-300 m. and so on have been grouped together as having fished within, 1 An underwater photograph recently taken from the French bathyscaphe F.N.R.S. 3 (Houot and Willm, 1955), depth and date not stated, shows a tight group of about 100 euphausians, all but two pointed in the same direction. ijg DISCOVERY REPORTS or mainly within, the broad limits of the 500-250 m. layer. In the same way horizontal nets actually fished at say 15 m., 22 m., 30 m., 38 m. or 49 m., have been taken as having together fished in the layer 50-5 m. Where so many different depths are involved and with a mass of material so large this is the most convenient way of marshalling the data and presenting the broader features of the distribu- tion from the bathymetric point of view. The vertical distribution of the staple food of the whales (euphausians over 20 mm. long), based on the total gatherings of the oblique nets, with the data from the surface (0-5 m.) gatherings of the horizontal nets added for comparison, is shown in Table 27. Table 27. Vertical distribution of the staple whale food based on the grand total captured in the oblique nets fished between the surface and 2000 m. and in the horizontal nets towed on the surface. Horizontal data in bold roman type, oblique data in italics Depth (m.) Total catch Number of hauls Average per haul S-o 572.730 412 i»390 loo-o 555,388 1,622 342 250-100 423 255 1 500-250 756 238 3 1000-500 36g 58 6 2000-1000 1 13 — Table 28. Vertical distribution of the staple whale food in the principal regions of its abundance and scarcity. Horizontal data in bold roman type, oblique data in italics Weddell drift. South Georgia and Bransfield Strait East Wind drift West Wind drift Depth Total Number Average Total Number Average Total Number Average [m.) catch of hauls per haul catch of hauls per haul catch of hauls per haul 5-0 497.532 27s 1,809 66,656 59 1,129 8,542 78 109 511,281 goy 563 43,158 272 158 949 443 1 00-0 2 250-100 103 101 1 283 44 6 37 110 500-250 69 90 — 641 64 10 46 84 1000-500 352 40 8 15 14 -' 24 2000-1000 1 2 — — 3 It is clear from these figures, even from the data provided by the oblique nets alone, that there must be a heavy massing of the krill at levels between 100 m. and the surface, the enormously high average catch-figure for the surface (0-5 m.) zone suggesting strongly that it must in fact be to depths con- siderably less than 100 m. below the surface that the bulk of the population is confined. While Table 27 presents the total picture, based on the vast bulk of the material available from the total area in which the whale food (p. 60, Fig. 5 a) is known to range, a regional presentation of the data (Table 28) based on the same material, demonstrates even more emphatically that whether the numbers involved be large or small, whether richly populated regions such as the Weddell system^ or East Wind drift, or sparsely populated regions such as the West Wind drift, be considered, the emphasis is always on the surface zone as being the principal locus of the congregating krill. Owing to the rather wide bathymetric horizons through which they are severally hauled the oblique nets cannot of course reveal the finer points of the vertical distribution at levels between the surface and say 250 m. However, the horizontal nets so extensively used on the South Georgia whalmg 1 That is, the Weddell drift proper together with the Bransfield Strait and the South Georgia whaling grounds to which it penetrates. THE OLDER STAGES 159 grounds and in the Bransfield Strait in 1926-7 (Discovery Reports, Station List, 1929) provide abundant evidence that between these critical levels (Table 29) a sharp decline in relative density takes place immediately below the surface, a decline that becomes rapidly more accentuated with depth. From the horizontal data in fact it would appear that the krill do not exist in any substantial measure of abundance beyond an exceedingly narrow zone that probably does not go deeper than 5 or 10 m. below the surface, such deeper concentrations of them as are encountered being evidently only of minor importance and extending down to only 50 m. or so, or perhaps a little below. Table 29. Vertical distribution of the staple whale food based on the gatherings of the three-level horizontal nets worked round South Georgia and in Bransfield Strait Depth {m.) Total catch Number of net hauls Avera^ per hail 5-0 50-5 100-50 284,781 6,198 11,208 153 30 188 1,861 206 59 250-100 2,557 144 17 Table 30. Orders of abundance of surface and subsurface catches recorded in and below the East Wind-Weddell surface stream Subsurface catches night Surface catches night and day ' Percentage ' Percentage Order of total net total net abundance Number hauls Number hauls 0 92 25 581 73 I-IO 74 20 139 17 10-100 69 19 45 6 I 00- I 000 71 20 26 3 >I000 S6 15 5 I Totals 362 99 796 100 Although our townets do reveal the existence of small to moderate concentrations at various levels below 5 m., principally, however, at subsurface levels above 50 m., the occasions when they have done so are far from common. It will be seen (Table 30) that of the grand total of 796 closing net hauls made below the East Wind-Weddell surface stream only 76, or 10%, produced catches of over 10 individuals, only 31, or 4%, catches of over 100, the vast majority, no less than 90%, pro- ducing negative results or catches of i-io. The corresponding figures for the surface (0-5 m.) layer are, catches of over 10, 54%, of over 100, 35 %, and of i-io or negative, 45 %. Together, the 76 sub- surface hauls with over 10 individuals accounted for 21,191 krill, only 392 short of the total euphausian catch of 21,583 obtained below 5 m. From this it follows that 720 subsurface net hauls, 90% of the total, captured between them only about half a euphausian per net, substantial evidence of how com- paratively rare, or if not rare at least widely scattered, subsurface concentrations of the krill must actually be. Moreover such few minor to moderately large subsurface gatherings as we have recorded (Table 31) are by no means exclusively daytime phenomena, there being little significant difference in the frequency of their occurrence whether it be light or dark. It will be seen too that both average catch and frequency of subsurface occurrence is conspicuously high at levels above 50 m., although even in this critical layer, close though it be to the zone of manifestly high surface night concentra- tion, our subsurface nets have not very successfully or very often it seems, been recording the really substantial concentrations that could reasonably be expected there. i6o DISCOVERY REPORTS Table 3 1 . Staple whale food gatherings of more than ten recorded at all subsurface levels examined belozv 5 m. Night gatherings in roman type, day gatherings in italics, the number of night and day hauls made at each depth interval being shozm in the column on the right Depth (w.) Night 2364, 46, 24 Day 1237, 990, 697, 697, go, 43 Night 1 108, 984, 642, 266, 70, 50, 32, 25, 18 Day 3570, 1500, 450, 426, 348, 2S9, 216, 202, 140, 137, 134, 124, 94, 92, 55, 38, 32, 25, 23, 22, 17 Night 151, 71, 53, 37, 37, 36, 29, 27, 27, 19, 17, 17, 16 Day 510, 454, 370, 336, 163, 123, 78, 73, 73, 60, 50, 41, 37, 32, 16,12, Night 324, 189, 27, 25, 24 50-S 100-50 250-100 500-250 Day 22 1000-500 Night 271, 55 Day — 2000-1000 Night — Day — % Frequency Frequency Depth Average Average of occurrence of occurrence (m.) night yield day yield night day 5<>-5 174 163 21 26 100-50 49 58 H 15 250-100 4 13 9 8 500-250 6 — S — 1000-500 10 — 6 — 2000-1000 — — — — 14 23 65 137 146 193 104 50 34 20 2 8 Krill in small numbers are occasionally taken in the 70-cm. diameter vertical nets, the frequency of their capture as might be expected being highest in the 50-0 m., 100-50 m. and 250-100 m. layers. The vast majority of these vertical gatherings, however, consisting as they do merely of one or tvs^o individuals,^ are too small to provide any reliable information as to the vertical distribution and only two of relatively outstanding size, out of the huge total of vertical samples examined, can be regarded as throwing any light on the matter. They were obtained at Station 1720, where 91 were taken in the 100-50 m. net, and at Station 356, where 95 were taken in the 250-100 m. net. While there is no knowing how the krill were actually distributed within these layers, if, like the surface patches (p. 149), they had been disposed in shallow rafts, no more than say a yard or two thick, through which the nets passed for only a second or two, then the numbers recorded at both stations might be regarded as representing clear instances of subsurface concentration in some unknown but perhaps not incon- siderable density. Finally, it may well be that the subsurface scarcity of the whale food we seem to find is even greater than our nets reveal. There can, for instance, be no absolute certainty when shooting open nets at night that a catch purporting to come from some deeper level, say 50 or 100 m., was not actually made, or at any rate largely contributed to, at the surface itself, as the net, shot open, let us imagine into a heavy concentration of krill, was momentarily checked to ensure that it was streaming well away and clear of the warp before being paid out to the depth required. For if with the momentary dipping of a hand net several hundred euphausians can be taken from a surface swarm it is manifestly possible that in the second or two, or possibly even longer, that the much larger stramin net must be ^ Referring to the standing crop of zooplankton in Weddell East Foxton (1956) remarks that except in its young stages E. superba is too 'large and active to be caught regularly by the N70 V, so that the volumes from stations in this region repre- sent the standing crop of zooplankton other than the older krill'. THE OLDER STAGES i6i held at the surface before being paid away, considerably larger numbers might be captured, an unknown proportion of which might well be retained as the net, still maintaining the measure of fishing strain necessary to keep it clear of the warp, was lowered to fish in deeper water. The practice in oblique fishing of shooting nets in pairs, a stramin net (N looB) with a silk net (N70B) immediately above it, renders the first net to enter the water, the N looB, still more liable to error through surface fishing. For in this operation the stramin net would be shot and kept fishing on the surface for upwards of 30 sec. or a minute,^ and sometimes for even longer if the operation were bungled, while the N70B was attached to its closing mechanism, shot and got clear away. It may be noted too that the numbers of krill caught in a subsurface net, should it happen to be the lowermost of a flight of three or more on the same warp, could be suspect on the ground that the catch could have been made or contributed to, not only at the surface, but also, before the net reached its intended level, possibly augmented on every occasion when the warp was halted for the attachment of the inter- mediate nets. o 10 40 70 C 130 O 150 190- 220- 250 r AVERAGE CATCH DAY 500 1000 I500 I I I 150 14 71 7a 43 40 22 10 2 AVERAGE CATCH NIGHT SOO 1000 I50O 2000 Fig. 17. Diurnal vertical distribution of the staple whale food, the number of net hauls made at each 10 or 30 m. depth interval being shown in the column of figures in the centre. Diurnal vertical distribution The diurnal vertical distribution of the staple whale food is shown in Fig. 17 in which the histograms representing the average catches at all levels investigated by day and by night have been constructed exclusively from data obtained at stations where open horizontal nets on the surface and closing horizontal nets in the subsurface strata were fished simultaneously. Here again for the subsurface nets, as in Tables 27-29, there has been a grouping of depths of tow into selected broad horizons, in this instance horizons of 30 m. vertical range. The salient features of the diurnal distribution thus disclosed are (i) the great abundance in which the krill are concentrated in the surface (o-io m.) stratum at night, and (2) their apparent absence, or virtual absence, from this stratum by day. At first sight, supposing there had been no simultaneous deep-level observations to accom- pany the surface series, this would suggest that E. superba was a typical vertically migrating animal, rising to the surface at night and seeking the deeper water by day. However, we see that the average daytime subsurface concentration does not at any level, down to the deepest we have examined, even remotely approach the average night concentration at the surface, and this suggests three possibilities : (i) the older krill are migrating by day to deeper levels than are covered by our horizontal observa- 1 In only 36 seconds, it will be recalled (p. 152), more than 140,000 large krill were taken by the boom (lateral) net used at Station WS 540. 17 ,^35 DISCOVERY REPORTS tions, (2) they are going down by day in isolated patches or pockets, chiefly it would seem to levels between 10 and 40 m., and, becoming scattered about in their patches over a moderate vertical horizon, are presenting a more difficult target for our horizontal nets than they do when massed at the surface at night, all along it seems, the same relatively narrow horizontal stratum, (3) there is definite avoidance by the older animals of the surface net by day. (i) I think can be ruled out, our deep observations with obUque nets below the East Wind-Weddell stream (Table 32) having failed conspicuously to reveal any substantial daytime concentration below 250 m. The scarcity of the staple whale food at the surface during the daylight hours would therefore appear to be due to a combination of (2) and (3). I return to this matter on pp. 268-78, where the whole question of the nature and extent of such vertical migration as may in fact be going on, and the behaviour of the krill that take part, is reviewed. Table 32. Daytime vertical distribution of the staple whale food at depths below 250 m. Depth Total Number Average (ffi.) catch of hauls per haul 500-250 38 50 076 1000-500 12 ao o-6o 2000-1000 I 8 012 If E. superba were in fact a rhythmic diurnal migrator, with a pronounced vertical movement such as for instance Hardy and Gunther (1935, p. 240, Fig. 125) have demonstrated in E. frigida,^ we should expect to find it during the daylight hours in far greater abundance and far more frequently in such abundance at depths of say between 500 and 100 m. than our 243 day observations in this particular horizon (Table 31) reveal it to be. Of these 243 net hauls only six in fact produced catches with even moderate claim to significance (that is catches of between 100 and 1000), the largest of them numbering 510 individuals, the remainder, 97% of the total, producing an average yield of a fraction over two euphausians per net. Mackintosh (1934) refers to the diurnal variation in numbers of E. superba taken in the upper (loo-o m.) oblique nets in the following passage: This is a difficult species to deal with owing to its extreme patchiness and tendency to form shoals. Many of the big shoals have been seen at the surface during both the night and the day, but the deeper shoals and the more scattered individuals might undertake vertical migrations. For the estimation of the average per haul at different hours, samples containing over 1000 E. superba have been disregarded. This should ehminate the disturbing influence of heavy shoal catches and give some idea of the general behaviour of the species. The resuking curve suggests only a minor degree of diurnal variation.^ At St. 461 the majority seemed to remain near the surface, while those living at greater depths gave some signs of moving up and down. At Sts. 618-25 there was quite a marked diurnal variation. The explanation would seem to be that while a section of the population of this species undergoes some vertical migration, the greater part remains at the same level [near the surface], especially perhaps when forming shoals. Even if a section of the population does so behave I question if such a movement could be described as an aspect of the general behaviour of this species. For in view of the lifelong swarming habit^ to which the krill (pp. 219-40) seem prone, a habit involving not merely isolated or occasional groups of individuals but it seems the entire euphausian population, it is now clear that to ignore the shoal 1 Even this movement Gushing (1955) suggests may prove to be more apparent than real, it being possible that this species also avoids the surface net by day. * Referring to various papers by M. E. Vinogradov, Bogorov (1957) states that 'it may be considered as determined that species of the surface zone do not perform diurnal vertical migrations, or their migrations are very limited'. ' I have already called brief attention to the swarming habit and suggested that it may persist from hatching onwards (Marr, 1955). THE OLDER STAGES 163 catches is to ignore a fundamental characteristic of its habit in the sea. Such catches in fact, far from being a ' disturbing influence ', are on the contrary normal, representing as they do the closest approxi- mation to the fiatural abundance of the krill in the sea our nets can provide. Clearly they cannot be disregarded, but must be accepted universally as the cornerstone of any presentation of the data that may be devised to show the vertical movements of the population as a whole. During the South Georgia plankton surveys of 1926-7 Hardy and Gunther (1935, p. 237) found that in general, although by no means always, far more krill were taken on the surface by night than by day and concluded that this species was in fact a vertical migrator although its behaviour in this respect seemed very erratic. In summer and autumn dense swarms of Meganyctiphanes norvegica commonly appear on the surface in the Gulf of Maine and Bigelow (1926) remarks that this phenomenon does not ' correspond to the diurnal vertical migrations shared in by many copepods, because the appearances of Meganyc- tiphanes at the surface appear to be independent of the time of day. Therefore, the actual captures so far recorded do not indicate any definite phototropism on its part, positive or negative, although it is doubtful whether it could long survive the full illumination of bright sunlight'. Paulsen (1909) con- cludes that since Meganyctiphanes is a microplanktonic feeder, when it comes to the surface it does so to seek food, not, as Bigelow (1926) comments, as a direct response to temperature or the state of the light. Bennett (1931) remarks that in bright daylight the whale food lives 'at some distance down in the water, but in dull weather and at night-time comes up nearer the surface, often to the top. At times', he continues, 'large patches of "whale food" appear on the surface, giving the sea a red colour'. The last sentence, of course, refers to daylight patches and it cannot perhaps be too strongly emphasised that it is not only in dull weather that such patches have been recorded, but also in brilliant sunshine (p. 155) to which the krill show no aversion, but on the contrary seem partial to a degree. It may be observed, too, that ' some distance down ' unless it be a statement based on assumption, would imply that the euphausians are visible at that level and therefore however bright the sun must still be rather close to the surface. A Norwegian gunner of long Antarctic experience once told me that, although he too had never seen krill on the surface in brilliant sunshine, when the sky was overcast he had often seen patches there. It may be, therefore, that in general the patches are easier to see when the weather is dull and that reflection of bright light from the surface tends to mask their presence there. ^ As Hart (1959) has recently said of blooms and surface discoloration in general, the overall visual effect is modified by the quality and direction of the incident light (sun's altitude) and by prevailing weather conditions such as cloud and state of the sea. While serving in U.S.S. ' Henderson ' in the New Zealand-Australian sector of Antarctica in 1947-8, Dietz (1948) found that the Deep Scattering Layer disappeared completely for 3 days after this expedition, southward bound, had crossed the 62nd parallel, and 'was rarely ever well developed in Antarctic waters'. In the same region a year later the French expedition in the 'Commandant Charcot' (Douguet, 1950; Tchernia, 1950) lost all trace of the layer after crossing the same latitude, repeated observation to the southward failing to reveal ' d'autres echos que I'echo du fond '.^ If the scattering layer, as has so often been suggested, is in fact caused by vertically migrating marine organisms these observations would seem to provide further evidence that the krill are never very far away from the surface of the sea. For if they were to sink daily in their massed fortnations to some deep subsurface level surely a pronounced indication of this phenomenon would have been detected by now on a modern echo trace. We cannot, of course, as yet say for certain whether these animals, how- 1 Gunther (p. 150) also records how conspicuous they can sometimes be under a grey sky when the sea is grey too. 2 LaFond, it will be recalled (p. 73), calls attention to the absence of acoustic scattering layers below the Arctic pack. 17-2 i64 DISCOVERY REPORTS ever densely crowded, would in fact act as scatterers of sound. Marshall (195 1), for instance, thinks that bathypelagic fishes, with their strongly reflecting swim bladders, are the primary agents, adding (Marshall, i960) that 'it seems likely that the most prominent of the fish sound-scatterers will prove to be myctophids'. Nevertheless, concentrations of euphausians have been found associated with scattering layers (Hersey and Moore, 1948; Saito and Mishima, 1953) and although small and lacking air bubbles they could well it seems be interceptors of sound beams especially if gathered tightly in swarms. The causes of scattering, however, remain obscure, and as Johnson (1957) pointedly remarks, ' Improved techniques of sampling with controlable gear are needed to solve the riddle'. Moore (1958) writes, 'From the physical aspects, fishes, with their highly reflecting swim bladders, seem the most likely [agents], although from the biological point of view much of the evidence points toward the euphausiids. . . . Whatever the source, these scattering layer records provide very valuable information concerning the details of diurnal migration '. The converse, that the absence of such records provides no such information, pointing it might be to an absence of diurnal migration, could equally be true, provided the phenomenon of deep scattering is of animal origin. Vertical distribution at Station 1835 Further evidence of the apparently incessant surface massing of the vast majority of the krill and of the absence in this species of any regular vertical movement is provided by certain operations that were conducted to investigate these phenomena in October 1936. The work to be described was carried out at Station 1835 in the western part of the Weddell drift, a locality which in October (p. 391, Fig. 133) is rich in swarms of 10 and 22 month old E. superba. The procedure was as follows: Beginning shortly after noon the vessel steamed seven miles along a south-west course towing one after another an ascending series of twelve i-m. diameter horizontal subsurface closing nets, the deepest at 250 m., the remainder rising step by step (Table 33) to a level 10 m. below the surface, the depth in each instance being measured by Kelvin tube. The towings were of 20 min. duration, each being accompanied by a simultaneous 20 min. horizontal towing (Table 34) with a net of the same kind on the surface. These simultaneous south-west-going surface and subsurface towings covered the hours of afternoon daylight and the twilight period of sunset and dusk. They will be referred to as series (a). On completion of this part of the programme the vessel was put about and, now in full darkness, steamed 4 miles back along the course she had taken, towing as before simul- taneous surface and subsurface nets, the latter, however, not rising but descending step by step from the 10 m. level to a depth of 128 m. The night towings will be referred to as series {b). In view of the extreme patchiness of the krill it is possible that had the ship continued along her south-west course on completion of series {a) she might have passed from an area rich in swarms to one where they were scarce or absent (or vice versa) and the resulting data, in so far as they could be used to compare night conditions with day, would have been valueless. To circumvent such a possi- bility, therefore, and to ensure that the population sampled by night was in fact the same as that sampled by day, the ship was deliberately put about, covering in series {b) a substantial part of the same water mass covered in series {a). For about a week prior to this investigation we had been striking dense swarms of young 10 month old krill both by day and by night in the surface waters through which we passed. Because of this certain precautions were taken in order to prevent, or at any rate reduce, contamination of the sub- surface samples with krill that might possibly (p. 160) have come from the surface, the long established practice, with its frequent warp stoppages, of fishing the nets in multiple flights, that is, several on the same warp, being abandoned in favour of a subsurface series shot singly, the nets in each instance being passed rapidly through the surface zone and paid away without interruption and at maximum THE OLDER STAGES 165 possible speed to their required level. Even so it will be seen such measures as were taken were not enough to prevent some sampling by the deeper level nets of both young and old swarms at the surface as soon as full darkness set in. Table 33. Measurements of the krill in the 10 and 22 month old swarms sampled by the subsurface closing nets at Station 1835, older swarms in roman type, younger swarms in italics Series ... (a) (b) State of light Depth (m.) Length range in 2-mm. groups No. measured Total sample No. measured Total sample 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 Broad daylight Sun- set Dusk Full darkness 250 210 180 164 150 128 84 77 40 30 20 ID 20 40 46 66 128 2 5 10 34 84 91 19 2 3 4 2 2 1 3 5 4 5 3 1 2 6 9 3 1 1 1 6 19 30 4 4 12 26 23 25 9 I 2 I I I I 3 4 8 10 23 16 10 6 4 I I I 9 13 20 21 21 16 16 12 8 3 4 16 36 23 21 4 2 I 3 3 4 247 4 g8o — 104 11 11 24 21 61 24 21 61 144 109 23 2 6 8 6 7 8 3 3 3 834 706 144 109 23 1 1 46 46 The numbers and measurements of the krill taken in both surface and subsurface nets at this station are given in Tables 33 and 34, which show that the young 10 month old swarms at least seem to be concentrated permanently in the surface (0-5 m.)^ stratum whatever the condition of the light. It will be seen, however, that the gatherings obtained late on in dusk and in full darkness are con- spicuously larger than those obtained under conditions of stronger illumination, and this no doubt is due to the measure of avoiding action the krill, even at this early stage of their development (p. 262), can take of the daytime surface nets. It will be seen too that while the young swarms were virtually absent from the daylight, sunset and dusk subsurface hauls (Table 33, series {a)) very small samples of them were apparently taken at 20 m. and deeper after full darkness had set in. This, however, might 1 The large catch that purports to have been taken at 10 m. in full darkness (Table 33, series {b)) also probably came from the 0-5 m. level, the net, owing to a miscalculation in shooting, having fished on or very near the surface for some minutes before the warp could be adjusted to its proper angle. i66 DISCOVERY REPORTS well in fact be only an appearance since in the darkness, in spite of all precaution, some euphausians might readily have got into the nets as, shot open, they were being passed through the densely popu- lated surface zone. Table 34. Measurements of the krill in the 10 and 22 month old swarms sampled on the surface at Station 1835, older swarms in roman type, younger swarms in italics Series . { «) I State of light Broad daylight Sun- set 0 Dusk Full darkness A Depth (m. r 6 8 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 _ 2 4 — 2 11 1 — — — — — — 10 6 30 6 13 .S 1 10 6 6 22 8 2 1 2 4 — 1 3 12 10 51 10 50 7 3 21 15 14 46 9 18 2 4 10 6 4 lb 14 22 77 7 73 7 2 30 22 22 105 lb 73 15 19 21 26 10 23 16 26 64 7 76 2 2 16 12 15 69 17 ibi 25 44 25 59 2b 64 18 17 55 6 85 2 1 11 7 8 62 14 2y2 40 62 30 68 56 92 20 1 14 — 25 1 — 1 2 3 16 — 100 15 18 11 12 4^ go 22 1 2 — 2 1 — 1 — 1 5 — 12 2 1 — — 19 31 24 — — — 1 — — — — 3 6 — — I — 2 I — 3 26 — 2 — — I — — — 7 19 — — 5 — 12 2 3 3 28 — 4 — — 4 — 2 — 17 57 — I 8 4 16 22 9 3 Length range in 30 — 8 — — I — 2 — 29 79 — — 17 12 22 35 16 7 2-mm. groups 32 — 26 — I 2 — 4 — 39 49 — — 25 15 II 40 32 35 34 — 12 — — 2 — — — 25 44 — — 34 19 10 32 23 43 36 — 8 — — — — — — 6 13 — — 8 20 II 10 17 33 38 — I — — — — — — 6 4 — — 3 14 8 9 15 23 40 — — — — — — — — — — — — I 13 2 — 10 9 42 — — — — — — — — — — — — I 2 4 — 3 7 44 — — — — — — — — — — — — I 2 2 — 4 4 46 — — — — — — — • — — — — — — I — — I 3 48 — — — — — — — — — — — — — I — — 2 2 SO — — — — — — — — — — — — — — — — I 4 52 — — — — — — — ' — — — — • — ■ — ~ ~ ~~ 54 — — — — — — — — — — • — — — ■ J6 — — — — — — — — — — ■ — — — — — — — No. measured 83 295 36 325 25 9 94 64 71 336 65 638 100 150 101 l^l lb 4 322 Total sample 83 295 36 325 25 9 94 64 71 336 65 638 783 logo 1506 1333 1314 1300 No. measured — 61 — I 10 — 8 — 132 271 — I 104 103 100 153 136 176 Total sample — 61 — I 10 — 8 — 132 271 — I 416 1503 789 1229 1084 2800 Other Discovery stations, notably Stations 207 and 461, at which consecutive nets were used, reveal an equally pronounced massing of the young krill at the surface regardless of hour or light (Hardy, iq;5; Fraser, 1936, Fig. 74). At Station 461, however, there seems to have been marked con- tamination of the subsurface samples (Fraser, 1936, Appendix II), again significantly the night samples, owing to the multiple net-flights used. The diurnal vertical distribution of the older swarms is evidently very much the same as for the younger except that the day and night massing of the former near the surface is much more difficult to demonstrate simply because the larger krill (p. 262, Table 53), in broad daylight especially, tend to avoid the surface nets much more easily than the smaller are able to do. It seems, however, that it must inevitably be concluded from the complete subsurface void revealed by the closing nets from daylight to dusk (Table 33, series {a)) that the 22 month old swarms must in fact have been present on the surface, not only from sunset to darkness (Table 34), but from beginning to end of the opera- tions conducted on this occasion, and that the negative, insignificant or minor catches recorded on the THE OLDER STAGES 167 surface from daylight to dusk (Table 34, series (a)) are simply an expression of the lessening measure of evasion that the older krill can achieve as the strength of the daylight wanes (see also p. 259, Tables 50 and 54). The large to very large catches obtained on the surface as soon as full dark- ness set in (Table 34, series (b)), obtained, it may be emphasised again, in the same v^^ater traversed by the last sbc hauls made in the waning hours of daylight before the ship was put about, provide convincing evidence of the extent of the evasion that must have taken place during the daylight, sunset and twilight hours. Turning again to Table 33, it will be seen that it is only in full darkness, and not, as might have been expected, by day, that there seems to have been some concentration of the I o 40 70 100 250 -• YOUNG SWARMS DAY 500 1000 13 3 a 6 V I 10 £ 40- g 70- 100 250 OLD SWARMS DAY 500 1000 13 3 2 6 YOUNG SWARMS NIGHT 500 _ 1000 ~l7 OLD SWARMS NIGHT 5O0 _ 1000 "?} 2 2 1500 10 40- 70 100- 250 TOTAL STOCK DAY 500 1000 13 3 2 6 I 10 S 70 100 250 TOTAL STOCK NIGHT 500 _ 1000 7J 2 2 Fig. 18. Diumal vertical distribution of the 10 and 22 month old swarms sampled at St. 1835, the number of surface and subsurface net hauls made in dayUght and darkness being shown as in Fig. 17. krill at depths immediately below the surface, ^ such concentration as was recorded, however, decreasing rapidly below the 20 m. level and virtually disappearing it seems at about 50 m. The complete subsurface void, however, encountered between 50 and 10 m. from sunset to dusk again places the reliability of these night subsurface samplings in some considerable measure of doubt, a critical assessment of their value again having to take account of the obvious difficulty of sending an open net through a dense surface population to a deeper level and ensuring that in the darkness it does not sample some of this population on the way. The diurnal vertical distribution of the first and second year krill taken at Station 1835, based on the figures given in Tables 33 and 34, is shown diagrammatically in Fig. 18 in which the histo- grams representing the average gatherings at subsurface levels are again, as in Fig. 17, based on a grouping of depths of tow into selected broad horizons. Ahhough Fig. 18 would clearly have carried more conviction had it been based on heavier daytime subsurface sampHng there are, never- theless, enough day observations below 10 m. to postulate that the apparent scarcity of both young and old broods at the surface by day was not due to their having been concentrated at this particular station at some deeper level between 10 and 250 m. If they had been it is surprising that not one of the eleven daytime subsurface nets, and notably not one of the six fished below 100 m., gives the slightest indication of their presence. It might again of course be objected that since our day obser- vations did not on this occasion go deeper than 250 m. it was somewhere below this level that the ^ The large catch at 10 m. (p. 165, note i) having probably been obtained at the surface itself. i68 DISCOVERY REPORTS krill were massed throughout the daylight hours. Our deep observations generally, however (Tables 27 and 28), provide ample evidence that they are never so massed, least of all (Table 32) by day. The accumulated evidence then from all sources reveals a heavy surface or near-surface massing of the krill, and that while its absolute vertical range may extend from the surface down to about 1000 m., in so far as it may be said to exist in concentrations dense enough to satisfy the needs of the whales, it is confined, especially at night, to an extremely narrow surface zone, no thicker it seems than some 5 or 10 m. By day, while some of it is on the surface, as both our eyes and nets so distinctly tell us, some of it too, principally it seems the older animals, is perhaps scattered about in patches chiefly at levels between 10 and 40 or 50 m. From this it would follow that the whales themselves must be largely surface or near-surface feeders, and that whether they normally range far into deeper water or not, it must be principally from the surface that these great animals gather the ample harvest necessary for their growth and development. On the assumption that heavy concentrations of krill, ' such as would concern the whale more than stray individuals', are found 'anywhere between the surface and 100 m. ', Laurie (1933) suggests that the lower limit of a blue whale's normal vertical range, in so far at least as feeding is concerned, may be 100 m. Even this estimate, however, again in so far as it concerns feeding alone, now seems unnecessarily deep. Referring to the diving and underwater feeding of penguins, Roberts (19406) states that it is improbable that either the Macaroni or Ringed Penguin, the latter a prodigious krill-eater, normally goes deeper than 5 m. This again suggests a massing of the krill at the surface and that in so far as its predators are concerned there is simply no need for them to go deep. In their study of the histological structure of cetacean lungs Haynes and Laurie (1937) find that alveolar epithelium is absent in the krill-eating fin, humpback and southern right whales, and suggest that the protection afforded by it is perhaps unnecessary to animals living in a dust-free environment. ' It is at least remarkable ', they add, ' that the thin membrane of the capillary can be subjected to the pressures involved in deep diving . . . without the protection of a lining of epithelium to minimize the danger of rupture'. It would not, however, be so remarkable if deep diving were unnecessary. Further evidence of surface feeding by baleen whales has recently come to light from a remarkable source. I quote from Chittleborough (1959 a) : ' On January 29, 1954, a member of a whaUng expedi- tion threw into the Southern Ocean a dentifrice tin containing a message bearing his name and address. The position of the vessel was then close to 64^° S., 92° E. On June 27, 1954, this tin and message was recovered from the intestines of a humpback whale which was being processed at the land station near Albany, Western Australia'. Since this obviously must have been a sealed tin it must have been swallowed at the surface. Moreover, it must have been swallowed in the Antarctic since it was released in a position well inside the krill-rich East Wind zone, from which it is virtually impossible that it could have drifted north, unless perhaps it had chanced to get into the cold krill-carrying surface stream that (p. 58, Fig. 4) flows north-westwards in this locality. An early, perhaps the earliest, account of surface feeding by baleen whales^ is recorded well over two centuries ago by Dudley (1725), who on the authority of an experienced whaleman of his acquaintance, tells how the North Atlantic right {Eubalaena glacialis) in still weather is seen ' skimming on the Surface of the Water, to take in a Sort of reddish Spawn, or Brett, as some call it, that at some Times will lie upon the top of the Water, for a Mile together'. Or to quote Collins (1886). 'They were all Finbacks, so far as I could tell. Their movements were sluggish, as they "played" back and forth in the tide rips, with their mouths open, the upper jaw just at the surface, scooping in " feed " '. Blue, fin, humpback, right (several species), sei, Bryde's and minke whales have all been reported feeding on the 1 See, however (p. 215), the remarks of Sagard-Theodat (1632). THE OLDER STAGES 169 surface.! In fact, as far as appearances go, of the known species of large, or moderately large, baleen whales only the Californian grey it seems could be described as a deep, or at any rate moderately deep, feeder, stomachs of this strictly neritic species having been reported to contain material from the bottom, amphipods, polychaetes, hydroid polyps, Buccinum, mixed in some instances with tiny pebbles and silt (Zenkovich, 1937; Tomilin, 1954).'^ It may be concluded too from this bathymetric survey that there can be no mass descent of the krill in winter from the mainly north-west-flowing Antarctic surface water into the mainly south-east- flowing warm deep water, a descent such as for instance Ommanney (1936) and Mackintosh (1937) have shown is annually undertaken by Rhincalanus gigas, Eukrohnia hamata and Calanus acuttis and is to be supposed as Mackintosh remarks to result in a large scale circulation whereby these important Antarctic species continue to maintain themselves within the limits of their normal distribution, their wintering in the counter-flowing deep current compensating for their northward drift at the surface. For as Table 35 shows, whatever the season of the year, and least of all in winter and spring, the older stages of the whale food are not encountered below 250 m., that is to say in the warm core of the deep current, except in negligible numbers. And even these it seems (p. 160), or at any rate some of them, might well have been captured at higher levels. Table 35. Seasonal abundance of staple ivhale food at depths below 250 m. Spring Summer Autumn Winter (Sept.-Nov.) {Dec.-Feb.) {March-May) (June-Aug.) Total Number Average Total Number Average Total Number Average Total Number Average catch of hauls per haul catch of hauls per haul catch of hauls per haul catch of hauls per haul 72 70 I 529 III 4 494 98 5 31 30 I Even supposing this species did in fact make a mass annual descent into the counter-flowing deep current the obvious question would arise as to what, it being a voracious herbivore, it would find to eat. It might not of course require to eat, or could exist on very little, but if it required to feed sub- stantially it seems doubtful if the rain of dead or dying phytoplankton from above would be enough to maintain it, unless it should prove (p. 355, note 2) to be a carnivore then. That the krill should be concentrated, as they manifestly are, so close to the surface is perhaps not after all very surprising when one considers that it is there, where the light intensity is strongest, that the diatoms on which they live are themselves most heavily massed. In northern European waters Gran (1912) concluded that 'the light optimum for far the greater part, if not for all, of our assimilating plankton algae is situated close to the surface, probably not as deep as 10 m., and we might perhaps even venture the assertion, that algae occurring in our latitudes with maximum deeper than 30 m. have never any optimum of development down here, but are in a relatively stag- nating period of life, either because they are removed from their true home, or because they have their period of growth at another season'. In conclusion, it may be observed, the foregoing takes no account of the ' escape factor ' (pp. 258-68) involved when sampling the surface layer, a factor operating by night as well as by day. If it were possible to put a figure for this, the sudden decline in density that has been shown to occur im- mediately below the surface would manifestly be revealed as a far more striking phenomenon than our nets are able to show. 1 Andrews (1909, 1916); Ingebrigsten (1929); Hjort (1933); Gunther (1949); Nemoto (1957, 1959) et al. ^ Among the large Cetacea as a whole the sperm whale is probably the only other species that on occasion visits the bottom to take demersal animals (Clarke, 1956). 18 DM 170 DISCOVERY REPORTS I have described the vertical distribution, and vertical movements, of the krill as I believe them to be revealed from our wealth of bathymetric data. Professor Hardy, with whom I have been in close contact, is inclined to stress the vertical migration rather more than I do and since his views involve an understanding of the whole phenomenon of swarming, which has yet to be described, I present and discuss them on pp. 268-78 at the end of the section dealing with the reaction of these animals, both as individuals and swarms, to the intrusion of ship and nets. Absolute density Although our townets (p. 266) cannot reveal it, some, if only approximate, estimate of the natural density of the krill in the sea can be formed (i) from the apparent density in a patch, and (2) from the number and approximate dimensions of surface patches observed in a given area. During her crossing of the Weddell Sea patchy region described on p. 150 observers in R.R.S. 'William Scoresby' recorded the approximate dimensions of 77 patches scattered about an area estimated to measure about 18,000 by 1000 yards. Assuming that all the patches there were actually accounted for, the density of the surface population could be expressed by the following equation, where v is the total volume of the 77 patches, obtained from the product of their area and thickness, n the number of krill in a cubic yard, and V the total volume of the surface stratum in which the patches were scattered about. The 77 patches covered a total area of approximately 13,010 square yards, their total volume, assuming each to be i yard thick, being approximately 13,010 cubic yards. Assuming a value of 46,656 (that is, one to the cubic inch) for n and that the krill instead of being concentrated in patches were distributed broadcast evenly throughout the whole i-yard thick surface stratum in which the patches were disposed, then ^ 13,010x46,61:6 , . 11- 1 D = ^ ~ — ^ = 34 euphausians to the cubic yard. 18,000x1000 J-r t- J This represents, in the immediate vicinity of the surface alone, some 140 million euphausians in a square nautical mile of patchiness, enough perhaps (p. 145) to provide more than fifty average-sized whales with a stomachful each. The estimate is conservative since some of the patches, especially towards the farthermost limit of vision,^ must inevitably have been missed. The weight of 8800 krill taken from a swarm on the southern edge of this patchy region (Station WS 558), drained of its formalin fixative and with the surface moisture removed by lightly drying between two layers of blotting paper, was 540 g., a measured sample (Table 36) showing that the swarm consisted of half-grown, approximately i-year old, euphausians, principally of the over 20 mm. or staple size. Assuming the same length frequency in all 77 swarms recorded, and taking the number of euphausians below every square yard of surface to be 34, then, expressed in terms of grams per square metre, 540x34x1-196 2.coe/m2 8800 5 ^'' ' which works out at 8502 kg. per square nautical mile. Now, while it is true that in this small part of the Weddell drift there was a most exceptional display of visible swarms on the surface, there is no reason to think that the actual density of the ' Throughout the 9-mile crossing of the patchy region the limit of effective vision was put at 500 yards to port and 500 to starboard. Length No. of {mm.) individuals i8 — 19 I 20 7 21 12 22 i6 23 36 24 24 THE OLDER STAGES '71 population there was of only local significance, or to suppose that our vessel was sampling and observing a population any richer than, for instance (p. 61, Fig. 56), that of the East Wind zone. The density estimates, therefore, could well I think be representative of the East Wind-Weddell surface stream as a whole. It is interesting, therefore, to compare them with the corresponding figures for whale flesh— 0-56 g./m.^ and 1908 kg. per square mile— recently given by Mackintosh and Brown (1956) for their inner (1° C isotherm to ice-edge) Antarctic zone which itself coincides broadly with the limits of the East Wind-Weddell stream. Table 36. Measurements of a sample of half -grown krill from a surface swarm at Station WS 558 Length No. of (mm.) individuals 25 23 26 15 27 12 28 4 29 3 30 — The density of the krill then from these figures would appear to be about 4^ times greater than the density of the whales, which does not in fact seem very much when one considers the incomparably larger number of other animals that also prey upon this species. Our figure, however, although a step in the right direction, may well fall far short of the natural density of the krill in the sea, it being based on the average weight of a single sample of immature euphausians that would not in nature be representative of the surface population as a whole. In February (p. 397, Fig. 136), when these observations were made, the surface population of the staple class consists of half-grown and adult krill assembled in a heterogeneous system of swarms in which the length frequencies of the individuals vary greatly from one swarm to another, the variation even in adjacent swarms (p. 234, Fig. 44) being sometimes quite pronounced. Among the half-grown swarms there would no doubt be many in which the average weight per individual would be say o-o6 g., as in the sample from Station WS 558, but among the older swarms there would be equally as many in which the corresponding weight would be very much higher. Heyerdahl (1932) gives the following averages for krill ranging from 28 to 65 mm. ^^^^^^ Average weight Length Average weight (mm.) (g.) {mm.) {g) 28 0-19 50 074 35 0-28 55 ^'12 40 040 60 I "48 45 0-56 65 i-6i From these figures, and from the figure provided by the sample from Station WS 558, the average weight per individual of the summer surface population might, therefore, be put as high as 072 g. and so D would equal 072 x 34 x 1-196 = 29-28 g./m.^ or approximately 50 times the weight of whale flesh instead of 4I. These are rough calculations and I present them with great reserve. Even so I believe the second equation, Weight of krill _ Weight of whale flesh ^ ' will in the long run prove to be the closer to reality. I believe, too, that the average weight of my 8800 half-grown krill may in fact be much too low, either because I over-dried my specimens or l8-2 172 DISCOVERY REPORTS because they came from a sample fixed in formalin 27 years ago and were not taken directly from the sea. Ponomareva (p. 145), using fresh Thysanoessa inermis, and estimating the wet weight from specimens dried off on filter paper, gives 0-57 g. as the average wet weight in samples ranging from 15 to 26 mm. long, and using this figure, assuming it to be the same in the corresponding length range in E. siiperba, the density of the krill would work out at 23-18 g-jm^, approximately 40 times the equivalent weight of whale flesh in the East Wind-Weddell stream. My estimates for the absolute density may be wrong, but at least they cannot be wildly wrong in so far as even the lowest result, 2-50 g./m.^, is still substantially higher than what it seems might be the corresponding density for whales. It would obviously be wildly wrong if it were less. Furthermore, Mackintosh and Brown's admittedly rough estimate for the Antarctic whale population has been thought to be too low (see p. 145); and if it should prove to be larger than these authorities have computed, and Vangstein (1956) seems to think it might be far larger, then the krill population that supports it would have to be correspondingly large, which suggests that of my three estimates for absolute density the two higher are the least likely to be wildly out. Even these, however, in the long run may prove to be low, perhaps much too low. They may refer in fact only to the immediate vicinity of the surface, perhaps only to the top 5 m. For as yet we have been able to form no clear picture as to how many, how spaced and how large are such patches as probably exist below the surface, beyond, although possibly (pp. 268-78) not very far beyond, the range of human vision. We must also allow for the possibility that many patches may be more than i yard thick. Gunther, for instance (p. 149), mentions 'a metre or two' and Matthews (p. 149) 2 fathoms. If there are in fact many such patches then obviously our highest estimate for absolute density, even for the surface zone alone, is much too low. Food As both Barkley and Hart (p. 45) have shown, Fragilariopsis antarctica is the most frequently occur- ring and abundant organism in the stomach contents of the krill. It would appear in fact from Barkley's findings that this species constitutes their staple diet, the average preponderance of this over all other organisms in the stomachs being as Si-i to 18-9%. Many other organisms or remains of other organisms, are recorded by Barkley, and a list of these is given below. I take this opportunity to publish at the same time a second list compiled from some field notes made by Dr Hart in 1937-8. In gratefully acknowledging this valuable source of information, I would point out that Dr Hart himself did not at the time consider his results of much value, since they were based on the examina- tion of only forty stomachs. In view of Barkley's findings, however, they prove to be of great interest. Neither list refers to any particular stomach, both showing the total organisms, other than Fragilari- opsis antarctica, recorded in the combined stomach contents of the krill these workers examined. The organisms considered by both Barkley and Hart as next of importance in the diet are shown in bold type. Broadly the two lists agree, both seeming to emphasise that small, spineless diatoms are the principal constituents of the stomach contents. It is perhaps a little surprising that Barkley does not record Thalassiosira, but possibly his ' Kleine Coscinodiscen ' would include this species. In summarising his results Barkley states that the dietary of all sizes of the krill ranging from 10 mm. long, which would include both Fifth and Sixth Furcilias, to 60 mm. long, is essentially the same. In his notes Hart observes a tendency for small animals to feed more voraciously than large ones and for adult males to feed more voraciously than gravid females.^ The occurrence of Cocconeis ceticola in three of Barkley's samples, one of them with 115 cells, is of great interest. Hitherto this species had been known (Bennett, 1920; Hart, 1935; Peters, ^ Bargmann (1945) states that the gravid females are unable to feed. THE OLDER STAGES 173 Organisms recorded by Barkley CHRYSOPHYCEA Not recorded COCCOSPHERES Not recorded SILICOFLAGELLATES Distephanus regularis ( = speculum) DINOFLAGELLATES Not recorded DIATOMS Actinocyclus spp. Asteromphalus heptactis Asteromphalus parvulus Asteromphalus regularis Biddulphia polymorpha Chaetoceros atlanticum Chaetoceros criophilum Chaetoceros dichaeta Chaetoceros radiculum Cocconeis ceticola Cocconeis imperatrix Corethron valdiviae ( = criophilum) Coscinodiscus spp. Small Coscinodiscus Dactiliosolen antarcticus Fragilaria curta Fragilariopsis sp. Navicula sp. Nitzschia seriata Rhizosolenia alata Rhizosoletiia bidens Rhizosolenia hebetata Rhizosolenia spinifera Rhizosolenia styliformis Synedra spathulata ( = pelagica) FORAMINIFERA Globigerina sp. RADIOLARIA Acantharidae Phaeodaridae TINTINNIDS Cymatocyclus onvallaria Cymatocyclus nobilis CRUSTACEA Not recorded Organisms recorded by Hart CHRYSOPHYCEA Phaeocystis sp. COCCOSPHERES Pontosphaera sp. SILICOFLAGELLATES Distephanus speculum DINOFLAGELLATES Peridinium antarcticum Peridinium sp. DIATOMS Actinocyclus spp. Asteromphalus hookeri Asteromphalus parvulus Asteromphalus regularis Chaetoceros criophilum Chaetoceros chunii Chaetoceros neglectum Chaetoceros sp. Corethron criophilum Coscinodiscus minimus Coscinodiscus minutus Coscinodiscus spp. Small or minute Discoidae Dactiliosolen antarcticus Fragilaria ? striatula Fragilaria spp. Fragilariopsis sublinearis Navicula spp. Nitzschia closterium Nitzschia delicatissima Nitzschia seriata Rhizosolenia alata Rhizosolenia hebetata Rhizosolenia sp. Synedra pelagica Thalassiosira antarctica Thalassiothrix antarctica Diatom ? spores Globigerina sp. FORAMINIFERA RADIOLARIA Not recorded TINTINNIDS Recorded CRUSTACEA Copepodites ? Copepod eggs 174 DISCOVERY REPORTS 1938; Karcher, 1940) only from the skin-film of whales, in which it had for long been recognised (Nemoto, 1958) as the dominant form, and had not been recorded free in the plankton. In view of these findings Barkley suggests that the Cocconeis cells may get into the water accidentally, it being conceivable that they sometimes get brushed off the whale ' purely mechanically ' through physical contact with the krill. He adds, however, that it might also be supposed that, since the whales from which his samples were taken were infected with Cocconeis, both frustules and spores of this species may occasionally be liberated into the sea. Hustedt (1958) also reports the rare occurrence of C. ceticola in the stomach samples he examined from south of Kerguelen. Hustedt (1958) also calls attention to the frequency of occurrence of Fragilariopsis antarctica 'und einige kleine zentrische Diatomeen ', notably Thalassiosira gracilis and Coscinodiscus lentiginosus, in the stomachs, and adds the following long list to Barkley's original findings : Diatoms from krill stomachs collected south of Kerguelen, the regularly or more commonly occurring species being shown in bold type Asterotnphalus hookeri Micropodiscus oliveranus Asteromphalus hyalinus Navicula criophila Biddiilphia weissflogii Navicula trompii Chaetoceros bulbosus Nitzschia harkleyi Charcotia actinochilus Nitzschia bicapitata Cocconeis gaussi Nitzschia sicula var. rostrata n.var. Coscinodiscus lentiginosus Pleurosigma directum Coscinodiscus pseudodenticulatus n.spec. Rhabdonema minutum Coscinodiscus stellaris var. symbolophorus RhapJioneis amphiceros Coscinodiscus tabularis Rhizosolenia styliformis forma longispina Coscinodiscus tumidus Thalassiosira antarctica Fragilariopsis cylindricus Thalassiosira gracilis Fragilariopsis obliquecostata Thalassiothrix antarctica Fragilariopsis rhombica Tropidoneis belgicae Fragilariopsis ritscheri n.spec. Tropidoneis glacialis Hyalodiscus scoticus While the predominance of Fragilariopsis antarctica and the minor degree of dominance of some other small, spineless diatoms in the stomachs seem to point to some measure of selective feeding, the proper interpretation of these findings, as Steuer (1910) and Hart (1942) point out, is rendered exceedingly diflRcult owing to the great differences that exist in the degree of silicification of the frustules of the various diatom species. The majority of the more commonly occurring stomach forms recorded by both Hart and Barkley are strongly silicified, forms that tend to persist in a recognisable condition in faecal pellets, bottom deposits and bird guano, being encountered, for instance, with particular regularity (Chun, 1903; Karsten, 1905; Neaverson, 1934) in the diatomaceous muds and diatom oozes from the shallow and deep-sea floor. It may be, therefore, as Hart remarks, that the more typically oceanic, less strongly silicified, forms are quite as important as food for the krill, but as a rule are digested too rapidly and too thoroughly to be readily identified in the stomachs. Recognis- able fragments of the more poorly silicified oceanic species such as Chaetoceros are rather rare, he states (Hart, 1934), even in the crop, the contents of which in krill from water in which such species are dominant presenting the appearance of a 'green porridge'. In regard to the Chaetocerids he remarks that spine fragments of the large C. criophilim indicate that the adult krill are capable of triturating and swallowing the larger diatom species in addition to possessing a fihering mechanism, while the several references in his notes to pieces of Thalassiothrix and the immensely long Synedra would point it seems to the same conclusion. Hustedt (1958) records C. criophilum fragments THE OLDER STAGES 175 regularly in the stomachs, Thalassiothrix antarctica fragments commonly, and Synedra as present but rare. During a brief but very intense period of blooming of the diatom Stephanopyxis turris, in the UUadulla region of New South Wales, Sheard (1953) noted that many specimens of Nyctiphanes australis were taken with their ' food baskets ' filled principally with faecal pellets, apparently those of C alarms finmarchicus, with which the flowering StephaJiopyxis was associated. He found too that, in a small collection of euphausians brought back by the B.A.N. Z. Antarctic Research Expedition, in the majority of the specimens of £^. superba, E. triacantha, E. spinifera, E.similis, Thysanoessa macrura and T. vicina he examined, the chief constituents of the food baskets were again faecal pellets. These findings led him to suggest that, although confirmation was required from larger sources of material, a prime food source of euphausians in areas of high diatom production might well be the faecal material of the grazing herbivores. If it were so, however, in so far at least as E. superba is concerned, it is surprising that neither Barkley, with his wealth of Antarctic material, nor Hart, makes any mention of it. Marshall (1954), referring to the intimate correlation between form and function in aquatic animals and their ways of seeking, finding and collecting food, notes that E. superba with its fine thoracic filter basket and strongly developed grinding mandibles is better equipped for deaUng with a phyto- plankton diet than, for instance, is E. hanseni, an omnivorous species, existing probably on copepods and a variety of other small animals, in which the thoracic filter is coarse and the molar parts of the mandibles small. He notes too other striking anatomical diflFerences between the almost exclusively herbivorous Antarctic krill and euphausians of the omnivorous or rapacious kind. The problem of the re-solution of silica in the sea is as yet little understood (Barnes, 1957). Cooper (1952) points to the evidence that the skeletons of many though not all diatom species are resistant to solution and for all practical purposes are insoluble while intact. However, he calls attention to Hart (1934, 1942) on the feeding of E. superba, noting how thoroughly and rapidly this species seems to be able to digest many of the more fragile and poorly silicified forms, adding that this provides strong circumstantial evidence that re-solution of diatomaceous silica is a rapid process. He notes, too, that freshly fractured surfaces of quartz and some mineral silicates are in a highly reactive state, readily yielding 'soluble silica' on contact with water, and that in the sea much the commonest way of exposing fresh silica surfaces by fracturing would occur in the guts of herbivores, such as E. superba, which grind or triturate their food. It would appear distinctly possible, therefore, that the grazing of the multitudes of the krill con- tributes something to the maintenance and recruitment of the enormous concentration of silicate (Clowes, 1938) in Antarctic waters. The phytoplankton, on which the krill are so admirably equipped to feed, contains /^-carotene (Kon, 1958), and this (Fisher, 1958) they are able to convert to vitamin A and store, principally in their eyes. It is from this preformed source that the whales, especially the blue whales (Braekkan, 1948), derive the rich vitamin A content of their livers. Nemoto (1959) reports that in the summer of 1958 the vitamin A content of baleen whale liver increased in the East Wind zone from west to east, being low between 60° and 100° E, gradually rising between 100° E and 160° W, and reaching a maximum in the former sanctuary between 160° and 100° W. He suggests as 'a simple and bold explanation' that the vitamin A content of Thysanoessa macrura, on which the whales (p. 48) are known to feed on the Pacific side, is higher than that of E. superba. Other factors, however, could also be involved. The vitamin values, for instance, might be expected to increase with the advance of the season as increasing nourishment is taken by the whales. Many of Nemoto's high values between 160° and 100° W are in fact in 176 DISCOVERY REPORTS March, while other moderately high March values occur between 130° and 140° E. The position of the ice-edge could also play a part, in so far as it could prevent ready access to good feeding in some sectors and facilitate it in others. And so as Hart (1934) has said, this small herbivorous crustacean forms a link in 'one of the simplest food chains possible, the building up of the vast body of the whale being only one stage removed from the organic fixation of the radiant energy of the sun ' by the diatoms on which it feeds. Faecal pellets In the vertical samples I examined from the Weddell and East Wind zones I often found masses of green faecal pellets ' packed with diatom skeletons, principally Fragilaria ', their enormous abund- ance in some samples suggesting strongly they were the recent excreta of densely packed shoals of E. superba. They occurred in more or less equal abundance from the surface down to the deepest level (1000 m.) examined from which it seems distinctly likely, the krill itself being massed principally at the surface, they were sinking. Moore (1933) examined a number of these pellets from the Bransfield Strait and concluded they were almost certainly those of E. superba, adding that they agreed ' in form with those in the gut of the animal ' as well as with others he had already described (Moore, 1 93 1 ) from other euphau- sians from the Clyde. He notes, however, that no pellets of this type were observed in our Discovery bottom deposits and this he suggests is ' probably due to their quick breakdown as was the case in the Clyde '. Many too (Sheard, 1953 ; Vinogradov, 1955), before they have sunk very far, must be devoured by other animals. It seems unlikely, however, that they contribute directly to the food of deep-water animals since they probably disintegrate (Bogorov, 1957) before reaching the levels at which they live. Similar pellets containing Fragilaria (dominant), Coscinodiscus, Rhizosolenia and Distephamis speculum were found all over the Ross Sea shelf, some of the samples running to ' millions '. No doubt they were the excreta of the other krill, E. crystallorophias, which swarms at the surface there, and from their distribution and relative abundance from the surface downwards they too were evidently sinking. Distephanus speculum, which is also eaten by E. superba, I found to be particularly abundant in these high latitudes. Spawning and hatching The act of spawning The fact that the deepest concentrations of larvae we strike are so often very largely, sometimes indeed all, at the same stage of development (Tables 13 and 14) suggests that such concentrations may spring from batches of eggs that all hatch simultaneously or within a very short space of time. This in turn suggests that spawning itself may be a phenomenon involving the mass shedding of eggs by swarms of gravid females (p. 250, Fig. 53) within an equally short space of time. Such phenomena are not unknown. I quote from the last paragraph of Korringa (1957) on Lunar Periodicity. To analyse cases of periodicity in reproduction, the annual rhythm (the length of the breeding season), the monthly rhythm (periodicity correlated with tidal sequence or moonlight cycle), and the daily rhythm (concentrating spawning or swarming to a certain well-defined hour of the day, or to certain phases of the tidal cycle) must be clearly dif- ferentiated. In extreme cases these three rhythms in combination may resuh in a complete concentration of repro- ductive activities in the species concerned, so that an entire population may spawn simultaneously during one single hour a year. In so far as the krill are concerned I would, of course, for Korringa's entire population substitute entire swarm. In her classic work on the male and female reproductive system Bargmann (1937, p. 348) counted over 1 1,000 eggs each from two gravid females. Even so, she adds, 'the process of laying is evidently a rapid one '. THE OLDER STAGES 177 Time of spmoning The northern or Weddell zone. From the occurrences of the eggs in the Bransfield Strait and in the adjoining waters of the western Weddell drift Fraser (1936, p. 113) has shown that spawning in this locality must last from November to March and is not the ' spontaneous ', that is short-term, outburst of Ruud who in the year of his expedition to the Weddell Sea (p. 44) supposed it, as a major event, to have been confined to the first half of January. The material examined since Fraser wrote, again yielded eggs throughout the period November to March and it seems clear that in the northern zone at any rate this is the maximum extent of the spawning season, the large numbers of spent and gravid females recorded in February and March (p. 180, Table 37) suggesting it is probably at its cUmax there in February. The southern or East Wind zone. Although little information about the spawning in the East wind drift can be gathered from the occurrences of the eggs themselves, their numbers (p. 182, Fig. 19) being generally so small, the occurrences of Nauplii and Metanauplii, stages which do not appear (p. 301, Fig. 70) in substantial numbers until February, suggest that in this relatively narrow coastal belt spawning does not begin on any appreciable scale until well into January, is probably in full swing by February and lasts as in the northern zone until March. It may be, however, that in these high latitudes it continues into April, for as late as the 20th of that month, at a station worked within sight of the Enderby coast, I found a number of recently spent females in which there were a few unshed, unsegmented eggs still clinging to the walls of the otherwise empty ovary. The later spawning characteristic of the East Wind zone and its compression in time is no doubt associated with the relatively short and generally unpredictable period there during which a blooming phytoplankton is offered to the gravid or near gravid females, the large-scale production of plant life in these high latitudes being much curtailed and hampered by the ice-cover that extends over them for the greater part of the year. As Hart (1942, p. 314) remarks, From the known climatic and ice conditions it is obvious that large-scale production [of the phytoplankton] can only begin when the first large areas of open water are formed in January, and as new ice begins to form in March it follows that the annual production must be crowded into three summer months with no possibility of a secondary autumnal increase. Our observations fully bear this out, the main increase evidently begins very suddenly in January and rises to a high maximum (as the oceanic values go) in February. A few moderately high values have been recorded in the early days of March, but taking that month as a whole the falling off was most marked. In far northern waters a blooming phytoplankton seems to offer a marked inducement to spawning among certain phytoplankton feeding bottom invertebrates, Thorson (1946) calling attention to the large-scale spawning of these animals that coincides with a sudden and heavy increase in plant production in the East Greenland fjords, Grainger (1959) to a similar phenomenon among certain herbivores of the plankton, especially copepods, of Arctic Canada. Law and Bechervaise (1957) give a vivid description of the brief opening up of the East Wind ice- field (oflF Mawson in 64° E) engendered by the onset of the ' sudden summer ' in these high latitudes and the almost continuous sunshine that goes with it. Blizzards had tailed off very suddenly in early December and, with the period of the midnight sun, long successions of peerless, clear days had arrived. Sometimes a whole week passed with no moment, day or night, when the sunlight was not gleaming somewhere on the great seaward icebergs. . . . Under these circumstances the heat radiated from the dark rock is astonishing. Men may sunbathe as in the high alps, even when the air temperature is below freezing. Lichens make colourful growth and all minor snow drifts rapidly melt away. For this one brief period in the year, liquid water is not a phenomenon. The disappearance of the sea ice may occur with dramatic suddenness. There comes a day when the sound of the sea lapping on the rocks is first audible. Any considerable swell is now able to crack the rotten ice floes and almost overnight the whole scene changes. The sea is mobile. 19 DM 1^8 DISCOVERY REPORTS They note, however, that the break-up of the winter ice is unpredictable. In 1956 it occurred in early February, but in 1954 it did not break out until March. They note, too, that the water did not remain unfrozen for long. In 1956, even before February was out, 'a skin of new ice was already forming again and the rocks edging the sea were white with frozen spray and tide- wash'. Depth of spawning In his study of the vertical net samples brought back by the 'Norvegia' expedition of 1928-9 Rustad (1930) found that the Nauplii and Metanauplii of Thysanoessa macrura G. O. Sars were con- fined exclusively to depths below 200 m., the majority to the 400-200 m. layer which, with one exception, was the deepest to be sampled by this expedition. In view of these findings he advanced the hypothesis that T. macrura spawns below 400 m., the resultant larvae rising and eventually reaching the surface when so far developed^ that they are able to feed on the phytoplankton. Owing to the poverty of his material he was unable to come to any definite conclusion about the spawning of E. superba. He seems, however, to have considered it unlikely that it would be a deep spawner from his remark, ' It must however be kept in mind that E. superba with our present knowledge must be regarded as a typical surface form'. Ruud (1932) rejected this hypothesis, objecting that the numbers of deep Nauplii and Metanauplii Rustad found, 135 in all, were inadequate and, like the very small numbers of krill Nauplii and Metanauplii he (Ruud) found at the same level, could only be regarded as stragglers that had wandered away from a much larger population which, along with the eggs he said, would be found, if a proper search were made for it, close to or right under the drifting ice. In other words he substituted his own hypothesis that it would be there, directly under the ice, that T. macrura, like E. superba, would spawn. In postulating this hypothesis Ruud seems to have been much preoccupied with the assumption that the water directly under the ice is difficult or impossible to explore. As Rustad (1934) quite properly points out, however, although there is considerable truth in this assumption, it cannot always be upheld. For even if there zvere concentra- tions of eggs hidden away beneath the floes they could hardly be expected to remain permanently inaccessible there, since the floes must sometimes shift their position relative to that of the water, in other words must sometimes travel through, as well as with, the water on which they float. A great deal of exploration has in fact been done both close to and inside the pack, particularly by the vessels of the former Discovery Committee, but so far, as Fraser (1936, p. 112) remarks, has failed completely to reveal any indication of surface spawning by E. superba in this particular environment. In his later paper Rustad (1934) extended his hypothesis regarding the deep spawning of T. macrura to include E. superba. He writes, ' It seems, however, that the opposite supposition [i.e. deep as opposed to Ruud's supposed surface spawning], viz. that E. superba and T. macrura spawn below the Antarctic surface layer — is more easily brought in accordance with the known facts. In this case the state of affairs will develop as follows : The developing eggs will ascend, and the major part will hatch, e.g. between 400 and 200 metres, where the youngest larvae are accordingly to be found. The larvae continue developing and ascending, and they finally reach the upper layers when they are able to feed'. While he was undoubtedly correct in supposing the newly hatched larvae of both species to rise, his nets, however, not going deep enough for him to visualise with any accuracy the actual level from which the ascent would begin, Rustad's ascending eggs must be regarded as somewhat hypothetical, for, as he admits himself, his material, a mere 16 eggs altogether, was far too scanty to permit him to conclude that any such thing could take place. Indeed, as will be shown presently, 1 That is, having reached the First Calyptopis stage. THE OLDER STAGES 179 there is every reason to believe (p. 184) that in so far at least as the krill are concerned the eggs unquestionably sink. In December 1930, at Station 540, Fraser (1936, p. 18, Table III) records that 2496 eggs of £". superba were taken at a depth of between 500 and 250 m., the vertical net in which they were captured having fished to within approximately 10 m.^ of the sea bed in the shallow shelf water of the western Weddell drift where it turns eastward away from the northern tip of Graham Land. At the same station the 250-100 m., 100-50 m. and 50-0 m. vertical nets yielded 478, 52 and 4 eggs respectively. As he remarks later (p. iii) the vertical distribution of the catch in this locality suggests a concentration of eggs near the bottom, a concentration which, having regard to the very near approach to the sea bed of the lowermost net at this station, leads one strongly to suspect, as Fraser himself suggests, that spawning in this instance took place on the bottom itself, the occurrences of eggs at the higher levels being possibly due to vertical mixing of the water column which in this region, as in all shelf seas, is known he says to be particularly effective. Fraser in general is in favour of deep spawning, pointing out, as already remarked, that in spite of intensive search, both inside and outside the pack-ice, the observations covering every month of the protracted breeding season of this species, not a single concentration of eggs has yet been recorded in the top 100 m. of the Antarctic surface layer. He remarks, too, on the occasional occurrence, together with eggs, of fully adult males and gravid females in deep hauls of the vertical nets, concluding with the guarded statement that since they should be found deep down at the same time as eggs are taken it might be inferred that their presence at these levels is connected with spawning. Such evidence, however, as he is able to provide in favour of deep spawning from the vertical distribution of the eggs themselves, is not very satisfactory, since apart from the rich haul at Station 540 the numbers he found elsewhere, like the vast majority of those subsequently recorded by Fry and me, are negligible. ' It is remarkable ', he writes, ' that compara- tively few eggs have been found in the samples analysed, and although Rustad's suggestion of circum- scribed shoaling "and the consequent concentration of the eggs in relatively small areas" may be partly the reason I do not think that it is completely satisfactory. E. superba must be reproduced in immense numbers to hold the key position it has in the ecology of Antarctic life. The yield of the 70-cm. vertical nets, with the possible exception of those at Station 540, is surely not indicative of the normal concentration'. In his concluding remarks on the purely bathymetric aspects of this problem Fraser (1936, pp. 1 12-13), while pointing out that much more positive evidence was required before any definite solution to it could be put forward, refers briefly to the possibility that the eggs may go through the major part of their development at depths beyond the reach of our deepest vertical nets,^ suggesting that the negligible numbers he found above 100 m. might represent 'the scattered product of dispersal of a much greater mass situated in still deeper water '. It may be he suggests that the eggs when laid usually sink below 1000 m., adding that the great abundance in the 500-250 m. net at Station 540 was probably due to the fact that this particular net struck a place where the eggs had become concen- trated through having been laid in somewhat shallower water than usual. If, however, there was nothing unusual in the discovery of this near-bottom coastal concentration, and we have no grounds for supposing there was, then the obvious inference to be drawn from it is that spawning in certain instances at least is a shallow water coastal phenomenon and that in such conditions it might even take place on the bottom itself. Bargmann (1945) remarks on the scarcity of her stage 7 A, that is, gravid or near gravid, females in the samples she examined, and since by far the greater part of her material came from the top 100 m. ''- The sounding at Station 540 was 510 m. ^ At the time when he wrote they had not, except in rare instances, been fished below 1000 m. 19-2 i8o DISCOVERY REPORTS Table 37. Monthly record of the spent and gravid females taken in the 100- and 200-cm. diameter stramin nets between the surface and 2000 m., catches from the surface {0-5 m.) layer being shown in bold type {For corrections applied, see pp. 5g and 282-3) 100-0 250-100 500-250 1000-500 2000-1000 f \ Month Station Gravid Spent Nov. 743 — 2 1046 4 8 1877 3 — Dec. 125 22 — 539 8 — 1050 4 4 1052 — — 1214 4 — 1216 4 — Jan. RS 18 I — WS373 16 8 312 • — 12 2168 1.329 — 2170 4 • — 2171 — — 2172 52 — 2183 12 — 2185 6 — 2188 12 — 2192 3 — 2194 24 — 2195 1,215 — 2196 2 — 2198 18 — 2199 12 ■ — 2200 — — 2543 3 — 2558 36 — 2567 97 — 2569 2 — Feb. WS 152 134 — WS915 3,186 — 351 — ■ 1,372 352 — 12 354 1,816 15.427 612 — 384 613 — 4 618 — 3.252 622 — 112 831 — — 1287 — — 1501 40 — 15" 3 — 1521 I — 1523 — — 1 541 36 — 1545 — 124 2588 I — 2589 6 — 2590 3 — 2592 I — 2593 6 — "^ r~ Gravid Spent Gravid Spmt Gravid Spent Gravid Spent 46 1 — 2 — — II THE OLDER STAGES i8i Table 37 {cont.) 100- -0 250-200 500-250 1000-500 2000-1000 A A A A Month Station Gravid Spent 1 1 r \ Gravid Spent Gravid Spent Gravid Spent Gravid Spent March WS917 23 24 368 638 640 642 643 644 646 658 6 28 68 408 3>704 12 100 12 3.048 76 S 64 1 137 — 12 — — 1138 Many Many — I 1141 — — — I 1 146 — 8 — — 1 147 1156 1158 17 I — — I 2 7 "59 — I 1 160 I — — — — — — — — — 1309 8 21 — I 1311 I 12 — — 1312 — — I — 1979 — 2 — — 1980 12 9 1982 — 13 ■ — — 2268 36 — 8 — 2271 32 I I — 2272 6 — 2274 I I I — — — 2594 — 8 — — — — 2596 2 — — — I — April 207 663 — "5 4 665 — I 1350 — I — — of the Antarctic surface layer she concluded that the scarcity of gravid females there seemed to point to their going deep to spawn. Subsequent analyses, however (Table 37), show that gravid females at the surface may not in fact be so uncommon as Dr Bargmann's analyses suggest. At Station WS 915, for instance, out of a total catch of 930 adults in the oblique (loo-o m.) stramin net 531 were females undoubtedly in this condition,^ while other though somewhat less substantial concentrations of this stage were found right on the surface in the 0-5 m. layer at Stations 2168 and 2195. Dr Bargmann herself kindly re-examined for me the females taken at these stations and confirmed that all unmistak- ably belonged to her category 7 A. It may be significant that all three records come from high latitudes in the East Wind zone, a region virtually unrepresented by the samples she looked at. Further evidence of massing by gravid females at the surface has more recently been obtained from another source. In January 1940 I found enormous numbers of them filling more than a quarter of the stomach of a fin whale killed in the eastern part of the Weddell drift. The krill were quite fresh, 1 Table 37 it will be seen shows Station WS 915 with 3186 gravid females in the surface layer. This, however, as explained on p. 59 and again on pp. 282-3, is a corrected figure. i82 DISCOVERY REPORTS and had obviously been very recently eaten, and since it may be assumed that the whales are surface feeders (p. i68) it may be concluded that this particular one was harpooned while feeding on a surface swarm of gravid E. superba} Again it may be noted this record comes from a region only represented in Dr Bargmann's material by samples obtained outside the breeding season. Finally, it may be noted that the scarcity of gravid or near gravid females upon which Bargmann comments may be associated in no small degree with the comparative ease with which these very large euphausians, ranging from 40 to over 60 mm. in length, can in broad daylight, and even in darkness (p. 265), avoid the surface nets. SHELF WATER TOTALS OCEANIC WATER TOTALS TOTAL PER 3444 i^ STATION 26- .66 53 52 25 118877666443222222221 I I I I I I I I I I I I I * ST 2594 * ST 1283 •{■ ST 540 J ST 2603 MANY 2X0 NAUPLII ACS ANTARCTIC CONTINENTAL SLOPE IN I500-I000M NET Fig. 19. Vertical distribution of every gathering of krill eggs recorded in vertical nets in shelf and oceanic water. Among the very large series of vertical net samples analysed since Eraser's work was published we found, as Fraser found, only one that yielded eggs in any real measure of abundance. This was at Station 2594 in the eastern part of the Weddell drift where the deep 1500-0 m. net, which as already mentioned (p. 100) I presume must have fished for some time in the cold bottom water, produced the very substantial total of 3444 eggs. The complete record of the occurrences of the eggs identified by Fraser, Fry and me is shown in Fig. 19 from which it will be seen that, apart from the notable exceptions provided by Stations 540 and 2594 and possibly also by Station 1283, the eggs for all practical purposes have never been encountered in anything other than negligible numbers. Our almost total failure to sample the enormous masses of them that must exist somewhere in the Antarctic seas is brought still closer to reality if one takes into account the very large number of occasions when during the long spawning season our nets were fished with negative results. For during that period, in the shelf and oceanic water of the East Wind- Weddell system, no fewer than 528 stations were made representing a grand total of approximately 2734 operations of the vertical net of which only 162, or just under 6 %, were positive. Our failure, all but twice, to strike the eggs in mass, leaves an unfortunate blank in our data which perhaps for long will remain unfilled. It might be suggested, however, for the sake of future investiga- tion, that having regard to the vertical distribution of those at Station 540 and of the Nauplii at * Whales have also been reported feeding on swarms of spent krill (Nemoto, 1959), from which a parallel conclusion can be drawn. THE OLDER STAGES 183 Station 2603 (Fig. 19), in so far at least as the shelf or slope regions are concerned, we failed because as a general rule we did not send our nets often enough close enough to the bottom. In the coastal waters of Antarctica and of Antarctic islands generally the sea bed is often very rough^ and ..oa)@®#0' so 100 250 500 750 1000 SHELF WATER U3 22 20 25 22 — 7 22 12 14 13 8 2 — OCEANIC WATER 41 20 7 — 49 44 40 10 28 73 70 33 — — I II B 40 Fig. 20. Developmental phases of the eggs. Diagram showing the approximate numbers and vertical distribution of the seven stages recorded between the surface and 1000 m. These figures are not exact counts, for the egg stages are often indeter- minate or doubtful, but they indicate the trend of depth ranges. For the eggs at St. 2594, see Fig. 22, p. 184. so 100 250 o 500- CX)00000(X)0 OOOCX)(X)Q)Q)« 00(D®@« 750 1000 SHELF WATER OCEANIC WATER Fig. 21. Vertical distribution of developing eggs in shelf and oceanic water, the developmental phases encountered at each depth interval being shown to the nearest 10%, the smaller marginal symbols indicating fractions of 10%. much encumbered with stones and erratic boulders, occasionally (Broch, 1951) supporting a rich and jagged Hydrocoralline growth, and in order to avoid damage to the costly vertical nets through coming in contact with it, it was customary to give it a wide berth, the practice in general, except when in very shallow soundings (150 m. or less), being to leave a gap of 50 m. or more between the bottom and 1 Recent soundings by the Russian research ship ' Ob ' reveal it to be distinctly ' hillocky ' (Lisitzin and Zhivago, i960). ,8^ DISCOVERY REPORTS the deepest net. Our failure to make a systematic exploration of this near-bottom stratum may in fact have been vital and it might have been more profitable it seems if we had taken greater risks, for it must surely be significant that the net which probably came within closer striking distance of the sea bed than any other of the shelf series, the 500-250 m. net at Station 540, should yield the second largest catch of eggs we have so far recorded. Failure to make systematic search in the near-bottom stratum of the slope waters of the continental land must also it seems be held to account for our failure there to strike them, the vertical distribution at Station 2603 (Fig. 19 and p. 93, Table 14) of that very near product of hatching, the Second Nauplius, distinctly suggesting that had we gone still deeper^ on that occasion we might well have struck them, lying possibly on the bottom itself. Finally, turning to the open sea (Fig. 19), it seems obvious, having regard to the apparent vertical distribution at Station 2594, that our failure (except once) to sample them in mass over deep water must simply be ascribed to our failure to carry out systematic observations (p. 98) at great depths below 1000 m. We may turn now to the possibility, originally suggested by Fraser, that although shed at higher levels they may sink before hatching to depths below 1000 m. In 3608 of the 4231 eggs identified by Fry and me from the vertical samples the state of develop- ment was clearly distinguishable, some being unsegmented or just beginning to divide, others fully segmented with relatively large or with very small cells, the most advanced showing varying degrees of naupliar development culminating in the easily recognisable form of the First Nauplius evidently near the point of hatching. A diagrammatic illustration of the developmental condition of those taken between the surface and 1000 m., based on the numerical record given in Fig. 20, is shown in Fig. 21 in which the conditions in shelf and oceanic water are compared. Taking the picture as a whole it will be seen that the eggs approach progressively nearer to a ripe condition as the depth increases, and the inference seems to be that those recorded in both shelf and oceanic water were in fact sinking eggs that were developing towards maturity as they went down. It would appear too, that they were Uberated in their unsegmented state, as Fraser (1936, p. 17) found with his laboratory-spawned specimens, that they were liberated in the Antarctic surface layer and that by the time they had sunk below 750 m. the majority had developed incipient or clearly recognisable naupliar forms. Of the total of 3444 eggs recorded at oceanic Station 2594 the developmental condition was recognisable in 2880 and is shown separately in Fig. 22. It will be recalled (p. 100, note 2) that the deep 1 500-1000 m. net at this station unfortunately failed to close, fishing from 1500 m. open to the surface. We cannot, therefore, be certain of the level, or levels, at which this exceptionally large catch was made. It might be supposed however from (i), the absence of eggs from all levels above 1000 m., from (2), the high proportion of the total catch, 75%, that contained developing nauplii, and from (3), the developmental condition of the deep eggs in oceanic water generally (Fig. 21), that an unknown but probably very considerable portion of it did in fact come from the 1 500-1000 m. horizon to which it is possible it may have sunk from a higher level.^ Recent laboratory experiments by Marshall and Orr (1953 «) indicate that the eggs of Calanus finmarchicus tend to sink more slowly in water at 0° C than they do in warmer water. It seems too that under natural conditions the eggs of E. superba behave in the same way, sinking more slowly in the colder waters of the shelf than they do in the warmer conditions of the open sea. It will be seen (Figs. 20, 21 and 23) that at all subsurface levels where significant numbers are involved, the 1 The deep 1400-1000 m. net at this station came within approximately 50 m. of the bottom in 1450 m. - See, however, p. 210. STAGE NUMBER % TOTAl 0 144 5 G) 144 5 @ ISA 5 ® 198 7 1 108 4 © 834 30 0 1296 45 Fig. 22. Developmental condition of the eggs at St. 2594. THE OLDER STAGES 185 developmental state of the shelf eggs is distinctly in advance of the corresponding state in oceanic water, and while this may merely mean that development on the shelf is more rapid than in the open sea^ a slower sinking rate in the shallower conditions seems a distinct possibility. As compared with the open sea, however, the shallow waters of shelf regions are relatively unstable, being subject to considerable turbulence arising through vertical mixing, tidal movements and so on, and it may be that this turbulence alone contributes to the near-surface occurrence of developmental stages that in oceanic water are more commonly encountered at deeper levels. As Fig. 23 shows, wherever advanced eggs containing developing nauplii were encountered in shelf water, they tended to occur nearest the surface where the water was shallowest. OCEANIC WATER -EAST WIND DRIFT >• < WEDDELL DRIFT- XL Fig. 23. Diagrammatic illustration of the developmental condition of the eggs in shelf and oceanic waters, the vertical egg columns representing the actual condition recorded at every station where observations on egg development were made. It seems clear then that such eggs as we have recorded must in fact have been sinking and were laid above 250 m. With the possible exception of those at Station 540 (Fig. 19), however, the numbers we find at this near-surface level are so consistently small that one hesitates to suggest that it is there, near the surface, whether in oceanic or shelf water, that spawning principally takes place. All in fact that can safely be concluded from the meagre results of our analyses is that some sporadic surface or near-surface spawning, evidently on a very minor scale, does take place in both shelf and oceanic water, and that the eggs so laid do in fact sink, developing as they go down, to hatch at deeper levels. Turning again to Fig. 19 and to such evidence as it may be said to provide as to what the depth of spawning might be, it will be seen that while it does emphatically not appear to be the near-surface phenomenon of Ruud there are two other distinct possibilities to be considered, (i) that the eggs might be laid on or near the bottom in Antarctic shelf or slope^ water (Stations 540 and 2603), and 1 This, however, seems unlikely. It is well known for instance from the work of Apstein (1909) and others that in many fishes the eggs develop much more slowly in colder than in warmer water. 2 Taking here the near-bottom occurrence of Second Nauplii at Station 2603 as indicating the presence, or very recent presence, of eggs. jg5 DISCOVERY REPORTS (2) that they might be liberated deep down in oceanic water (Station 2594) far away from land. At all three stations, however, the possibility that they might have sunk to their locus of deep concentration from a considerably higher level must always be kept in mind. Although little satisfactory information about the depth of spawning can be gathered from the vertical distribution of the eggs themselves, the vertical distribution of the breeding adults— the pairing sexes and the spent and gravid females— points to it being a surface rather than a deep phenomenon. The complete record of the adult males and females taken in the vertical nets since these investigations began, based on the analyses of over 3000 samples from the East Wind-Weddell zone, is shown in Table 38. It is a meagre record for all depths, revealing the deep occurrence of adults, which Eraser (p. 179) originally supposed might be connected with deep spawning, as a phenomenon of exceptional rarity, to which, as the negligible deep catches of spent and gravid females bear further witness, we cannot it seems attach very much significance. A single net, however, out of this enormous series of virtually negative hauls, the 250-100 m. net at Station 356, in which 39 gravid and three spent females were taken, does it seems provide some evidence that spawning, although not necessarily a very deep operation, might take place some little distance below the surface. Clearly this net must have struck an isolated subsurface pocket of adult females, producing by horizontal standards a small, but by vertical standards a distinctly large, sample which, again (p. 160) having regard to the low sampling power of the vertical net and the distinct possibility (p. 149) that the pocket itself might have been one of shallow draught, might in fact have represented a subsurface congregating of spawners in some unknown, but perhaps not inconsiderable measure of abundance. Since, however. Station 356 was made on the South Georgia whaling grounds, a region it will be shown presently (p. 190) where virtually no spawning takes place whatsoever, the occurrence of gravid females a short distance below the surface there, in so far at least as the major aspects of the egg-laying are concerned, seems of doubtful significance. Perhaps not too much weight can be attached to the evidence from the vertical nets, because the scarcity of any of the larger forms, even near the surface, shows their low catching capacity. The larger towed nets, however, give more significant results. The vertical distribution of the adult males and females based on the much more voluminous data from our towed stramin nets is shown in Table 39, the figures being derived in a large measure from the published records of Bargmann^ (1945, Appendbf), but including much additional information gathered from the field observations of other members of the Discovery staff. In this table, which is based exclusively on our obser- vations in the principal region of euphausian abundance (the East Wind-Weddell system) the figures for the surface (loo-o m.) layer show the average monthly catch, that is, the total monthly catch divided by the total monthly net hauls, the figures for all levels below 100 m. showing the total monthly catch, not the average, since the latter it will be seen proves to be so small that in the vast majority of instances it can only be expressed as a very small fraction. Furthermore, owing to their extreme smallness, the catches of adults in the large 2-m. diameter TYE nets so extensively used below 250 m. have not been expressed in terms of standard i-m. diameter stramin net catches as described on p. 157. Such subsurface catches of adults as were recorded in the larger nets were in fact so small that they have been entered in Table 39 as they stood, regardless that is of the size of the apparatus used or of the period of towing involved. In order, however, to give full expression to their smallness, in every instance where the large net was used it has been regarded as having represented four hauls of the smaller net, when fished for 30 min., or as having represented t^/T hauls of the smaller net, if fished for longer, where again (p. 157) t is the actual time of fishing of the TYE, and T the standard time of fishing of the i-m. diameter stramin net. 1 Corrected to conform with the total catches of the nets from which her samples were derived. THE OLDER STAGES 187 Table 38. Monthly record of adult {spermatophore carrying) males and females and of spent and gravid females taken in the yo-cm. diameter vertical nets Month Nov. Dec. Jan. Feb.* March April Nov. Dec. Jan. Feb.* March April Station WS480 480 537 538 540 760 764 2137 2168 2192 WS376 WS38S WS389 WS390 WS504 356 1114 128s 1541 1545 1965 1 142 2268 2295 663 8SS WS480 480 537 538 540 760 764 2137 2168 2192 WS376 WS38S WS390 WSS04 356 1114 128s 1541 IS4S 1965 1142 2268 2295 663 855 50-0 1 00-50 250-100 Adult Adult Gravid Spent 3 Adult Adult Gravid Spent 3 Adult Adidt Gravid Spent 39 500-250 A , 750-500 1000-750 3 — * At Station 2594 in February seven adult males and three spent females were taken, but at an unknown depth since they were in the net hauled without closing from 1500 m. to the surface (see p. 294). i88 DISCOVERY REPORTS Table 39. Monthly vertical distribution of the pairing sexes and spent and gravid females, average catch-figures of less than one being indicated by an asterisk and the number of net hauls in each vertical horizon by the figures in italics. For fuller explanation see text {For corrections, see pp. 59,157 and 282-3). Spring Summer Autumn Winter Depth , * , , * > r * V < * > (m.) Sept. Oct. Nov. Dec. Jan. Feb. March April May June July Aug. Males carrying spermatophores loo-o no 43 4 23 10 295 14 I — — — 8 42 48 68 70 136 123 135 6g 18 17 18 48 250-100 — — — I 4 — 26 — — — — 11 13 23 30 40 17 51 29 8 4 10 9 500-250 — — — — 5 I 17 — — — I 16 25 7 13 29 36 26 26 5 9 14 20 1000-500 — — ■ — — — ■ — 21 — 4 9 8 3 10 4 26 5 — 14 2 2 2000-1000 — — — — — — — —6—4892—7 — — Females carrying spermatophores lOO-O — * I 27 8 120 2* 42 48 68 70 136 123 135 69 18 17 18 48 250-100 — — 2 3 29 5 y____ — 11 13 23 30 40 17 51 29 8 4 10 9 500-250 — I- — I I I 18 — — — — 16 25 7 13 29 36 26 26 5 9 14 20 1000-500 — — -I — 2 — I — — — 4 9 8 3 10 4 26 5 — 14 2 2 2000-1000 I — — — — — — —6—4892—7 — — Gravid females loo-o __* *23 34 i_____ 47 63 117 87 122 153 167 72 18 30 18 55 250-100 — — — 3 48 I — — — — — — 7 14 38 40 45 20 52 31 8 11 56 500-250 — — — — — 3 II — — ■ — — — 16 25 7 13 29 36 26 26 5 9 14 20 1000-500 — — — — I — I — — — — 4 9 8 3 10 4 26 5 — 14 2 2 2000-1000 — — — — — — — —6—4892—7 — — Spent females loo-o — — * * * 13s 45 2 — — — — 47 63 117 87 122 153 167 72 18 30 18 55 250-100 — — — I — 2 2 — — — — — 7 14 38 40 45 20 52 31 8 11 5 6 500-250 — — — — — II I — — — — — 16 25 7 13 29 36 26 26 5 9 14 20 1000-500 — — — — — — — — — — — 4 9 8 3 10 4 26 5 — 14 2 2 2000-1000 — — — — — — — —6—4892—7 — — THE OLDER STAGES 189 It is evident then, from the data so presented, that neither pairing, as shown by the vertical dis- tribution of the spermatophore-carrying males and females, the stages recognised by Bargmann (1945) as 6 and 7 in males and 6 in females, nor spawning, as shown by the vertical distribution of the gravid and spent females (Bargmann stages 7 A and 7B), can be described as anything other than a surface phenomenon.^ In other words, as might be expected, there is nothing to suggest that the vertical distribution of the breeding adults is in any way different from that of the whale food as a whole. It might, however, be argued that spawning takes place in the still greater depths that as yet remain unexplored, in the great deeps where it seems (p. 98) the eggs themselves are hatching. While this is a possibility that cannot be ignored our failure to capture anything but negligible numbers of spent, gravid or recently paired females at intermediate depths seems to point to its at least being improbable. For if the gravid females go to such enormous depths to shed their eggs, unless they do so so widely scattered as virtually to defy capture even by our largest nets — a wholesale scattering which while possible is wholly inconsistent with the Hfelong swarming habit (p. 230) to which this species seems prone — it is astonishing that not one of our hundreds of towings, many of them with nets of the largest size, worked below 100 m. during the spawning season^ has yet revealed the slightest evidence of a concentration of them on their way from the surface down. Localities of spawning In so far then as it can be gleaned from the vertical distribution of the ripe or spent females the evidence in favour of deep spawning in oceanic water is to all appearances negative. At the same time the evidence in favour of large-scale oceanic surface spawning, although apparently positive in so far as the vertical distribution of the spent and gravid females is concerned, is patently contradicted it seems (Fig. 19) by the vertical distribution of the eggs themselves, the numbers of the latter recorded in the Antarctic surface layer being far too small to warrant the conclusion that it is there, near the surface, that spawning over deep water is actually taking place. Although females then apparently ready to spawn and females that have in fact spawned are both encountered near the surface in oceanic water we cannot altogether disregard the possibility that of the gravid, or apparently gravid, and spent females recorded there, the former may fail to reach the actual point of spawning while the latter it seems might have laid their eggs in some other and possibly distant place. While it is difficult to judge from these contradictory appearances what is in fact taking place in or over deep water, it seems that, in so far at least as the major aspects of the spawning are concerned, the open sea can be playing at most only a secondary part. It is suggested, therefore, that the enormous annual output of larvae necessary to replenish the whale food population may have its origin somewhere in or near the coastal water of the Antarctic continental land where as already seen (p. 168, Fig. 19) the second largest of the only two major concentrations of eggs to be recorded was originally found and where one of our only three major occurrences of a recent product of hatching, the Second Nauplius, was encountered (p. 93, Table 14, Station 2603) very close to the bottom. Moreover, the fact that we have repeatedly struck the eggs with far greater frequency in or near the shelf water of continental Antarctica than we have in the open sea provides further ground for supposing that it must be in the coastal or near-coastal waters of high or relatively high latitudes that spawning is principally taking place. In the shelf and slope waters of the Bransfield Strait, North Graham Land and the East Wind zone 98 stations were made, covering, if all the years these investigations have lasted be 1 From the pronounced abundance of paired females he finds in whales' stomachs Nemoto (1959) also concludes that the krill copulate on the surface. ^ Expressed in terms of i-m. diameter stramin nets the actual number so worked in the period November to March was 384- igo DISCOVERY REPORTS Table 40. Monthly record of the total eggs and larvae taken on the South Georgia whaling grounds based on the catches of the jo-cm. diameter vertical nets, the number of stations made monthly {or bi-monthly) being shown on the right in bold Calyptopes Furcilias Depth First Second Meta- , * , , * ^ {m.) Eggs NaupUi Nauplii nauplii 1 231 23456 Month Nov. Oct. Sept. Aug. May- April March Feb. Jan. Dec. 250-0 500-250 750-500 1000-750 250-0 500-250 750-500 1000-750 250-0 500-250 750-500 1000-750 250-0 500-250 750-500 1000-750 250-0 500-250 750-500 1000-750 250-0 500-250 750-500 1000-750 250-0 500-250 750-500 1000-750 250-0 500-250 750-500 1000-750 250-0 500-250 750-500 1000-750 I — 1 — 4 — 2 — 585 148 2 — 8 14 — — 32 I 32 4 17 2 4 4 10 I 92 6 187 5 36 6 I — 41 51 4 16 12 10 3 2 4 iS I 33 49 49 14 reckoned with, the whole protracted span of the spawning season, November to March. Eggs occurred at 47, or 48%, of them. In the same period, again reckoning all the years the work has lasted, 442 stations were made in the oceanic water of the Weddell and East wind zones and in the coastal and oceanic waters of the South Georgia whaling grounds. Eggs occurred at 44, only 10% of them. Covering this period in the South Georgia area alone 196 closely spaced stations were made repre- senting a grand total of 880 operations of the vertical net. Yet for all this intensive local work only eleven eggs were found, seven at one station and one each at four others, a result so meagre that it is difficult to avoid the conclusion that the spent and gravid females found in this krill-rich northern field contribute little or nothing locally to the annual recruitment of the euphausian population and that these island waters therefore are not concerned in the spawning except perhaps on an insignificant scale. THE OLDER STAGES I9i Even if the pronounced scarcity of eggs at South Georgia were by itself not enough to warrant this at first sight paradoxical conclusion, for there is always the possibility that they might have been concentrated there at some critical depth, for instance, on or near the bottom, where our sampling is inclined to be less thorough than it is in the upper layers, our almost complete failure to find any evidence of hatching, to which the absence of Nauplii and the capture of only two Metanauplii (Table 40) bear ample witness, points to its being the most reasonable inference that can be drawn. As Marshall and Orr (1955, p- 70) have said of Calanus finmarchicus, 'the presence of males and the presence of females with spermatophores have both been taken as signs that active breeding is going on but events in East Greenland show that they are not necessarily so. The only safe indication is the presence of eggs, nauplii or young copepodites'. Simpson (1956) writes in similar terms on the spawning of fish, noting that the 'systematic determination of the distribution of the youngest pelagic stages [including the eggs] of any fish . . . provides a means of delimiting with very considerable precision both the time and position of spawning'. It might be observed too that our virtually com- plete failure to find eggs in these island waters cannot be ascribed to a wholesale neglect of the levels, perhaps the critical levels, adjacent to the floor of the sea. For although the near-bottom stratum in the deep oceanic water of the South Georgia area was in fact left entirely unexplored the corresponding stratum closer inshore was by no means so completely neglected. Between November and March 90 stations were made in water less than 1000 m. deep, deep nets (Table 41) coming within 50 m. of the bottom at 79, or 88 %, of them, the average proximity of approach being as little as 22 m. Table 41. Nearness of approach of deep nets to the bottom in the shallow waters of the South Georgia whaling grounds Nearness of approach No. of nets (m.) I-IO IS 11-20 31 21-30 10 31-40 13 41-50 10 Not only does the absence of any substantial deep living larval community indicate how insignifi- cant in this field the scale of the hatching must be, but it also suggests, and suggests very strongly, that such few surface larvae, Calyptopes and early Furcilias, as from time to time have been recorded here from January to March (Table 40) must get carried to the island in the surface stream. Through- out the breeding season our vertical gatherings of young surface forms in this neighbourhood have consistently been small or negligible, only one in fact of the relatively enormous series of net hauls we made yielding a really substantial catch, this single gathering, a March one, of 707 Calyptopes, accounting for nearly 60% of the total Calyptopes and early Furcilias recorded round the island since work there first began. Even this, for South Georgia, exceptionally large vertical gathering is unlikely to have been produced in situ, to have sprung in other words from a local hatching or local spawning, it having been obtained in the surface stream to the south-east of the island, well away from the land and well into the tongue of Weddell water (p. 58, Fig. 4) that approaches it from that direction. Ruud (1932) is also of the opinion that in so far as the major aspects of the spawning are concerned the South Georgia region does not come into the picture and that the krill population there cannot, therefore, be of local origin but must spring from surface-borne incursions of larvae born in higher latitudes in the Weddell Sea. He writes, ' If E. superba spawns under and near the drift ice, its proper spawning ground is the zone of the pack-ice, and as a rule the conditions off South Georgia are not 192 DISCOVERY REPORTS suitable for its spawning. The stock of krill at South Georgia must therefore be carried thither by the current in the Antarctic surface layer '.^ It may be that temperature plays a major part in determining the distribution of the spawning krill. I quote from Moore (1958): ...there are likely to be marginal zones around the area inhabited by a particular species in which temperatures are suitable for survival but not for breeding. Such areas depend for repopulation on a supply of immigrant larvae or adults.'^ An animal with a short life span may not have time to be carried far from its breeding area before it dies and may, therefore, not attain a region where temperature would limit survival. Euphausiids, with their relatively long life span, however, have the opportunity of being carried long distances by the currents, and Einarsson (1945) has shown that they may have extensive fringes of non-breeding population. As Einarsson himself remarks, with marine animals generally, the spawning area is more restricted than the whole geographical range, the distribution of the freshwater eel {Anguilla anguilla L.) pro- viding a classic example.^ He notes, too, that whereas with fishes, in so far as oceanic dispersal is concerned, the larvae alone are influenced by the currents, euphausians ' are subjected to the influence of the currents throughout life ', deliberate wanderings towards certain spawning areas such as are undertaken by some fishes being ' out of the question because of the feeble swimming power of the animals. Consequently ', he continues, ' the individuals which are carried out of the reproduction area will seldom contribute to the maintenance of the stock '. Paulsen (1909), referring to the wide temperature range of Meganyctiphanes norvegica, and other boreal euphausians, observes that they are not ' strict in their claims on the surroundings ', adding that if it proves correct that they spawn in warm and not in cold water, ' diflFerent claims on the surroundings may be made for propagation than for the maintenance of life '. In his survey of the zoogeography of the Atlantic Hydromedusae, Kramp (1959) writes on similar lines : ' How far away from their place of origin the neritic medusae may be carried along with the currents depends on the velocity and direction of the currents and the duration of life of the medusae We must also bear in mind that the geographical distribution of a medusa does not necessarily coincide with its native habitat, because it may have been carried to an area where it may continue its swimming existence and take nourishment for its own maintenance, but is unable to propagate'. Or to quote Tucker (19596) on his new eel hypothesis, 'Parallel cases of expatriated populations failing to breed are the British Octopus vulgaris, the Norwegian Palinurus elephas and the Lagos Branchiostoma nigeriense, all of which are maintained by immigrations of larvae bred elsewhere and so represent similar cases of wasted reproductive potential '. Or as Gunter (1957) has put it in his remarks on temperature and the distribution of marine life as a whole, ' Species restricted by reproductive stenothermy may live in other areas but must be recruited every year from a spawning center '. In his recent account of the zoogeography of the pelagic polychaetes of the South Atlantic Tebble (i960) finds that juvenile Tomopteris carpenteri (specimens less than 12 mm. long) are plentiful round South Georgia but apparently absent from the deep oceanic water near the South Sandwich Islands and along the meridian of Greenwich. He suggests that T. carpenteri may breed round 1 See, too, the passage from Risting quoted on p. 43. ^ The italics are mine. * According to the new hypothesis of Tucker (19590) it would seem that the Leptocephali that regularly cross the Atlantic from the Sargasso Sea, and eventually as elvers ascend the European rivers, are now to be regarded as of North American origin, and that the supposed return of the European females to spawn in the Sargasso area does not in fact take place. Even so the European eel still provides us with a classic example of discrete spawning coupled with wide geographical range. The distribution of larval and adult Swordfish {Xiphias gladius) in the North Atlantic (Taning, 1955) points to a similar phenomenon. THE OLDER STAGES i93 South Georgia but not in the open sea, and if this is so here also perhaps we may have an instance in which a plankton animal of wide adult geographical range is spreading from a restricted spawning area. A still higher shelf sea egg frequency than is recorded above (p. 189) is obtained if the Bransfield Strait-Weddell West region be considered by itself, it being there that the majority, 82 in all, of these high latitude shallow water stations were made. Eggs occurred at 41 of them, a frequency of 50 "0. But by far the highest frequency we have recorded was encountered during the survey of this region that took place in February 1933 (Discovery Reports, Station List, 1931-3) when eggs were taken at no fewer than 14, or 82%, of the 17 stations that were made. That the frequency should apparently be at a maximum in February is perhaps not surprising, for it is during this month, judging from the particularly large numbers of spent females that have been encountered both then and in March (p. 188, Table 39) that spawning may be said to be at its height. In striking contrast to the pronounced regularity with which eggs were encountered in shelf water in February 1933 it may be noted that out of 129 stations made in the oceanic water of the Weddell drift, the East Wind drift and the South Georgia region in the same month eggs were present at only ten — and this at the supposed peak of the spawning season. In view of the pronounced regularity of their occurrence in shelf water it is possible that the eggs recorded in the upper strata there (p. 182, Fig. 19), small or negligible in number though we have found them to be, may represent Eraser's ' scattered product of dispersal ' of much larger masses located below, located perhaps, as the eggs at Station 540 so distinctly suggest, on the bottom itself. Not a little of the evidence then, particularly the evidence provided by the near-bottom concentra- tions of eggs and Nauplii recorded at Stations 540 and 2603, suggests that the females find in the cold shelf or slope waters of the continental land, in high or relatively high latitudes, more suitable conditions in which to deposit their eggs than are to be encountered in the warmer oceanic water farther off the coast. It suggests too that spawning in the shelf water of the Bransfield Strait-Weddell West region may be taking place on a bigger scale than elsewhere in the circumpolar sea. In other words, in so far at least as the Atlantic sector is concerned, there would appear to be a distinct possibility that the krill, /ar/rom depositing their eggs anywhere throughout the vast area in which they are known to range, may in fact have a localised spawning place from which the resultant eggs and larvae spread away to populate the whaling grounds through the agencies already described on pp. 99-105. Passing to a critical examination of this suggestion it will be convenient first to consider to what extent, if any, it may be said to be upheld by the regional distribution and relative abundance of the spent and gravid females in the circumpolar sea. This, based on the catches of the i-m. diameter stramin nets for the period November, when the first eggs appear, to April, when the last spent females are found, is shown in Figs. 24 and 25. It would appear from these figures that the principal spawning areas are (i) high latitudes in the East Wind drift, (2) the Bransfield Strait, (3) the Weddell drift, notably Weddell West, and (4) the tributary offshoot of that surface stream which reaches the island of South Georgia. It would appear, too, in fact seems clear, that an enormous area of the Antarctic Ocean, the vast extent of the circumpolar West Wind drift, carrying as it transpires neither spent nor gravid females, is not a region where the eggs are laid. Although the distribution of these stages does not at first sight point anywhere to any particularly localised spawning ground, except in so far as the narrow coastwise current of the East Wind drift may be said to be localised, the pronounced abundance of spent females that appears in the western part of the Atlantic sector (Fig. 25) suggests strongly that there at least, somewhere in the general area Bransfield Strait-Weddell West-South Georgia, the eggs are shed on a vastly greater scale than they are elsewhere in the circumpolar sea. Narrowing down this region still farther by excluding the 194 DISCOVERY REPORTS 1201 130° W 180-E iso° Fig. 24. Distribution of gravid females, November to April. THE OLDER STAGES I9S I20p lSO° W180°E 150° Fig. 25. Distribution of spent females, November to April. 196 DISCOVERY REPORTS ® 100-1,000 -, ® 1.000-10,000 10,000-100,000 Fig. 26. Summer distribution of the surface larvae (stramin net hauls). Ice-edge February mean. ? J THE OLDER STAGES 197 120 l'l'rlMMilM'\il ■! MM'Vmm ■ !■ I.'VVM ' l't'\^' ^ ' k ' ^ ■ k ' V S ■ \ ■ 7 V ' / ' ' ' ; ' / ' J ' M > ' J i / W ' f M ' I ■ ^ I ' I ' I ' I ' f ■ ] U ' I ' ] ' I 1 150° W 180-E 150° Fig. 27. Summer distribution of the surface larvae (vertical net hauls). Ice-edge February mean. iqs discovery reports South Georgia whaling grounds, where spawning it seems (p. 190) is of negligible account, we are left with the shelf water of the Bransfield Strait and the shelf and oceanic water of the western Weddell drift as being to all appearances the main locus of the spawning krill. Since, however, in spite of the abundance of spent females in the oceanic water of the western Weddell drift, there is no satisfactory evidence (p. 189), either from the surface or the deeper layers, that it is in or over deep water that they actually lay their eggs, it is difficult to avoid the conclusion, already alluded to on p. 189, that the spent females found for instance between the South Orkneys and the South Sandwich Islands (Fig. 25), and doubtless those at South Georgia as well, get carried there as spent females in the surface drift, having laid their eggs some time before somewhere to the westward of where they were captured. In other words, highly localised, in terms of Antarctica as a whole, as the Bransfield Strait- Weddell West region may be said to be, it would appear that, disregarding for the moment the obvious potentialities of the East Wind drift, in the Atlantic sector the spawning of the krill, in its major aspects, might well be confined to within still narrower geographical limits, and be in fact the pheno- menon, the remarkable thing, I have suggested it to be, a phenomenon peculiar, perhaps for all practical purposes exclusive, to the shallow shelf and slope waters that in this sector lie in the far western and south-western reaches of the Weddell drift. The region of circumscribed spawning to which our evidence seems to point is not an easy one to explore and perhaps for long, if not always, will remain undefined, the south-western reaches of the Weddell drift being particularly difficult and hazardous of access owing to the piling up of the pack-ice there. Further exploration, however, in the more readily accessible westerly reaches of the current, for instance, in the extensive area of shelf water lying to the south of the South Orkney Islands and in the shallow water associated with the Clarence Island-South Orkneys and South Orkneys-South Sandwich Islands ridges, might reveal that they too are involved in the spawning. The distribution of the larvae that reach the surface in summer (January-March), those, that is, that have but recently accomplished the long developmental ascent, provides perhaps the most con- vincing evidence we have (Figs. 26 and 27) that somewhere not far from where I have supposed it to lie there must exist a relatively discrete spawning area where eggs are deposited on an immense scale, a scale, moreover, to all appearances incomparably greater than anything to be found elsewhere in these far southern waters. Fig. 26 shows the summer distribution of the Calyptopes and Furcilias based on the surface (0-5 m.) gatherings of the towed stramin nets, and while the catch-figures in- volved in its construction include very few of the First Calyptopis, a stage which owing to its small size (p. 360) readily escapes through the meshes of the stramin net, it nevertheless presents a most striking picture of the major spawning that seems to be taking place somewhere in the western part of the Weddell Sea. The summer distribution of the recent risings based on the material from the vertical nets, in which the First Calyptopis is always well represented, is shown for comparison in Fig. 27, and here again the emphasis is on Weddell West as being to all appearances, in the northern zone at any rate, the principal locus of the spawning krill. It must be significant, too, that throughout March, when the vast majority of the eggs have probably been laid, in Weddell West, but there alone, we struck the larvae at one level or another in our vertical hauls at every station zve made. It is true that in April (p. 311, Fig. 75) we find something rather different, for then we have the Calyptopes well represented as far east as 0° and 20° E in Weddell East. These late risings to the east, however, will be dealt with later (see pp. 313-14). Fig. 27, however, points clearly to the existence of another, perhaps secondary, spawning ground in the coastal waters of the East Wind drift, a spawning ground which, oddly it would seem, the stramin nets with their much greater catching capacity completely fail THE OLDER STAGES 199 to reveal. Their failure, however, can readily be explained, for, as Table 42 shows, it is simply due to the fact that whereas in the rich northern zone the older Calyptopes and the Furcilia stages are found in the surface from January onwards, reaching a maximum in March, in the East Wind drift, even as late as March, the young surface population is not very large and still dominantly First Calyptopis, a stage that (see above) is virtually unrepresented in the catches of the stramin nets except possibly when present in exceptional abundance. The relative scarcity and backward development of the larvae in the East Wind zone is again probably to be ascribed (p. 177) to the shortness of the summer there and the brief blooming of the phytoplankton in these high latitudes. Table 42. Composition of the larval populations in the Weddell and East Wind zones in summer, Weddell larvae in roman type, East IVitid larvae in italics Calyptopes Furcilias Month Second Nauplii Meta- nauplii 1 2 3 1 2 3 4 5 6 No. of stations March — 2135 8467 1516 3215 2460 606 145 302 246 45 26 528 1046 7S8 18 — — 1 — — — — 25 Feb. 13 106 792 240 222 255 154 18 I — — 27 1 206 256 1 — 1 — — — — — 37 Jan. 338 12 81 21 51 8 10 4 8 3 — 19 39 In a recent paper on some problems in larval ecology Wilson (1958) observes that spawning seasons in general may be related and adapted 'to particular biological characteristics of a water normally occurring at a certain time of year, just as we know that they often are related and adapted to tempera- ture '. It could well be that it is temperature above all that is determining the geographical distribution of the spawning krill, the virtually always sub-zero temperature of the coastal waters of the continental land. Time of hatching From the occurrences of the Nauplii (p. 90, Tables 13 and 14) it is clear that hatching begins in November and continues throughout summer until autumn, probably ending, as a major phenomenon, in March. It seems likely, however, that it may continue sporadically and on a much reduced scale until later, the several rich hauls of Metanauplii (Tables 13 and 14) recorded in late April in the Weddell and East Wind zones suggesting that in both regions some hatching must take place in the earlier part of that month. As with the spawning (p. 177) the hatching in the East Wind drift must be late, not beginning on any appreciable scale (Table 14) until January. Depth and localities of hatching Although much of the evidence (again having particular regard to the rich haul of shelf-laid eggs recorded at Station 540 and to the rich coastal haul of Nauplii recorded near the bottom at Station 2603) seems to point to the spawning of this species being, in its major aspects at least, a shallow water coastal phenomenon, taking place on a particularly large scale in the far western reaches of the Weddell drift and in the coastal waters of higher latitudes, it is a remarkable fact that our data provide very little evidence that hatching on any major scale takes place in the shallow conditions in which it seems the eggs are laid. All our observations in fact point to the contrary, that it is deep down in oceanic water, principally in the Weddell and East Wind zones, either close to the shelf regions on the Antarctic continental slope or at some considerable distance away from them, that any such large- DISCOVERY REPORTS 60 80 120 Fig. 28. Distribution of deep larvae in relation to the 1000 fathom line, substantial numbers being represented by the larger black circles, insignificant numbers of less than 10 by the smaller. For explanation of the circumpolar coverage figures, see pp. 287-9. 20I THE OLDER STAGES scale hatching occurs. This phenomenon is illustrated in Fig. 28, which clearly shows that where- ever substantial numbers of deep rising larvae, Nauplii and Metanauplii, were encountered, they occurred either right on the Antarctic continental slope or in water exceeding 1000 fathoms (1829 m.) at variable distances out to sea. It shows too that wherever these early stages were encountered in the shallower shelf sea areas, they never occurred there other than in negligible numbers. Since the presence of large or small numbers of Nauplii or Metanauplii must indicate at still deeper levels (p. 98) the presence or recent existence of correspondingly large or small numbers of hatching eggs, it seems clear enough that it must be at enormous depths, in the open sea, and not in shelf water, that hatching takes place on the immense scale necessary to maintain this teeming population. ACS ANTARCTIC CONTINENTAL SLOPE * NO OBSERVATIONS Fig. 29. Numbers and vertical distribution of Nauplii, Metanauplii and First Calyptopes, recorded in shelf and oceanic water showing the enormous scale of the hatching that takes place in the deep ocean as compared with the shallower waters bordering the continental land. Note that the numbers plotted above 1000 m. in oceanic water (e.g. 440) represent the total Calyptopis catch for every 4 stations made beyond the slope (see footnote i). The relatively enormous scale of the hatching that takes place in oceanic water as compared with that so far discovered in shelf waters is shown more realistically in Fig. 29 which illustrates diagram- matically a shelf sea, such as exists in the Bransfield-Weddell West region, with a continental slope and deep oceanic water beyond. All the larval occurrences from shelf, slope and oceanic water plotted in Fig. 28 are represented on it, the gatherings of Nauplii, Metanauplii and First Calyptopes recorded at each station, or, as for the Calyptopes in oceanic water, every four adjacent stations,^ being shown as indicated in the key. As far as the bathymetric scale has allowed, the stations are arranged so as to appear over the actual depth of water in which they were made. Clearly, as this diagram shows, it is the Antarctic continental slope, and the deep oceanic waters of the East Wind 1 Owing to the great abundance of First Calyptopes in oceanic water, especially between 750 m. and the surface, the catch- figures for this stage, because of overcrowding, cannot be shown clearly in this diagram for each individual oceanic station. They have accordingly been shown for every four adjacent stations made beyond the slope. 203 DISCOVERY REPORTS zone and Weddell drift that lie beyond, that are par excellence the principal spheres of hatching of the eggs of this euphausian. Indeed, in so far as the slope waters are concerned, it may well be that there they hatch on the sea bed itself, an inference that as already suggested (p. 184) may readily be drawn from the vertical distribution of the Nauplii at Station 2603, the slope station in Fig. 29 where over 400 Second Nauplii and 75 Metanauplii were recorded in a net that came within 50 m. of the bottom. It has been shown then, that in spite of the frequency with which eggs have been encountered in shelf water and in spite of the concentration of them found at Station 540, the production of Nauplii, Metanauplii and First Calyptopes that results from shelf hatching, when compared with the cor- responding production resulting from deep oceanic hatching, is to all appearances negligible to a degree. It still remains to be considered, however, to what extent this disparity, so obviously apparent in the very early larval stages, may be said to be true of the larval production as a whole. The total monthly larval gatherings from vertical hauls made in the shelf waters of the Brans- field Strait, Weddell West, Ross Sea and Bellingshausen Sea, and from similar hauls made in the oceanic waters of the Weddell and East Wind zones are brought together for direct comparison in Table 43. It will be seen from this presentation of the data that, having regard to the total number of stations operated monthly in both shelf and oceanic regions, from January, when the climbing larvae first begin to come within range of our deepest vertical nets in substantial numbers (p. 90, Table 13), to April, the last month in which we have observations in Antarctic shelf water, the productivity of the shelf regions is consistently low, and in general does not even remotely approach the corresponding productivity in the open sea. There is evidence, it is true, of some minor concentration of larvae in the shelf regions in March and April, but since all the shelf larvae recorded in these months were in fact recorded from the Bransfield Strait, and since all except the single Metanauplius in March are essenti- ally surface forms, they need not necessarily have been produced in situ but could have drifted into the strait with the Weddell or Bellingshausen Sea water which, as Clowes (1934) has shown, enters it, the former from the east, the latter from the west. Further evidence of the negligible scale of the shelf hatching, again based on the vertical data, is provided by Table 44 in which the disparity in productivity in shelf and oceanic water is shown on a vertical scale. In the construction of this table, which is based on the data from the vertical stations mentioned in the preceding paragraph, the months from which material for direct comparison is available are treated not separately as in Table 43, but as a whole, the production figures, restricted in this instance to the months in which the eggs and succeeding larval stages may be said both severally and by groups to have their optimum range in the plankton, having been obtained in each instance by dividing the total optimum-period-catch of the eggs and the twelve separate larval stages in any particular vertical horizon by the total number of net hauls made during the optimum period at that particular level. Thus (see again Table 43), the optimum range of the eggs and Nauplii has been put at as from November to March, of the Metanauplii from December to April, of the Calyptopes from January to April, and of the Furcilias from February to April. ^ From this presentation of the data, in which fractional results have been disregarded throughout, it will be seen that whereas in oceanic water the eggs and larvae are found down to the greatest depths examined during the spawning season, the larvae in the surface ranging from the First Calyptopis to the Fifth Furcilia stage,- the only substantial product of the shelf seas in the same period is that of eggs alone. 1 Strictly these ranges apply only to the northern or Weddell zone of larval abundance. They do not apply to the southern or East Wind zone where the spawning is late (p. 177) and the larval growth-rate (p. 355) exceedingly slow. * The Sixth Furcilia (Table 43) is, of course, present as well, but in this presentation of the data appears as a fractional result. THE OLDER STAGES 203 Table 43. Vertical net hauls. Monthly production of larvae in shelf and oceanic water compared, the total oceanic catch in roman type, the total shelf catch in italics First Second Meta- Calyptopes Furcilias Nauplii NaupUi nauplii , * v , * ^ No. of Month 1 2 3 123456 stations April — — — 4 55 72 11 25 g8 23 — — 14 — — 1,867 954 1,740 1,162 514 527 98 33 22 8 36 March — — 1 6 sg 205 43 — — — — — 8 — 528 7,467 11,212 1,824 3.264 2,473 625 157 304 246 46 80 Feb. 9 10 26 25 — — — — — — — — 49 — 1,435 2,936 1,037 244 223 256 154 18 I — — 71 Jan. 1 11 20 8 3 — — — — — — — 25 — 347 96 60 19 IS 3 2 — — — — 43 Dec. 2 4 ^ — — — — — — — — — 19 — 13 9 16 I— ______ 25 Nov. — 4 — — — — • — — — — — — 30 — — — — — — __ — — — — 27 Table 44. Vertical net hauls. Larval production in shelf and oceanic water compared, figures for shelf water in roman type, for oceanic zvater in italics. For further explanation see text Calyptopes Furcilias Depth First Second Meta- , ^ ^ , * ^ (m.) Eggs Nauplii Nauplii nauplii 123123456 50-0 — — — — 67642111 — 100-50 — — — — 9 3 9 S — — — — — 250-100 — — — — 32 543 3 — — — — 4— — __________ 500-250 — — — — 5________ 20 — — — — — — — — — — — — 750-500 — — — 2^j — — — — — — — — 1000-750 — — 3 22 — — — — — — — — — 1500-1000 112 — 49 14 — — — — — — — — — The actual depth of hatching in oceanic water is still a matter for conjecture. It is clear, however (p. 98), that it must be very deep, taking place normally far below the range of our deepest vertical nets. Our repeated failure, except apparently once (p. 90, Table 13, Station 2594) to capture eggs and Nauplii below 1000 m., in spite of systematic search covering the maximum possible extent of the hatching period, November to April, leaves little doubt (p. 102) that it is at depths below 1500 m., the bottom limit of our deepest observation, that the new-born krill make their earliest appearance. Between 1500 and iooo m. we have rarely, it is true, struck larvae of any kind, and although it is distinctly surprising that the Second Nauplius and Metanauplius at least (p. 90, Tables 13-15) have so seldom been encountered there, I have little doubt that the answer to this is that we have hardly if ever sampled this deep level thoroughly enough, in the right places, at the right time of year. We have no deep hauls (below 1000 m.) for instance in Weddell West, where, in January and February 204 DISCOVERY REPORTS (p. 301, Fig. 70 and p. 310, Fig. 74), the earliest risings are known to be taking place on a massive scale and where deep, very recently hatched, larvae would then I think be expected, much more regularly perhaps, to be taken in our nets. Even so this is not to say that hatching is not a very deep Fig. 30. Development of sinking eggs in shelf and oceanic water showing how hatching in the shallower conditions gives rise to occurrences of Nauplii and Metanauplii unusually close to the surface. The Nauplii and Metanauplii are partly diagrammatic. event. On the contrary I believe that, as a major phenomenon, it must occur far below 1500 m. and would put it in the cold deep stratum below 2000 m. Our repeated failure to strike even rare specimens of the First Nauplius between 1500 and 1000 m. provides strong additional evidence of how deep in fact the hatching must be. The full significance of this deep, and for all practical purposes exclusively oceanic, hatching will be dealt with presently. In the meantime attention is directed to certain unusual features presented by the occurrence and vertical distribution of the very few Nauplii and Metanauplii that have thus far THE OLDER STAGES 205 been recorded in the shallow water of the Antarctic continental shelf. In the first place it will be seen (p. 201, Fig. 29) it is there, and there only, that the rarely seen First Nauplius has ever been en- countered, and secondly that all three stages, the First Nauplius, the Second Nauplius and the Metanauplius, in their vertical range consistently approach far closer to the surface than their counter- parts in deep water, the Second Nauplius and the Metanauplius, are ever found to do. Obviously, as already stated (p. 98), the absence of the First Nauplius from our oceanic stations is simply because there it occurs beyond the range of our deepest nets. It seems equally obvious that its rare occurrence in shallow water is simply because the few eggs that hatch there, and hatch it would seem on the bottom, must sometimes, in shallow soundings, inevitably be so near the surface on hatching that the First Nauplius cannot fail to come within striking distance of it almost immediately it begins to climb. Bottom hatching in shallow water must also of course explain the occasional near- surface occurrence in shelf seas of the Second Nauplius and the Metanauplius. A more realistic illustration of the near-surface occurrence of the Nauplii and Metanauplii that results from the occasional hatching that takes place in shallow water, as compared with the deep-sited occurrence of these stages that results from hatching in the open sea, is given in Fig. 30, which, again representing a shelf sea with deep oceanic water beyond, shows what evidently must happen to eggs which, supposedly laid not far below the surface in both shelf and oceanic water, subsequently sink, developing towards maturity as they go down. Fig. 30 is based somewhat broadly on Fig. 20 and it will be seen that I have not used a true vertical scale, nor have I pictured the developmental phases of the eggs quite as they would appear if based exactly as shown in the earlier diagram. However, the main purpose of this figure is to emphasise again (i) that such eggs as we have recorded are evidently sinking, (2) that they develop progressively nearer to maturity as they go down, and (3) that in shallow water those that complete the final stages of their development and eventually hatch, it seems on the bottom, cannot help giving rise to Nauplii and Metanauplii much closer to the surface than the Nauplii and Metanauplii in the deep ocean could ever be expected, and are ever in fact found, to be. In fact, since the trend in Fig. 20 shows the penultimate egg stage at depths round about 500-750 m. we must assume that the eggs in the final stage (with fully developed nauplius) must sink from somewhere near 1000 m. to about twice this depth before they hatch. Influence of the sinking shelf water on the liberated eggs Unless some major blunder has been made in the interpretation of the data — as, for example (i) too much having been assumed from the frequency of occurrence of eggs in shallow water, (2) too much from the solitary discovery in a shelf sea of any concentration of eggs at all, (3) too much from our failure to find, except in the surface layers, spent or gravid females in the concentration in which we should expect them to be if spawning were to take place at great depths, and (4) too much from our failure to find evidence of large-scale hatching other than deep down in the open sea, then it seems clear enough that, although the shelf and possibly slope waters of the far western and south-western reaches of the Weddell drift are apparently a major locus of the spawning krill, the bulk of the eggs deposited there, instead of hatching and producing large numbers of Nauplii and Metanauplii in situ, do so on the contrary in the open sea at great depths and at varying distances from the shelf. If, as Fraser (p. 179) originally suggested, the eggs when laid do in fact normally sink below 1000 m. before hatching it would follow that those that are laid in shelf water, as they continue to sink, must inevitably come in contact with the bottom itself, there presumably to lie until they hatch. Since, how- ever, such as do hatch there are so rare it seems possible that all but a very few do not remain in the places where they fall but get carried away to hatch at deeper levels. There is only one movement known 2o6 DISCOVERY REPORTS to Antarctic hydrology that could be held responsible for such a phenomenon — the sinking of the cold highly saline shelf water which as Deacon (p. 99) has shown is formed in winter, on a major scale in the south-western part of the Weddell Sea and on a lesser scale between the South Orkney and the South Shetland Islands, and the subsequent spreading of this water northwards and eastwards below the Weddell drift as the Antarctic bottom current. This, for the time being at any rate, hypo- thetical conception of transport from shelf to deep water is illustrated diagrammatically in Fig. 3 1 , which as before shows a shelf sea with deep oceanic water beyond, the sinking from the shelf of the cold water supposedly carrying eggs, the hatching in deep water and the subsequent developmental ascent. It is true that this mechanism cannot easily account for the frequency of eggs and scarcity of early larvae we record in the Bransfield Strait, sometimes 100-200 miles away from the oceanic depths outside. There is something yet to be explained here,^ but it is a mechanism which may well apply to eggs laid on parts of the Weddell shelf and slope which are near enough for transport into oceanic depths before they hatch. The eggs for instance recorded near 55° W beyond the eastern end of the Bransfield channel (p. 298, Fig. 68) are not in fact very far away from the deep sea. Speculative although this hypothesis may seem it is nevertheless not entirely without substance, for at least we have reason to believe (p. 100) that the cold bottom water flowing eastwards under the Weddell Sea is carrying eggs, and doubtless First and Second Nauplii as well, and carrying them it seems (p. 213) with considerable velocity. It seems, too, to have some measure of claim to be a working hypothesis, providing as it does a not unduly far-fetched explanation, in so far at least as one can search in Antarctic hydrology for one, of our otherwise inexplicable failure to find substantial evidence of hatching in the shelf waters in which, or close to which, it seems so much of the spawning takes place. Whatever substance however it may eventually prove to have one thing seems certain. In view of the lack of evidence of mass sinkings of eggs (p. 185) to hatch at deeper levels and of our failure so far to find the spent and gravid females at the great depths where clearly the eggs are hatched, only intensive exploration of the cold bottom water below the Weddell and East Wind zones and the subsequent discovery, not once or twice, but consistently and over a wide area, of substantial numbers of spawning females there, will prove it to be substantially wrong. In the meantime, it may be observed, the distant off-shore occurrences of deep recently hatched larvae recorded below the Weddell drift and the coastal or near-coastal occurrences of the deep stages recorded below the East Wind drift (p. 200, Fig. 28) are not inconsistent with the part I have supposed the bottom water to play in the dispersal of the eggs and naupliar stages. For as Deacon in the passages already quoted has shown (pp. 99-100) the cold water does not sink everywhere from the shelf to flow away as a major current — it does so only in the Weddell Sea, the higher salinity and temperature of the warm deep water elsewhere forming all round Antarctica an effective barrier to such a movement. If, however, the shelf water, supposedly carrying eggs, were in fact sinking all round the Antarctic continent and flowing away to the north we should expect to find, not as we do find in the Weddell region only, but all round Antarctica, deep concentrations of Nauplii and Metanauplii much farther away from the land than our observations have shown them to be. The inshore or near-coastal distribution of the deep larvae in the East Wind zone, therefore, is only it seems what might be expected if the conception of transport from shelf to deep water is correct and the sinking Weddell shelf water the instrument engendering the eastward dispersal of the shelf and slope laid Weddell eggs I suspect it to be. By whatever chain of events the eggs and Nauplii come to be in the bottom water, deep transport to the east of course implies that both must spend some considerable time there if they are to be carried any distance along. There are some grounds for supposing they do. Heegaard (1948), for instance, finds that in Meganyctiphanes norvegica the period, spawning to hatching, might last for ^ It may be (see p. 121) that the hydrostatic pressure at the bottom of the shelf seas is too low for successful hatching. THE OLDER STAGES 207 5 or 6 days/ and for all we know in the sub-zero conditions in which it seems the krill eggs are laid the corresponding development might take longer. We have to consider, too, the period hatching to Metanauplius or First Calyptopis. We do not know this in E. superba, but in M. norvegica it seems (Heegaard, 1948) it might take up to 15 days or even more. Again, in the krill a longer period might elapse. Thus an unsegmented egg, liberated say in the slope waters of the south-western Weddell Fig. 31. Supposed influence of the sinking shelf water on the Hberated eggs. The Nauplii, Metanauplii and First Calyptopis are partly diagrammatic (see pp. 205-8). Sea, might spend a minimum of 21 days in the bottom water before, as a Metanauplius or First Calyptopis, it came within reach of our searching nets which in Weddell West and Weddell Middle did not go deeper than 1000 m. If the speed of the deep current were 2 knots^ the eggs and resultant larvae could, therefore, it seems, be carried at least 500 miles to the east. Although we do not know the speed of the bottom water there is some historical evidence (pp . 2 1 2- 1 5 ) that it may be higher than is generally supposed. Bruce, for instance, was particularly impressed by the strength of the deep current that on occasion seemed to be sweeping his trawls, already heavily overweighted, off the bottom below the Weddell Sea, and he calls attention, too, to the famous occasion when Ross failed to strike bottom in this region, his sounding line evidently having been swept away by a current of similar force. It is possible in fact, as Deacon (1959) has recently said, that the bottom water has a transport comparable with that of the Gulf Stream. With fishes, as Simpson (1956) observes, an important source of error in positioning spawning from captures of eggs or larvae lies in the drift of the eggs and larvae away from the places where the eggs are laid. He notes that where currents are strong the error will be particularly large for many winter- 1 In the eggs of Thysanoessa inermis and T. raschii the same period is said to last from 6 to 7 days (Ponomareva, 1959). ^ Such a speed is by no means impossible. The recently discovered Cromwell Current, the great underwater stream that flows below the equatorial Pacific, is said to have a maximum velocity of 3 knots at a depth of 500 ft. (Knauss and King, 1958 ; Boehm, 1959; Deacon, 1960(2; Charnock, 1960). 2o8 DISCOVERY REPORTS spawned eggs and larvae, where development is slow, but small for fast-developing eggs such as, for instance (Ahlstrom, 1943), pilchard. In the high latitudes where the krill are spawning, in so far at least as temperature goes, winter is virtually perennial, and so all in all it seems the southern stage might be well set for a wide margin of drift between captures of Nauplii or Metanauplii and the spawnings from which they spring. Although over the great expanse of shallow soundings at the head of the Ross Sea (p. 125, Fig. 13) an immense body of intensely cold, highly saline water must be formed in winter, just as in the Weddell Sea, it is unlikely as Deacon (1937, p. 115) has said that it can escape to the north to flow away as a bottom current because of a ridge (Pennell Bank) extending across the northern entrance to the gulf. Herdman (1948 a), however, has shown that close to the Victoria Land coast there is probably a gap in the ridge through which the cold water it seems might in fact escape, although if any does so, as Fleming (1952) suggests, its effect is likely to be small. Our repeated failure to strike the Nauplii and Metanauplii in the deep oceanic water to the north of the Ross embay- ment during the spawning time, particularly in January, February and March, may well, therefore, be associated with the absence of any major outflow of bottom water in this locality. Deep spawning reconsidered Although our observations point to a preponderance of spent and gravid females at the surface (p. 188, Table 39) and to their virtual absence far below^ we cannot altogether disregard the possibility that in oceanic water the gravid females may sometimes sink or swim down to great depths in order to lay their eggs. For if this be ignored it is difficuh, if not impossible, to explain, in terms of deep transport in the bottom water from a distant locus of spawning in Weddell West, the substantial occurrences of Nauplii and Metanauplii we record in Weddell East (p. 90, Table 13) at Stations 2594 and 2346 in February and April. The difficulty is obvious, for Station 2594 in 00° 1 17' E lies approxi- mately 1500 miles, and Station 2346 in 19° E approximately 2000 miles, east of the Graham Land coast, and if deep transport of eggs from this distant locality was in fact responsible for the presence of the Nauplii and Metanauplii at, for instance. Station 2594, it would have to be assumed— supposmg the larvae to have had their origin in a far western spawning that took place in mid-season, say about the middle of February— that since Station 2594 fell on the last day of that month, the bottom water was travelling at about 4 knots, approximately 100 miles a day, which while possible (pp. 212-15) is difficuh to believe. It is even more difficult to explain, in terms of deep transport, the occurrence still farther east of the Metanauplii at Station 2346, for in this instance, again supposing a mid-season spawning, it would have to be assumed that, since this station fell on 27 April, the bottom water had been carrying eggs that did not hatch, or if they did hatch, without the resultant larvae rising, for a period of close upon 2 months if not more. This again seems hardly credible. In short, these oc- currences of deep larvae so far from Weddell West could only be explained in terms of deep transport if the bottom current were travelling at some phenomenal speed or again if the life span of the developing eggs, Nauplii and Metanauplii were spread over an incredibly long period and the de- velopmental ascent took equally as long. In so far, however, as the deep larvae in 0° are concerned there might it seems be other suitable shelf or slope regions, as for instance off the Princess Martha- Coats Land coast (p. 58, Fig. 4), over which the krill might be spawning and away from which the cold water might be sinking, localities from which it might well be that the eggs and developing Nauplii could be carried as far east as Station 2594, if not in fact beyond, without assuming an unduly fast rate of travel or an abnormally long lapse of time. In regard to this suggestion it may 1 The net result of well over 1000 oblique and vertical hauls made below 500 m. during the breeding season (Tables 38 and 39) was the capture of 14 adult females of which four were gravid and one spent. THE OLDER STAGES 209 be of considerable significance that it is in the neighbourhood of the Princess Martha-Coats Land coast (p. 213) that particularly strong movements in the cold deep stratum have been reported. Although our nets so far have not revealed it happening, the migration of the gravid females to great depths would appear to be well within the powers of this large and vigorous euphausian. It seems in fact it could happen quite quickly, Hardy and Bainbridge (1954) having shown experi- mentally that its lesser counterpart in northern waters, Meganyctiphanes norvegica, is able to swim vertically upwards at a speed of 92-8 m./hr.^ for i hr. and vertically downwards at a speed of 128-8 m./hr. for the same period. In so far as the much larger krill are concerned, if the downward speed recorded for M. norvegica could be sustained in nature, sustained that is over long periods, it would mean that the gravid females could swim down to say 2000 m., spawn, and return spent to the surface in a matter of about 32 hr. Indeed, if the sustained horizontal speed of the krill against the tidal stream directly observed by Gunther (p. 155) is any indication of their capabilities in the vertical plane, it would appear that they could accomplish the same journey inside 8 hr. In other words they would go down and up in a night. If a deep spawning migration did in fact take place so rapidly, and above all if it were undertaken by single individuals and not by the massed formations of females to which our evidence seems to point (pp. 233-4), it could conceivably have escaped our notice simply because the area in which we have operated is so vast and our observations in it, comprehensive though they be, are yet so widely scattered. We are always, however, confronted with the fact that such a rapid descent as we are postulating would only be possible if the gravid females were utterly indifferent to the enormous changes in pressure involved. For if they do go at high speed to great depths to spawn, down to say 3000 or 4000 m., they would encounter quite suddenly pressures of from 300 to 400 atmospheres, and although Regnard (1891)2 has shown experimentally that Cyclops and Daphnia have recovered after 5 min. at 600 atmo- spheres ( = a depth of about 6000 m.), we do not know if predominantly surface living animals like the Antarctic krill could equally endure and survive such enormous environmental change. That they must both endure and survive it, if the deep spawning we are postulating is a fact, is clear from the pronounced occurrence at the surface (Table 39) of those that have shed their eggs.^ They have in fact even been reported (Bargmann, 1945) actively feeding there.* In the present state of our know- ledge, therefore, the most we can say is that deep spawning, if it takes place, and above all if it takes place rapidly, postulates in the gravid females an astonishing measure of accommodation to gross environmental change, and that whether or not they can in fact adjust themselves to it perhaps only future experiment, backed by the extensive use of large, very deep, closing nets, will eventually show. In the meantime it might be noted, in the face of the Regnard experiments, as Russell (19276) remarks, pressure does not seem to be a factor of very great importance in the vertical distribution of plankton animals. Finally, it might be observed, although the large haul of eggs at Station 2594 is suggestive of deep oceanic spawning, their developmental condition, the majority of them containing advanced or very advanced Nauplii (p. 184, Fig. 22), points rather to their having been laid much nearer the ^ An even higher experimental rate of cUmb is claimed for this species by Bainbridge (1953) who states that it is known to be able to swim vertically upwards at a speed of about 128 m./hr. for several hours on end and to be capable of bursts of climb as high as 271 m./hr. We must always bear in mind, however, that although these phenomena have been observed in the laboratory it does not necessarily follow that the same would occur in nature (Moore and Bauer, i960). ^ Cited by Russell (19276). ^ The swarms of spent females recently reported by Nemoto (1959) from whales' stomachs would also postulate that if they do go deep to spawn they must return alive to the surface. * In his field notes Mr Fry records that a gravid female taken on 19. i. 38 was placed that day alive in a jar on the deck of 'Discovery 11'. On 22. i. 38 it shed its eggs. It was alive the night of 25-26. i. 38 when it moulted. It was killed and pre- served with its moult on 26. i. 38. 23 DM 210 DISCOVERY REPORTS surface, subsequently sinking and developing towards maturity as they went down. For in the unlikely event of the sinking eggs becoming neutrally buoyant at say 1500 m., and continuing to develop there without further sinking, we should have expected them to be, if in fact shed at this level and captured immediately or very shortly after, largely unsegmented (p. 184) and not in the very advanced state of development in which so many of them were found. It is obvious, of course, from their condition that the majority must have been captured long after they were laid since the eggs when first laid are unsegmented. If then they did not as suggested sink from a higher level, but were in fact laid where captured, it must be supposed that they sank still farther, subsequently being returned to the 1500 m. level, many of them now practically ready to hatch (p. 102), by the marked upwelling of bottom water that had evidently taken place in this locality. Conclusions and future research Although certain facts seem clear, as for instance (i) the enormous scale of the spawning that appears to be associated with the shelf and slope waters of the far western Weddell drift and, in a slightly lesser degree in higher latitudes, with the shelf and slope waters of the continental land, (2) the absence, or virtual absence, of both spawning and hatching on the South Georgia whaling grounds and in the West Wind drift, and (3) the enormous scale of the deep hatching that occurs in the oceanic water of the Weddell and East Wind zones, much obviously has yet to be done before the whole complex problem of the spawning can finally be solved. Our failure, for example, to find satisfactory evidence of large-scale spawning deep down in oceanic water, despite the scale of the hatching there, our failure equally to find satisfactory evidence of large-scale hatching in shelf seas, and above all the influence it seems the sinking shelf water and resultant bottom current may have on the liberated eggs, are all matters that call for much further investigation. It seems clear that future work should include the following : (i) Systematic exploration from November to April, with vertical closing nets, of the cold bottom water at great depths below the Weddell stream, special attention being given to depths between 3000, or even deeper, and 1500 m.^ This it seems certain would fill in the missing detail of the commencement of the development ascent, bring to light the masses of hatching eggs that surely must exist somewhere between these levels, reveal the true bathymetric range of the First Nauplius and demonstrate whether or not its existence was ephemeral. (2) The repetition of these observations, with the same objects in view, in the oceanic water of the East Wind drift. (3) Systematic exploration from November to April, with large towed closing nets, at great depths, between 3000, or again deeper, and 1500 m., of the oceanic water of the Weddell and East Wind zones. This, if done thoroughly enough, would no doubt demonstrate conclusively whether or not the females go deep to spawn, and if they do so as scattered individuals or as swarms. (4) Systematic exploration from November to April, with both vertical and towed nets, in the shelf and slope waters of the western Weddell drift and of the Antarctic continental land at all levels from the surface to the bottom, particular attention being paid to the sea bed where it seems the eggs might be laid and where, in slope waters at least, they would certainly appear (p. 202) to be hatching. ^ To save precious ship's time, and achieve rapid serial sampling at these great depths over as wide an area as possible, a prime essential will be the development of a multiple-sampling vertical net on the lines described by Motoda (1953). Using such a net, with small terminal store-nets revolving into position, Motoda was able to obtain five separate, successive vertical samples in a single haul. A multiple vertical plankton sampler, employing different mechanical principles, has more recently been described by Be, Ewing and Linton (1959), and Motoda (1959) produces many other ideas. Indeed, as Sysoiev (1957) observes, one of the major tasks confronting oceanographical instrument-makers today is the construction and perfecting of apparatus for observations at great depths. THE OLDER STAGES 211 Intensive exploration of the shelf waters at all levels, although primarily directed towards determining whether the spawning that takes place there is a near-surface, bottom or near-bottom phenomenon, might at the same time help to clear up the mystery of our failure so far to produce evidence of large- scale hatching in these shallow conditions, a failure which suggests the possibility that the majority of the shelf or slope laid eggs might get carried away. (5) Direct measurement, using the neutral-buoyancy floats that Swallow (1955) has recently employed with such success below the Gulf Stream (Swallow and Worthington, 1957; 1961), of the rate at which the bottom water is travelling below the Weddell stream. No doubt there are other matters that might seem to call for attention, such as, for instance, our failure to find anything but the most meagre evidence of both spawning and hatching on the South Georgia whaling grounds. Much of what still remains in major doubt, however, will, I feel sure, be largely cleared up as a result of the measures suggested. Finally, in view of the immeasurable abundance of the krill in the Antarctic seas and the fantastic numbers of eggs that must be liberated annually to maintain such a population, I feel sure there must be some simple explanation of the so far almost complete failure of our nets to sample the eggs in the massive concentrations in which they must naturally occur in the sea. Our failure, except once, to find the hatching eggs is of course easy to explain, for they, except it seems for very rare instances, lie far beyond the range of our deepest vertical nets. But if they are laid near the surface, where the mass of the gravid and spent females seems to be, then surely somewhere in the enormous area we have explored, we should have struck them there, or struck them, as they went down, at some deeper level. It is not that they are too small to be readily taken in the vertical net, for twice it has sampled them with at least some measure of success. Nor it seems can it be that they are so widely scattered that only negligible numbers can be captured, for again everything points to the contrary, that they occur naturally in concentrations, as for instance at Stations 540 and 2594. The gravid swarm in fact (p. 250, Fig. 53) must liberate an enormous mass of highly concentrated eggs, the ripe females (Bargmann, 1945) being capable of producing over 11,000 ova each, while the concentrations of Nauplii and Metanauplii so regularly encountered deep down in oceanic water suggest strongly that neither can spring from a scattered stock. I feel sure, therefore, that the mass of the eggs must be laid somewhere where we have not explored at all, or not explored very thoroughly, and the inaccessible south-western side of the Weddell Sea and the slope waters of the continental land seem the most likely places. For it must be significant, in so far as the slope waters are concerned, that on the rare occasions when we approached the continental land, as for instance at Stations 1671, 1713, 2600 and 855 (p. 93, Table 14), Metanauplii and First Calyptopes, obviously sprung from recent deep hatchings, were encountered rising towards the surface, and that on the still rarer occasions when we worked right on the slope itself (Stations 1662 and 2603) we struck the Nauplii and Metanauplii very close to the bottom. To maintain the krill population at its existing fantastic level, having regard to the enormous wastage it must suffer during the larval phase, the number of eggs liberated annually must in fact be astronomically large.^ Thorson (1946), for instance, suggests that the 478 or more million eggs released in under 5 months by the Californian sea hare (MacGinitie, 1934, 1935) are produced with the expecta- tion that only one of the resultant larvae will survive to reach maturity. Somewhere, therefore, in these southern waters, at certain times and in certain as yet it seems unexplored situations, there must be vast masses of eggs, presenting an easy target for the right apparatus if used at the right depths, in the right places and at the right time of year. Whether in fact our apparatus is the right one or not, ^ Parrish, Saville, Craig, Baxter and Priestley (1959) estimate that a single patch of herring eggs measuring 320 by 320 m., distributed in an almost continuous carpet over the sea bed, at its maximum might contain 258,000 million eggs. 23-2 212 DISCOVERY REPORTS certainly it seems it has only been used once or twice (with some small success) where these masses perhaps may be. No doubt the shortness of our vertical hauls contributes something to the scarcity of the eggs we find, and it may be that the gatherings of our towed silk nets, comparatively few of which have yet been examined, will reveal substantially larger numbers. I did, however, make a special point of examining the catches of our oblique silk nets (N70B), wherever five or more eggs were encountered in the vertical series, in the expectation that these longer-term hauls, lasting 20 min. in the surface ( 1 00-0 m.) layer and for half-an-hour at deeper levels, would give better results. They have not in fact done so. Some it is true have produced larger samples, but not significantly larger, none, how- ever, throwing any further light on where the masses of the eggs must be and none yielding the sub- stantially larger samples one would expect from nets fished for so long. The majority in fact show fewer eggs than the vertical nets, or none at all, and since the oblique nets do not fish precisely the same water as the vertical nets, there would appear to be a distinct possibility that such few sinking eggs as we have been sampling are rather limited in lateral spread. At Station 540, for instance, where we struck one of the largest recorded concentrations of eggs in our vertical nets, there were four at 50-0 m., 52 at 100-50 m. and 478 at 250-100 m. In the obhque (155-0 m.) net fished at the same station, but not quite in the same water, there were only 15. If the eggs sampled here in those 250 sec. had in fact been widely distributed, both horizontally and vertically, then the oblique net fishing for 1 2 times as long in a substantial part of the same vertical horizon would surely have sampled them with conspicuously greater success. A possible flaw in this argument, however, is that they might have been largely con- fined to a narrow horizontal stratum and so would not in any case have offered all that much better a target for the oblique than for the vertical net. SPEED OF THE COLD BOTTOM WATER It has long been thought that the movement of water in the deep sea, especially near the bottom, is virtually absent or extremely slow. The recent work of Swallow and Worthington (1957; 1961), however, below the Gulf Stream has revealed, on the bottom itself, a surprisingly strong movement, and although as yet no direct measurements have been made below the Weddell Sea there is historical evidence that there too the bottom water may be travelling faster than has hitherto been supposed. In March 1843 Sir James Clark Ross in the 'Erebus' and 'Terror' (Ross, 1847, vol. li, p. 363) and in the same month 61 years later W. S. Bruce in the ' Scotia' (Bruce, 191 1, pp. 156-7) both experienced what clearly must be supposed to have been abnormal interference with their deep-sea gear, Ross when sounding near 68° 14' S, 12° 20' W, Bruce when dredging near 71° 22' S, 16° 34' W. Both positions are located in deep water in the eastern part of the Weddell Sea, not far from the Coats Land-Princess Martha coast and directly over where some of the coldest bottom water is now known to lie. Ross, the pioneer, recounts his experience in the following passage. 'After a gentle air from the S.W., which dispersed the clouds, it fell perfectly calm; and the swell having subsided, the boats were lowered to try for soundings. Owing to our having always struck ground in less than two thousand fathoms in other parts of the Antarctic ocean, we, unfortunately, had only four thousand fathoms of line prepared, the whole of which ran off the reel without reaching the bottom'. He did not suspect that anything had gone wrong and for long this sounding of '4000 fathoms no bottom', or Ross Deep as it came to be known, was accepted as the deepest ever to be recorded in the Antarctic.^ It was not 1 In a world map compiled by J. G. Bartholomew for Mill (1903) the Ross Deep, on the evidence of this single sounding, appears as an enormous area of water deeper than 4000 fathoms far south in the Weddell Sea, covering 20° of meridian between 65° and 70° S. It appears again conspicuously as the 'Fosse de Ross' in Richard's bathymetric chart of the Atlantic published in 1907. SPEED OF THE COLD BOTTOM WATER 213 until 61 years later that it was finally disproved by Bruce who, when sounding over the same position in March 1904, got 2666 fathoms, ' blue mud '. While paying characteristic tribute to the pioneer work of Ross, Bruce (1905, 191 1, p. 173) points out that the cumbrous hempen line Sir James had used had on this occasion ' evidently sagged, after the weights had touched the bottom — if they touched at all — the Une being carried away by the strong currents that exist in that region ', and from Bruce's own account of this locality which I quote below there can be little doubt that the strong currents to which he alludes are in fact to be identified with the movement of the Antarctic bottom water. It seems fairly certain that Bruce's explanation of this now famous hydrographic failure is correct, and that the more obvious one that Ross had simply paid out too much line through failing to notice his sinkers had reached the bottom does not apply. For Ross, now nearing the end of his long circumpolar voyage, was by this time well practised in the use of his primitive sounding gear, with which, despite the immense labour^ involved in its operation, he had hitherto been highly successful. Conditions, moreover, were ideal for sounding and there seems every reason therefore to suppose that but for some such unusually strong interference by undertow as Bruce himself was later to encounter, Ross would have been aware he had struck bottom that day just as he had been so often before. That in fact he had was demonstrated long afterwards by Gould (1924) who, from data supplied to the Admiralty by Ross in July 1843, plotted the times taken for each 100 fathoms of line to run out, showing from the sudden and pronounced increase in the retardation of the speed of the descending line at 2200 fathoms, that it was clearly at about that level that contact had been made. Both Brown (1923) and Gould (1924), however, call attention to the deep interference Ross must have encountered. Brown remarking, 'Thus ended the mythical Ross deep. Undoubtedly the strong under-current had caught Ross's lines and misled the world for sixty years '. Bruce's own experience in this region is equally interesting. On 19 March 1903 he records that 'in a depth of 122 1 fathoms, the trawl was lowered, putting out 2000 fathoms (= 2\ miles) of cable, but it did not touch the bottom '. Here again it must be emphasised that since Bruce, like Ross before him, was by this time already well practised in his craft, having developed a highly efficient deep- water trawling technique, it is unlikely that his failure to reach bottom in this instance was due to any mishandling of his gear. The only way it could be accounted for he says was 'that there were strong under-currents sweeping the trawl off the bottom '. It was not, moreover, an isolated experience for it had 'occurred more than once in this locality' and this in spite of the fact that the trawl warp, against this very contingency, had been heavily overweighted with ' four furnace bars, each weighing 22 lbs., and two olive-shaped weights, each of 20 lbs.'. Pirie and Brown (1905), referring to the diffi- culties of sounding in Ross's time, also call attention to the strong undercurrent found by the ' Scotia ' below the Weddell Sea, stating that it 'was a source of great trouble, and on two occasions prevented the trawl from reaching the bottom, while on a third we had to pay out about 1000 extra fathoms of cable before this was effected '. These incidents seem to have left a profound impression on Bruce's mind, for in a later chapter (pp. 183-7), speculating on the possibility of a northward spread of benthic life from high latitudes through the agency of a strong bottom current having its origin in the Antarctic, he writes, ' That there is a strong underflowing current south of 70° S., in the vicinity of Coats Land is certain, for on three occasions the " Scotia's " trawl was prevented from reaching the bottom evidently having been swept [away] by such a current'. 1 In a letter to Bruce, quoted by Rudmose Brown (1923), Sir Joseph Hooker refers to the 'almost superhuman attempts' of his old commander to sound at great depths, adding that he was not surprised that now and then he should have failed. Ross's deep soundings were carried out from two open boats moored fore and aft, the forward boat keeping the after, which carried the sounding reel, head on to wind and sea. After the line had been run out it had to be laboriously manhandled back on to the drum. 214 DISCOVERY REPORTS Referring to the absence or great scarcity of diatoms in the blue mud deposits of the Weddell Sea and their immense abundance in its surface waters Pirie (19056) suggests tentatively that the sinking cells may be swept away by the bottom current off to the north and east to be laid down as the great belt of diatom ooze that in the South Atlantic (Murray and Philippi, 1908; Hough, 1956) is found principally between 50° and 60° S. Drygalski (1928), somewhat more emphatically, expresses the same opinion. In more recent times we of the Discovery Investigations recall vividly the unusually loud and clear- cut echoes we used to receive when sounding over the very region traversed by Ross and Bruce, fairly conclusive evidence that the bottom there is hard, and hard perhaps because it is swept clean of echo-deadening ooze by a current of considerable power. Moreover, since the bottom water can be traced over an immense distance to the north, Wiist (1933) having shown that in the western Atlantic basin it flows for nearly 7000 miles directly to 40° N, it seems reasonable to conclude that in the high latitudes near its source it is moving with considerable velocity.^ In the Indian Ocean too it travels a long way, Sewell (1948) suggesting that, as a carrier of certain Antarctic copepods, its influence may extend as far as the Bay of Bengal. If it does in fact extend so far^ this would again imply rapid transport, for copepods are short-lived and to survive carriage over such an immense distance they would have to travel quickly. In his recent review of the principal Antarctic water movements and their climatological and zoogeographical significance Deacon (1959) has written: We have a fair knowledge of the water masses and a qualitative idea of their movements but cannot get much farther without direct measurements of the currents and the water circulation at all depths By analogy with what we begin to know about the North Atlantic ocean the deep currents probably travel i or 2 miles a day, but there is some evidence from the behaviour of sounding lines and dredges and from the movements of plankton that the bottom current from the Weddell Sea may be stronger: and, judging from its effect on the bottom temperature in all the deep channels leading from the Antarctic ocean, it must have a transport comparable with that of the Gulf Stream.^ Perhaps there are other phenomena contributing to violent bottom movements below the Weddell Sea. It is becoming widely recognised (Heezen, 1959) that turbidity currents engendered on the continental slope may now and then travel down the slope with tremendous force, their influence extending far out over the abyssal plain. Below the great expanse of shelf water to the north of the Ross Ice Barrier the bottom is covered by a deep glacial mud of such exceedingly fine texture that it feels soft as velvet to the touch. A similar deposit (Wordie, 1921 a) is found at the head of the Weddell Sea where, by virtue of the proximity of the vast Antarctic ice-sheet and its enormous capacity for depositing terriginous material, it could be building up until its load became unstable and so from time to time (Heezen, 1959) trigger off a turbidity current on the adjacent slope moving away at enormous speed. Such a current was triggered by the earthquake which shook the continental slope south of Newfoundland in November 1929, Heezen and Ewing (1952) calculating that it had moved off down the slope at 50 knots and was still travelling at well over 12 knots far out over the abyssal plain well over 450 miles away.* 1 A recent re-examination of the data obtained by the 'Meteor' expedition reveals unexpectedly high velocities in the bottom water as it flows along the foot of the continental slope off South America, Wust (1955) calculating that there it may be reaching a maximum speed of 8 nautical miles a day or about a third of a knot. 2 It seems unlikely in fact that it docs, Sverdrup (1954) pointing out that although the deep northward flow is particularly conspicuous in the Atlantic, where it can be traced to at least 30° N, in other oceans it reaches only to about 20° S. ^ Surface velocities of as much as 4^ and 5 knots have been recorded in the Gulf stream near Cape Hatteras (Malkus, 1953 ; Stommel, 1958). * Bruun (19576) suggests that such currents, starting perhaps as slides on the slopes, may carry small animals in suspension for long distances, while Laughton (1959), in a recent review of the forces at work in moulding the face of the deep-sea floor, calls attention to their far-reaching influence as sediment carriers and to their enormous initial velocities. SPEED OF THE COLD BOTTOM WATER 215 A combination of several physical and topographical phenomena is required to produce a turbidity current. I quote from Heezen (1959). Turbidity currents first of all require sediment, just as rivers require water, without it they do not exist. Turbidity currents require a trigger mechanism ; this can be supplied by earthquakes, hurricanes impinging on the shore, high bedload discharge of rivers; and lacking a trigger mechanism, turbidity currents probably can occur after long continued deposition simply by slope failure resulting from over-steepening of a depositional slope. Finally, turbidity currents require a slope. The optimum conditions for the generation of a turbidity current probably would be a large body of fine, well-sorted sand triggered by an earthquake on a steep slope. Some of these requirements exist at the head of the Weddell Sea. There is a wide shelf covered by an ultra-fine terriginous deposit, the hurricanes are there, there is a steep slope, and above all there is continuous, high and probably rapid bedload discharge (Lisitzin, i960) from the ice-girt land that could well, it seems, lead to a slope failure such as Heezen describes. CAUSES OF PATCHINESS Some early writers on whales and whaling have sought to link the accumulation of the feed into surface patches with deliberate action on the part of the whales themselves. Perhaps the earliest and most fanciful story of such action is recorded by Sagard-Theodat (1632) in his account of the whales of the Gulf of St Lawrence. I quote from True's translation (1904). 'I was very much astonished by a Gibar which with its fin or its tail (for I could not well discern or recognize which it was) struck so terribly hard on the water, that one could hear it for a long distance, and I was told that it was to surprise and mass together the fish, in order afterwards to swallow them '. Or again, to quote Dudley (1725), fin whales, 'with a short turn, cause an Eddy or Whirlpool by the Force of which, the small Fish are brought together into a Cluster; so that the whale with open Mouth, will take in some Hundreds of them at a Time'. Later writers (Buchet, 1895; Racovitza, 1903; Millais, 1906) repeat this story, Millais stating that whales when feeding make 'one or two subsidiary circles to drive their prey together '. Close observers of the habits of whales in more recent times however do not record this practice and it must be regarded as a fanciful belief which probably arose as Gunther (1949) suggests as a fisherman's explanation of the patches and their frequent association with ' boltering '^ whales. In any case, as Gunther points out, pronounced and widespread patchiness has been observed in areas marked by a conspicuous absence of whales, which need not, therefore, be instrumental in bringing it about. Although it may have nothing to do with the formation of patches, circling it seems is in fact occasionally practised by feeding whales, the following passage suggesting that in existing patches it could cause still denser crowding of already densely congregating krill. I quote from Ingebrigtsen (1929) on the humpback. It employed two methods of capturing 'krill' when the latter was on the surface of the water. One was to lie on its side on the surface and swim round in a circle at great speed, while it lashed the sea into a foam with flukes and tail and so formed a ring of foam. The frightened ' krill ' gathered together in the circle. This done the humpback dived under the foam-ring and a moment later came up in the centre to fill its open mouth with krill and water, after which it lay on its side, closed its mouth, and the catch was completed. The other method was to go a short distance below the surface of the water, swimming in a ring while at the same time it blew off'. The air rose to the surface like a thick wall of air bubbles and these formed the ' net '. The ' krill ' saw this wall of air bubbles, were frightened into the centre, and then the manoeuvre of the first method was repeated.^ '^ ' Boltering. ' A name given by Norwegian whalers to describe the habit, peculiar to some whales, of swimming or turning on their sides when feeding. Andrews (1909) gives a particularly good description. - This appeared in the reputable Rapport et proces-verbaux des reunions, and although to many it may seem a remarkable and somewhat fanciful account, no one it seems has yet questioned its authenticity. 2,6 discovery reports Effect of animal exclusion In his Hypothesis of Animal Exclusion, Hardy (1935, pp. 273-356 and 1936c) suggests that vertically migrating plankton animals get carried away in patches from zones of dense phytoplankton into zones where the phytoplankton is poor through becoming involved in subsurface currents travelling at different speeds or in different directions from those of the currents on the surface. I quote from page 353- Thus we shall have a series of parallel belts of plankton at intervals at right angles to the current. Now this is a phenomenon which we have observed from the 'Discovery' on a number of occasions; particularly striking were parallel belts oi Pyrosoma stretching across the sea, and possibly the patches oi Euphausia superba, Parathemisto, etc., observed in the consecutive net series^ (Figs. 133 and 134), were in reality belts at right angles or making an angle with our direction of steaming. He adds that if the organisms migrated upwards at 2100 hr. and downwards at 0300 hr., and the water into which they penetrated under moderate phytoplankton was travelling 200 yards a day faster than the surface water, and that into which they penetrated under the heavier phytoplankton i mile a day faster than the surface, then the belts of plankton would be 150 yards across and | mile apart. Such a disposition in the patchy krill has in fact been observed in the sea. Professor Hardy, however, tells me that he did not imply that this hypothesis would explain the swarms of krill we see so often on the surface, and it could, therefore, equally well be, I believe, that the patchiness of this species revealed by his consecutive nets, so strikingly illustrated in his Fig. 134, was simply due to their random hitting, or missing, such swarms, and he readily agrees that this must in fact have been what happened. ^ We must not, however, overlook the possibility that the dynamics involved in exclusion might serve to produce patchiness among the larvae, for they, unlike the adults, are passive enough and as such would be readily susceptible to any subsurface movements that might be operating to displace them away in patches from a supposed exclusion area. If patchiness does so arise, so early in the life-history of this species, then it would seem to follow that the dense concentrations of krill we see on the surface are concentrations of long standing, of a year or more in age. But even in the larvae (p. 109, Figs. 9-11) such vertical movement as there is seems neither to be rhythmic enough nor, for the vast majority of the population, to go deep enough, for exclusion dynamics to be effective. As Hardy remarks an inverse relationship between density of zoo- and phytoplankton has been recognised for many years and there can be little doubt that the grazing of the animals on the plants must in part at least be responsible for it. In the cyclonic movement claimed to have been discovered by the Russians off Enderby Land, however (p. 47), Beklemishev gives an instance of such a relationship which he states was due neither to the grazing of the zooplankton nor to animal exclusion. In the centre of this supposed vortex it will be recalled a dense local accumulation of E. superba, with patches 100 m. across, was found at the surface over a considerable area in which the phytoplankton was poor, Beklemishev ascribing the poverty of the plants to the rapid upwelling of diatom-poor water from below, and the local richness of the animals to the fact that they, too, had been brought to the surface from some deeper level. He adds that this was a good example of the inverse correlation between animal and plant plankton 'unconnected with the displacement of animal by plant life postulated by some writers'. Or to quote him again (Beklemishev, i960) in a more striking passage, 'The barren upwelled "new water" carries along the phytoplankton cells to the • Conducted at Stations 150 and WS 53 on the east side of South Georgia in January 1927. 2 Many years after his Hypothesis appeared Hardy (1956, 1958) writes that his attempts to test its challenging ideas experimentally have so far failed. Perhaps he may yet succeed. CAUSES OF PATCHINESS 217 peripheral region of the cyclone, thus forming a "clearing" in the diatom bloom', in which, he continues, the rising euphausians, ' the later Furcilia larvae and young adolescent krill ', assisted by the upwelling water, accumulate from below to gather in a plantless zone. An interesting hypothesis, but can it after all be reconciled (pp. 157-70) with the general subsurface scarcity of older animals, more particularly of late Furcilias and early adolescent forms, we so repeatedly record, especially in the deep intermediate layer from which (p. 47) Beklemishev distinctly suggests this rising popu- lation may be coming? Beklemishev (1959) has recently invoked this hypothesis to explain gross irregularities in the distribution of the euphausian population as a whole. He starts with the assumption that the krill live deep in the warm intermediate layer, but get carried to the surface by local upwellings in the centres of isolated cyclonic movements along the East Wind- West Wind boundary zone. Thus, since it has been suggested (Ivanov and Tareev, 1959) that such upwellings are in turn generated by atmo- spheric cyclones centred over them, he concludes that the grosser irregularities of the distribution of this species spring ultimately from atmospheric phenomena. This again is an interesting hypothesis but it does not fit the facts, {a) because the krill it is now clear do not live deep in the warm inter- mediate layer, and {b) because our net-hauls in aggregate do not reveal any discrete accumulations of the population along the East Wind- West Wind boundary zone, but suggest that the swarms, larval, adolescent and adult, without any gross irregularity or conspicuous discontinuity, are spread more or less uniformly throughout the East Wind surface stream (p. 60, Fig. 5 a) and that their principal locus of abundance (outside the Weddell current) lies (Fig. 56) to the south of the divergence zone. As Nansen did (p. 72) so many years before him, Bogorov (1938) attributes the intense spring flowering on the fringes of the Arctic basin to the great build-up of nutrients that takes place below the solid pack in winter. Various factors he observes contribute to this massive accumulation, the decay of plankton, the enriching of the surface waters by nutritive material carried up from the bottom, the rich salt load brought down by the great Arctic rivers and above all because, since there is no, or virtually no, phytoplankton below the ice to consume them, the accumulating salts can go on accumu- lating. During this period of ' biological winter', as he has called it, 'when darkness and ice dominate over the sea ', the plankton as a whole is poor, the quantity of zooplankton, however, prevailing a little over that of plant plankton. With the spring break-up and melting of the sea ice and the exposure of the surface to sunlight, the phytoplankton, feeding without stint on the wealth of nutrients accumulated in winter, begins to multiply, rapidly reaching a massive maximum. This intense flowering, or 'biological spring', occurs principally along the ice-edge, in recently opened spaces among the floes and in 'water breaches' in the fast ice, and while it lasts he finds the plant plankton may increase until it outweighs the zooplankton sometimes 120-fold. Soon, however, the position is reversed. The plants, having quickly consumed the salts, begin to die oflF and develop at a slower rate, their decline being accelerated by the massive assault of the rapidly accumulating herbivores feeding on them. Thus, by the time large areas of open water appear ('biological summer'), the zooplankton has become established as dominant. As Bogorov points out, the spring break-up of the Arctic pack does not occur simultaneously all over the polar basin. It may be late in one area and early in another and thus biological spring, possibly even in areas adjoining one another, could be correspondingly late or early. In the area of late break-up phytoplankton would dominate, in the area of early break-up the dominant plankton would be animal. Thus, he concludes, the wide- spread inverse correlation between relative abundance of plant and animal plankton observed in the Arctic seas has its roots primarily in the seasonal and regional history of the polar pack. In other words it seems, if I follow him correctly, the vagaries of the ice-cover could be producing biological winter in areas adjoining others already in a condition of biological spring or even in the sununer state. 24 DM 2i8 DISCOVERY REPORTS Long after he produced this interesting hypothesis the author (1957) referred to it in the following passage. ' We have believed . . . that the absence of large quantities of zooplankton often observed in areas with mass "blooming" of phytoplankton, is a result not of the frightening or "exclusion" of animals but of the seasonal plankton development. If this is true, the plant community is, naturally, first to develop '. The same concept has recently been invoked by Beklemishev (19586) to explain direct and inverse relationships between diatom and copepod abundance in the Okhotsk and West Bering Seas, the major factors involved ' being the movement of biological seasons from the south northward and the difference of seasonal states of adjacent plankton communities '. Many years ago Fish (1925) wrote, 'the great numbers of diatoms filling the water apparently cause conditions unfavourable for animal life of any sort. The macroplankton seems to be literally choked out '. He was not, however, altogether convinced that such a phenomenon could in fact be taking place and merely offered this as a possible explanation of the inverse relationship between density of plant and animal plankton he had observed in the Woods Hole region. Although possibly unfavourable for some animal life, dense concentrations of diatoms may not, however, be unfavourable for euphausians. Lasker (i960), working with the swarming Euphausia pacifica from the Californian coast, finds that if healthy specimens of this species be put into a heavy suspension of algae (enough cells to tint the sea- water green), they ' quickly filter out enough algae to completely fill the intestine, digestive gland, and crop ', suggesting rather strongly that a rich phytoplankton need not be harmful to the southern krill, but rather the reverse. Effect of pack-ice Dense concentrations of larval and adolescent krill are frequently encountered at the ice-edge, particularly in winter and spring, and Fraser (1936, pp. 158-9) suggests two explanations to account for this phenomenon. First, he says, the cutting down of the sun's rays by the ice, and the interruption of the rhythmic vertical movements of the larvae which follows, lead to a concentration of young krill at the surface whatever the hour of the day or night. Secondly, he suggests that the state of the illumination of the winter sea, already poor and now aggravated by the presence of the pack, and the pronounced scarcity of diatoms that goes with it, will result in a movement of larvae from deeper inside the ice-field to areas of richer grazing at its periphery. Such a movement, he says, to some extent at any rate, would explain the dense concentrations of young krill so often met with there. Whether this or the interruption in diurnal rhythm does in fact explain them is very hard to judge, for no one can yet say for certain what effect, if any, the pack-ice may have on the horizontal or vertical movements of the larval and adolescent krill. I believe it has nothing to do with the forma- tion of these dense concentrations whatsoever, since there now, it seems, can be little doubt (p. 75) that the apparent congregating of both larvae and adolescents at the ice-edge is simply due to the periodic freezing over of the open water of regions already rich in larval krill. It has sometimes been suggested that the krill in an ice-infested sea might congregate underneath the floes, subsequent melting of the ice leaving the sea covered with a haphazard system of surface patches reflecting the patchwork pattern of the open pack. Whether they do in fact so congregate is difficult to ascertain. So far as can be seen, however, it is to the cracks and open spaces between the floes that the krill prefer to keep rather than to the poorly illuminated conditions directly below them. However, neither congregation between, nor under, the fioes would provide a satisfactory explanation of patchiness since in the end a system of patches would be formed far more closely knit than direct observation (p. 148) and the evidence provided by towed nets (p. 148) have shown it to be. It would be difficult, too, to explain the widespread patchiness that has been encountered hundreds of miles from the ice in terms of a patchiness having its origin in the pack. CAUSES OF PATCHINESS 219 The patches as swarms Since the enormous variation in gatherings of the larger krill revealed by our stramin nets is a measure of the pronounced patchiness that is know^n to exist among the older stages of this species, so it may be said the equally great variation in the larval gatherings of the vertical nets, and in the larval and post-larval gatherings of the stramin nets, reveals a patchiness no less pronounced in the younger stages as well. Hardy and Gunther (1935, p. 211) record the patchiness of very young E. superba (which subsequent analyses have shown to be the Calyptopis and Furcilia stages, but principally the former) revealed by a series of consecutive surface net hauls in the Bransfield Strait in April 1927; while Gunther (1936, p. 173) records a similar tendency to form patches among the very young stages of another euphausian, so far unidentified, encountered near San Juan in the Peru Coastal Current in June 1 93 1. In this locality streaks of a reddish surface discoloration proved to be caused by 'an almost incredible quantity of Euphausian cyrtopias . . . affording a parallel to the swarms of Eiiphausia superba which sometimes colour the Antarctic with patches varying from ochre to brick red '. It would appear, in fact, from the enormous variation in catch-figures (Tables 45-7) revealed by the random sampling of our vertical and stramin nets, that E. superba is prone to patchiness Table 45. Vertical patchiness. Patchiness in larval E. superba at depths below 250 m., the figures in bold type indicating patches supposedly sampled with maximum effect. N2, Second Nauplius; M, Meta- nauplius; Ci, First Calyptopis. Stations prefixed by an N or a V are {p. 52) from Norwegian sources 1500-1000 looo-yso 750-500 500-250 • ' « , ' , , * , , * , Month Station Date N2 M C 1 N 2 M Ci N2 M C 1 N 2 M Ci Dec. 762 8 66 — — — — — — — Jan. N 42 18 — — 92 N43 19 _ _ 75 V 1 1 22 — — 2213 1492 23 —243- — — — — — 823 27 327 25 — — 18 2 — — — V 12 28 — — 243 1662 28 5 10 — — — — — — — Feb. 1671 2 — — — — 10 I — — — V13 12 _ _ g 1965 15 _ 61 59 — 2 32 — — 58 N 19 17 — — 19 618 18 — — — — 20 4 — — — 620 19 — 46 — 2 10 — — — 1545 28 _ 138 _ _ 57 78 — — 36 2594 28 1000* 400* — 434 2240 — — — — — — — March 1138 2 — 911 36 — 604 245 — — 485 2600 3 — — — — 39 1—67 5 — — lis 1142 4 _34__i8 — — 5 2603 4 510 94 — 18 387 — — 443 22 — — 20 "44 S — 8 — — 549 53 — — 51 636 8 — — — — 171 — — 31 637 8 — 10 4 — 24 2 — — 36 2610 8 — 12 — — 14 2 — — 10 638 9 — — — — 32 — — 5 647 12 — 557 7 — 3702 282 — — 150 April WS 197 17 _ 115 _____ _ 28 854 20 _ 137 5 _ 47 7 _ _ 35 855 20 — 185 36 — 332 67 — — 13 2346 27 — 830 10 — 221 41 — — 39 * Estimated numbers (see p. 89, footnote 3). 24-2 220 DISCOVERY REPORTS ■Ac o 2: in -« t^ E-H 1:^ (3 & -« e, p^ J^ o O) ^ s -« -Si •Ki VJ t ^ ^ s a H a, 53 3 CO rJS ^t^ w •*-<» « •^ «:) ^ Q «H "«» Q •i^ •53 .^ ^ vO •<*- JJ 2 ca h «0 O N O t^ O 00 so " On U1 N O , 00 M o o f*i r* r^ lo N I N ^0 -^ M I ^ O^oo IH N M 0>vO " N M lO vO vO n M 1^00 OnO N inirjOO M N ro-tf^'1""^r^00000000 O^O^O^OnO N MCO O-O hh n c4 f^^Mi^^ a »> fOf^^r^^nf^f^ c*lf*^mK,vO'OvO ^ mcou^ r^f "^ cvo^ i-.\0 r-^vO olsO fO\0 O^OnO^O O^ror^roc^lt^N O N O '■^ 14 rf CAUSES OF PATCHINESS 221 tS 5 tS I < o IN I; O O o CO in O H "S O N s o O vO >0 M lO Tf t> t^ O 00 H •* ON M M in o> N o m m N I r»J in I ■t 00 o -too t^ "-I t% f^ ro r^ f*^ H O* ro l^ N -t I -t- N N M o invo 00 o ©"Ml- ^0 O M t^ M O \0 N N M o 00 CO Nc*i-1-r^l^00 Of^ooa-MMro-t-fOvOOOOI^OOOv-^ir^O^w^OOTt-MtO ^in\Dr^ Ompju-iON^ >/ioO« ro"-" ro-^ ^ ^ ^^^ ^^ ^ ^^ ^ & S > o 222 DISCOVERY REPORTS Table 47. Horizontal patchiness. Patchiness in larval, adolescent and adult E. superba as revealed by the samples from the stramin nets towed in the surface {100-0 m.) layer, the figures in bold type indicating patches supposedly sampled with maximum effect {Corrected figures, see pp. 5g and 282-3) January Station Date Larvae 11-20 mm. Over 20 mm. Station Date Larvae WS 497 I — SS 26 4 — 575 8 - WS 338 8 — WS901 9 — 348 594 2547 22 546 579 10 — 2168 10 — WS53 II - 2170 II — 2171 II — SS 29 13 — 150 15 — 813 16 - 2185 16 — — 160 2561 27 171 602 19 — 1270 19 — 2195 19 — WS 63 20 — 818 20 — 2198 20 — SS34 21 — WS 372 21 — SS 42 I — WS 552 2 — 1673 3 — 1675 5 — 1507 7 — 351 9 — 352 9 — 354 9 — WS 908 9 — WS 378 10 — WS 909 10 — 1511 II — WS 933 II — WS 380 12 — WS9II 12 — WS38I 14 — I5I7 14 — WS95S 15 - WS385 i6 — WS 386 16 — WS 387 16 — WS 388 16 — WS 390 17 — WS 391 17 — WS 393 17 — 617 18 — WS 956 18 — WS 144 19 — WS 14s 19 — 1-20 mm. Over 20 mm. Station Dat — 444 WS373 21 — 16,629 820 21 — 408 2199 21 — 13a 305 22 348 594 2547 22 — 144 WS535 23 — 2,961 SS36 24 — 2,016 WS904 24 — 924 WS536 25 — 200 WS9SI 25 — 864 WS537 26 — 29,250 WS538 26 288 — 2560 27 — 160 2561 27 — 376 WS541 28 — 748 WSs42 28 — 1,506 2562 28 — 388 WS543 29 — 2,612 WS930 29 13s 495 WS547 30 — 152 1665 30 — 7,468 2567 30 February — 5.624 WS395 19 168 — WS396 19 — 1,116 WS397 19 — 488 619 19 — 8,544 WS399 20 — 15,600 WS147 21 — 384 WS 149 22 — 87,570 WS 152 22 — 17,076 2814 22 — 756 627 23 — 108 1534 23 — 168 1535 23 — 20,000 2244 23 — 204 2815 23 — 258 WS565 24 — 648 1285 24 — 168 1537 24 — 948 2250 25 — 280 WS156 26 — 320 1 541 26 — 1,084 2590 26 — 588 WS159 27 — 22,080 1543 27 — 168 WS162 28 — 144 1544 28 — 264 1545 28 — 1,212 WS163 29 — 14,536 WS164 29 — 18,000 WS915 29 1-20 mm. Over 20 mm. — 3,702 — 460 — 5,544 — 148 2,922 — — 190,000 — 2,828 — 210 — 126 — 948 — 688 — 30,960 264 — 759 108 — 684 745 30,148 280 — — 1,602 — 312 — 642 — 609 ~ 2,892 16,800 — 124 — 4,200 — 816 — 628 — 2,736 — 300 — 2,368 — 1,554 — 20,998 — 140 — 7,212 — 278 — 9,600 — 4,784 — 3,176 — 5,538 — 136 — 160 — 1,032 — 1,084 — 264 — 555 — 720 — 4,912 — 6ai — 416 — 33-925 — 5,544 CAUSES OF PATCHINESS 223 Station WS 165 WS 167 1136 1547 2594 365 1138 H 2600 19 22 1980 2268 2603 WS 917 2606 WS169 WS 172 1 146 1 147 368 639 640 641 643 WSI94 644 "54 23 24 663 1331 1332 1333 207 WS196 2316 2322 WS198 855 1354 1355 1356 1357 1359 Date Larvae Table 47 (cont.) March 11-20 mm. Over 20 mm. Station Date Larvae 2 2 3 3 4 4 4 4 4 5 5 6 6 6 7 8 9 10 10 10 II II 12 14 14 4 4 5 5 7 H H 17 19 20 I 2 3 4 6 6.750 409 870 7,248 150 2,256 8,400 58,488 1,152 2,046 1,086 1-392 240 480 1779 I 1780 2 178 1 2 1782 3 1783 4 1784 4 1786 6 1787 6 1789 8 * The round figures — 1.564 2000 16 — — 7,584 WS937 16 — — 10,248 31 17 — — 2,992 32 17 — — 207 2004 17 — — — 372 18 2,550 — 1,420 38 19 — — 720 373 19 1,710 — 212 374 20 19,200 — 124 375 21 180 — 154 1713 22 — — — 1715 23 — — 10,600 2013 23 — 145 303 2014 23 — 489 120 2295 23 — 1,032 4.452 WS572 24 — — 520 1717 24 — — 9,600 1719 25 — — III WS575 26 — — 107 658 26 — — 21,060 1720 26 — — 4,848 2299 26 — — 132 2300 26 661,992 — 3.552 40 28 — — 10,668 41 28 — — 168 1326 29 2,004 — 120 1327 29 294 — 434 WS527 30 1,720 — 2,576 1328 30 15,648 " 6,140 1330 April 31 6,672 — 360 1342 21 1,104 — — WS202 23 720 — — 1346 23 1,162 — — 1350 26 720 — 385 2344 26 35.520 — 400 1351 27 42,240 — — 861 28 380 — 105 862 28 212 — 1352 28 18,000 ~ 284 1353 May 30 109 — — 1361 8 140 — — WS434 18 — — — WSiio 26 119 — — WS112 27 416 4.704 WS114 June* 28 1,112 — — 1790 8 10,000 — — 1791 9 5.000 — — 1792 9 10,000 — — 1793 10 10,000 — — 1794 10 10,000 — — 1795 II 1,000 — — 1796 II 500 — — 912 24 1,616 — — 2843 24 316 11-20 mm. Over 20 mm 260 — 270 — 197 — 162 1,227 363 2,278 7.440 118 996 200,160 103 100 354 717 108 14,420 1,122 3.877 528 2.499 168 370 148 4,060 120 420 1,388 10,000 10,000 10,000 10,000 10,000 10,000 10,000 10,000 in hundreds and thousands are estimates of the ' large ', ' very large ', or ' enormous ' numbers recorded. 224 DISCOVERY REPORTS Station Date Larvae 11-20 mm. Table 47 (cont.) July Over 20 mm. Station 2851 2852 2853 2854 2855 2362 2856 2857 2363 2364 2392 2393 1388 2394 2395 2396 2397 2398 2399 2400 2405 2406 2407 2408 2409 1403 1404 140S WS276 WS277 1408 1409 1411 WS278 WS279 WS280 2447 1823 WS29S WS296 WS297 1824 WS298 WS299 WS 300 WS301 1437 1825 WS302 WS 303 WS304 1826 WS30S WS306 1833 9 10 II II 12 12 12 13 13 17 17 18 18 18 18 19 19 19 20 21 22 22 22 23 3 3 4 5 5 6 8 12 13 17 3 4 4 4 4 5 5 5 5 5 S 6 6 6 6 7 7 9 7,091 14,976 22,272 1.047 6,192 5.414 1,128 16,233 8,582 15.232 476 1,167 493 8,849 3.229 570 1,260 18,029 741 177 180 5.914 2,871 12,902 10,666 542 300 170 320 100 320 708 36 840 190 402 1,812 104 108 378 •55 748 118 149 15.848 17,664 349 336 601 4.263 954 1,692 328 808 123 6,172 2.235 394 873 949 544 360 134 160 730 172 5.202 2,052 1.552 218 10,872 1,264 1,022 194 134 2,268 366 596 194 181 488 1.939 5.520 402 408 2365 2859 2861 2369 2370 2863 2372 2864 2374 237s August 2410 2411 1393 2412 1394 1395 2413 2414 2415 WS259 WS261 WS263 WS264 WS267 WS268 September 772 3.460 4.756 4,696 660 304 180 WS281 WS282 WS283 WS285 WS286 WS288 2432 2443 2444 244s 2446 October 2,610 288 612 2,720 640 3.768 1. 144 6,504 1834 183s 1836 1837 1839 453 454 4SS 459 460 461 462 463 464 465 2465 467 2467 469 Date 13 13 14 15 15 15 16 16 19 20 23 23 24 24 25 25 25 25 26 28 28 28 28 29 29 17 17 17 18 18 19 25 29 30 30 30 10 10 II 12 13 16 17 18 19 20 21 23 24 26 26 27 28 28 29 Larvae 11-20 mm. Over 20 mm. 2.557 420 134 269 350 336 15.360 48,000 2,760 1,660 5,280 10,997 10,392 ",657 2.436 1,306 4.032 2.432 139 262 598 1,150 122 284 170 452 330 2,880 1.592 13.490 116 182 644 103 4.366 767 189 1,038 1,809 208 2.572 342 4,766 766 3.844 227 152 102 284 180 2,598 207 326 132 1,258 1,306 2,000 1,106 17,170 226 1,548 1,537 1,740 3,645 3,446 1,990 26,530 3,918 162 9,278 2,058 196 23,594 152 415 214 144 356 1,960 4,320 5,400 7,880 244 132 220 « 4,410 216 260 3.000 100 352 CAUSES OF PATCHINESS 225 Table 47 (cont.) November Station Date Larvae 11-20 mm. Over 20 mm. Station Date Larvae 11-20 mm. Over 20 m 471 I 5.770 16,152 — WS486 21 — 502 5.484 472 I 592 1,124 — 500 21 — 516 — 2477 I — 464 — WS487 22 256 7.352 1.376 1003 2 — 156 — WS488 22 100 — 724 2478 2 6,368 11,610 — 502 22 — 288 2479 2 — 2,179 150 2093 22 80 680 — 1854 3 — — 670 WS491 23 — — 1,168 1009 6 — — 904 503 23 — 604 — 1866 9 — 192 — 507 23 818 9.07s 1,132 1868 10 185 560 — 2094 24 — 120 114 187s 13 — 252 1.452 511 25 — — 104 481 14 — 358 — 517 26 — 1.316 4,420 483 H 11,764 40,868 6,192 1039 26 — — 1,092 484 16 11,910 60,209 14.704 2100 26 — 617 258 485 16 — 1,260 252 WS94S 26 — 200 487 16 164 1,104 120 518 27 — 184 316 2743 16 — 698 — 1632 28 — 1,203 367 491 17 206 756 — 523 29 — 19.152 12,319 492 i8 10,624 45.536 15.540 1046 29 — — 108 493 19 — 210 — 1634 29 — 252 — 494 19 15,206 46,570 30,220 2112 29 — 192 2,064 495 19 — 116 — 2113 29 — 356 — 1029 19 — — 148 1636 30 — 406 124 December 1050 I — — 264 WS30 19 — — 212 2119 I — 996 — 125 19 — — 481 1640 3 1. 179 4,029 — 129 19 — — 5.207 1640 3 — 120 — 537 19 — 9.496 36,008 757 4 — — 204 538 19 — — 112 2137 7 — 420 102 539 19 — — 340 2139 8 — 120 146 132 20 — 126 8,092 2513 9 — 160 5,888 133 21 — — 4.299 WS947 10 — — 2,096 136 21 — — 961 SS15 II — — 408 548 21 — — 480 769 12 — ■ — 252 549 22 — — 460 530 13 — — 100 139 23 — — 285 WS322 16 — — 136 WS896 28 — — 216 WS26 18 — — 42,500 RS 14 31 — — 1,064 throughout every known phase of its development from its earliest deep-living Naupliar and Meta- naupliar appearance until it reaches the adult state. Among the material taken in the vertical nets there are 308 samples of larval krill gathered at all levels traversed between 1500 m. and the surface and covering all twelve stages of the larval life-cycle. The number and stage frequency of the larvae in each sample have been determined and the principal results of these determinations, disregarding in most instances samples totalling only one or two larvae, are shown in Figs. 32-7. Throughout these diagrams, which are arranged in order of advancing time, the stage frequency is expressed as a percentage of the total larvae in each sample, the number of individuals^ upon which the frequency in each instance is based being shown by the figures below. '^ In most instances this number was in fact the total sample, although in some few, where the gatherings were very large, it was based on a fraction. 25 DM 226 DISCOVERY REPORTS In general it will be seen a single developmental phase (or at most two adjacent developmental phases) makes up the vast bulk of any larval sample, many of which indeed, the majority in fact of those from the deeper levels, consist of larvae all, or virtually all, at the same fixed stage of grovrth. It will be seen, too, that the stage frequency throughout is normal, the mode, in so far as a single stage may be said to represent a mode, in almost every instance being sharply defined and of high, in many instances of the highest possible, value. The number of stages found together in any one sample rarely exceeds four and when five or more are encountered this happens only at levels above 250 m. In the deeper strata (Figs. 32 and 33) it is rare in fact to find more than two, the samples from below 500 m. being represented exclusively by one, or by one and by one other immediately succeeding larval stage. The position, however, in the 500-250 m. layer is perhaps the most striking, for there it will be seen in all but three instances the samples consist exclusively of First Calyptopes, a phenomenon that springs from the fact that the metamorphosis of the Metanauplius (p. 97) is accomplished well below the 500 m. level, the resultant First Calyptopis during the final phase of the developmental ascent traversing the 500-250 m. layer without moulting except in such rare instances as for example are provided (p. 90, Table 13) by the enormous concentration of larvae encountered at Station 1138. The samples from the upper strata (Figs. 34-7) are also represented quite frequently by one stage alone although there just as frequently by two, three or four. The frequency of occurrence of multiple stage samples, however, would naturally be higher in the upper than in the lower strata, for it is at the higher levels near the surface that the larvae first begin actively to feed, and to grow and moult on a massive scale. The dominance, repeatedly encountered in our samples, of one, or of one and one other adjacent larval stage, and the high modal values that go with it, can only it seems have one explanation, namely, that the patchiness of the larvae that persists from their earliest deep appearance until the completion of their life-cycle in the surface, arises directly from their being assembled in a system of swarms in which the separate individuals in each instance are of the same or much of the same age and which from hatching onwards seem to keep together until the end of their natural span. There is nothing new in such an idea. Gurney (1924), for instance, in his account of the Decapod larvae of the British Antarctic ('Terra Nova') Expedition, writes, 'It seems probable that the larvae of Decapods and Euphausiids are not so much at the mercy of the currents as might be supposed. It is not very unusual to find swarms of the larvae of one species in different stages of development, which seems to indicate a power of keeping together from hatching onwards, or of collecting in a suitable locality '. Again, Russell (1927a), recording a dense swarm of Corystes cassivelaunus zoeas in the Plymouth area, and 6 weeks later and in the same position a ' great shoal ' of megalopas of the same species, writes of the latter, ' It would be interesting to know whether this was the same shoal that was met with on April 8th, when the Corystes were then in the later zoea stage. It is, at any rate, an indication that the larvae must have some considerable powers of keeping together if, when they have passed through all the stages of their pelagic existence, they can form so large a shoal as that met with on May 19th, 1925 '. What these powers could be, unless it simply be that these creatures, having originally been brought together through the hatching of dense concentrations of eggs, prefer thereafter always to keep in close contact with their own kind, will perhaps always be a matter for conjecture. As Bainbridge (1953) remarks, 'there may be many means by which animals keep together in swarms, such as sight or a rheotactic sense', while Hardy it will be recalled (p. 154) suggests there might be some ' guiding principle — almost as if there were some leader in command of the whole ' which keeps the swarms intact, and it may be that this principle is pure instinct, 'an inborn capacity for doing apparently clever things, suffused with dim awareness and backed by strong CAUSES OF PATCHINESS 227 FURC 5 5 „ 4 3 " 2 " I CAL 3 2 I META NAUP 2 N5 EXAM -INED DEC JANUARY FEBRUARY MARCH 8 9 3 IBI9202223 27 272B28 28 2 2 15 15 15 17 19 19 I928282828{2 2 2 762 1947 N43 VII 823 VI2 1662 1671 1965 N19 620 1545 1545 1138 1138 764 N42 300 1492 823 1662 1671 1965 1965 618 620 1545 2594 1138 DEPTH □ 500-250 ^ 750— 500 ■ 1000-750 WEODELL SWARMS EAST WIND SWARMS 1=^ Scale per cent o so 100 wmmm i 12 13 81 9 369 255 6 2 34 19 lO 138 36 260 623 3 92 15 27 17 15 11 I20 58 24 12 134 438 318 DEPTH 6 FURC 5 4 3 2 " 1 " 3 CAL 2 u 1 r. META 2 NAUP n; EXAM- -INED Fig. 32. The patches as swarms. Stage frequency in larval samples, showing high modal value of stage in samples from below 250 m. Based on vertical samples (see p. 225) from oceanic water and continued in Fig. 33. 3334444444555BB88 8 8 8 8 10 12 12 122222 2226|l7 17 17 202020202020 27 27 27 26OO260O1I42 2603 2603 1144 1144 636 637 2610 2610 647 647 1713 1720 W197 854 854 855 2346 2346 260OII42 1142 2603 2503 1144 635 637 637 26IO 1312 647 1713 1713 W197 W197 854 855 B55 2346 DEPTH FURC 6 • 5 " 4 > 3 " 2 1 CAL 3 • 2 1 META NAUP 2 N? EXAM -INED DEPTH Q 50O-25O ^ 750-500 ■ 1000-750 WEODELL SWARMS EAST WIND SWARMS Scale per cent o so too o so loo DEPTH fi^cSJB»«H=gife^lPHa5ar 40 115 IS 501 445 8 52 38 26 12 10 585 150 2 17 2 142 35 399 840 39 72 7 5 387 20 167 18 14 66 16 6 3984 3 4 115 28 54 221 13 262 6 FURC 5 « 4 " 3 2 I 3 CAL 2 1 META 2 NAUP N2 EXAM -INED % SAMPLE FROM 1500— lOOO Fig. 33- The patches as swarms. Stage frequency in larval samples, showing high modal value of stage in samples from below 250 m. FURC 6 ■ S ■ 4 • 3 > 2 ■ 1 CAL 3 • 2 " 1 META NAUP 2 N° EXAM- -INED FEBRUARY MARCH IB 19 21 21 22 232830|4 12 15 17 17 17 17 18 18 19 19202025 25 2528 2B2b| 2 222234444555 VF 303 VII V12 VH 1965 N19 NI9 618 620 622 362 1545 1545 365 1138 2600 1142 2603 1 144 VG 304 1492 320 V13 VI4 N19 618 620 622 361 362 1545 365 1138 1547 1142 2603 1144 1144 452 133 * * 22 6 59 5 304 59 30 92 116 3 25 572 169 222 39 169 13 30 54 99 20 512 39 217 148 12 328 249 19 617 810 B 59 52 27B 16 MONTH DEPTH 6 FURC 5 " 4 • 3 " 2 " I 3 CAL 2 » I META NAUP 2 N2EXAM- -INED * NUMBER NOT ASCERTAINED Fig. 34. The patches as swarms. Stage frequency in larval samples, showing high modal value of stage in samples from above 250 m. Based on vertical samples from oceanic water and continued in Figs. 35 and 36. 25-2 228 DISCOVERY REPORTS MARCH FURC 6 5 4 3 • 2 I CAL 3 2 I META NAUP 2 888899'3';99 9 10 10 12 12 12 12 12 19 202122 2222 2223 25 2626 26 27 28282830303030 368 637 369 539 1987 1988 I990 647 1994 373 375 1713 2293 2298 2300 2302 194 W527 1328 636 2610 639 1987 1987 1312 647 647 1994 374 1713 1713 2014 2017 I720 193 194 W527 1328 STATION ImnMAJUidmdJA o so 100 k'./,,J EAST WIND SWARMS Scale per cenl 6 FURC 5 ' 4 > 3 " 2 • I 3 CAL 2 " I META 2 NAUP N2EXAM- -INED 50 21 44 156 47 43 56 20 87 36 65 119 80 III 287 15 89 490 353 8 12 115 545 113 9 94 367 325 48 97 90 35 39 I 119 98 108 148 N2 EXAM- -INED Fig. 35. The patches as swarms. Stage frequency in larval samples, showing high modal value of stage in samples from above 250 m. See legend to Fig. 34. Samples at stations 193 and 194 are from the Bransfield Strait. i APRIL MAY JUNE 2 3 4I4I4I5I5I6I517 19 20202020202020 21 22 22 22 23 25 25 26 27 27 27 28 30 I 2 3 4| I 2 3 5 6 8 9 9 lO W529 1331 2315 2318 2320 WI98 854 854 855 W200 2335 1346 1350 I3S1 2345 1353 1355 1357 1781 1785 1790 1792 198 2315 2318 2320 WI97 WI99 854 855 855 W20I 2335 1348 23442346 1352 1354 1356 1779 1782 1790 1792 1794 DEPTH FURC 6 > 5 „ 4 • 3 " 2 " I CAL 3 • 2 • I META NAUP 2 STATION DEPTH o so 100 Stole per ce .'■'-'- ^1^ V-^j _ .3.M.. '""'""' I I _ ^^ I 1 1 M ^^^ ° '°'°° HH m WEDDELL SWARMS Y ' '^gj EAST WIND SWARMS 6 FURC 5 • 4 " 3 ■' 2 " I 3 CAL 2 » I META 2 NAUP N2EXAM- - INED 90 68 1057 532 158 25 5 108 55 154 15 84 177 100 159 72 100 143 152 344 90 289 72 209 220 167 64 45 43 98 II 19 45 14 34 53 lOO 100 80 25 28 79 79 201 N9EXAM- -INED Fig. 36. The patches as swarms. Stage frequency in larval samples, showing high modal value of stage in samples from above 250 m. See legend to Fig. 34. Sample at Station 198 from the Bransfield Strait. MONTH JUNE SEP MONTH DATE STATION DEPTH FURC 6 5 • 4 ■ 3 • 2 I CAL 3 " 2 I META NAUP 2 14 27 28 27 28 24 24 228022801 383 383 861 862 | 887 888 | 912 2843|2879 I n Scole per c«nl o so too 1^ ti- Jim ^k I i^b o so too ■I I ■r-i—i I o so 100 DEpTi # DEPTH r~] 50- o ^ 100- 50 ■ 250-100 o so loo o so lOO DATE STATION DEPTH 5 FURC 5 4 3 2 I 3 CAL 2 > 1 ' META 2 NAUP NS EXAM- -INEO 34 376 620 168 230 72 56 25 1480 17 12 N» EXAM- -INED Fig. 37. The patches as swarms. Stage frequency in larval samples from cold surface water assumed to be deflected from the East Wind drift, showing high modal value of stage in samples from above 250 m. CAUSES OF PATCHINESS 229 endeavour'/ combined perhaps with that 'little dose of judgement or reason' that, as Darwin in his Origin of Species has said, 'often comes into play, even with animals low in the scale of nature '. The behaviour of Gunther's swarm at the jetty (p. 155) does seem to suggest a marked posi- tively rheotactic sense, although whether the reaction of the krill in this instance to the current, if it was in fact to the current, was dependent on a tactile sensitivity to it, or on visual stimulation by the jetty or its shadow, or even by the sunlight itself, perhaps only experiment will show. Certain aquatic insects it is known (Schulz, 1931; Tonner, 1935) show a positive rheotaxis that apparently springs from tactile perception of the current by the antennae or from straightforward vision, or from a combination of both. Other invertebrates, both terrestrial and marine (Carthy, 1957), seem to be attracted to shadow (positive skototaxis), and it would seem that our krill swarm was repelled by it (negative skototaxis). It might be supposed, too, that the individuals of a swarm contrive to keep heading in the same direction (p. 151) by eye alone, as a body of troops does when marching in line abreast. As Thorpe (1956) has written of the remarkable and so far little understood regimentation in schooling fish, 'Whatever may be the function of such extraordinary social organization, there is no doubt that it must be partly learned and based on a series of extremely precise visual orientations '. In the open sea with no landmarks to guide them it is not so easy to see how the krill, if they do, as seems possible, orientate themselves visually on external objects, could maintain themselves for long over a given position, or point themselves in one direction rather than another. But here again sight or a delicate tactile perception of each other would help them at least to keep contact and to point, too, perhaps in the same direction. Even in the open sea, however, there is at least one 'landmark' by which they might orientate themselves visually. By day, for instance, there is the sun and recent researches on terrestrial arthropods suggest the possibility that it may exercise some influence on the behaviour and movements of pelagic animals. It has become widely known since the pioneer work of Frisch (1948) on bees that many terrestrial arthropods have eyes that are highly sensitive to linearly polarised light, using this sensitivity to navigate by means of sky polarisation. Waterman (1958) summarises the literature that has recently been published on this phenomenon, observing that the widespread occurrence of polarised-light behaviour in these animals makes it most likely to be present in a variety of aquatic forms as well, adding that, if it is, 'the complex patterns of light polarization known [Waterman, 1955] to be present will be bound to influence the movement of plankton and other pelagic animals in the photic zone'. He continues : Although a variety of physical and chemical clues to orientation have been proposed for their environment, animals living in open water would appear to have but sparse information from which to direct their horizontal migrations. Particularly when well below the surface, but still far from the bottom, they might have little or no obvious sensory data permitting them to gauge geographical direction. Yet if they are sensitive to the underwater polarization patterns, a polarized-light compass could be available to them throughout a considerable part of the photic zone and thus function much as a sun or sky compass does for terrestrial animals. * The late Sir J. Arthur Thompson in his first-year lectures. A simple definition indeed, yet it embodies the essence of leading ethological thought. I quote from Thorpe (1958). ' Lorenz' main argument is that in each example of true instinctive behavior there is a hard core of absolutely fixed and relatively complex automatism, an inborn movement form. This restricted concept is the essence of the instinct itself; it is now usually referred to as th^ fixed action or the. fixed-action pattern. Such action patterns are items of behavior in every way as constant as are anatomical structures, and potentially just as valuable for phylogenetic and systematic studies. Where such action patterns constitute an end point or climax of either a major or a minor chain of instinctive behavior, they have come to be known as consummatory acts. The internal coordination mechanism of these consummatory acts is assumed to be generating some Specific Action Potential (SAP). This is the internal drive, which is of the very essence of the word instinct.' 230 DISCOVERY REPORTS Summing up he notes that most of the underwater Unear polarisation springs from the scattering of refracted sun's rays by particles in the water, a phenomenon affecting much of the photic zone. The pattern of polarization therefore is dependent on the sun's position in the sky and hence changes markedly with time of day. Since a number of arthropods are known to be highly sensitive to linearly polarized light, it is suggested that the behavior of certain pelagic forms may be determined in part by this environmental parameter. The most likely significance of this factor in nature would be in directing local diurnal movements of populations and in pro- viding a sort of celestial compass that could be used in orienting horizontal migrations. Although linearly polarised light is the principal component of natural underwater illumination, elliptically polarised light is also present as a definite but less prominent component (Ivanov and Waterman, 1958). It will be interesting to determine, as these authors remark, whether this too is of any biological significance. It is obvious, however, that much experimental work is needed, together with further observation on the swarms in their natural state, before we can even begin to understand what keeps them together and behave as they do. To quote Tinbergen (1953) on the behaviour of congregating animals: Once the animals have come together, we see co-operation of numerous kinds. The simplest kind of co-operation is 'doing the same thing' as others. When one Herring Gull flees, the others flee as well This principle of 'sympathetic induction', as McDougall has called it, can be seen at work in many social animals, Man included. We yawn when we see another yawn, we get scared when we see signs of intense fear in another man. It has nothing to do with imitation ; the reacting individuals do not learn to perform certain movements by watching others perform them, but they are brought into the same mood, and react by making their own innate movements. Watching the flight manoeuvres of a flock of Starlings or waders reveals another type of co-operating. Such animals not only fly when the others fly, but they direct their flight to that of the others. It is highly fascinating to see how thousands of Starhngs, flying round above their roost on a winter evening, turn as if at command ; left, right, up, and down. Their co-operation seems so perfect that one forgets the individuals and automatically thinks of them as one cloud, as one huge 'super-individual'. The swarming krill, with their superlative co-operation or capacity for doing the same thing, may be Ukened to Tinbergen's starlings, for their mass behaviour, too, seems to be that of a super-individual. Beebe (1932) gives a superb description of the schooling of the aj-inch long Bermudan fish Atherina, a passage recalling in almost every line Hardy's classic account of the incredible capacity for concerted action, the incessant fluxing or amoeboid movement, the mass awareness, sense and intent, exhibited by the swarming krill. A few feet away from me was a concentrated school of small fish. When they were drawn out into a long ribbon I counted fifty, and a conservative estimate of the whole was fourteen hundred — and this was an extremely small school, as schools of Little Arrows go. Slowly the school swung first in one, and then in the opposite direction, it changed from a circle to a dumb-bell, to beads on a string, to a crescent, and now and then I halved it momentarily, when it rejoined and became some unnamed figure. It had length, breadth, and depth, intentions, achievement, attempt, refusal, impetus — but one never thought of individual units. As I watched, it became more and more a sentient individual, with particular desires and activities; in the little harbour beneath me were snappers, and squirrel-fish, cow-fish and bream and — it. It kept halfway from the bottom, and its shadow on the bottom was the verisimilitude of a slow creeping worm, which flowed over pebbles, reached fingers out sideways, drew back and then slipped out of sight, as the sun went under a cloud. I could do strange things with this Atherina organism — like boring three holes at once through it. I waited until it ! became a broad oval and then tossed three tiny pebbles simuhaneously. As they sank, there opened beneath each of them a round hole, and soon they all dropped through three separate wells and the trio of apertures closed up again. There was no especial fright, simply a very reasonable withdrawing from the unusual phenomenon of pebbles dropping from the sky. The reaction of the krill it will be recalled to Gunther's dropped lead on line was precisely the same. In its beginnings then, the inherent capacity of this species for maintaining itself in swarms is a , purely larval phenomenon. It can be traced, however, from the time when the larvae first begin to become adolescent (Fig. 38), through the whole of adolescence and up to the adult state, the measure- CAUSES OF PATCHINESS 231 1" 2 12 w ,0 S = I 4 JULY AUGUST SEPTEMBER OCTOBER 9 21 22 22 24 26 2928 29 29 2912 4 13 17 17 17 IB 18 19 19 25 25 29 30 2 3 3 3 4 4 4 S 5 5 5 6 6 6 6 7 7 8 2854 2856 2365 2963 2864 2405 2408 «26l W264 W267 I403 W279 W262 W285 W287 2431 2443 2447 W294 W295 W297 W299 W30I W302 W304 W305 W3I0 2655 29S7 2959 2372 2398 2406 24(2 W263 W266 W26e W276 W260W293 W266 W268 2432 2445 1821 1823 W296 1824 W3O0 1825 W303 1826 W306 r-m I r-m O SO JOO 11 I O so 100 T T I I 1 I I I , o so roo 80 47 40 32 56 ISO SO lOO 45 IOC 362 lOO lOO lOO lOO 62 192 236 24 100 71 lOO 61 lOO lOO lOO 56 33 51 58 7 lOO lOO 166 lOO 52 59 59 50 lOO lOO lOO 405 90 58 291 lOO 498 lOO 321 77 224 64 |»ffp|^ff^*^'f™»!"f% -TffWP Willi DATE ST 22 20 Fig. 38. The patches as swarms. Stage and length frequency in larval, or mixed larval and adolescent, samples from the Antarctic surface layer. Based on all measured samples of 10 or more from the northern or Weddell zone and continued in Figs. 39-42. The specimens, measured to the nearest millimetre, are plotted in 2-mm. groups, the operative figure in the length range being the second of the group, e.g. 12 includes specimens of 11 and 12 mm. DATE ST OCTOBER NOVEMBER 9 10 10 10 10 10 10 10 10 10 10 10 10 10 to II 12 16 17 18 I9 2020 21 21 21 22 22 23 25 26 26 26 26 28 28 28 29 31 | I I 2 2 2 2 4 6 13 14 16 16 17 17 1832 1835 1835 1835 1835 1835 1835 1635 1637 454 459 1846 461 461 462 464 2465 467 2468 2475 472 2477 2479 1O09 493 467 491 1835 1835 1835 1835 1935 1835 1635 1936 453 455 460 461 461 461 463 2463 465 2467 469 471 1003 2478 IBS6 461 484 468 AOOL FURC6 (l)(l)(0(l)CI) 50 295 94 71 638 ISO 171 322 lOO lOO lOO 57 SO SO lOO 103 606 97 39 73 lOO 66 60 106 lOO lOO lOO 83 325 64 336 348 lOO 164 lOO 50 100 45 SO 50 50 46 46 52 30 no lOO 121 3IO 42 lOO 95 75 TffffTfflrpffwifffliin f-wffir DATE ST 34 32 30 28 26 24 22 20 AOOL 6 F Fig. 39. The patches as swarms. Stage and length frequency in mixed larval and adolescent, or purely adolescent, samples from the Antarctic surface layer. See legend to Fig. 38. SS6 8 52 32a S24 NOVEMBER DECEMBER JANUARY FEBRUARY 18 19 19 21 21 21 222222222323232324252626262627272629 292930|6 9 lOII II 13 19 19 |l025 26 27 27 27 27 27 27 2930 7 6 9 9 9 10 11 492 I029 500W4B7 2093 503 2094 1624 517 1628 519 2109 2112 2116 2SI3 S S30 539 V(S37 W54OW54OWS40W542 I507 351 354 ISM 494 W466 2091 502 2093 507 2094 210O 1039 2103 1630 S23 2113 2510 S 527 537 S W536 W540W540 924 2567 349 352 612 ^ SAMPLE FROM SIGHTED SWARM lOO 200 lOO lOO I02 SO 251 64 lOO 96 74 52 65 87 189 99 23 61 116 568 904 604 lOO lOO 55 92 SO lOO lOO 112 67 129 ICO 92 594 197 76 62 lOO 290 27 57 49 379 325 66 522 539 95 619 35 60 517 AOOL FURC6 TfW W -: ""T™ T ADOL 6FURC Fig. 40. The patches as swarms. Stage and length frequency in mixed larval and adolescent, or purely adolescent, samples from the Antarctic surface layer. Specimens plotted in 4-mm. groups, the operative figure in the length range being the fourth of the group, e.g. 28 includes specimens of 25, 26, 27 and 28 mm. 232 DISCOVERY REPORTS DATE ST 64 lA 60 ^5' g?= I" '40 2 36 UJ32 O z2e < a 24 l20 B 16 5 12 FEBRUARY MARCH APRIL MAY JUNE JULY AUGUST SEPTEMBER OCT IS 2021 2222 23 26 26 1 8 10 12 14 14 19 26 I S 7 7 7 7 7 14 15 I7 26 27|l8 2 4 9 10 II II 12 16 16 17 19 2224 27 2828 2 3 5 5 17 27 2 3 618 624 833 WI56 2S94 643 23 38 42 207 207 207 2318 2344 W434 * 2853 2855 2864 2393 2408 2874 W264 I405 W277 1813 IB23 622 626 627 2590 368 2614 24 658 663 207 207 WI96 665 1351 1781 2852 2854 * 2391 2399 2412 W264 I403 W277 W282 W290 Scale per cent (2) (2)(2)(2)C0 (2Kll o so lOO (3) (2) (2)0) (i)0)(a(i)(3)(iKi)(iKiHi)(i)C2) (a (3) (2) (2) (3) (3) (3) (2) (2) C3) (3) (4) (2) (4) 100 71 lOO 118 176 lOO 75 100 91 30 69 33 57 370 20 35 23 50 31 45 92 44 50 lOO 49 92 580 lOO 28 lOO 388 63 47 84 231 41 I02 20 60 66 25 lOO 232 25 14 23 46 73 63 lOO 47 47 62 DATE ST 6S 64 60 56 52 4S * LENGTH FREQUENCIES BASED ON AGGREGATE CATCH FROM SEVERAL NET HAULS Fig. 41. The patches as swarms. Length frequency in adolescent samples from the Antarctic surface layer. See legend to Fig. 40. OCTOBER NOVEMBER DECEMBER JANUARY FEBRUARY MARCH APRIL 6e u>64 %60 g" 0 52 5 48 S44 40 z 28 2 24 4 5 5 5 8 10 10 10 10 10 10 10 10 10 13 17 2022 3 6 8 13 19 1 4 12 18 19 19 21 21 6 7 8 22 2428 30 9 9 10 1022 8 lO I023 7 7 7 7 27 W295W298 1825 1835 1835 1835 1835 1835 454 461 1009 1875 I050 769 125 548 795 799 312 2567 354 612 368 643 207 207 2346 1824 W298 1831 1835 1835 1835 1835 1839 1844 1854 1864 494 757 535 539 549 797 W373 825 351 356 WI52 640 2295 207 207 (51 (3) (3) (4) (4) (41 (4) (4) (4) (5) (6)(5)(6)(S)(5i Scale per cent (7) (8) (6) (6) (61 (7) (6) (8) (7) (8) (6) (8) (8) (8) (5) (8) (6) (8) (91 (5) DATE ST 68 64 60 56 52 46 44 40 36 32 2S a« 20 16 12 8 40 32 107 61 271 103 153 176 49 216 102 45 27 132 I04 lOO 136 171 17 II 58 120 18 42 99 97 37 55 5 79 702 159 32 U 71 61 100 171 I 160 33 I09 39 5 50 lOO 17 184 52 Fig. 42. The patches as swarms. Length frequency in adolescent and adult samples from the Antarctic surface layer. See legend to Fig. 40. JANUARY FEBRUARY MARCH APRIL 6 7 8 8 10 10 19 19 21 2224 2830| 9 9IOI022294 5 5 6 810 10 15 21 23 23 24 7 7 7 7 7 7 2026 27 795 575 580 602 820 312 2567 354 612 W9I5 2271 2274 640 20002295 1717 207 207 207 2344 797 799 2168 2195 W373 825 351 356 WI52 2268 2606 368 643 1711 1715 207 207 207 855 2346 EAST WIND SWARMS Scale per cent (6) (7) (8) (8) (5) (6)(7)(6X7) (6)(8)(6K7)(6)(8)(7X8)(6X7) (7) (7)(a)(8X8) (5) (8)(8)(8)(8)(8)(8) (5) n; meas- ■URED 702 69 23 63 77 32 146 III I60 830 60 33 17 153 184 134 25 52 39 370 61 159 87 167 tOO 171 36 SO lOO 471 342 33 109 91 57 39 50 56 53 III N° MEAS- -URED F'g- 43- Final condition of the adult swarms in the East Wind-Weddell surface stream. See p. 233 and legend to Fig. 40. I CAUSES OF PATCHINESS 233 ment and staging of a very large number of samples covering the later developmental phases (Figs. 39-42) demonstrating in sample after sample a normal length or stage frequency in which the value of the mode, as in the purely larval samples, is persistently high. As with the larval samples. Figs. 38-42 are again arranged in order of advancing time, beginning with July, when the first adolescents appear, and ending with April, 21 months later, when the last spent females are found in the plankton, the material on which the length and stage frequencies are based having been chosen from the Weddell or northern zone, since it is there we have obser- vations and measurements covering every month of the year. The length frequencies, expressed in 2-mm, groups from July up to the beginning of the second half of November, and in 4-mm. groups from then onwards, appear in the upper part of each figure, the stage frequencies, where determined, in the lower. The number of individuals measured, or both staged and measured, from each sample is shown by the large unbracketed figures, the figures in brackets showing, where determined, the dominant stage of the female in each sample. Thus, following Bargmann (1945, pp. 1 13-15), i, 2, 3, 4 and 5 indicate females in from very early to late adolescence, 6 indicating paired females, 7 ( = Barg- mann 7 a) gravid females, and 8 (= Bargmann 7B) spent females. The normal length frequencies and high modal values repeatedly encountered in the Weddell samples are, of course, equally phenomena of the East Wind drift and of the cold northward encroaching tongues of water to which it gives rise. In fact wherever and in whatever developmental phase this species had been fairly sampled the samples obtained, as the many developmental diagrams presented in the distributional section show,^ point everywhere to its congregating in discrete groups in which the length frequencies are persistently normal and the modal values persistently high. The final condition of the 24-28 month old swarms, incorporating the data from the East Wind drift, is shown in Fig. 43 in which the length frequencies reveal a tendency for the animals in the Weddell swarms to be somewhat larger than their contemporaries in the East Wind zone, sug- gesting that, taking the krill population as a whole, it is in the lower latitudes that the adults attain their greatest length. Even Fig. 43, however, probably does not reveal the true condition of the adult swarms. It is based on townet samples and since there is an escape factor involved, particularly (p. 262) where the largest animals are concerned, it may well be that in both Weddell and East Wind samples we have repeatedly been recording modal lengths of distinctly lower value than perhaps would be found in the swarms in their natural state. This is distinctly suggested by the condition of the adult samples (p. 143, Fig. 14) collected by Ruud and Crimp from whales' stomachs. Turning again to the larvae, it will be seen that with the advance of the spawning season the age pattern, or modal values, of the surface swarms (Figs. 34-7), particularly in those encountered in the Weddell stream from March to June, becomes increasingly complex, the surface population as a whole becoming more and more heterogeneous as successive batches of eggs hatch out and the existing swarms become augmented by others of later generation. In the East Wind drift on the other hand, because of the compressed spawning season and very slow growth-rate (p. 355) characteristic of these high latitudes, the corresponding age pattern from January to April is largely homogeneous. It would appear then, from the random sampling of our nets, that the patchiness of E. superba arises from the congregating of this species from hatching onwards in a system of swarms, each discrete and complete in itself, the individuals of it, especially as larvae, all, or very largely, in the same developmental phase. Or to put it in another way, the enormous variation in catch-figures our nets reveal, as Motoda and Anraku (1955) have put it, springs from the 'statistically non-random distribution ' of the organisms we have been sampling. It would appear, too, that at any given time or place an individual swarm need not necessarily be composed of euphausians of the same age as that ^ See particularly p. 397, Fig. 136. 26 nM 234 DISCOVERY REPORTS of its contemporary or neighbour, a phenomenon only to be expected in view of the protracted period over which the eggs are laid. The data provided by the special method used for sampling the surface patches in daylight confirm these appearances. This method, which is described on p. 152, was first used at Station WS 540, where in the short period of 35 min. five samplings, each producing an enormous quantity of euphausians, were carried out on five separate but adjoining patches, each lying NON -IDENTICAL ADJACENT SWARMS IDENTICAL ADJACENT SWARMS Fig. 44. The patches as swarms. Length frequency in five separate adjacent patches sampled at St. WS 540, showing the distinct variability of age pattern that can be exhibited by neighbouring swarms (Length scale in 2-mm. groups, e.g. 30 = 29-30 mm.). approximately 400 yards from its neighbour. The length frequencies of these gatherings, based in each instance on the measurement of over 600 individuals, are shown in Fig. 44. It is clear from these graphs that in three instances (samples B, D and E) the patches represent swarms of practically identical age and that in two instances (samples C and A), they represent swarms, C older and A still older, than B, D or E. In terms of spawning A would represent the earliest, C a somewhat later spawning, and B, D and E a later still. It is unlikely that some chance advantage of environment or feeding could be responsible for the advanced condition of A and C since all five swarms are located too close to each other in space and time for that to be possible. If the krill were not so discretely disposed, but scattered broadcast throughout the whaling grounds, the persistently recurring normal length frequencies characteristic of our samples, and the high modal values that go with them, would both disappear, and the length frequencies, ceasing to have normality, would exhibit random and inconsequential modes instead. In view of the astronomically large numbers of larvae that must be produced to maintain the older population at its existing fantastic level our vertical gatherings of the young stages are on the whole surprisingly small. So enormous in fact must the larval population be, that I should have expected it to have been sampled with much the same success as for instance (Hardy and Gunther, 1935, Appendix 11) certain copepods are, but manifestly in the main (Table 48 and Figs. 45-8) it has not. A possible explanation of this anomaly is that the larval swarms, like the older swarms we see CAUSES OF PATCHINESS 23S C.PROPINQUUS DEPTH 75- 1 — 10 10—100 lOO — 1000 1000— 10000 loooo-iooooo Fig. 45. Orders of abundance of E. superha Calyptopes and Calanus propinquus (as tabled by Hardy and Gunther, see text) expressed as percentages of the total samples from the Antarctic surface layers (vertical net hauls). so-o Fig. 46. Orders of abundance of E. superba Calyptopes and Calanus acutus expressed as percentages of the total samples from the Antarctic surface layer (see legend to Fig- 45)- % E. SUPERBA % D.PECTINATUS 50- \ \ ~~ ■°-, 25- ^~-^^ y -c' ^^^^^^—^ ~-~ Q 1 1 1 1 1 — 10 10-100 100— 1000 1000-10000 lOOOO-lOOOOO Fig. 47. Orders of abundance of E. superba Calyptopes and Ctenocalanus vaniis expressed as percentages of the total samples from the Antarctic surface layer (see legend to Fig. 4s). Fig. 48. Orders of abundance of E. superba Calyptopes and Drepanopus pectinatus expressed as percentages of the total samples from the Antarctic surface layer (see legend to Fig. 45). 26-2 236 DISCOVERY REPORTS on the surface (p. 149), although perhaps of considerable lateral extent, are disposed in shallow rafts or plates no more perhaps than a metre or two thick in which the vertical nets, fishing only momen- tarily, capture far smaller numbers than they would obviously do if the swarms were densely disposed in depth. The orders of abundance of the Nauplii, Metanauplii and First Calyptopes taken at deeper levels are also (Fig. 49) inclined to be low, suggesting that from the outset the larvae are assembled in pockets of shallow draught. °/ '° DEPTH 100 • 7S- 1000 - 750 75- T 1 r 10-100 100-1000 1000 -10000 lOOOO-IOOOOO Fig. 49. Orders of abundance of deeper living E. superba larvae and Ctenocalanus vanus expressed as percentages of the total samples from the Warm Deep current (see legend to Fig. 45). Further investigation, involving a comprehensive and elaborate programme of staging and measuring, may reveal that other plankton animals, like the krill, throughout much of their lives, are assembled discretely on or near the surface in swarms, and I believe that a great deal of the patchiness of the plankton as a whole^ will eventually be explained in terms of this phenomenon. In so far as our Discovery collections are concerned, one is struck by the frequency with which the surface (0-5 m.) nets produce enormously larger samples of certain species than are produced, for instance, by oblique nets hauled open to the surface or by horizontal or oblique closing nets fished at subsurface levels. One is struck, too, by how often these enormous surface samples are severally composed of individuals that to the eye at least appear to be all of much the same size and in much the same developmental phase.- Among the euphausians E. crystallorophias (p. 124) is a known surface swarmer, and Thysanoessa macrura may well prove to be another, the recent discovery by the Japanese (Nemoto and Nasu, 1958) that this species sometimes contributes substantially to the diet of baleen whales 1 As Cassie (1959) has recently written, 'It is now well known and generally accepted that many, if not most, plankton organisms show a marked departure from randomness in their horizontal distribution, even when samples are taken at relatively close intervals'. 2 Baker (i960) has just called attention to the shoaling of young squid {Ommastrephes pteropus Steenstrup) at the surface at night, noting that ' those tending to form shoals were mainly in the smaller size groups and their size did not appear to vary greatly within any one shoal'. See also Zelikman (i960) on the swarming of euphausians in the Barents Sea. CAUSES OF PATCHINESS 237 suggesting strongly that at times at least it must congregate, densely concentrated, at or near the surface. Nemoto himself (1959) concludes that from the 'dominant appearance' of T. macrura in the stomachs of some whales it must form ' large swarms in the sea ', and it must be to surface swarms that he refers, for nowhere in his recent account of the feeding habits of baleen whales (Nemoto, 1959) does he suggest that the southern species are anything other than surface feeders. We may include too in this category Eiiphausia vallentini, the small krill reported to be eaten by the ' Pigmy ' blue whales recently discovered (p. 48, note i) near Kerguelen. As for other southern plankton forms, outward inspection of our specimen jars alone is enough to show repeatedly that the large tomopterid, T. carpenteri, the calanoid, C. propinquus, the amphipod Parathemisto gaiidichaiidii, the pteropods, Limacina balea and Cleodora sulcata and the salp, Salpa fiisiformis aspera,^ to mention only a few of the more obvious cases, are all species that tend to be massed in enormous numbers on the surface, and massed, moreover, to all appearances discretely in swarms. Hardy and Gunther (1935, Figs. 131-5), from their consecutive net hauls off South Georgia, have already demonstrated the extreme patchiness of some of these forms and of several others I have not mentioned, their results strongly suggesting that it is in surface swarms, each, as the krill swarms seem to do, behaving very much as a single organism, that many plankton animals exist in the sea. The truly remarkable patchiness they demonstrate in Parathemisto gaudichaudii,^ at two widely separated stations where consecutive nets were used, leads us strongly to suspect that this is in fact how they live. In his massive account of the plankton of the Gulf of Maine, Bigelow (1926) gives many instances of conspicuous swarming at the surface by plankton animals. The following are among the more striking: RADIOLARIANS Copepods Acanthotnetron sp. Calanus finmarchicus Temora longicornis COELENTERATES Ceniropages typicus Hydromedusae Metridia lucens Melicertum campanula Amphipods Staurophora mertensii Euthemisto bispinosa juv. Phialidium languidum Euthemisto compressa juv. Scyphomedusae Euphausians Cyanea capillata var. arctica Meganyctiphanes norvegica Amelia aurita Thysanoessa inermis Thysanoessa raschii Ctenophores Decapods Pleurobrachia pdeus Larvae of Cancer spp. Beroe cucumis p^g^^jj ^^^^^^ CHAETOGNATHS PTEROPODS Sagitta elegans Limacina retroversa Clione limacina CRUSTACEA Cirrepedes APPENDICULARIANS Nauplius and Cypris larvae Oikopleura sp. ^ Now Salpa thompsoni (Foxton, 1961). ^ The recent discovery (p. 48) that this large pelagic amphipod contributes substantially to the diet of sei whales provides further evidence that it congregates densely in surface swarms, Hardy and Gunther's results suggesting that such swarming is equally a phenomenon of the daylight as of the night-time hours. 238 DISCOVERY REPORTS Table 48. Vertical net hauls. The orders of abundance in which samples of E. superba Calyptopes and certain copepods occur in the Antarctic surface layer Depth Order of Calanus Calanus Ctenocalanus Drepanopus E. superba abundance propinquus acutus vanus pectinatus Calyptopes I-IO 23 21 10 6 33] 10-100 43 23 26 9 16 100-1,000 8 9 24 12 14 1,000-10,000 4 I 8 13 I 10,000-100,000 I — — — — . I-IO 12 17 9 7 38] 10-100 39 33 10 7 IS 100-1,000 31 23 41 9 10 1,000-10,000 2 3 21 15 I 10,000-100,000 — — — 4 — . I-IO S 9 I 4 56^ 10-100 32 19 6 6 27 100-1,000 33 34 39 8 7 1,000-10,000 4 II 27 12 2 10,000-100,000 — — I 6 — 50-0 100-50 250-100 Many years ago when I visited the Arctic I made the following note (Marr, 1927) on the oc- currence of plankton at the surface during the daylight hours of the summer and early autumn of 1925- August 30th, 80° 13' N, 51° 34' E, i-metre cheese-cloth tow-net. Duration of haul, | hr. Surface. Surface temp., 28-8° F. Calanus, Amphipods, Medusae, Scyphomedusae, Gymnosomata (Clione) in large numbers; Clio innumerable. The sea appeared to be very dirty and the net was brown with Algal debris, which on examination proved to consist of many species of Diatomaceae and broken down vegetable (Algal) matter. The net was choked with it. It had a sickly, horrible smell, as of rotting seaweed. September 9th, 78° 38' N, 36° 28' E, i -metre cheese-cloth tow-net. Duration of haul, | hour. Surface. Surface temp., 34° F. Tens of thousands of Calanus and Clio ; Amphipods, Sagitta, Clione, in huge numbers ; and a perfect maze of Medusae, Scyphomedusae, Pleurobrachia, and Beroe. As regards the larger surface animals, the Ctenophores, Pleurobrachia and Beroe, were constantly observed in the still, crystal-blue water of the pack, particularly on calm days when the sun was brilliant and the temperature high. In the same note I wrote, ' On several occasions I have hauled in a [surface] net literally black with thousands of the little Pteropod, Clio, and swarming with active Calanus '. This also refers to daylight gatherings. I can, however, recall no occasion when we made comparable daylight gatherings in the Antarctic when fishing conventional stern nets on the surface of the sea. In fact for our daytime surface observa- tions I find repeatedly in our records the entry ' catch negligible, thrown away '. It may therefore be significant (see p. 265) that these enormous Arctic gatherings were got from a sailing ship, unac- companied by vibrational disturbance such as might be produced for instance by a revolving screw or by dynamos or other machinery. I record these rough notes, made when first, as a naturalist, I went to sea, because I believe that for long we have been putting too much trust in what seems to be revealed by apparatus, and not enough in what we can actually see. It is probable that the swarms in which it seems so many of these animals exist keep discretely apart in the sea. In other words, swarms of different species do not mix. If they did we should expect to find in whales' stomachs heterogeneous collections of plankton animals and not, as we CAUSES OF PATCHINESS 239 do, pure cultures of euphausian species such as E. superba or Thysanoessa inermis. Aron's observa- tions on feeding salmon (p. 148, note 3) also suggest that swarms of different species keep discretely apart. The 'Walfischfrass ' (food greedily eaten by the whale) of Crantz (1765), which according to Massy (1932) seems to have been the pteropod Scoresby (1820) figures under Clio borealis (now Clione Umacijia Phipps), might also it seems be a species that lives much at the surface in swarms. Scoresby lists it among the plankton devoured by the Bowhead or Greenland Whale, Bigelow (1926) calling attention to the vast shoals in which it occurs in the Arctic seas and the ' bounteous food supply ' it provides for right whales in this northern field. ^ In a note on the euphausian population of the Patagonian shelf, of which E. vallentini, E. lucens, E. similis, E. longirostris, E. triacantha and Thysanoessa gregaria are the principal species. Hart (1946) states that none ' have been observed to form dense swarms discolouring the surface of the sea, as E. superba commonly does in the Antarctic '. It is perhaps too early yet, however, to accept this as conclusive, for as yet, except for some of the larvae of E. triacantha (Baker, 1959), our extensive collections of euphausians from this wide shelf region have not been worked up. In regard to the northern euphausians, Hjort and Ruud (1929) state that they are convinced from their researches off More that they swarm only ' during the spawning time ', in other words only as ripe adults. Einarsson (1945), however, questions this view, noting that immature Meganyctiphanes norvegica have been found swarming in August, after the spawning is over.^ 'I am inclined to favour the hypothesis ', he continues, ' that hydrographic conditions, especially current conditions, are mainly responsible for the phenomenon of swarming, although I must admit that evidence is extremely scarce and inadequate. Further investigations for the elucidation of the factors concerned with the swarming are therefore warmly to be recommended'. Macdonald (1927) suggests that the swarming of M. norvegica at the surface is caused 'partly by the presence of predatory fish and tidal currents ', and in an early account of swarm formation and the occurrence of surface swarms Vanhoffen (1896) associates these phenomena with ' Stromkabbelungen ', the turbulent and complex water movements found along the margins of great ocean currents or where two converging currents meet. The German word can best perhaps be rendered as 'stream- quarrelling '. The swarming of the Antarctic krill at least is clearly not confined to the spawners. Nor is it due to any local or particular trick of the currents, for it occurs widespread throughout the circumpolar sea all over the faces of its great surface streams. In other words it is not a manifestation of any specialised or short-term behaviour. On the contrary, as Hurley (1959) has said of the dense con- ^ I have never, however, been wholly satisfied that the Bowhead did (or does) in fact eat this swarming pteropod. Scoresby himself, it is true (Scoresby, 1823), records it in 'vast quantities' off the Greenland coast, includes it (Scoresby, 1820) among the ' Medusae and other animals, constituting the principal food of the whale ', and remarks on its regular occurrence in water frequented by feeding whales. But, he continues, in the very few whale stomachs he had in fact opened ' squillae or shrimps were the only substances discovered ', adding that in the mouth of a fresh-killed whale he once found ' a quantity of the same kind of insect'. And the 'insect' he figures looks to me uncommonly like Thysanoessa inermis, the species devoured in such quantities (p. 138) by the North Atlantic right. Professor Jameson of Edinburgh, where Scoresby was a student, also (Jameson, 1823) refers to the 'different species of shrimps' that with other animals 'seem to contribute the principal food of the mysti- cetus', while the swarming 'insects' of John Gravill, captain of the 'Diana', convey the distinct impression that it is upon A jointed animal, rather than a naked mollusc, that the northern right whales feed. Nothing however is known for certain. Lilljeborg (1886) suggests that Calanus and Thysanopoda might be among the more important animals on which they live, Southwell (1881) stating that it is 'a kind of shrimp, found in great abundance in the Arctic seas', that provides the major portion of the plankton they devour. ^ It will be recalled too (p. 137), that other observers have reported immense shoals of both M. norvegica and Thysanoessa inermis at the surface in August, September and October. 240 DISCOVERY REPORTS centrations of brittle stars recently photographed in Cook Strait, New Zealand, it would seem to be 'an aspect of the normal behaviour of a naturally gregarious group of animals'.^ Johansen (1925) also calls attention to the probability that swarming or schooling in marine animals, particularly in fish, is a long-term phenomenon, having its origin far back among the larval forms. He notes, for instance, that many fishes, even while still young, move about in shoals, adding that the highly irregular yield of cod and herring larvae from his serial (consecutive) net hauls in the Kattegat distinctly suggested that the phenomenon was ' manifested already ' in the larval phase. As Cassie (1957) has recently said, 'One of the simplest assumptions which can be made as to the pattern formed by organisms in space is that they are randomly distributed. In practice randomness is fairly uncommon, the usual pattern being an "aggregated" one in which organisms tend to cluster in certain regions. . . . Sometimes the reason for aggregation is obvious. Shade loving plants cluster in shady niches. Young animals like to stay near their parents. On the other hand there is no obvious reason why fully grown plankton should not be randomly distributed, since there seem to be no specially favourable niches in the ocean '. Proneness to lifelong swarming such as seems to be mani- fested in the Antarctic krill, however, would provide a simple explanation, even in such a relatively changeless environment as the open sea. Colebrook (i960 a) calls attention to the highly non-random horizontal and vertical distribution of the zooplankton in Windermere, describing (19606) a swarm of Daphnia hyalina at the surface there as 'fairly distinct, about i m. in diameter and confined to the top 10 cm. of the water'. He notes that the vast majority (82%) of the individuals of this swarm were in the same developmental phase and suggests that such a gathering might be produced and maintained partly by water turbulence and partly, if not indeed entirely, by ' social activity of some kind implying purposeful movements by the animals '. IMPACT OF THE SWARMS ON THE OCEAN PASTURES With such a vast herbivorous population so discretely disposed, and the enormous capacity for grazing concentrated in a single swarm, it is not surprising that in nature we should find, side by side, places where the phytoplankton is rich and others where it is scanty or poor. In general, one would expect, the gaps or 'voids ' separating the swarms, so long as they remained untenanted, would be richer, indeed immensely richer, than the water occupied by the congregating krill, an average-sized large swarm (p. 148), say 60 by 40 yards in surface area and i yard thick, containing, at 40,000 to the cubic yard,^ 96 million euphausians, the whole comprising a compact grazing unit of enormous local destructive power. One could well in fact imagine it making a clean sweep of the diatoms among which it swam, then moving on, for the swarms it seems (p. 156) do move bodily about, to wreak equally effective havoc among the richer pastures of the voids. As Bigelow (1926) has said of Calanus finmarchicus, a pelagic swarmer ranking high in ecological importance with the Antarctic krill, it probably 'makes greater inroads on the planktonic plants on which it preys than do all other copepods combined, and conceivably it may practically exterminate them locally and temporarily'.^ No doubt salps too wreak similar destruction, Bigelow (1909) recording that they sometimes occur in the Gulf Stream ' in hordes, and on such occasions strain the water bare '. The swarming salps of the Antarctic could equally well I believe bring about the local extermination of the plants on which they graze (Foxton, 1961). ^ Cf. the marked gregarious habit in Ophiothrix quinquemaculata recently demonstrated by Czihak (1959) in the Adriatic Sea, or Thorson (1957), who remarks that brittle stars occur in such dense swarms that 'every centimeter of surface must be swept by their arms'. - Approximately (see p. 151) one to the cubic inch. ' Cf. Bainbridge (1953, 1957) who has said, 'grazing and migration into a large uniform mass of phytoplankton could result in the eating out of lines of plants'. IMPACT OF THE SWARMS ON THE OCEAN PASTURES 241 A parallel phenomenon is presented by the impact of ctenophores on copepods and other plankton animals. Oi Pleurobrachia pileus, for instance, Bigelow (1926) has written, 'Wherever these cteno- phores swarm they sweep the water so clean and they are so voracious that hardly any smaller creatures can coexist with them. Copepods in particular are locally exterminated in the centers of abundance for Pleurobrachia, though in their own turn they may swarm nearby'. In his recent studies on the distribution of the standing crop of zooplankton in the Southern Ocean, Foxton (1956), working with the material from our vertical nets, arrives at the somewhat unexpected result that the crop (excluding that of the staple whale food which cannot, or virtually cannot, be sampled by the vertical nets) reaches its maximum in the krill-poor West Wind drift, being actually from two to three times heavier there than in the East Wind-Weddell zone where the krill are so abundant. What can lie behind such a remarkable difference? It could, I suggest, spring directly from the impact of the multitudes of grazing swarms on the pastures of the circumpolar sea, at its massive maximum in the East Wind-Weddell stream, where surely it must be operating to the disadvantage of other herbivores, at its minimum in the West Wind zone where equally it must be operating to their advantage. As Graham (1956) has said of the 'world of plankton', it is a ' dynamic world, in which the plants are as directly removed by the animals as is the grass by herbivores of the plains. Within that world there are side-relationships, one species depending on another or adversely affecting another, perhaps by competition for nourishment, perhaps by secretion of harmful or beneficial metabolites '. HAPPENING OF THE WHALES UPON THE SWARMS In krill-rich regions such as the Weddell drift, where as we have seen (p. 148) the swarms may be encountered every two or three hundred yards or so, or even closer together, and may be scattered thus in all directions over hundreds of square miles of sea, it might be supposed that the questing whales simply come upon their food by chance, for in the midst of such profusion they could hardly it seems fail to strike a swarm, or at any rate come close enough to see one, in whatever direction they moved. In other words, as Beklemishev (i960) suggests, having come upon this richly spread table, they probably start seeking their food at random. Obviously, however, though chance must play a part, equally obviously, the swarms being often massed at the surface and distinctly visible there, in daytime at least, eyesight must also come into play. For as Eraser and Purves (1959) observe there is evidence that the visual sense can be normally developed and even adapted to underwater vision in conditions of good visibility.^ Hjort (1933), too, remarks on the keen eyesight of fin whales, observing that no sooner do they catch ' sight of a patch of food than they are upon it in a flash'. It may be, too, that the incessant movement of the limbs, pleopods and mouthparts of the concen- trated myriads of a swimming and feeding swarm creates a vibrational disturbance in the water^ to which the whales are acutely sensitive, the wax plug in the external auditory meatus, as Eraser and Purves (1954) and Purves (1955) have recently discovered, being a good conductor of sound, ' especially ^ MacGinitie and MacGinitie (1949) remark on the poor quality of whales' vision, stating that 'most whalebone whales have poorly developed eyes and have little use for sight'. Breathnach (i960), after a comprehensive review of the literature, states that until much further work is done ' it is not possible to form a final opinion on the status of vision in Cetacea. Observa- tions from life, Mann's findings [see p. 143], and the condition of the subcortical centres suggest, however, that it is much more acute and discriminatory than previously thought'. ^ Strange echo records have recently been detected in the North Sea (Gushing and Richardson, 1956). They are described as 'noisy' and seem to have been produced by densely packed euphausians (Nyctiphanes couchii) just below the surface. 27 DM 242 DISCOVERY REPORTS of high-frequency vibrations'.^ This again, especially at night, if they feed at night, would help them to locate their prey. Fraser and Purves (1959) also call attention to the recent discovery that whales use echo-location.- They write: ' The echo-location method consists of the transmission and reception of an intermittent pulse with a variable recurrence frequency comparable with that used by bats. There is evidence also that cetaceans can discriminate the quality of echoes with a high degree of accuracy. Judged by the great development of the brain of these animals it would not be surprising if it is found that they have a complete sound picture derived from purely auditory sensations'. In effect, as Gray (1953) has remarked of bats, the whales 'have evolved a very efficient echo-sounding equipment', and as Griffin (1955) more recently has said, its employment 'either to maintain orienta- tion with respect to the bottom, surface, or large obstacles, or possibly to locate fish or other prey in the water as bats appear to do in the air ', seems a distinct possibility. From the results of his recent experiments with bottle-nosed dolphins Kellogg (1958) concludes that the ' location and discrimination of submerged objects by reflected sound signals is without doubt a necessary and fundamental perceptual avenue for these cetaceans', and there seems every reason therefore to believe, as Jonsgard (1959) has just written, that 'both the sounds which cetaceans produce and their well-developed auditory organs play a large part in the welfare of this group of animals '. Lillie, who originally described the wax plug in 1910, suggests that the 'whale probably receives sound-vibrations by means of vibrating bony surfaces, after the manner of fishes ', the tympanic bulla, fastened to the periotic bone by two thin pedicles, being a relatively dense and heavy sounding-box that could easily be set in motion. In a recent review of sensory perception in the Cetacea Slijper (1961) concludes that it may well be that 'whales and dolphins in search of food are attracted by the sound of their prey'. Hardy (1956) states that sometimes in the Antarctic, large swarms of krill show up in the dark as phosphorescent patches in the water, adding that 'perhaps the whale is more easily able to find the best feeding place when it is lit up like a restaurant ! ' I question however if the night swarms normally light up the surface of the sea, personal observation over many years and a search through the litera- ture suggesting that such a phenomenon, although it may happen, is far from common. I have never seen the krill behave so myself nor have I heard mention of such behaviour by the whalers, and they perhaps would be the most likely to know about it. Antarctic literature contains surprisingly few references to night phosphorescence whatsoever, and fewer still, if indeed any, to massive displays such as might be expected from large swarms of krill. The 'uncommonly white ' water (p. 40, note i) reported by Cook in February 1775 could it is true have been caused by a luminescent krill patch^ and so too perhaps could the 'gleaming' sea reported by Bellingshausen (1945) south of Kerguelen in February 1820. Luminescence in the krill was observed by Murray (1895) in the 'Challenger' in February 1874 and by Wilton (1908) in the ' Scotia' in February 1903, and they, so far as I am aware, are the only observers (other than Hardy) to have recorded it. Murray writes, 'The above surface organisms were collected by Mr Murray in a boat, Eiiphausia superba being especially abundant (the supplementary eyes of which were in the evening observed to be phosphorescent) '. There is some doubt however as to whether this refers to a natural display in the sea or to an exhibition of lumi- nescence by a sample brought back alive on board. For if Murray was close enough to see it was the 1 The whale's ear in fact is said to be so sensitive that it can even detect the ultrasonic beam emitted by the modern ' echo whale-finder' (Crisp, 1954). It is interesting, therefore, to record that during recent survey operations by a Royal Naval Ilydrographic unit in West Graham Land over-inquisitive killer whales were readily 'discouraged from taking an unhealthy interest in the motor-boat' by switching on the echo-sounder (Wynne-Edwards, i960). - Klumov (1961) suggests that they come upon their food mainly by echo-location. ^ It was sighted round about midnight, near Candlemas Island in the South Sandwich group. HAPPENING OF THE WHALES UPON THE SWARMS 243 light organs of his specimens that were glowing he must I think have had them before him in a dish. Gunther it will be recalled (p. 153) also mentions them shining in the dark, but his specimens were in contact with the deck of a vessel, not in their natural element. Wilton's observation on the other hand clearly refers to a natural display. 'Many Euphausia', he writes, 'were frequently observed, and it was noticed after dark that these caused a phosphorescence on the ice and in the water'. This was recorded in 60° 10' S, 42° 35' W shortly after the 'Scotia' entered the pack-ice of the Weddell Sea. A year later in the same waters, but much farther to the south and east, he records other instances of night phosphorescence but gives no hint as to the cause. Some he describes as ' large blobs ' on the surface about the size of a Royal Quarto page. Much in fact of Wilton's phosphorescence might have been produced by organisms other than E. superba.^ The absence of night phosphorescence in Antarctica is mentioned I see by Webster (1834), an early observer of natural phenomena in these high latitudes of particular accuracy and diligence. He writes, ' At no time did I observe the least tendency to scintillation in the waters of the southern parts; and, although the nights were very dark when we left, the sea was not phosphorescent'. Carrington (i960) it is true has said that after dusk the bright red euphausians so abundant in the colder parts of the southern seas ' cause the whole surface of the ocean to glow with a ghostly radiance ', but he gives no authority for this. It is possible that the krill do not luminesce spontaneously, but perhaps only when agitated as, for instance, when colliding violently with the deck of a ship or perhaps when in contact with ice. Clarke and Breslau (i960) find that among certain Dinoflagellates, while spontaneous luminescent flashing is rather weak, their flashing increases strongly on agitation. IMPACT OF THE WHALES ON THE SWARMS Since they vary enormously in size, from some only a foot or two across to others as much as half an acre or so in surface area, it is obvious that while in certain instances a whale could dispose of an entire swarm at a single mouthful in others the demoHshing or gross depletion of a patch or local accumulation of patches would be far from being such a catastrophic occurrence. In general, single whales, or small groups of them, would by themselves no doubt take some considerable time to bring about a serious local depletion among the swarms on which they fed and would, therefore, tend to remain for some time in the area in which they were feeding. Large schools on the other hand, if they kept together, would make shorter work of their prey and would, therefore, tend to keep on the move more readily than the scattered whales. It is easy to imagine, for instance, the 100 or perhaps 200 whales Gunther describes feeding on p. 150 wreaking such wholesale devastation among the local concentration of swarms on which they were centred that in a relatively short time they would be compelled to move on to find, and continue their work of destruction in, fresh fields as yet perhaps unravaged by their kind. Indeed it will be recalled Gunther 's foraging whales already seem to have left their mark upon the swarms on which they preyed. And they do not, it seems, move by any means entirely at random, the results of whale-marking distinctly suggesting that having wrought destruction, or at any rate gross depletion, in one area they tend to move away towards richer, or shall we say as yet unravaged, fields against the main trend of the surface stream that is carrying their food as it flows, o-group whales,^ for instance, marked near South Georgia (Rayner, 1940), in a remarkably large number of instances have been killed in Weddell West, suggesting that the foraging whales as they continue to feed move in the main against the surface ^ Hodgson (1905), for instance, records that during the winter in McMurdo Sound 'hauling the tow-nets always provoked a brilliant display of an emerald green phosphorescence, chiefly from the contained Copepods and Ostracods'. 2 Whales into which marks have been fired and recovered later in the same season. 27-2 2^ DISCOVERY REPORTS drift. Other, although far fewer, o-group whales marked in more easterly parts of the Weddell current seem to have behaved in the same way, moving westwards into, or towards, Weddell West counter to the surface stream. Moreover, whales of groups 1-4I all show latitudinal or meridional dispersal that again suggests that their principal movements while feeding are against rather than with the currents, the East Wind whales tending to move east, the Weddell whales west and the South Georgia whales south or south-west. We cannot, of course, be sure that the apparent movements of the later group whales, especially the lateral movements, are real, or even that they take place in Antarctica itself, because as Brown (1954) points out, a whale may have done much travelling between the positions of marking and recovery, even whales killed a few weeks after marking possibly having wandered a long way off the direct lines joining the marking places and those where recovery was made. Even so these resuks seem significant, especially in so far as they seem to indicate a definite anti-current trending in feeding o-group whales. It must be significant, too, that the majority of whales of all groups marked in the krill-poor West Wind beh have been killed farther south in the well supplied feeding-grounds of the Weddell and East Wind zones, further indication of pronounced trend towards regions of richer browsing. Definite evidence of upstream (easterly) movement in the west-going East Wind drift is scarce, such indications as there are coming mainly from i, 2, 3 and 4-group whales. Brown (1956), however, gives two instances involving o-group whales where easterly movement is marked, in the same paper pro- ducing further evidence of due south movement in o-group whales from the krill-poor West Wind zone. A particularly notable instance of upstream lateral movement in the krill-rich Weddell drift is given by Ravninger (1955). A blue whale marked from the 'Enern', with two marks, one protruding, on 28 November 1954 in 54° 12' S, 16° 02' E, was recognised later by the captain in 60° 46' S, 11° 23' W on 10 February 1955. Brown (1956) records two even more striking cases where whales, moving upstream, travelled a long way to the west in this krill-rich zone. A fin whale marked in 54°o6'S, 07°03'E on 15 November 1955 was killed loi days later in 55° 54' S, 33° 50' W, 1380 miles away to the west, and in the same season the mark from a blue whale shot in 59° 24' S, 41° 13' E on 2 November 1955 was recovered 107 days later in 61° 37' S, 08° 09' W, 1470 miles away to the west. Together these records show that while feeding, or on the search for food, some whales in the Weddell stream undergo a resuhant westerly movement of from 11 to 14 miles a day. Marks recently recovered from o-group whales in the Pacific sector, reopened, after a long interval, to whaling in 1956, reveal there (Brown, 1957), a restless movement among the whales, with a distinct westerly trend. This is especially noticeable in the narrow strip of West Wind water lying to the south of 65° S, where some whales, both fin and blue, in the course of their summer feeding have travelled almost due west for many hundreds of miles. Marks from i -group whales suggest a similar movement. In this former sanctuary for whales (p. 394, Fig. 135 ; p. 410, Fig. 143) their food is scarce, it being in this very region (p. 48) that they have been turning to Thysanoessa macrura to eke out it seems a scanty diet. These extensive lateral summer movements, therefore, could well spring from a restless search for food and are probably aggravated by the fact that large-scale dispersal to the south is denied them because of the vast and impenetrable ice-sheet on the Pacific side (p. 49) that covers so much of the East Wind zone. It is interesting too that so many have gone west, against, not with, the surface stream. Finally, the marking results seem to suggest that in regions well supplied with food the whales do not undertake extensive lateral movements but tend to' remain where they are where the feeding is good. This seems to apply with particular emphasis, for instance, to Weddell West (p. 394, Fig. 135), a region offering a rich table for the whales and one, moreover, that even while suffering 1 Whales into which marks have been fired and recovered from one to four years later. IMPACT OF THE WHALES ON THE SWARMS 245 depletion is continually being replenished (p. 384) by fresh influxes of euphausians from the East Wind drift as it sweeps coastwise up from the Weddell Sea.^ It cannot, it seems, be without some significance that marks fired into o-group whales in this krill-rich field, with two exceptions, have (where recovered) all been recovered inside it, always allowing of course (Dawbin, 1959) that the distribution of the recoveries in this region was something more than simply a measure of the distribu- tion of the marking and chasing vessels. In his recent study of the distribution and movements of blue and fin whales in the Atlantic sector Arsenev (i958^>) states that during the summer the herds are always on the move in search of food, their position at any time being fixed by the Euphausia they eat. He finds too that in the Weddell drift their principal movement while feeding is from north-east to south-west, that is, counter to the surface stream. It has been assumed for some time (Chittleborough, 1958) that the stocks of humpback whales in Antarctic areas iv and v are distinct and that they do not intermingle. Recently, however, from an analysis of the data forwarded to the Bureau of International Whaling Statistics, Chittleborough (19596) has demonstrated what seems to have been a large influx of humpbacks from area v westwards into area iv in the summer of 1958-9, and recent recoveries of marks from this species (Brown, 1959) suggest the movement may be of regular occurrence. Regarding what seems to have happened in 1958-9, Chittleborough remarks, 'it seems reasonable to suggest that an environmental change (possibly of a temporary nature) resulted in some change in the distribution of euphausiids, enabling the humpback whales in the Group v population to feed farther west than usual '. It is possible, I suggest, that they were obliged to feed farther west than usual, easy access to the krill-rich fields of the East Wind zone perhaps having been denied them owing to a northerly, and perhaps only local, extension of the ice-edge, compelling them to move west, counter to the West Wind stream, in search of their summer feed. Similar West Wind movements, apparently skirting the northern periphery of the East Wind zone, seem to have been undertaken in the same season by blues and fins (Brown, 1959). In his most recent account of the feeding and movements of baleen whales Nemoto (1959) suggests the interesting possibility that if rich Antarctic feeding-grounds are already occupied in ' biological strength ' by the early migrating blues, the late comers, the fins, might be obliged to seek their food in other waters, and this he concludes would account for the segregation of the larger southern feeding whales into groups of fins and groups of blues. Between 1945 and 1957 a total of 960 blue, fin and humpback whales was marked in the Antarctic (Clarke and Brown, 1957) and the work is still going on. It will be interesting to see if future recoveries provide further evidence of the feeding movements the marking experiments so far seem to suggest. INTIMATE STRUCTURE OF THE SWARMS Sex ratio Table 49 shows the proportion of males to females in samples from swarms in which the sex ratios have been accurately determined. The figures are based partly on the sex determinations published by Bargmann (1945, Appendix, Table 19) and partly on the field observations of other members of the Discovery staff. Only samples of fifty individuals or over have been included. In the majority of instances, it will be seen, these almost exclusively random gatherings show repeatedly that the swarms are composed of males and females in approximately equal numbers, the three instances of non-random sampling, namely, where nets were deliberately towed through swarms seen on the surface, also showing a sex ratio which for all practical purposes is unity. There are comparatively few instances where the males vastly outnumber the females or vice versa, and it is 1 Especially, it seems (pp. 384 and 390), in summer, from January to March. 246 DISCOVERY REPORTS Table 49. Sex ratio in swarms of E. superba Per- Per- No. of No. of centage centage Station males females male female Remarks Per- Per- No. of No. of centage centage Station males females male female Remarks January April WS373 48 51 46 52 — 42 64 27 70 30 — 575 22 47 32 68 — 207 41 70 37 63 1 Females spei — 169 156 52 48 Sample from 207 12 40 23 77 / swarm seen on WS 196 33 28 54 46 — the surface 665 33 33 50 50 — 2346 56 55 50 50 — 602 33 30 52 48 — WS537 60 62 49 51 — May 797 25 38 0 40 60 — 1359 30 52 37 63 — 799 103 81 56 44 — 818 157 270 37 63 — June 820 54 23 70 30 — 82s 136 208 40 60 — 1781 24 75 24 76 — 2168 8 399 2 98 Females gravid August 2195 76 378 17 83 Females gravid 2567 86 91 49 51 — WS 264 37 33 53 47 February 2408 2412 31 20 59 34 34 37 66 63 WS152 70 36 66 34 — WS156 68 50 58 42 — September 351 22 33 40 60 — WS 277 34 16 68 32 _ 352 354 30 53 30 29 50 65 50 35 — I8I3 37 55 40 60 — 354 356 73 4 38 45 66 8 34 92 Females gravid October 612 231 302 30 70 — WS290 27 35 44 56 — 612 86 58 60 40 — WS298 42 35 55 45 — 618 46 54 46 54 — 454 95 90 51 49 — 622 59 41 59 41 — 454 21 29 42 58 — 624 46 25 65 35 — 455 22 28 44 56 — 627 54 46 54 46 — 461 93 123 43 57 — WS915 215 531 29 71 Females gravid 463 30 22 58 42 — March 1823 1839 35 84 37 87 49 49 51 51 23 42 33 56 44 — 24 5 79 6 94 — November 38 368 643 643 50 38 40 35 50 32 60 74 50 54 40 32 50 46 60 68 — 484 492 494 43 28 36 36 30 61 54 48 37 46 52 63 56 — 658 92 139 40 60 — 523 95 121 44 1156 1711 35 51 23 40 60 56 40 44 — 1009 1854 57 36 51 75 53 32 47 68 — 1715 1717 23 80 34 84 40 49 60 51 December 2268 53 212 20 80 ' Males and females im- mature — 29 28 51 49 1' 47 • 51 samples from 2268 2268 2268 12 37 22 60 97 84 17 28 21 83 72 79 • 125 52 49 47 50 53 49 swarms seen on surface 2271 22 37 37 63 Females gravid 537 216 163 57 43 — 2295 65 35 65 35 — 539 29 31 48 52 — 2295 5° 19 72 28 — 539 17 36 32 68 — 2295 44 18 71 29 — 548 43 36 54 46 — 2606 82 69 54 46 — 549 36 35 51 49 ^ INTIMATE STRUCTURE OF THE SWARMS 247 LENGTH RANGE MONTH APR SO- lOO- FEB SO- -too- DEC 50- -100- DEC 50- -100 JAN 50- -100- NOV 50- -lOO- OCT 50- -100- OCT 50- -100- OCT SO- -100- OCT 50- MONTH 8 12 16 2024 28 32 36 404448 52 56 60 64 STATION WSI96 627 S37 492 454 459 463 WS295 STAGE 9 I 2 3 4 S 6 7 1 I I ! I I r II I I 1 I -| — I — I — I — r STAGE o* I 2 3 4 5 6 7 1 — I — I — I — I — I — r ^ — I — I — I — r^ — I — I — r -\ — I — 1 — I — r 1 — I — I — I — I — r -T — I — I — I — I — r 2 3 4 5 6 7 T — I — i — I — r 1 I I I 1 r -T — I — I — I — r 3 4 S 6 7 T — I — I — I — I — T^ — I — I — I — I — r -T — I — I — I — I — r^sn — I — I — I — I — r [ 1 — I — I — I — I — r™T — I — I — I — I — r 8 12 16 2024 28 32 36404448 52 S6 6064 I 2 3 4 5 6 7 STAGE 9 I I 2 3 4 5 6 7 STAGE (j' ^ SAMPLE FROM SIGHTED SWARM Fig. 50. Sexual dimorphism and developmental condition of the sexes in svfa.TmsoiE.superba,showmg growing dominance of the male. Arranged in ascending order of modal size. Length range in 4-mm. groups, e.g. 16 includes specimens of 13, 14, 15 and 16 mm. difficult to say whether or not all may be simply due to inadequate sampling. There appears to be some evidence, however, that subsequent to pairing (see Stations 2168, 2195, 356, WS 915, 2271 and 207) the swarms may in fact consist predominantly, and sometimes almost exclusively, of females, all either gravid or spent, suggesting that when the males have transferred their spermatophores they begin to die off earlier than the females, the latter persisting for some time as the dominant, and eventually perhaps as the exclusive component of the swarm. In most animals that have been studied it may be remarked (Comfort, 1956) the male sex is the shorter lived. 248 DISCOVERY REPORTS MONTH lOO MAR SO- too- MAR so lOO- FEB SO -lOO- FEB SO lOO MAR 50- lOO FEB SO -lOO APR SO 8 12 16 2024 28 32 36404446 52 566064 STATION 368 -lOO- APR SO- -lOO- FEB SO- lOO FEB 50- MONTH 38 351 354 24 352 42 207 618 622 STAGE $ 2 3 4 5 6 7 — 1 — 1 — r-1 — 1 — r J i 1 — ; — I — r 1 I I I — I — r -I — 1 — 1 — r 2 3 4 S 6 7 STAGE cf" 12 3 4 5 6 7 T — I — I — I — I — I — r -^ — I -^ — I I — r^^^T*- r^ — I — -^ — I — r "1 — I I I I — I — r I 2 3 4 5 6 7 J 1 I I I — I — r i 4 1 — I I I I I I M 1 1 I 1 1 1 r J 1 I I T" 8 12 16 2024 28 32 3640444852 566064 I 2 3 4 5 6 7 fe. STAGE 9 2 3 4 5 6-^ STAGE cT LENGTH RANGE Fig. 5 1 . Sexual dimorphism and developmental condition of the sexes in swarms of E. superba, showing growing dominance of the male (Continuation of Fig. 50). Relative size of the sexes In her growth curve illustrating the average monthly length of the larvae, adolescents and adults Bargmann ( 1 945 , p. 1 29, Fig. 3) shows that from the earliest stage at which the sexes can be distinguished (when the krill have reached the Sixth Furcilia stage) the females are consistently, if only slightly, smaller than the males. Using largely the measurements given in Table 19 of her appendix an attempt has been made in Figs. 50-52 to illustrate this slight measure of sexual dimorphism in so far as it can be expressed in terms of the length frequencies of the sexes that have been recorded in our INTIMATE STRUCTURE OF THE SWARMS 249 LENGTH RANGE Fig. 52. Sexual dimorphism and developmental condition of the sexes in swarms oi E. superba, showing dominance of the male giving way to dominance of the female after pairing has been accomplished (See legend to Fig. 50). samples of the younger and older swarms. In general it will be seen that, up to a point, the dominance of the male, detectable even in swarms consisting of Sixth Furcilias and very early adolescents (Fig. 50, Stations WS 295, 463, 459, 454 and 492), tends to become more pronounced as development proceeds. It does not, however, persist throughout the life of the swarm, tending to disappear when pairing (Fig. 52, Stations WS 152 and WS 373) has been successfully accomplished and the majority of the females, being impregnated, have reached stage 6. From then until it becomes gravid the female rather than the male would appear to be the dominant partner. 28 2S0 DISCOVERY REPORTS The dominance of the adult female is particularly pronounced, for instance, at Station WS 915 where a full grown swarm consisting of stage 7 males and gravid females was encountered in the East Wind zone near Enderby Land in February 1936. The sample obtained consisted of 215 males and 531 females all of which were measured to the nearest millimetre. Their length frequencies in 2-mm. groups, plotted on a somewhat finer scale than has been used in Fig. 52, are shown in Fig. 53. SO-i 40 ;30- '20 10 ■ 100 SC^ 32 34 36 38 4042 44 46 48 SO 52 54 5658 60 62 64 66 LENGTH RANGE I 2 3 4 5 6 7 STAGE d* I 2 3 4 5 6 7 STAGE $ Fig. 53. Sexual dimorphism and developmental condition of the sexes in a fully adult swarm of E. superba, showing dominance of the female in a sample from St. WS 915. Developmental condition of the sexes In Table 19 of Bargmann's appendix it is repeatedly shown that, except in the very young mixed larval and adolescent swarms encountered from August to November, the dominant state of sexual development in the male is normally one or two stages, sometimes more, ahead of the corresponding state in the female. The advanced state of the male, as Figs. 50-2 show, persists throughout the greater part of the life of the swarm up to the point of pairing, when as might be expected swarms in which both males and females are dominantly at stage 7 are not uncommonly encountered. An ingenious device has recently been constructed by Vittorio and Livia TonoUi (1958) and used to study the form and internal structure of plankton patches in the Italian lakes. It could well, it seems, be employed with advantage in future studies of the biometrics of krill swarms, with especial reference perhaps to the larval swarms. A rather long plankton net ends in a polysthyrol tube that reaches the boat. The tube is here connected to a turbidi- meter, from which another tube runs into a metallic box, where a controlled aspiration can be exerted. The water, containing plankton filtered by the net, ascends through the polysthyrol tube, and, flowing through the turbidimeter, enters the vacuum box. In this way, by regulating both the speed of the boat and the aspiration to a constant value, we can obtain, through the variation of the turbidity values, an indication of the density of the particles suspended in the layer where the net is fishing. In the vacuum box we have twenty vials, each provided with a filtering window. When we are warned that the net has reached a patch of plankton, i.e., when turbidity increases, we can defiect the current in which the plankton is suspended successively into each vial. In this way we can collect successive uninterrupted subsamples along the same patch. INTIMATE STRUCTURE OF THE SWARMS 251 By means of this apparatus we are investigating whether the structure of a patch is in any way orderly Above all, we are considering the food availability in the interior of the swarm, and the varying distribution of sex-ratios and developmental stages. In this way the form and internal structure of a larval swarm might perhaps be accurately mapped together with the form and internal structure of adjacent swarms and the distances separating them from each other. We already indeed find that such swarms, in so far as their stage frequencies go, are anything but disorderly. 1 I I [ 1 1 r DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT T 1 f 1 1 r NOV DEC JAN FEB MAR APR Fig- 54- Growth curves of E. superba after Ruud and Bargmann. GROWTH AS INDIVIDUALS AND SWARMS As INDIVIDUALS In illustrating the growth-rate of plankton animals, it is customary to plot the average monthly or half-monthly length of as large a series of measured specimens as possible. Thus Ruud (1932), mainly from specimens collected from whales' stomachs, and Bargmann (1945) from specimens collected from the plankton, obtain the growth curves reproduced in Fig. 54. Before one year's growth is over, when the adolescents (p. 339, Fig. 95) first appear in substantial numbers in August, Bargmann is able to distinguish between males and females and accordingly from August onwards her growth curve divides into its male and female components. As Fig. 54 shows the growth- rate in the two sexes is very similar, the two curves, although the females are consistently smaller than the males,! following approximately the same course. Ruud does not treat the sexes separately and since he only had substantial larval samples for January, and little or no larval material covering the rest of the year, his curve for the greater part of the first year's growth is largely conjectural, and, as Bargmann (1945, p. 123) points out, in its early stages at any rate, much too steep. In certain broad essentials, however, the two curves agree, both showing at the end of the first year the increase in growth that accompanies the spring blooming of the phytoplankton and both a distinct slowing up 1 See again, however, Fig. 53 which clearly suggests that in a final burst of growth the female overtakes the male. 28-2 252 DISCOVERY REPORTS of growth beginning towards April of the following year with the onset of the winter phytoplankton mini- mum. Surprisingly, however, neither show any marked acceleration at the end of the second year of life, when the phytoplankton blooms again. ^ It will be seen that in December and January at the end of the first year, and again in January and February at the end of the second, Ruud shows higher average monthly length figures than Bargmann does either for her males or females. This discrepancy may well have arisen through the curves having been based on the one hand on stomach samples, and on the other on plankton samples, the whales (p. 1 43 , Fig. 1 4) being distinctly more efficient samplers of the larger and more active krill than our stramin nets. Professor Hardy, with whom I have discussed this matter, points out that the different effects of digestion and fixation upon the living krill could also account for the difference, or contribute to it, digestion tending to relax the animals, fixation to contract them. Comparing his growth curve for Euphausia triacantha with Bargmann's curve for E. superba Baker (1959) notes that whereas E. triacantha stops growing in the first half of the second year of life, E. superba is still growing during the early part of the third. The greater size attained by the adult krill would, there- fore, he concludes appear to be reached by a longer period of growth rather than by a faster average rate. The material on which the curves in Fig. 54 are based comes almost exclusively from the Weddell drift, the Bransfield Strait and the South Georgia whaling grounds, the resultant curves, there- fore, referring essentially to the relatively warm northern zone of euphausian abundance and not to the colder slow-growing region of abundance in the East Wind zone. Bargmann's material however does in- clude a few specimens from three samples obtained in the East Wind zone (Stations 575 and 602 in January and Station 1359 in May) and it is interesting to note that the average length of the males and females in all three falls below, in two instances well below, the natural trend of her male and female growth curves for the northern zone.^ As SWARMS Since it now seems certain that E. superba from hatching onwards spends its whole existence as a member of a swarm, itself (p. 230) virtually an individual organism, it has been considered worthwhile trying to express the growth-rate in terms of the development of the swarms themselves. Without having recourse to the massive labour of working out for every swarm the average length of the euphausians in it, this has been done by plotting the sharply defined modes into which the length frequencies of the larval, adolescent and adult swarms so consistently fall on a time-mode-value scatter diagram, the value of the mode in each instance being taken as a simple and convenient measure of the developmental condition of the swarm as a whole. In larval swarms, for instance, in which the mode is represented by a stage and not by a length frequency, I have expressed the modal value in terms of the average length of the dominant stage, a swarm for example in which the dominant stage is the First Calyptopis being given a value of 2, the average length to the nearest millimetre of the First Calyptopes measured by Eraser (1936, p. 25, Table viii). Thus, taking the larvae as a whole and still following Eraser, the dominance of any particular stage in any individual swarm has been expressed to the nearest millimetre, or as for the Fifth and Sixth Furcilias to the nearest two millimetres, in terms of modal value as follows : Modal value Modal value Dominant stage (mm.) Dominant stage (mm.) Nauplii or Metanauplii i Second Furcilia 6 First Calyptopis 2 Third Furcilia 7 Second Calyptopis 3 Fourth Furcilia 8 Third Calyptopis 4 Fifth Furcilia 10 First Furcilia 5 Sixth Furcilia 12 * But see, however, p. 253, Fig. 55. ^ Since this was written Nemoto (1959) has produced a growth curve based on material from the East Wind zone. This does in fact show that throughout life both males and females in this far southerly coastal belt are distinctly smaller than their contemporaries in lower latitudes. GROWTH AS INDIVIDUALS AND SWARMS 253 Concentrations of eggs have also been given a modal value of one. In swarms, other than purely larval swarms, where the modal value has been obtained directly from measurement, it is expressed by 2-mm. groups up to 44 mm., and thereafter by 4-mm. groups up to 60 mm. The resuks of this treatment of the data are shown in Fig. 55 in which the open circles represent the swarms in the East Wind drift and the solid circles the swarms outside it, the vast majority of the latter from the Weddell or northern zone. Taking first the position in the north it will FEB APR NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT 1 1 1 1 t 1 1 1 1 1 1 1 NOV 1 DEC 1 JAN t FEB MAR APR MAY JUN JUL AUG II SEP OCT 1 1 NOV DEC JAN MAR 1 1 1 1 60 u 56 •• a r.^K •• 52 ^ •Im a. :k Uil- o- 48 / ** \ .~ ^ •• •• HSV»» 44 * * J * • • • 0*0 S 42 \«» • / K • oo 40 •• it •• ** * ** is " 38 • • *** 0* * * • • • o o 36 ••• •'•*• S •• •• • • o 34 • • • ••• • • 32 • W ; 's* o • a 30 • ••■ •A ooo o • :k THIRD YEAR 28 • •• • S o • SWARMS 26 y ^ I r\f^^ vr^Af^ ^\ti A m tc ^— ^_^^»^^^__ ^ .* •- JW BOO OO • 24 < FIRST YEAR SWARMb * •• "* OOO ooo 22 M. :: %• CO oo 20 .... X o.'oo o 18 :«;« WEDDELL SWARMS :V::V: ••;.?• VA S8 16 • ••■M» •.• ••• SSS^SS EAST WIND SWARMS ... .•.•.*.'.'.. ••• 14 ::: ::::::: •V.' 12 •• •••••% ill ••• • ■> 10 :: • • ;:: ::: ■< -SECOND YEAR SWARMS — > 8 7 ■•■ . • 6 * •*.*■*• .V. u 5 • •. ••• • 4 ::" :!tJ ••• 3 •• .. :: v 2 •• :::::s-SssgH"H' stage $ (i) (i) (i) (0 (0 (0 0-2) (2) (2) (2) (3) (3) (3) (3) (4) (5) (5) (6) (7) 1 . ::: :/. ::::: :;;;s;: -. ., ^ (0 d) (0 (1) 0-2) (2) (2) (3) (3) (3) (4) (4) (4) (5) (6) (5) (7) (7) (5) 1 1 [ 1 I 1 1 1 1 1 1 1 NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT 1 NOV 1 DEC 1 JAN 1 1 1 1 1 1 1 FEB MAR APR MAY JUN JUL AUG 1 1 SEP OCT 1 1 1 1 NOV DEC JAN MAR FEB APR 60 56 52 48 44 42 40 38 36 34 32 30 28 26 24 in 22 3 20 § IB 16 < 14 o 12 2 10 8 7 6 5 4 3 2 I Fig. 55. Growth of the swarms expressed in terms of modal value, showing the persistent lag in developmental stature that is found in the East Wind drift. Vertical scale from 10-44 rrini- inclusive in 2-mm. groups (e.g. 10 = 9-10 mm., and 44 = 43-44 mm.) and thereafter in 4-mm. groups (e.g. 48 = 45-48 mm.). For further explanation see text. be seen that spawning, in so far as it is represented by concentrations of eggs, Nauplii or Metanauplii, begins in November, when the first eggs are found, and continues throughout December, January and February until March. It is possible that it may last even longer, into April when the last sur- viving young swarms consisting dominantly of Metanauplii are represented in our samples. From January onwards, until June, they exhibit monthly a heterogeneous pattern of ages, the result of the successive releases of batches of eggs over a period lasting at least 5 months. The majority encountered from January to March, however, are represented principally by the First Calyp topis stage. The first swarms in which the Fourth and Fifth Furcilias are dominant appear in March, swarms dominated by the last larval stage, Furcilia 6, although present in June, not appearing on a substantial scale until July. As a result of the slowing up of the growth-rate that accompanies the winter phytoplankton minimum, the pronounced spread of the modal values encountered from February to June undergoes a considerable degree of compression, which, becoming first apparent in July, lasts virtually to the end of October. In November, however, when the phytoplankton blooms there is a renewed spreading out of the modal pattern, the rapidity of the growth that takes place from then until March being very striking, the modal values, concentrated about 16 in October, rising to a maximum of 46 by the end of the second year's summer. The decline in phytoplankton values that 254 DISCOVERY REPORTS follows, however, again appears to be accompanied by a marked slowing up of the growth-rate, the vast majority of the second year swarms showing little upward trend in the scattered values of their modes from April to October. Thereafter, when the phytoplankton blooms again, there is a final and abrupt acceleration to the full adult stature. It will be seen that there are seven adolescent swarms, enclosed in Fig. 55 by a circle, that fall outside the natural trend of the modal values exhibited by the swarms as a whole. They were recorded at stations made in February and March at the beginning of the second year of growth, and although their modal values lie above the general trend for this period, the developmental condition of the sexes in them, the males dominantly in stage 3 and the females dominantly in stage 2,^ suggest that their position in the plot is correct. For if, as Bargmann (1945, p. 124) has shown, each develop- mental stage probably lasts 2 months in the male and 2^ months in the female, it is obvious that not one of these seven swarms would reach full maturity for at least another 10 months or more. Possibly, however, they are exceptional, representing perhaps the maximum possible growth attainable during the first 15 or 16 months of larval and adolescent life. It is perhaps significant, too, that all seven come from the South Georgia whaling grounds where in summer (p. 79, Table 5) the krill encounter, and perhaps grow most rapidly in, the warmest conditions of their temperature range. There may, however, be another explanation of this apparent anomaly. Having regard to the large size of the individuals of them it is possible that all seven swarms may after all be at the beginning of their third year of life, not, as plotted in Fig. 55, at the beginning of their second, 26-27 month old swarms in fact that, having failed to ripen to maturity, will not spawn perhaps for another 12 months. In other words, it is suggested that the life-span of this species, normally 27-28 months, may occasionally, if rarely, be protracted for yet another year.^ In E. triacantha Baker (1959) finds a similar possibility. This is in fact an interesting point, because it provides a loophole for there being interbreeding between broods that would otherwise appear to have been completely iso- lated in time. Without some such interbreeding one might have expected that such isolation, over a vast period of time, might have produced two distinct subspecific races. Although as yet no one has examined our material minutely enough to say if there are in fact any differences in the characters of the krill spawned in alternate years, such a study, it is possible, might throw some interesting light on the processes of speciation. One would not, of course, expect there to be much difference between the two races because selection by the environment would be the same for both. The figures in brackets at the bottom of this diagram show the prevailing monthly condition of the sexes in the northern or Weddell swarms.^ They are based throughout on the sex stage determinations of Bargmann (1945, Appendix, Table 19) and have been worked out by taking the dominant stage of male and female sexual development in the samples she examined and expressing it as a monthly average. Little further comment need be made on these stage figures here, the developmental condi- tion of the sexes in the swarms having already been discussed on p. 250. It may be noted, however, that the dominance of the male, in so far as sexual development goes, which is detectable from December of the first year onwards, gives way in March-April of the third year to a dominance in the female. Occurring as it does at the very end of the life-span of the swarms this reversal of dominance can probably as already mentioned (p. 247) be ascribed to the fact that the fully adult males, having transferred their spermatophores, have already begun to die off, leaving the swarms consisting largely of spent or gravid females. 1 Following Bargmann (1945, Appendix, Table 19). - See also p. 402. ^ Although our field observations contain much general information as to the developmental condition of the swarms in the East Wind drift, noting, for instance, the presence or absence of spermatophores in the males and females, and whether the latter were gravid or spent, there are little precise developmental data, based on dissection, from these high latitudes. It has accordingly not been possible to work out exact figures for the prevailing monthly condition of the sexes for the swarms in the East Wind zone. GROWTH AS INDIVIDUALS AND SWARMS 255 Although the East Wind drift is virtually closed to vessels except from January to May there are enough data from this short period to show that, following the late spawning and subsequent slow larval growth-rate in these high latitudes, the mode values of both second and third year swarms, at their maxima and minima, tend to fall persistently below the corresponding values for the northern DEPTH FURC 6 . 5 . 4 3 2 I CAL 3 2 • I META NAUP 2 N9 EXAM- -INED 8 762 JAN 28 28 FEBRUARY 2 4 17 17 17 17 282B28282825 MARCH APRIL 2020202020202021 22 22 23 3333244444488889 22 22 22 2026 25 1662 1671 N19 NI9 1545 1545 1545 2600 260O 355 2603 2603 2603 26IO 2610 1713 1713 I720 854 854 855 855 2335 1346 Vi2 VH NI9 VI4 1545 1545 361 2(00 2600 2503 2603 2603 358 26IO 359 1713 374 2300 854WI99 855 W20O2335 TiMffcWft^cfUVW D DEPTH 250-0 5OO-250 1000 -500 REGION ^ WD 1 JL 13 T ^ ^ W ° ''°'°° _lT 1 T 15 II 19 39 138 35 19 40 115 617 387 20 39 22 59 512 134 116 328 72 167 SOI 445 52 SO 10 97 90 17 142 108 221 98 15 84 16 44 119 48 287 35 45 13 154 45 5 FURC 5 4 3 2 I 3 CAL 2 I META 2 NAUP N5 EXAM- -INED •*■ SAMPLE FROM 1500 - lOOO Fig. 56. Developmental condition of larval swarms in the East Wind drift, showing the corresponding condition of the oldest larval swarms found in the Weddell drift at the same, or approximately the same, time. 68 u) 64 => 60 g 56 " 52 2 48 1 44 40 S 36 ixJ 32 i 28 2 24 20 f 16 i ' = N? MEAS- -URED NOV DECEMBER JANUARY FEBRUARY MARCH APRIL MAY 2 2 28 29 3 7 8 91919 10 10 2021 22 27 27 28 3030 3 7 5 8 17 10 16 18 22 2023 23 27 25 I I 4 8 12 14 17 19 26 25 8 7 I003 1632 I540 2139 RS9 2168 2197 2547 2561 1665 1573 1675 2225 2226 2814 1535 1543 1297 2503 1154 2004 I720 1351 2478 523 S 2513 537 S 305 W540W542 2567 IS07 349 512 616 522 527 25902594 368 24 36 656 207 (3) '^^ o 50 100 f.",'" i WEDDELL DRIFT ^H EAST WIND DRIFT n (0(0 (0(1) (0 (I) (2) (2) (2) (2) Scale per cent (0 (2) (0(0(0 121 I570 1637 256 46 61 174 245 345 205 186 82 85 200 146 ISO 100 75 416 269 527 95 lOO 310 100 57 189 379 325 20 522 lOO 818 lOO 35 517 lOO 100 lOO 388 176 63 84 lOO 231 102 STATION 66 64 60 55 52 48 44 40 36 32 28 24 20 16 12 8 N?MEAS- -URED ^ SAMPLE FROM SIGHTED SWARM Fig- 57- Developmental condition of adolescent swarms in the East Wind drift, showing the corresponding condition of adolescent swarms found in the Weddell drift at the same, or approximately the same, time. Note vertical scale in 4-mm. groups, e.g. 20 = 17-20 mm. zone. During the long period, June to December, when the East Wind drift is virtually closed to navigation, the modal values must be correspondingly low, a single observation in December^ giving some indication of how low they might be. In other words it is clear from Fig. 55 that in the East Wind zone although some of the swarms attain the modal stature of their northern counterparts, a high proportion of them, to all appearances throughout life, never do. It is also clear that the avail- able feeding-stuff for the whales that reach these high latitudes must often consist of considerably smaller animals than they would normally encounter in the northern zone. More striking illustrations of how the East Wind swarms persistently lag behind their northerly contemporaries, in both stature and development, are given in Figs. 56 and 57, the former 1 Obtained from a factory ship which (p. 124) had forced a passage through the ice to a position not normally attainable by a vessel such as 'Discovery II'. 256 DISCOVERY REPORTS showing the remarkably backward developmental condition that persists in the East Wind larval swarms from January to April, with the corresponding condition of the oldest contemporary larval swarms from the Weddell zone added for comparison, the latter how the northern adolescent swarms between November and April, that is on the threshold, and in the early part, of their second year of life, tend persistently to be encountered in a more advanced condition (in so far at least as length frequency may be said to indicate advancement), or at any rate to be composed of larger animals, than their contemporaries in the southern zone. A corresponding disparity in size has been found in Thysanoessa inermis which grows to 27 mm. in the Japanese Sea, but only to 22 mm. in the colder Barents Sea (Ponomareva, 1954). % SEP lOCT 1 NOV 1 DEC i JAN 1 FEB IMAR 1 *f R ImAY 1 JUN 1 JUL 1 AUG 2S- ©©@@(g)@@@@@00 25- ©©@@@@@@®@0e 25- ©©@®®®@@®®00 25- ©©®®@@@@®@00 25- © ®®[3@®@@ ® @ 0 0 25- ©®®@®®@@@®®0 SEP [OCT [nOV 1 DEC | JAN | FEB ImAR | APR [mAY [ JUN | JUL [ AUG Fig. 58. Moulting. Monthly frequency of occurrence of cast skins between the surface and 1000 m., the en- circled figures showing the numbers of observations on which the frequency in each instance was based. SEP lOCT |NOV I DEC I JAN | FEB |MAR | APR |MAY [ JUN | JUL | kijQ "j) (a) i,?g> (15) (47) (74 i g; (\BJ ijQ) [\2) ( 0 ® (g) (@) (^3; 64) ^ (18) ^ @ 0 0 Fig. 59. Moulting. Monthly frequency of occurrence of cast skins between the surface and 1000 m., the three hauls from 250 m. to the surface being treated as one. In the East Wind zone the sun's rays strike obliquely with the result that they are deficient in intensity, much ultra-violet radiation is lost and the total available heat is low. All this, quite apart from its adverse effect on photosynthesis, may contribute to the slow growth-rate we find among the high southern krill, much as it affects the annual growth of many deciduous and evergreen plants in the Arctic (Dansereau, 1955). MOULTING In the vertical samples I examined I frequently noted complete casts of euphausian skins, or fragments of euphausian skins, which from their large size could readily be referred to E. superba. Only casts that could obviously be referred to this species were recorded, very small casts which, though possibly those of young adolescent krill, might also have been confused with those of other and smaller species, MOULTING 257 being disregarded. They occurred, as might be expected in gatherings of such short duration, in very small numbers, singly or in twos or threes, at all depths from the surface down to looo m. The monthly frequency of their occurrence between these levels (Figs. 58-60), based on samples from the East Wind drift, the Weddell drift and the Bransfield Strait, seems to throw a certain light not only on the probable depth at which these animals cast their skins, but also on the period during which the older stages of the surface population are feeding and growing at their maximum intensity. Let us consider first the probable level or levels at which the moulting takes place. At first sight it would appear from Fig. 58 that there is a marked increase in the frequency of the moult with increasing depth suggesting that the krill at a time when they must be % assumed to be most vulnerable to their surface predators might seek the comparative safety of deep water in order to shed their skins. Since, how- so ever, our observations at all levels (pp. 157-70) are directly at variance with the existence of such a movement there must be some other explanation of these apparently contradictory appearances. Obviously it must be that 2s- the skins, being dead things, are sinking, and that in consequence of the variable duration of our vertical hauls they tend to appear with greater DEPTH 25- O 75 50 75 50- 25- 75- 50- 25- O -o- 34 ^" 500- ® frequency in the three deep 250 m. hauls than they do in the shorter ones °- above 250 m. For example, if there were one sinking skin located some- where in the 250-0 m. layer it is obvious that a net fishing in only part of this layer, say from 100 to 50 m., would have less chance of capturing it than Fig. 60. Moulting. Frequency one fishing at the same time through the whole 250-0 m. water column. Thus of occurrence of cast skins if the three uppermost net hauls, 250-100 m., 100-50 m. and 50-0 m., be between the surface and 1 000 m. , . 111-11 /T^. ^ in the Bransfield Strait, represented as one contmuous 250 m. haul, then it would appear (rig. 59) that the frequency of occurrence of casts in the four now equally distributed horizons between the surface and 1000 m. is for all practical purposes the same. And from this it could be argued, the casts being dead and sinking, that the moulting takes place somewhere near the surface, or at any rate above 250 m., where the main concentrations of the whale food have been shown to be. The vertical frequency of occurrence of casts in the Bransfield Strait for the months of November and February combined is shown separately in Fig. 60. From this the same general inferences can be drawn as from Fig. 59, namely, that the sinking skins are more or less equally abundant at all depths covered by our observations and that they are probably shed at the surface. Turning now to the distribution of the casts in time it will be seen that on the whole there are enough observations from September to June to indicate a period of intensive growth lasting from November to March, marked by an abundance of cast skins in the plankton, followed by a period of slackened growth, marked by the absence of cast skins from the plankton, the latter period extending from April to June and probably, if we had observations to show it,^ covering July and August as well. As Bargmann (1945) has noted, the periods November to March and April to June correspond respectively although somewhat roughly with the times of the phytoplankton maximum and post-maximal decrease. It is not surprising, therefore, that the former should be found to correspond exactly with the period of rapid growth she records (p. 251, Fig. 54) between October and March, when this species in its second year practically doubles its average length, and that the latter no less exactly should correspond with the period of arrested development indicated by the flattening out of the growth curves of her second-year euphausians between April and June. It is interesting too to note that the period November to March coincides with the known protracted spawning season of the krill. ^ I did not examine our samples for July and August. 29 DM 258 DISCOVERY REPORTS In the lined shore crab, Pachygrapsiis crassipes Randall, Hiatt (1948) finds pronounced variation in the frequency of the moult with temperature, and well-marked periods of inhibited ecdysis coinciding with the winter months. The myriads of chitinous integuments shed by the moulting krill, not to mention the massive rain of organic debris they must produce in death, no doubt (Wolff, i960) provide an important food source for benthic animals living at abyssal depths. REACTION OF LARVAL, ADOLESCENT AND ADULT KRILL TO SHIP AND NETS In studying the distribution, relative abundance and absolute density of plankton animals, we have to rely mainly on townets which sample them for the most part at random and at more or less widely scattered points. And since, as Hardy (1936a and 1936^) has shown, the plankton as a whole is inclined to be patchy, or very uneven in its distribution, in any large-scale investigation such as this a very large number of samples are required before any reliable estimate of the distribution, relative abundance and absolute density of some selected species can be formed. This is obvious and applies to all plankton work of the kind in which we have been engaged. It applies, however, with particular emphasis to E. superba, an exceedingly patchy species, more patchy perhaps than any other, forming in the surface zone dense swarms of exceedingly active animals, swarms separated from one another by considerable 'voids' in which the krill are very scarce or absent. It is of manifest importance, therefore, that being now about to consider the distribution and relative abundance of these animals as revealed by the random sampling of nets, we should consider in some detail how and under what conditions our nets may be sampling the swarms and how and under what conditions the swarms themselves, or the individuals of them, behave in relation to, or react to the intrusion of, both nets and ship. These questions will be considered under four headings, (i) the powers of the krill to evade as individuals, (2) the capacity of the individuals of a swarm to act in unison, (3) the mechanical effect of scattering by the ship, and (4) the inferences that may be drawn from (i), (2) and (3). Powers of the individuals to evade We have already seen (pp. 154-7) that in daylight, especially in bright sunlight, the krill, both young and old, are exceedingly sensitive to outside stimuli, reacting immediately to sudden disturbances and taking effective measures to avoid them. The first and most obvious thing to consider, therefore, is to what extent and under what conditions the younger and older stages of the surface population may take similar action when confronted with the warp and bridles of an approaching net. That they take such action is certain, the comparative ease with which they have been seen, with a sudden backward kick, to avoid the vastly more rapid approach of a fish for example, suggesting it must be extremely effective. How effective, in daylight at any rate, may be gathered from the following passage from Mackintosh (1934). 'Active avoidance of the net must sometimes take place, especially by such species as Eiiphausia superba. I have been able from the ship's side to watch a net being towed a few feet below the surface and passing through a shoal of this species. The Eiiphausians could clearly be seen to leap backtoards out of the way of the approaching net.'^ Clearly this represents an instance of 'evasion by sight', combined to some extent perhaps with evasion stimulated by touch or vibration. Our observations repeatedly show that the measure of sight evasion achieved can be correlated, ^ The italics are mine. REACTION OF KRILL TO SHIP AND NETS 259 either directly or inversely, (i) with the intensity of the light or altitude of the sun, (2) with the degree of turbulence of the sea, in so far as that may be said to unsight or confuse the surface individuals, and (3) with the size and vigour of the krill themselves. Altitude of the sun The relationship of the sun's altitude to the gatherings of the larger krill we took at the surface in our stramin nets is shown in Table 50 which reveals such a clear-cut inverse correlation between catch and light values, that there can be little doubt that as the sun sinks and the strength of the daylight wanes the krill find it more and more difficult to avoid the net, presumably because they find it more and more difficult to see it. In broad daylight in fact they would seem to see it so clearly that avoid- ance is virtually complete. It could of course be objected that when the sun was high, or at any rate above the horizon, most of the krill had migrated to some deeper level and so could not in any case have appeared in the surface catch, except perhaps in the negligible numbers recorded. A discussion of this possibility is given at the end of this section, where it transpires that both avoidance and some degree of vertical migration are perhaps equally involved. Table 50.* Varying light intensity and its relationship to the catches of the staple whale food in the horizontal {0-5 m.) stramin nets Night-time in high latitudes {sun low on horizon, at Twilight sunset or {sun below Broad daylight sunrise) horizon) Full darkness Total catch 1,123 701 7,698 299.995t Number of hauls 396 22 28 357 Average catch 3 31 275 840 * Note this and Tables 51-6 which follow were drawn up before all the material, upon which, for instance, the vertical distribution is based, was finally examined. f Excluding one enormous catch of over 200,000 at Station 2014. In so far as it affects the reliability of our data, evasion by sight operates principally in high lati- tudes. In the neighbourhood of South Georgia, for instance, and in the Weddell drift generally, there are always, both in summer and winter, ample dark hours to work in. Farther south, however, in the coastal waters of the continental land, which are only accessible in summer or early autumn, much of the work has to be carried out in more or less continuous daylight, the sun, particularly in January and February, never setting or dipping only for short periods below the horizon. Turbulence of the sea The state of the sea, whether it be calm, slightly disturbed or boisterous, seems to have a marked effect on the capacity of the krill to dodge the nets. This phenomenon is shown in Tables 51 and 52 in which the measure of turbulence is set down according to the Douglas Scale as follows : Calm to smooth Force o-i Slight to moderate Force 2-3 Rough to very rough Force 4-5 High to very high Force 6-7 Catches where the real state of the sea was masked or damped down by the presence of pack-ice have been excluded from both tables which refer exclusively therefore to the open ocean. 29-2 26o DISCOVERY REPORTS Table 51. Varying turbulence of the sea and its influence on the catches of the horizontal {0-3 m.) stramin nets. Night-time data in roman type, daytime data in italics State of sea Calm to smooth Slight to moderate Rough to very rough High to very high Total catch 4,022 64 88,874 5(>,59B 88,528 117,169 3" 4 Under 16 mm. • Number of hauls* 33 31 88 log 100 121 II 17 Average catch 122 2 1,010 519 885 968 28 Total catch 161 64 12,968 2,334 4,170 8,194 1 16-20 mm. < Number of hauls 14 25 48 62 30 57 S 8 Average catch II 2 270 37 139 143 — Total catch 13.353 22 45.454 281 65,863 60s 163 Over 20 mm. < Number of hauls 47 45 117 172 117 163 9 20 Average catch 284 388 1 563 4 18 Sum of average catche groups combined s of the three size 417 4 1,668 557 1,114 46 * Note this number throughout is restricted to the periods during which the three broadly grouped developmental phases severally have their optimum range in the plankton. Thus the under 16 mm. group ranges from March to December, the 16-20 mm. group from September to December and the over 20 mm. group all the year round. It is clear from both tables that the surface population, whatever the state of its development/ can be sampled more effectively when the sea is rough or disturbed than when it is flat, calm or smooth. This is particularly evident for the daylight hours but is also, and perhaps unexpectedly, true for the dark hours as well, although in a slightly lesser degree. Apparently, therefore, because their vision is blurred, or perhaps simply because they are confused, or again perhaps because even although they may see the net turbulent eddies carry them into it, the krill, both young and old, and by night and by day, are far less able to avoid the nets when there is surface disturbance than they are in calmer conditions. The extreme poverty of the material taken in high to very high seas is not, however, to be ascribed to any particular action on the part of the krill themselves, but rather to the decreased efiiciency of the net when used under such rough conditions. For these were the conditions which marked, and often indeed went beyond, the extreme limit of endurance of our apparatus, conditions with winds at gale or near gale force when the nets would become split, or get torn bodily from their rings, or get lost, complete with bridles, altogether. In such conditions, even if they survived intact, they would generally be moving about so violently that such krill as they might chance to capture would no doubt largely get lost through wholesale spilling of the catch. Normally we fished the stern nets with the wind on the port beam, the beam wind filling the nets, and, blowing them to leeward, 1 The apparent diminution of the oblique night catches of the 16-20 mm. class with increasing turbulence (Table 52), which in view of the day and night data provided by the horizontal nets (Table 51), and the day data provided by the oblique nets (Table 52), is obviously anomalous, can be attributed to the fact that the night stations at which these particular oblique gatherings were obtained chanced to fall in places where 16-20 mm. euphausians were scarce. REACTION OF KRILL TO SHIP AND NETS 261 Table 52. Varying turbulence of the sea and its itifluence on the catches of the oblique {100-0 m.) stramin nets. Night-time data in roman type, daytime data in italics Calm to Slight to Rough to High to State of the sea smooth moderate very rough very high Total catch 16,016 150,616 113.239 — 6,445 75,589 71,781 888 Under 16 mm. ■ Number of hauls* 85 197 192 15 72 22"] 202 23 Average catch 188 764 589 — 89 333 355 38 Total catch S.632 5.029 1,796 3 1,516 44,723 115,713 — 16-20 mm. ■ Number of hauls 42 106 77 7 44 133 118 10 Average catch 134 47 23 — 34 336 980 — Total catch 13.197 51.295 221,580 105 3,842 85,994 192,573 56 Over 20 mm. ■ Number of hauls 124 309 265 24 160 473 339 36 Average catch 106 166 836 4 24 181 568 1 Sum of average catches of the three size 428 977 1,448 4 groups combined 147 850 1,903 59 * See note below Table 51 . keeping them from fouling the warp as they entered the water. In heavy weather however, to safe- guard life, this practice was abandoned and we towed head on to wind and sea. In other words we were hove to, the vessel plunging violently as the stern rose and fell, 20, 30 feet or more, in the cfests and troughs of the waves. The sudden downward movement of the hull would give the nets a sharp backward thrust and it was probably this more than anything else that led to the spilling of the catch. An interesting general inference that may be drawn from Table 52 is that with every developmental phase of the population it portrays it provides substantial support for our evidence that the main concentrations of these animals are never very far away from the surface. For if it were not so and such concentrations had a much wider bathymetric range, extending downwards into the tranquil water below the relatively narrow zone of surface turbulence, then the influence of the state of the sea on the catches of the upper oblique nets would not, either by day or by night, be nearly so pronounced as it transpires to be. Activity and vigour of the krill themselves As might be expected, the measure of evasion achieved is greater in the older and more vigorous animals than in the younger and weaker, presumably because the latter, although they may be able to see just as well as the adults, are unable to dodge away so rapidly or so far. The mounting capacity to evade developed by the krill as they grow is illustrated in Table 53 in which the three broadly grouped developmental phases portrayed may be described in general terms as follows: the under 16 mm. group, vast majority early and late surface larvae with some very early adolescents; the 16-20 mm. group, vast majority early adolescents; the over 20 mm. group, older adolescents and 262 DISCOVERY REPORTS adults. It is clear from this presentation of the data that while the older krill in the over 20 mm. class practically elude the surface net altogether during the day, the younger and not so active animals do not by any means so readily escape capture. In fact the surface larvae and very early adolescents in the under 16 mm. class can it seems be taken in more or less equal abundance whatever the condi- tion of the light, although the early adolescents in the 16-20 mm. class already clearly seem to have developed a distinct capacity to evade. The turbulence data given in Tables 51 and 52, notably the daytime data, also reveal a tendency for the older krill to be the more successful escapers, those in the over 20 mm. class tending to elude capture, in calm to moderate seas, in greater numbers than the smaller animals are evidently able to do. Table 53 . Mounting capacity to evade developed by the krill as they grow, from data based on the catches of the horizontal {0-5 m.) stramin nets. Night-time data in roman type, daytime data in italics Sum oj average catches of the Size group (mm.) Total catch Number hauls of Average catch three size groups combined Under 16 193.347* 176,443 243 286 796 617 16-20 17.311 11,616 94 156 184 74 1,820 695 Over 20 299.99s 1,824^ 357 418 840 4 * Excluding one enormous catch of over 600,000 at Station 2300. ■j- Including total catch and net hauls from second column of Table 50. Table 54. Varying light intensity and its relationship to the catches of the largest (over 30 mm.) krill in the horizontal {0-5 m.) stramin nets Night-time in high latitudes {sun low on horizon, at Ttmlight Broad sunset or (sun below Full daylight sunrise) horizon) darkness Total catch 296 300 2.173 89.795 Number of hauls 396 22 28 297 Average catch 4S 200- 00 ° 3 0 0 EVENING AND DARKNESS 4-HOUR PERIOD IBOO- 559 64 6 3 — 12,864 2,770 57 10 3 — 11,177 1,727 57 3 3 — 10,666 6,071 56 4 3 — 10,124 716 53 7 3 — 5,464 1,338 45 — 3 — 4,497 522 35 6 3 — 4,032 180 32 10 3 2 4,018 482 31 6 3 I 2,871 75 28 I 3 — 2,841 375 25 — 3 — 2,760 825 24 — 3 — 2,432 90 18 6 2 2 1. 975 960 18 — 2 — 1,660 562 17 2 2 1 964 562 15 3 2 — 616 477 12 — 2 1 592 112 II — 2 2 480 6 9 — 2 I 454 16 9 I 2 I 409 45 9 — 2 — 342 3 7 2 2 — 26g — 7 — 2 — 211 8 7 — — 200 112 7 — — 198 12 7 — — 182 18 6 I — 180 30 6 2 — 177 4 6 — ■ — 139 6 5 — — 136 I 4 — — 132 I 4 — • — 109 24 4 1 — 105 5 4 3 — 77 6 3 — ■ — 1,864 26,912 S8 140 1 9 451 20,652 119 136 — 7 1,216 13-674 27 58 — 6 250 4,624 24 43 3 4 254 1.392 8 15 — 2 990 1,086 1 15 — 1 27 354 6 13 — 1 3 322 — 10 — 1 59 240 4 9 — 1 28o DISCOVERY REPORTS Table 58. 16-20 mm. class. The 111 simultaneous samplings by the surface {0-5 m.) and oblique {100-0 m.) stramin nets, the 102 instances zvhere the surface catch was the larger, or not the less, in the upper part of the table, the nine instances where the oblique catch was the larger below. Night hauls in roman type, day hauls in italics NiooH NiooB NiooH NiooB NiooH N looB 0-5 m. 100-0 m. 0-5 m. 100-0 m. 0-5 m. 100-0 m. 7,854 555 24 3 3 — 2,046 1,154 — — — — 1,224 1,041 24 I 3 — 1,128 332 22 — 3 — 974 534 19 — 2 — 78s 33 16 — 2 — 420 24 13 1 2 — 381 I 13 — 2 — 288 6 13 S 2 1 280 3 12 1 2 — 253 1 11 1 2 — 220 40 10 — 2 — 188 12 9 — 2 — 160 I 9 — 2 — 138 99 8 1 I 107 I 7 — I 72 13 7 — — 65 3 6 — — 63 — 5 — — 61 13 5 — 1 60 25 5 — — 60 24 5 3 — 57 — 5 — — 56 — 4 I — 53 7 4 — — 45 1 4 — — 42 6 4 — 1 40 — 4 — ■ 1 38 II 4 — — 35 I 3 — — 35 — 3 — — 33 — 3 1 — 29 1 3 — I 29 — 3 — — 25 4 3 — — — 30 70 — 9 — 3 1 34 4 6 1 3 — 10 — 6 — I That our day and night surface gatherings of larvae and early adolescents and our night surface gatherings of older adolescents and adults should so often be larger than our oblique gatherings is only to be expected, since with a population such as this, the greater part concentrated at or near the surface in a system of swarms, a net towed on the surface for half an hour is manifestly far more likely to strike a swarm than one that is hauled, as the oblique net is, only for a few minutes through the surface zone. Indeed, it is abundantly clear from Tables 57-9 that the oblique net must often break surface in the ' voids ' or places of euphausian scarcity that separate the swarms. It is clear then from these tables that the more reliable estimate of the relative abundance of this species in the plankton is to be obtained from the data provided by the surface nets and that the distributional data as a whole will acquire increased validity if some correction be applied to the CORRECTIONS APPLIED TO THE CATCH-FIGURES Table 59. Over 20 mm. class. The 208 simultaneous night samplings by the surface {0-5 m.) and oblique {ioo~o m.) stramin nets, the 168 instances where the surface catch was the larger, or not the less, in the upper part of the table, the 40 instances when the oblique catch was the larger below NiooH 0-5 m. 11,500 10,600 10,248 7.440 7,212 5,888 5.538 4.704 3.176 2,992 2,278 1,420 987 717 621 sss 502 434 420 370 367 363 303 283 278 260 259 212 207 203 197 168 168 146 135 124 118 114 III III 107 102 3 12 16 136 32 58 4 48 III 18 N looB 100-0 m. SO 2,412 408 484 2,272 10 2,664 475 SI 43 35 48 27 8 49 477 61 10 286 io8 308 277 87 14 IS 30 95 13 36 3 25 40 13 12 7 I 3,680 2,892 272 268 162 154 122 105 103 88 NiooH 0-5 m. 72 70 69 65 58 57 56 55 52 45 44 43 41 40 40 40 39 38 36 36 33 33 32 30 29 28 27 25 24 24 24 23 22 20 18 18 18 16 16 16 16 15 44 13 10 8 I 3 NiooB 100-0 m. 7 34 2 9 9 6 I I 6 23 5 22 10 I 3 19 10 4 18 4 3 16 3 5 7 I I 3 4 9 II I 6 I 10 79 69 50 31 25 13 13 12 II 10 NiooH 0-5 m. 15 15 15 14 14 13 13 13 12 12 12 II II 10 10 10 10 9 8 8 8 8 8 7 7 7 7 6 6 6 6 6 5 5 5 5 5 5 5 4 NiooB 100-0 m. II I 4 6 I 4 3 I 5 4 4 4 3 I I I 4 10 10 10 9 6 6 4 4 4 4 NiooH 0-5 m. 4 4 4 3 3 3 3 3 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2 2 NiooB 100-0 m. 32 282 DISCOVERY REPORTS catch-figures of the oblique nets, in the many instances where such catch-figures alone are available, in order to bring them more into line with the heavier gatherings from the surface nets. It will be seen from Table 60 that the ratios of the total horizontal (0-5 m.) to the total oblique (loo-o m.) catches of the under 16 mm., 16-20 mm. and over 20 mm. classes of the surface population are roughly 2, 5 and 4 respectively, and, accordingly, in the absence of any more satisfactory method (see below), the multiples 2 for the under 16 mm. class and 4 for both the 16-20 mm. and over 20 mm. classes have been appHed to raise the oblique (lOO-o m.) catches of these classes (in all instances where such catches alone are available) to the orders of abundance they would evidently severally have been found in had they been obtained in the surface net, or, as in the case of the adults alone, the night surface net. The corrections principally affect the catch-figures of our very large collection of oblique (lOO-o m.) day samples of the larger (over 20 mm.) krill, these animals, as already noted, being virtually impossible to sample with the surface net except when used in the dark. In the result, it will be seen, although not mathematically exact, in so far as they may be said to operate to the advantage of the distributional data, they appear on the whole to be reasonable and to operate tolerably well. Table 60. Total catches of the horizontal {o-^ m.) and oblique {100-0 m.) stramin nets when fished simultaneously Size group in mm. Total horizontal catch Total oblique catch Number of simultaneous samplings Under 16 16-20 Over 20* 157.515 15.672 84,411 98,895 2,950 19.585 138 III 2o8 * Night data only In plotting the distribution and relative abundance of the surface population on our circumpolar charts I have used for the stramin net samples a conventional system of circles of increasing diameter representing catches grouped in the following orders of magnitude, i-ioo, loo-iooo, 1000-10,000 and over 10,000. Referring again to Tables 57-9 and assuming all negative catches to have been in fact very small ones, of the order of say i-io, it will be found that in 331 cases out of the grand total of 381 simultaneous samplings where the surface catch was the larger, if the oblique catch be multiplied by 2 or 4 as required, the figure we arrive at in each instance falls within the order of magnitude represented by the horizontal catch or does not deviate materially from it. The corrections applied do not, of course, always achieve this result. For example, in the sixth sampling of the staple population shown in the first column of Table 59 we have a clear instance where the surface net must have struck a swarm and sampled it with considerable eflFect, whereas the oblique net obviously missed it, or virtually missed it, altogether, producing so small a catch that it makes no difference whether the cor- rection be applied or not. However, they work so frequently in the right direction, bringing the oblique catch-figures so often into line with the orders of magnitude represented by the surface samples, that they can manifestly be used with advantage in the construction of the distributional charts. Even in the relatively few instances where they do not operate to full advantage, as, for instance, where the surface catch was 11,500 and the oblique only 50 (Table 59), it is obviously better, assuming there to have been no surface net at this station, that the oblique catch should be represented by a minor (lOO-iooo) order of magnitude rather than by one of insignificant size. Again, taking another example from Table 59, where the horizontal catch was 10,248 and the oblique 408, it is far better, again assuming absence of surface observation, that the latter should be raised to the second highest order of magnitude rather than left as it stands. In the many instances where the oblique net failed to take CORRECTIONS APPLIED TO THE CATCH-FIGURES 283 a single euphausian the correction of course cannot be applied. However, this does not matter a great deal because as our tables show this always happened when the surface population was itself evidently very scarce or represented at most by samples of minor abundance, by gatherings in fact rarely exceeding 100 and never more than 500. Although these are rough calculations, any more accurate estimate of the factor, or factors, required to bring the oblique into line with the surface catch is precluded by the wide scatter of the oblique and horizontal catch-figures. An estimate (for the over 20 mm. class) based on the method of least squares leads to a value of 3, but the error involved is so large that there is really nothing to choose between this figure and the simple ratio of 4. These standard corrections have not been applied in cases where the oblique catch was already in excess of 10,000,1 nor in such instances where simultaneous sampling revealed it to be larger than the surface, all such instances being accepted as representing the true relative order of abundance of the surface population wherever they chanced to be recorded. Nor has any correction been made for the state of the sea. This is hardly necessary since, except in the shelter of the pack, these southern waters, as Tables 51 and 52 show, are generally in the state of agitation in which the krill are most easy to capture. The great majority of our samples were in fact obtained in such conditions, and with a mass of data so large the relatively few occasions when the nets were fished in calm or smooth water, to the manifest disadvantage it is true of the catch, are not likely to affect the overall picture of the distribu- tion to any material extent. Some allowance, however, has been made for the persisting daylight in which a number of our night surface hauls were made in high latitudes during the Antarctic summer, the catch-figures for all such nets being multiplied by 3, since the average catch obtained when the sun was a little below the horizon (Table 50) was approximately three times smaller than the average catch in full darkness. This obviously seems too small a correction for the surface catches obtained when the sun was low on the horizon, or at sunset or sunrise, the average catch in these conditions being 27 times smaller than the average catch of the dark hours. It so happened, however, that at each of the 22 stations worked in this failing state of the light the oblique catch was so correspondingly small that the application of any larger factor than 3 would not I think in these instances have been justified. Finally, where necessary, corrections have been made for time, all towings being resolved into standard hauls of ^ hr. duration (see p. 59), and also for net aperture (see p. 157). Where relevant the above corrections have been used in the construction of all distributional charts based on the data from the stramin nets, and, as explained on p. 59, in certain tables and figures which appear on earlier pages. Although arbitrary they are obviously necessary and can be justified on the ground that some such correction, even if uncertain, is better than blind acceptance of catch- 1 Although this is statistically inconsistent, it is in fact the only reasonable thing to do. The highest order of abundance in which we normally sampled the staple population (krill over 20 mm. long), with either the oblique or surface net, was 10,000-100,000. Very rarely, in fact once in the oblique and once in the surface net, this order was exceeded and gatherings of 100,000-1,000,000 were obtained. Only seven, however, of the 1700 odd oblique nets we shot produced catches of 10,000- 100,000, all from the richly populated East Wind-Weddell surface stream. These seven gatherings, therefore, must represent it seems instances where the surface net would not normally have done any better, and by leaving them as they stand all I have done has been to keep the highest order of abundance our stern nets can reveal within its normal limits. As I see it, biologically, there could only have been one strong reason for multiplying these seven oblique gatherings by 4, and that would have arisen if the surface net had consistently been capturing 100,000-1,000,000 euphausians of the over 20 mm. class. In plotting the distribution of plankton animals we cannot use mathematics on our charts. All we can do with our conven- tional systems of circles or squares of increasing size is to show broad differences, and the whole idea is to show these dif- ferences as clearly as possible, choosing a biggest circle that is not too large to obscure the rest of the data and yet not so small that the differences are not well defined. Sometimes it is true my largest circles do obscure a number of negative stations, but this is an unavoidable imperfection associated principally with the closely worked South Georgia whaling grounds, a relatively insignificant part of the vast area we covered, and even there it does not matter very much since the facts of the distribution in these island waters are shown later on separate seasonal charts drawn to a much larger scale. 32-2 284 DISCOVERY REPORTS figures that in most instances are manifestly far from representative of the natural abundance of the krill in the sea. As Hardy (19586) has said, ' The use of nets in the sea is really an experiment, an experiment to see if the net, by being used in a certain way, and at certain intervals of time and space, can fix an adequate series of samples which will represent the actual distribution of the animals concerned in nature '. It could be argued that my factor for the oblique-horizontal ratio is too low, that since the hori- zontal net spends ^ hr. at the surface, as against only about 2 min. by the oblique, it should be 1 5 instead of 4. This could well be, but only if (i) the krill were scattered broadcast evenly throughout the surface zone, and not, as they are, in rather widely separated swarms, and if (2) the surface net during its ^ hr. in the water struck a narrow swarm a mile or so long and fished all along, not across, it. That this should happen, however, is extremely unlikely, the swarms that have been seen as a rule being of vastly smaller dimensions. Moreover, if 15 is indeed the ratio we should expect a figure of this order to show up when the average catch-figures for an enormous number of oblique and surface net hauls were compared. We find, however, looking at the figures for the staple whale food, that the average catch of 732 night oblique hauls is 391, of 1013 day oblique hauls 278 (Table 56) and of 357 night surface hauls 840 (Table 50). This does not suggest that the ratio need be very high or indeed that the surface net is all that vastly more efficient than the oblique. Even if the enormous catch of over 200,000 (Table 50) be included in the night surface total, the average, 1397, still works out at only three and a half times the average oblique catch by night and only five times the average oblique catch by day. But perhaps the overriding consideration is this. Whether the factor (or factors) used to express the oblique in terms of the surface catch be high or low, in the end it only serves to heighten the overwhelming importance of the East Wind-Weddell surface stream as a carrier of the larval, adolescent and adult krill. I found this repeatedly during my early attempts to portray the distribution, when, before all the data finally came to hand, I was using a factor of 8. The same overall picture, massive East Wind-Weddell abundance and scarcity in the West Wind zone, is obtained without using any factor at all, as I again found when I originally began plotting our data using the oblique catches alone, and as they stood. A striking example of this is provided by Fig. 157 (p. 425) which shows the gross distribution and relative abundance of the krill based exclusively on oblique (loo-o m.) gatherings. Above all, turning to the situation in the East Wind zone, the correc- tions applied to the catch-figures there have their own especial value in so far as, I feel sure, they help greatly to offset the heavy disadvantage (p. 59) under which so many of our samples were collected in these high, virtually nightless, summer latitudes. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL Introduction and plan of presentation In the following section the horizontal distribution of the krill, already broadly outlined on pp. 57-64 (Figs. 5 a, 56 and 5 c) and the factors controlling it are described in detail, the matter presented covering every phase of the developmental history from the egg to the adult state. The distribution is illustrated on two sets of circumpolar charts, one based on the data from the vertical nets, the other on the data from the horizontal (0-5 m.) and oblique (loo-o m.) stramin nets, the gatherings of the latter with the standard corrections applied where necessary as described on pp. 282-3. The vertical net series deals exclusively with the eggs and both deep and shallow living larval forms, the stramin net series with the surface larvae, the young, very small and occasionally eaten adolescents (the small whale food) and Avith the older adolescent and adult krill (the staple whale food) which constitute the bulk of the diet of the baleen whales. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 285 Numerous although our observations in the south have been, the major problem of periodic coverage of so vast an area as the circumpolar sea could only it was found be resolved by an arbitrary grouping of months regardless of the years in which they fell, it being manifestly impossible for a single vessel, or even group of vessels, to have been everywhere in so large an area in every month of a single year. Thus the distribution of the young stages, based on the data from the vertical nets, is presented by two-monthly (in some instances three-monthly) periods, and the distribution of the young, adolescent and adult stages, based on the data from the horizontal and oblique nets, by three- monthly (seasonal) periods throughout the year. Attempts were in fact made at first to produce monthly distributional charts, but the majority broke down owing to the size of the area under investigation. It will be seen, however, that in the result the grouping I have used has provided a measure of periodic coverage which at all material times permits certain broad inferences to be drawn, both as to the facts of the distribution and its dynamics, inferences it seems, arbitrary though the presentation of the data has been, that have substantial claim to reality. In marshalling the data from the vertical nets it appeared best, after careful consideration and much experiment, to deal with the larval stages in groups rather than to produce separate bi-monthly or tri-monthly charts for each individual larval stage. The eggs alone have been given charts to them- selves. Thus the grouping of the eleven^ larval stages has been done as follows: group (i), the deep living Second Nauplii, Metanauplii and First Calyptopes that in oceanic water^ occur between 1000 (or 1500) and 250 m., group (2), the shallow living First, Second and Third Calyptopes, group (3), the shallow living First, Second and Third Furcilias and group (4), the shallow living Fourth, Fifth and Sixth Furcilias, groups (2), (3) and (4) being confined for all practical purposes to depths between 250 m. and the surface. Group (i) is a natural grouping in as much as it brings together the total larval community rising from deep down to populate the surface waters above. Moreover, it is a group confined exclusively to a single water mass, the warm deep current, and in consequence every member of it can be assumed to be in a current with a southerly component except (p. 123) below perhaps the greater part of the Weddell stream. Groups (2), (3) and (4) are also natural groups for they too spend virtually their whole life in a single water mass, the Antarctic surface layer, where again the horizontal movement is constant, in the East Wind drift to the west, in the Weddell drift to the north and east. Furthermore, it so happens that in nature (pp. 225-8) groups (2), (3) and (4) do not as a rule occur mixed indiscriminately in a single haul. The extent to which they tend to be segregated or mingled can be judged by inspection of Figs. 34-7 and others on later pages which show the developmental condition of the population in different seasons and places. While there may be occasional intermingling it will be seen between two groups, one, however, always preponderating over the other, in most instances where our nets have encountered them they have encountered them singly, one stage or another in any one of the three as a rule being strongly dominant. And so for this reason too groups (2), (3) and (4) may be said to be natural ones. The enormous accumulation of data gathered is presented on the following plan. Taking first the material from the vertical nets and beginning with the eggs, the charts show in turn the distribution and relative abundance of each larval stage,^ or group of larval stages, from the time when it first appears in the plankton until the end of its natural span.* Thus, by two-monthly steps from November 1 Including the First Nauplius there are in fact twelve. The earliest larval stage, however (p. 98), is so poorly represented in our samples that its distribution, although it can be surmised, cannot be plotted. ^ But not (p. 205) in shelf water. ' For convenience the egg is regarded here as a larval stage, though strictly it is not. * In so far, however, only as that end can definitely be ascertained in regions accessible to vessels throughout the year. It will be shown later that in the East Wind drift, which is closed to shipping for the greater part of the year, the larvae in all phases of their surface development persist longer in the plankton than they do in lower ice-free latitudes. 286 DISCOVERY REPORTS Up to the end of June, and by three-monthly steps thereafter, the individual months, however, being distinguished by appropriate symbols, they show, first the distribution of the eggs from November to April, followed by that of the deep living forms, the Second Nauplii, Metanauplii and First Calyptopes, for the same period. Then come the charts showing the distribution of the grouped surface dwellers, first that of the three Calyptopes from November to June, then that of the early Furcilias from January to June and finally that of the late Furcilias from March to December. The plan in short constitutes an attempt to survey in as orderly a manner as possible the whole period of the larval existence in the plankton, beginning with the time when the eggs first appear and ending with that, more than a year later, when, in the Weddell drift at least,^ our townets have shown, the last few Sixth Furcilias^ surviving from the protracted spawning season moult and become adolescent. The four bi-monthly sets of charts portraying in turn the distribution of (a) the eggs, (b) the deep- sited Nauplii, Metanauplii and First Calyptopes, (c) the shallow living First to Third Calyptopes, and (d) the shallow living First to Third Furcilias, are severally followed by composite end-charts showing, by monthly symbols, the gross distribution of these stages, or more correctly groups of stages, throughout their protracted life-span in the plankton. A similar end-chart concludes the bi-monthly and tri-monthly series presenting (e) the distribution of the late Furcilias. The main purpose of these charts, it will be seen, is to show at a glance and on a single page some rather striking features of the larval distribution and its dynamics, which, although already apparent, are not so obviously apparent, in the serial presentation of the data that precedes them. The detailed presentation of the data just described is followed by a summarised version presented on two groups of charts, the first showing the bi-monthly distribution and relative abundance of the total eggs and larvae (the latter combining both deep and shallow living forms) from November to April, the second the bi-monthly distribution and relative abundance of the total shallow living larvae from January to June. In regard to the second group, it will be seen, charts of the distribu- tion and relative abundance of the total surface population will in certain cases already have been presented in the detailed series, the total surface larvae for November-December, for instance, when the eggs first appear, being represented by the Calyptopes alone and the total surface larvae during the following winter and spring (July to December) represented exclusively by the late Furcilias. The distribution of the larvae based on the vertical data ends with a single chart showing the gross or absolute distribution of the total egg and larval population all the year round, wherever and when- ever, that is to say, it has been encountered in the vertical nets. In other words it ends with a picture which constitutes the record of the horizontal distribution and relative abundance of the eggs and larvae, the latter regardless of stage, that has been revealed by the whole series of vertical net samples examined since these investigations first began. The charts based on the gatherings of the towed stramin nets follow each other on much the same lines except that, as already noted, in marshalling the data the year has been divided into three- monthly, or seasonal, periods throughout. They deal first with the distribution and relative abundance • Actually in the far colder conditions of the highest East Wind latitudes we have explored our stramin nets show (p. 372, Fig. 120) that among the slow-growing, late-spawning population there some Sixth Furcilias do persist without moulting over December, surviving in this state throughout January and February, and even into early March of the following year. ^ The recording of 'very large numbers' (p. 138) of Sixth Furcilias from the stomachs of blue and fin whales examined in the late spring of the pelagic season of 1939-40 suggests, however, that in certain, possibly exceptional years, the last larval stage may survive in Weddell East in some considerable measure of abundance until the second half of December. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 287 of the surface larvae en masse, beginning with the period January to March, or summer,^ when the young swarms first begin to be taken in quantity in the stramin nets, and ending with the same period 15 months later when the last few Sixth Furcilias surviving in the East Wind zone finally disappear from the plankton. Next come the charts showing the distribution of the young adolescents, the very small whale food between 1 1 and 20 mm. long, a group for which the most convenient starting-off point is winter (July, August, September) since it is then, notably in July, and again it must be emphasised probably in the Weddell drift alone, ^ that the Sixth Furcilias that spring from the earliest hatching, or hatchings, first begin to moult and become adolescent. From winter then the young adolescent distribution is traced through spring and summer until autumn (April, May, June) a year later, when the small whale food seems practically everywhere to have outgrown itself and is no longer to be found on the feeding-grounds. The charts showing the seasonal distribution and relative abundance of the staple (over 20 mm.) whale food complete this series, beginning with spring (October, November, December), when the whales begin to arrive in quantity on the feeding-grounds, and ending with winter (July, August, September) a year later, when the whaling season is long past and the whales themselves have com- pleted their northward migration. These final charts, being of greater concern to the industry than the others, were to have been accompanied by the repetition, on a reduced scale, on the pages opposite those on which they severally appear, of the corresponding charts for the massed larvae and young adolescents, thus providing at a glance an overall view of how the larval, small and staple sizes, in other words the total euphausian surface population, are severally distributed and rival one another in dominance from season to season throughout the year. Unfortunately this plan was too late for publication and the reader is referred back to the originals instead. As in the presentation of the vertical data composite end-charts showing as before, but here by seasonal symbols, the gross distribution of the size groups they follow, severally complete the three sets of charts based on the gatherings from the stramin nets. This series, like the vertical series, ends with a single chart based on the analysis of every sample examined from the surface waters since these investigations first began. It shows the gross or absolute horizontal distribution and relative abundance all the year round of the total euphausian surface population, presenting the catches of the stramin nets in their entirety, the gatherings of larvae, adolescents and adults at each station being combined into one, frequently enormous, whole. The order of presentation of the distributional charts, with the pages on which they appear, is as tabulated on p. 288. The distributional charts throughout are provided with monthly, or where necessary, seasonal, circumpolar coverage diagrams showing the number of vertical or towing stations made every 30° of meridian south of 50° S. This, it will be seen, is a critical parallel, virtually enclosing the whole horizontal range of the krill. It is, however, an arbitrary limit, sometimes well outside the northern ^ The seasons into which I have split up the year are not the true astronomical seasons of the southern hemisphere. The true astronomical southern summer, for example, is December, January and February, not January, February and March. Yet the latter three, although collectively not the true southern summer, may for all practical purposes be regarded as such in these high latitudes. More appropriately perhaps, they might be said to represent the true geographical summer since it is then that the pack-ice is at its farthest south and an all round coverage of the circumpolar sea is permitted to vessels, and to whales, up to the highest latitudes. Since, moreover, the whole of the modern pelagic whaling season now takes place during January, February and March the grouping of these three would seem from every point of view to be a more natural and practical one than that of December, January and February. 2 Not necessarily, however, in the higher latitudes of the East Wind drift. Judging from the persistence there of a sub- stantial very young adolescent population (p. 397, Fig. 136) until well into March, a population that clearly must have been derived from a spawning that took place a year or more before, it seems highly probable that even if the Sixth Furcilia does exist in these high latitudes in July, it does not moult on a major scale until some considerable time afterwards. 288 DISCOVERY REPORTS Charts based on the vertical data {a) The eggs November-December (p. 291) January-February (p. 293) March-April (p. 295) Gross distribution (p. 297) (b) The deep Hving Nauplii, Metanauplii and First Calyptopes, or total deep living larvae November-December (p. 299) January-February (p. 301) March-April (p. 303) Gross distribution (p. 307) The shallow living First to Third Calyptopes November-December (p. 309) January-February (p. 310) March- April (p. 311) May-June (p. 317) Gross distribution (p. 322) {d) The shallow^ living First to Third Furcilias January-February (p. 324) March- April (p. 325) May-June (p. 328) Gross distribution (p. 329) {c) {e) The shallow living Fourth to Sixth Furcilias March-April (p. 332) May-June (p. 333) July-September (p. 336) October-December (p. 337) Gross distribution (p. 341) (/) Total eggs and larvae November-December (p. 343) January-February (p. 344) March-April (p. 345) {g) Total surface or shallow living larvae November-December — as in (c) January-February (p. 346) March-April (p. 347) May-June* (p. 348) July-September* — as in (c) October-December* — as in (e) (Ji) Gross distribution of total eggs and larvae all year round (p. 350) *The total surface larvae for these periods being also the total larvae Charts based on the oblique and horizontal data (a) The massed surface larvae January-March (p. 359) April-June (p. 363) July-September (p. 366) October-December (p. 369) January-March (p. 372) Gross distribution (p. 373) {b) The small whale food (11-20 mm.) July-September (p. 375) October-December (p. 377) January-March (p. 378) April-June (p. 380) Gross distribution (p. 382) (c) The staple whale food (over 20 mm.) October-December (p. 389) January-March (p. 394) April-June (p. 399) July-September (p. 404) Gross distribution (p. 407) {d) Gross distribution of total euphausian surface population all year round (p. 410) limit of the krill so that a certain number of stations are counted into the ' coverage ' figures which are not strictly within its habitat. Ideally, only those stations within the total habitat should be included, but since there is no clear, and above all no concentric, northern boundary to the distribution of the species it is hard to say which should be included and which should not. However, in its absolute geographical range (p. 60, Fig. 5 a) the krill does extend northwards to the 50th parallel (and even beyond) in the Atlantic sector and coverage south of this latitude does show that it is there, where the hydrology is dominated by the Weddell stream, and there alone, that it ever extends thus far. In other words the coverage figures I present demonstrate conclusively that it is the strength and movement of the great Antarctic surface streams that in one sector or another determine the horizontal range of this species, and produce its essential eccentricity,^ rather than temperature or some other environmental * parameter '. Above all, in so far as they apply strictly to the West Wind drift, they show repeatedly See (Fig. 56) the even more striking eccentricity of the circumpolarity of its major concentrations. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 289 that the scarcity of the krill we record in this great circumpolar belt is in fact a very real phenomenon and not merely an appearance that could spring from inadequate sampling. I have not included in my circumpolar coverage all the closely spaced stations we made round South Georgia^ and in the Bransfield Strait (p. 51, Figs. 2 and 3), because this would have given a very exaggerated picture of the density of our sampling between 60° and 30° W as compared for instance with that of Weddell Middle and Weddell East, and of course of other sectors of the circumpolar sea. In both heavily sampled regions I have limited the coverage figures to one station per month, thereby bringing the overall oceanic coverage between 60° and 30° W more into line with that of other sectors. In the result it will be seen the coverage figures, although based on a somewhat arbitrary northern limit, will help the reader to decide at a glance to what extent the observed abundance, presence or absence of the various stages of the whale food, from place to place, from month to month or from season to season, could be said to be real or to what extent it might be questioned on the ground that whereas in some areas sampling was heavy in others it was negligible or absent. To avoid interference with the mass of other plotted detail, much of which is often very crowded, the distributional charts are without annotation throughout. For all place-names, hydrological boundaries and other physical features mentioned in the text, however, the reader is referred to Fig. 4 on p. 58. It is important to note that, throughout the distributional charts, the monthly positions of the pack- ice are mean positions so that except where instanced to the contrary stations which appear to lie well inside it need not necessarily have done so. The vast majority of such stations would in fact have been located at the ice-edge, if not outside it, at some greater or less distance away. Although the vast bulk of the data comes from Discovery Investigations sources, in the construction of some of the horizontal and vertical series, but principally the latter, I have used additional data based on material collected by the 'Norvegia' expeditions of 1927-8, 1929-30 and 1 930-1, by the floating factory ' Vikingen ' in 1 929-30 and by the Australasian (B . A.N.Z. A.R.E.) expedition in 1 929-3 1 . The Norwegian material was worked up by Rustad (1930 and 1934) and Ruud (1932), the results I am glad to acknowledge, more especially the negative results, helping considerably to confirm and extend my own findings on certain important aspects of the distributional problem. The Australian material, which comes principally from the i-m. diameter stramin nets, was examined by Dr Keith Sheard of the Commonwealth Scientific and Industrial Research Organisation whose records of the occur- rence of the over 20 mm. or staple whale food class were kindly communicated to me. Again, as with the Norwegian data, it is the negative rather than the positive results of this important expedition that prove to be the most interesting. The eggs and larvae from the vertical samples The eggs Taking the egg charts as a whole it cannot perhaps be too strongly emphasised, even at the cost of repetition, that except for a single instance in December (Fig. 64) and for another two in February (Fig. 65) our gatherings, contrary to all expectation in view of the vast scale (p. 211) on which the krill must multiply, have throughout the spawning season repeatedly been exceedingly small, and for this reason it must be acknowledged that in so far at least as the eggs are concerned the distributional charts are not altogether satisfactory. They cannot, for instance, be said to carry the conviction of those portraying the distribution of the larvae, which on the whole are based throughout on very substantial samples. Clearly, except for the three instances mentioned above, the mass of the eggs 1 For the local distribution of the krill on the South Georgia whaling grounds and the density of the seasonal sampling there see pp. 417-20, Figs. 148-55. 33 "« 290 DISCOVERY REPORTS which must exist has never been properly sampled, a fact which, there is good reason to believe (p. 1 82), may be ascribed (i) to our persistent, and understandable, disinclination to hazard our apparatus too close to the bottom in Antarctic coastal waters, or (2) to the possibility, somewhat remote (p. 189) though it appears to be, that the eggs, being sometimes laid deep down in oceanic water, are always in such circumstances beyond the reach even of our deepest nets. Even if spawning near the surface were followed by rapid sinking of the eggs to such deep levels, it is hard to see how they could so rarely be sampled if there were widespread oceanic spawning. Unsatisfactory, however, though they are, they do it seems provide some indication that spawning on a major scale in or over deep oceanic water far away from the continental land is not a very common event. For out of 858 vertical stations^ made over deep water during the spawning season, the vast majority of them extending down to the 1000 m. level, but including 64 down to 1500 m. or more, only two can be said (Fig. 65) to give any indication that such a spawning could have taken place. November-December. The first small, though perhaps significant, capture of eggs, as Fig. 64 shows, is recorded from the Bransfield Strait in November. In the same month small to negligible numbers, that is, less than 10 per station, were recorded at six other stations in this shelf region. Apart from these occurrences there are no other November records throughout the circumpolar sea. In December, at the extreme western end of the Weddell drift, where its course passes across the eastern opening of the Bransfield Strait, we struck the second largest concentration of eggs, 3030 in all, that has so far been recorded. The vast majority were obtained in a 500-250 m. net haul at a station where the depth was only 510 m., an occasion when the net had probably begun fishing closer to the bottom, without actually fouling it, than on any other occasion either before or after. Elsewhere, beyond the north Graham Land shelf region, on the South Georgia whaling grounds, where the heavy negative coverage is particularly to be noted, and in fact throughout the entire oceanic water of the circumpolar sea, there is still little indication in December that spawning has taken place, and it is perhaps significant that such scanty indications as there are should be largely confined to the Western Weddell drift not very far from the place where a major spawning seems to have occurred. There are, however, two very small occurrences, both in the last 3 days of the month, from the shelf water of the East Wind drift near the west Graham Land coast. Two major facts then seem to emerge from the November-December egg chart, and while both have been fully discussed in an earlier section (pp. 176-212), they are important enough to merit repetition. They are (i) the spawning of the krill, at its commencement at least, is to all appearances a relatively high latitude, shallow water, coastal phenomenon, and (2) having regard to the extensive and on the whole convincing negative coverage of other parts of the circumpolar sea, and especially of the coastal and oceanic water round South Georgia, it might in fact at the same time be an event peculiar to the north Graham Land shelf region (and no doubt to other adjacent shelf areas as yet unexplored) and not as has hitherto been supposed a general and widespread occurrence all round Antarctica. The first egg chart in short seems to justify the view that the existence of a discrete spawning ground in the far western (or south-western) reaches of the Weddell stream must be more than mere hj^pothesis. Moreover, the obvious objection to this, that spawning on an unknown but presumably comparable scale might simultaneously be taking place under the vast ice-sheet which throughout November until about mid-December still extends over the circumpolar sea, can readily be answered. For if such a thing were happening, as Ruud (p. 44) once suggested it might, we should expect to find, when * Representing over 5000 operations of the vertical net. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 291 60° 9(1 120 (S) • 10-100 (•) 0 100-1,000 (•) ^ 1.000- 10.000 Fig. 64. Distribution of eggs. November and December. Ice-edge late December mean. 33-2 293 DISCOVERY REPORTS the winter ice cleared away, the earUest surface product of this hypothetical spawning, that is, the Calyptopis stages, spread by January all the way round Antarctica and not, as they are (p. 310, Fig. 74), confined, still for all practical purposes exclusively, to the western Weddell drift. jfaniiary-February. The January-February egg chart (Fig. 65) again seems to emphasise, even if only in a minor degree, the importance of the north Graham Land shelf region as a spawning ground and gives some indication too that the shallow water associated with the South Orkney-South Sandwich Islands ridge may also be a region where the eggs are laid. The minor concentrations of eggs recorded in the north Graham Land area, six in all of which only four are plotted, were all obtained in February. There are no vertical observations in these waters in January. Elsewhere, apart from two notable instances to the contrary of which more will be said presently, our vertical hauls above 1000 or 1500 m. provide no indication that any major or widespread oceanic spawning has occurred. The South Georgia area with its heavy coverage is barren and the whole circumpolar sea down to the deepest levels^ we have probed is either equally barren or at best carrying only negligible numbers of eggs. There is a conspicuously barren area extending into very high latitudes in the Pacific sector though here the East Wind drift was poorly sampled between 90° and 140° W. Such very small occurrences as there are elsewhere are confined exclusively to high latitudes in the East Wind drift and again, in some measure of concentration as in December, to the western reaches of the Weddell stream. Although, however, the East Wind occurrences are negligible this again it seems must be attributed to our wholesale failure to sample the main mass of the eggs which in these high latitudes must by now have been laid. For when we turn to the distribution of the early deep living larvae (p. 301, Fig. 70), the distribution in other words of a very recent product of hatching, it becomes at once clear that spawning has already taken place in the coastal or near coastal waters of this west- going stream, beginning on a reduced scale in January and reaching major proportions by about the middle of the following month. The very large haul of eggs in Weddell East and the other moderately large one in the East Wind drift on the northern outskirts of the Ross Sea are, as already mentioned, the only instances of what could be described as major spawning in or over deep oceanic water we have encountered. The Ross Sea haul at Station 1283 in late February consisted of 263 eggs disposed vertically in a 1000 m. water column as follows: Depth (m.) Eggs 50-0 — 100-50 10 250-100 84 500-250 168 750-500 I 1000-750 — Although not a very large haul- it must be regarded as representing a clear, if possibly exceptional, instance where spawning had taken place in the surface layers over deep oceanic water (3950 m.) and far away from land. A long way from any coast, however, though Station 1283 undoubtedly is, it is perhaps worth noting that it lies at no great distance from shallow soundings, the 750 fathom line, as far as our imperfect knowledge of the bottom topography here permits us to judge, being no more than 70 odd miles away to the west. The larger haul in Weddell East presents a more difficult problem. The sounding there was ^ In general our deepest vertical net was fished between 1000 and 750 m., but there are many instances, particularly along the meridian of Greenwich, where the 1500-1000 m. layer was also covered. ^ Compare, for instance, the relatively enormous gatherings of larvae (p. 90, Tables 13 and 15) obtained at Stations 1138, 1328 and others. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 293 }(? 120 150° irL:^L±illI_UJ^b-LL^LXLL_l,:_lM'v\'V'\'\'V\ i' n i . T^. ^. > . W 180°E 150° Fig. 65. Distribution of eggs. January and February. Ice-edge February mean. 294 DISCOVERY REPORTS 5400 m. and the catch obtained, at Station 2594 in the last hour of the last day of February, consisted of 3444 eggs presumed to be disposed in a 1500 m. water column as follows: Depth (m.) Eggs 50-0 — 100-50 — 250-100 — 500-250 — 750-500 — 1000-750 — 1500-1000^ 3444 Although the significance of this occurrence has been discussed on other pages (pp. 100, 184, 208-10), for the convenience of the reader I shall review briefly the facts and the main points of the argument presented there. The deep net at this station it will be recalled, fishing for some unknown part of its course in the bottom water that comes from the west, must clearly have taken a substantial proportion, if not all, of the eggs it captured from the cold deep stratum itself. It will be recalled, too, that a high proportion, 75%, of the eggs obtained were in an advanced developmental phase, many apparently about to hatch, if not actually on the point of, or in the process of, hatching. Three alternatives were put forward to explain their presence in this deep situation: (i) they sank there from a spawning that took place at a higher level, (2) they were there because the females sank or swam down there to spawn, and (3) they got carried there from a distant spawning in the far western or south-western reaches of the Weddell drift in the fast moving bottom water which in the neighbourhood of Station 2594 appears to have upwelled to a point abnormally close to the surface. Of the three alter- natives the first was considered the most likely and the third, in view of the immense distance involved, the least likely or impossible. A point perhaps in favour of the second alternative is the fact that in the same net there were nine adults (an exceptional number for a vertical net) which included three spent females. There were, however, many adults including spent females at the surface and it is possible that the deep open vertical net had sampled this population. While virtually rejecting the third, however, it was pointed out that there might be other places where spawning was taking place much closer at hand (within about 500 miles), in the shelf or slope waters for instance of the Coats Land-Princess Martha Coast, near the very region in fact where Ross and later Bruce (p. 212) recorded such violent interference with their bottom apparatus, from which the eggs could well have been carried in the bottom water eastwards to 0° and even beyond. On reconsidering the whole matter there might yet it seems be a fourth alternative. Station 2594 although itself in very deep water lies little more than 1 20 miles north by west of the shallow but as yet imperfectly delineated and sparsely sounded Maud Bank where the minimum depth so far recorded (U.S.N. Hydrographic Office, 1955) is only 656 fathoms. If spawning takes place over this bank it is conceivable that the eggs might get displaced obliquely away from it by the bottom water as it comes up against it from the west. In support of this suggestion it might be noted that Deacon (1937, p. 109), in his classic description of the bottom water, remarks that sudden changes in depth such as, for in- stance, occur in the neighbourhood of submarine ridges like the Scotia Arc ' are certain to give rise to turbulent movements in the bottom currents '. March-April. The captures of eggs for March-April (Fig. 66) are small or negligible and, such as they are, are exclusively March records. While this may mean that spawning is now, that is in March, on the wane, other considerations, based as before on the deep larval records, suggest that * It will be recalled (p. 100, note 2 and p. 184, Fig. 22) that this net did not close and fished open to the surface. The absence of eggs at all levels above 1000 m. and the advanced developmental condition of the eggs in the deep open net suggest strongly, however, that they did, or the vast majority of them did, come from below 1000 m. See also p. 183, Fig. 21. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 295 50° O I ■ I ■ i ■ I ■ I ' 1 1 M i I ' 1 1 M . M . I . j ,^-. AJy. uj.ui:zxrf:zr. 30° 120 ® lO-lOO '<: (•) loo-i.ooo :■ (•) i.ooo-io.ooo ISO- W180°E 130° Fig. 66. Distribution of eggs. March and April. Ice-edge March mean. 296 DISCOVERY REPORTS again we may have failed completely to strike the main mass of the eggs. The March capture, for instance, in the East Wind drift of a very large number of Second Nauplii very close to the bottom on the Antarctic continental slope (p. 93, Table 14 and p. 303, Fig. 71) suggests that there at any rate, in these high latitudes, spawning might still be in full swing. Evidently, however, judging from the unbroken negative coverage extending in April from 90° W eastwards to 90° E, it ceases every- where on a major scale before the end of March. It may continue on a much reduced scale in the East Wind zone into April, for although no free eggs are recorded then, as late as the 20th of the month (p. 177), at a station near Enderby Land, a recently spent female was captured in which, still clinging to the otherwise empty ovary, were a few unsegmented globular eggs of the same size, colour and translucency as those recorded from time to time from the upper layers of the plankton. There is still one free record, although a minor one, from the coastal waters of north Graham Land suggesting that spawning there may be continuous from November to March. For the rest it will be seen the occurrences of eggs, such as they are, are confined as before exclusively to the Weddell drift and to the East Wind drift, in the latter locality mainly near the land. The South Georgia area, in spite of the thoroughness of the sampling there, is again for all practical purposes barren, and so too, it would seem, is the Pacific sector up to very high latitudes, though here again the East Wind drift is scarcely sampled. The principal inferences to be drawn from the distribution of the eggs, a comprehensive view of which is given in Fig. 67 (see also p. 301, Fig. 70 and p. 303, Fig. 71), may be summarised as follows : (i) Although the main mass of the eggs has hardly if ever been properly sampled, spawning it appears may be a phenomenon associated in its major aspects with the shallow shelf or slope water of the north and east Graham Land peninsula^ and with the corresponding shelf or slope waters of continental Antarctica in the higher latitudes of the East Wind drift. Yet paradoxically both spent and gravid females have in fact been recorded (p. 194, Figs. 24 and 25) far from the slope and coastal waters of the continental land. (2) There is an early spawning, associated with the shelf or slope waters of the north Graham Land region, in November-December. (3) There seems to be a discrete or restricted spawning there of which we have continuous evidence from November to March. (4) There is a later spawning associated with the shelf or slope waters of the East Wind drift which, beginning on a reduced scale in January (or rarely perhaps in very late December), is in full swing in February and March and continues possibly even into April. (5) Throughout the spawning season the mass shedding of eggs in the upper layers of the oceanic water far away from land seems to be an uncommon event. Yet, again paradoxically, the vertical distribution of the spent and gravid females (p. 188, Table 39) suggests that spawning is in fact a surface phenomenon. (6) Spawning at deeper levels, if it takes place at all, must take place far beyond the reach of our deepest, 1 500-1 000 m., vertical nets. (7) Spawning in general, whether it be a deep or shallow phenomenon, is restricted to the East Wind- Weddell surface stream although virtually none it seems takes place on the South Georgia whaling grounds. The overriding conclusion suggested in (i), that spawning is a phenomenon associated with shelf or slope regions, is strongly supported by the distribution of the eggs in, and in the neighbourhood of, 1 And doubtless, too, with other shelf or slope regions in this locality, but farther south, as yet unexplored. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 297 ISO W180°E 130° Fig. 67. Gross distribution of eggs. Ice-edge March mean. Large circles indicate catches of more than 10. 34 2^8 DISCOVERY REPORTS the Bransfield Strait, which, for the period November to March, is shown in Fig. 68. There, it will be seen, such significant concentrations as were encountered, without exception, occur on the continental slope just on, or at no great distance beyond, the 500 m. line. In the very shallow water of the shelf itself, however, and in the deep water beyond they are absent or conspicuously few. 62 60° 58 56° DEPTH KEY I : LESS THAN soom;: MORE THAN SOOM:: 62 es 54° '"."^ W 162 C' ^ (:/'; ■■'■■ ■ j~^-^^^'- 100-1,000 1,000-10.000 ■■/'''■'■'''' ISO- ■W 180°E 150° Fig. 71. Distribution of deep larvae. Nauplii, Metanauplii and First Calyptopes. March and April. Ice-edge March mean. 304 DISCOVERY REPORTS Coast, the occurrences are exclusively of the Metanauplius and the First Calyptopis stage. While this may again simply be due to the fact that in most instances our nets did not go deep enough to sample the Nauplii, fifteen negative observations down to 1500 m. along the 0° line, nine in March and six in April, rather seem to indicate that in the Weddell zone at least the Second Nauplius is now^ perhaps approaching the end of its natural span. The chief point of difference between this and the January- February distribution is that whereas in January-February there was a marked crowding of the main concentrations of the deep larvae in a circumscribed region of the western Weddell drift, in March-April this has given place to a system of rather widely scattered concentrations which by April stretch all along under the drift from west to east. The wide scatter of the more important occurrences might again it seems be taken to indicate that some or all of the deep larvae are now, particularly in April, becoming scarcer in the plankton. The January-February chart (p. 301, Fig. 70) shows a pronounced concentration of the deep larvae in Weddell West. By March this concentration has spread to about half-way along Weddell Middle, a west to east shift that is more likely, I think, to be due (p. 100, Table 17) to the eastward movement of the bottom water carrying the hatching eggs and developing Nauplii, rather than to an oceanic spawning having occurred in Weddell Middle some time after the major larval outburst farther west. In view of its great distance from the north Graham Land coast the major April occurrence of deep larvae towards the eastern end of the current (Fig. 71, St. 2346 in 20° E) could hardly, however, be ascribed to such a movement. On the contrary it seems it must be taken as repre- senting an instance, uncommon although our net hauls suggest such instances to be, of oceanic spawning far away from land where the females had either shed their eggs in the upper layers of the plankton, the eggs subsequently sinking, or had sought great depths to spawn. In either event this spawning could not have occurred at (or over) the place where the deep larvae were found, it being obvious that if it had the interplay of the successive water movements in which the eggs and resultant larvae subsequently became involved would have led to the displacement of the latter possibly far away from it. Apart from one exceptional occurrence with which I shall deal presently, there is no major evidence of hatching anywhere outside the Weddell stream except again, as in January-February, in the slope, coastal or near-coastal waters of the East Wind drift. Two Metanauplii it is true, were taken on the South Georgia whaling grounds and one in the Bransfield Strait (all March records), but otherwise both heavily sampled regions are barren. Such considerable part, too, of the Pacific sector as has been examined is also barren. Of the four major East Wind occurrences there is one in March of exceptional interest. At Station 2603, it will be recalled (p. 93, Table 14), a very large number of Second Nauplii along with some Metanauplii were discovered very close to the bottom in the slope water off the Princess Martha Coast. The sounding here was 1450 m., the deep net being sent down to 1400 m., a depth which, although supposedly only 50 m. from the bottom, might well in fact have been closer to it since, while the vertical work was in progress, the ship for all we know might have drifted into shallower water. No eggs were found at any of the seven levels examined at this station. The total deep larval population, however, was disposed vertically as follows: Depth (m.) Second Nauplii Metanauplii First Calyptopes 50-0 — — — 100-50 250-100 z 500-250 — — 20 750-500 1000-750 1400-1000 18 510 443 387 94 22 HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 305 Now it might well be supposed from such a disposition that the missing eggs at this station were in fact lying hatching on the bottom itself and that the First Nauplii to which they were giving rise had, as they began to climb, all moulted and become Second Nauplii before reaching the level traversed by the deep 1400-1000 m. net. If this were so, and if it be assumed that the newly hatched Nauplius immediately begins to climb, in view of the very short distance it would have had to travel in this instance before reaching the 1400 m. level, there would seem to be good grounds for supposing, as Fraser (p. 119) originally suggested, that the life span of the First Nauplius in the plankton is in fact exceedingly short. On the other hand, if it did not rise immediately but lay for some time dormant on the bottom, or again if the rate of climb were very slow, as it might well be for such a puny creature, it would follow that its existence might not after all be so ephemeral as Fraser imagined. I can suggest three possible reasons for the presence of these very recently hatched larvae so close to the bottom: (i) the eggs sank there from some higher level before hatching, (2) the females spawned on or near the bottom, and (3) the eggs, although laid perhaps at some higher level, got carried to the bottom by local sinking of the adjacent shelf water. The occurrence of the Second Nauplius at Station 2603 is the only record of this stage we have for March-April and it is significant and perhaps to be expected that it should come from the East Wind drift where the main spawning takes place so much later than elsewhere. Whether or not it survives there into April is harder to say, since none of our East Wind April nets go deeper than 1000 m. At two stations, however, off Enderby Land (p. 93, Table 14) substantial numbers of Metanauplii were recorded on the 20th of the month and in view of the lateness of the date it (the Second Nauplius) might well it seems survive in these high latitudes until at least over March. The one exceptional occurrence of deep larvae to which I have already alluded, exceptional in the sense that it belongs neither to the East Wind nor to the Weddell stream, is of a relatively enormous number of Metanauplii and deep First Calyptopes in the oceanic water of the Scotia Sea at a point more than 100 miles north-north-west of the northern boundary of the Weddell drift. The occurrence could in fact be described not only as exceptional but as highly anomalous, for the locality in which it was recorded falls strictly speaking within the southern limits of the West Wind drift, a region already shown throughout its vast circumpolar extent, to be barren of eggs or of recently hatched larvae in the deeper layers. It was recorded on 12 March at Station 647 and is represented in Fig. 71 by the large black and white plot to the north of Graham Land, a little eastward of meridian 60° W. The deep larvae were disposed vertically as follows : First Depth (m.) Metanauplii Calyptopes 50-10 — — 100-50 — — 250-100 — — 500-250 — 150 750-500 3702 28a 1000-750 577 7 Clearly somewhere far below the 1000 m. level in this locality, although not necessarily (p. 304) directly below it, there must have existed, in all probability some time previously, a mass of hatching eggs, resulting, it might be supposed, from another instance of large-scale oceanic spawning in either the deep or surface layers far away from land. Probable, however, though it may seem, this explana- tion of the anomaly is not altogether satisfactory, for not only have we failed consistently to find any evidence of large-scale spawning in this locality either from our records of spent and gravid females or from our records of eggs, but throughout these investigations have repeatedly found the area to be barren, or virtually barren, of a potential breeding stock as well. In fact the net yield of breeding 35 DM 3o6 DISCOVERY REPORTS females from fifty stations operated during the spawning season and covering the whole West Wind region of the Scotia Sea, or more precisely that part of it lying between the northern boundary of the Weddell drift and the Antarctic convergence, was twelve gravid and two spent, the average yield per station of potential breeders, say euphausians over 30 mm. long, being only two. In view of these virtually negative findings we must again it seems turn to the distribution and movements of the bottom water for a more satisfactory explanation of the presence of these deep larvae so far outside their normal locus of abundance in this sector. Between the South Orkney Islands and the South Shetland Islands and on either side of the relatively shallow ridge of the Scotia Arc, Deacon (1937, p. 108) has shown that some bottom water is formed by a convection current which reaches from the surface to the bottom in winter. Some of this water (Deacon, 1937; Clowes, 1938) spreads westward for a short distance into the Drake Passage and, assuming it to be carrying hatching eggs and developing Nauplii, the former shed say somewhere in the slope waters associated with the Scotia Ridge, this westerly movement might conceivably be held responsible for the presence of the Meta- nauplii at Station 647, where, it is interesting to record (p. 90, Table 13), an isolated patch of bottom water was in fact encountered at a depth of a little over 3000 m. The principal inferences to be drawn from the distribution of the deep larvae, a comprehensive picture of which is given in Fig. 72, may be summarised then as follows : (i) Although the main mass of the Nauplii, represented exclusively by Second Nauplii, has only been sampled on a few occasions, because in the vast majority of instances our nets, in the right places and at the right time of year, did not go deep enough, from the distribution of the Metanauplii, of which we have substantial gatherings, it is clear that hatching, in its major aspects, is a phenomenon of the deep oceanic water of the Weddell drift (where it probably takes place in the bottom water) and also of the slope or coastal waters of the continental land. In shallow shelf regions it only takes place on a negligible scale and virtually never it seems happens on the South Georgia whaling grounds. (2) There is an early hatching deep down in the oceanic water of Weddell West in December and although it appears so far only to have taken place on a very small scale much deeper observations than we have at present will probably reveal it, then if not before, as a major event. (3) There is a later hatching associated with the slope, coastal or near-coastal waters of the East Wind drift, which, beginning on a reduced scale in January, is in full swing by February. The East Wind hatching, however, seems to be confined to the New Zealand, Australian, Indian Ocean and Atlantic sectors. (4) The deep larvae spread eastwards in the bottom water below the Weddell drift and by early March have reached a point about half-way along Weddell Middle. (5) There is again evidence (e.g. at Station 2346) that spawning must sometimes, if rarely, take place in or over deep oceanic water far away from land or banks. (6) Hatching in general is restricted almost exclusively to deep water below the East Wind- Weddell surface stream, but in rare instances it seems hatching eggs and developing Nauplii may be carried through the movement of the bottom water some distance outside this system into the West Wind region of the Scotia Sea. (7) In the Pacific sector there is no evidence of any hatching, but it can be seen from Figs. 70 and 71 that the pack-ice there mostly extends to the northern boundary of the East Wind zone so that all but a few of our samples are in fact from the West Wind drift. If any large-scale hatching does occur in 1 this region, it must take place in the inaccessible waters beyond the 70th parallel, though even thisi (p. 124) seems improbable. (8) Since the occurrences of the deep climbers must coincide approximately with those of the eggs] from which they spring, in other words must betray the existence, or recent existence, at still deeper] HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 307 ISO" W 180-E 150° Fig. 72. Gross distribution of deep larvae. Ice-edge March mean. Large circles indicate catches of more than 10 (except in December where smaller gatherings of less than 10 have been plotted to show the restricted distribution of the earliest deep occurrences). 35-a 3o8 DISCOVERY REPORTS levels, not necessarily however (p. 304) directly below, of hatching eggs, then perhaps the overriding inference to be drawn from Fig. 72 is that the major hatchings from which the krill population springs take place (i) in Weddell West, especially between December and March, and (2) in the coastal or near-coastal waters of the East Wind zone from the Ross Sea westwards to the Princess Martha Coast between January and April. The shallow living First, Second and Third Calyptopes November-December. The larvae of this group have now completed the developmental ascent, and as Fig. 73 shows in November-December some have already reached the surface, although so far only in small numbers. The occurrences, such as they are, are mainly December ones, there being only two in November, one of a Second Calyptopis on the South Georgia whaling grounds and the other of a First Calyptopis (a Norwegian record) in Weddell Middle. Both are probably highly exceptional. It will be seen that the two December occurrences of any note are located in the oceanic water of the western Weddell drift^ more or less directly over the region where in the same month the first deep stragglers from the rising Naupliar and Metanaupliar population were recorded. Elsewhere, apart from a few widely scattered Calyptopis stragglers, there is no indication yet, that is, in December, of any large-scale accomplishment of the developmental ascent throughout the circumpolar sea. The heavily sampled shelf waters of the Bransfield Strait-North Graham Land area and the equally heavily sampled South Georgia whaling grounds (apart from the November record) are barren of these stages and so virtually also is the whole circumpolar sea east about from 30° W to 60° W. In the East Wind drift the occurrence of a single First Calyptopis (a Norwegian record) is to be noted off Enderby Land. January-February. It is evidently not until January (Fig. 74) that the Calyptopes in Weddell West begin to reach the surface on a really substantial scale and by February it will be seen they are found spread across this sector as far east as 30° W.^ Some now, both in January and February, appear on the South Georgia whaling grounds, but in view of our failure to find either eggs or deep recently hatched larvae in these island waters from November right through to February (Figs. 64, 65, 69 and 70) it seems certain that these could not have sprung from any local hatching but must have got carried there in the surface stream. Apart from one substantial February occurrence in the slope region of the East Wind drift off the Princess Martha Coast (near where as Fig. 70 shows the eggs must also have been hatching) the surface waters of the circumpolar sea, whether shelf or oceanic, are barren, or virtually barren, of Calyptopis forms. The negligible scale of their occurrences in the heavily sampled shelf waters of the Bransfield Strait-North Graham Land and Ross Sea areas is again particularly noticeable, further evidence (though of course from the same stations as in Fig. 70) of the relative unimportance of the hatching (p. 201) that must take place in these shallow conditions. The concentration of the vast majority of the earliest surface arrivals in Weddell West, and their virtual absence at this time from other parts of the circumpolar sea, again strengthens the evidence that a major spawning takes place somewhere in the western reaches of the Weddell stream and at the same time demonstrates conclusively that the November-December spawning there (p. 290 and Fig. 64) could emphatically not have been taking place simultaneously, undetected all round Antarctica, under the vast winter ice-sheet which still girdles the continental land. 1 One of them in fact appears just outside the Weddell drift, but this may only be an appearance since the water boundaries plotted in these charts can only be described as approximate and no doubt vary considerably from month to month and from year to year. ^ The two January plots of the highest order of magnitude are from Norwegian sources. They were recorded by Ruud (1932) in ' Vikingen' on 22 and 28 January 1930. The locality in which they occurred was not examined by us that year. I HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 309 60 120 \> ' > ' / ' f ' > ' IZZ3±^ ISO" •W 180°E 150° Fig. 73. Distribution of surface larvae. First, Second and Third Calyptopes. November and December. Ice-edge late December mean. 3IO DISCOVERY REPORTS 9 J 120 1.000-10,000 150° W 180°E ISO" Fig- 74- Distribution of surface larvae. First, Second and Third Calyptopes. January and February. Ice-edge February mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 311 60 90 120 ^ ■ ■ ■ ^ ' ^1 150° W 180°E ISO- Fig. 75. Distribution of surface larvae. First, Second and Third Calyptopes. March and April. Ice-edge March mean. 312 DISCOVERY REPORTS From the dates of the major January recordings of these surface forms some estimate can be formed of the duration of the developmental ascent. All, both Norwegian and British, fall late in the month, from the 22nd onwards, and since the earliest recorded major spawning in Weddell West is in December (at Station 540 on the 19th) and since there is no evidence of any large-scale populating of the surface layers there throughout that month, it would appear probable that the bulk of the developing eggs and ascending larvae spend at least 30 days in their deep environment before the latter, as First Calyptopes, reach the surface on any substantial scale. What part of this time is spent in the actual ascent is harder to judge for we do not know how long the eggs, after being shed, take to develop to the point of hatching.^ We do know, however, that hatching must be taking place deep down through- I out December (p. 300) and on that ground it might be supposed the developmental ascent takes the better part of a month, if not longer. J From the East Wind zone we have a single mid- February observation off the Princess Martha Coast " to suggest that at its earliest the arrival of the First Calyptopes at the surface there takes place about a month later than it does in Weddell West. The preponderance, however, of negative, or virtually negative, observations in the former region, especially between 30° W and 60° E, suggests that it ■ may not in fact be until March (p. 311, Fig. 75) that the accomplishment of the developmental ascent becomes general in these high latitudes. There is one anomalous occurrence of three Calyptopes, two First and one Second, which lies actually outside the Antarctic convergence immediately to the west of the S-shaped bend in it that occurs about half-way between the Falkland Islands and South Georgia. No doubt this anomaly can be ascribed as already recorded (p. 76) to the complex and somewhat unstable hydrological conditions that seem to exist in this locality. March-April. In general the March-April distribution of the surface Calyptopes (Fig. 75) follows closely that of the deep larvae for the same period. The whole of the Weddell drift from west to east is now carrying large, sometimes enormous, numbers, and in the East Wind drift, notably again in the Australian-Indian Ocean- Atlantic sectors, it is evident that, although our observations there are rather few and widely scattered, there is an almost equally substantial surface population in the slope, coastal or near-coastal waters from March onwards. Turning again to Figs. 74 and 75 it will be seen that as with the deep larvae there is again a west to east shift involving by March the extension of the principal concentrations of the surface Calyptopes, confined in January-February to Weddell West, to a point about half-way along Weddell ' Middle, an extension that at first sight seems to be simply the result of the easterly movement of the surface stream. While the surface drift is manifestly involved it is not, however, the sole instrument of easterly dispersal, comparison of the developmental condition of the surface swarms in Weddell West in February with the corresponding condition of those of Weddell Middle in March, and further j comparison of the developmental condition of the Weddell Middle March swarms with the cor- responding condition of those encountered still farther east in April, suggesting very strongly that an equally important transporting agent is the cold bottom current out of which the new-born krill seem | to be rising. This is illustrated graphically in Fig. 76 which shows the stage frequency of the total j surface larval catch at eighteen stations covering the Weddell drift from west to east, Weddell West in February, Weddell Middle in March and Weddell East in April. The stations are arranged in the order of their progression towards the east, that is, in the direction of flow of the surface stream, the j gatherings of the several stages obtained at each station being expressed in each instance as a per- centage of the total catch in the surface (250-0 m.) layer. The number (not the percentage) of deep 1 In the eggs of Meganyctiphanes norvegica, Heegaard (1948) gets the impression that the period, spawning to hatching, , might last for 5 or 6 days. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 313 larvae recorded at each station is also shown. The diagram is based practically throughout on observa- tions where the total surface catch was large or very large, at sixteen of the stations shown it being of the order of loo-iooo or above, at the remaining two (Stations 368, 369) of the order of 50-100. Stations at which the total catch was less than 50 have not been included. On this basis the graphical expression of the stage frequency presented here may be said to carry the maximum significance our data permit. Taking first then the position in February and March, if the larvae were spreading from a major locus of spawning in Weddell West eastwards into Weddell Middle entirely in the surface stream, as they manifestly appear to be doing in Figs. 74 and 75, we should expect to find that, since they LONGITUDE DtX OF MONTH STATION FURCILIA SECTOR and MONTH 60W CALYPTOPIS WEDDELL WEST FEBRUARY 30W 47° 44° IS IS I96S 618 40° 19 620 34° 17 VI4 30° 25 361 30° 25 362 DEEP LARVAE 213 24 WEDDELL MIDDLE MARCH WOE 27° 2 365 27° 8 358 27° 9 369 26° 23° 12 4 1994 1142 18° 5 1144 Seal, per c^" =° "° r^^, r^7„„ I I 213 24 24 - - - I - - - II 30 661 I - - - - 1141 WEDDELL EAST APRIL 30E OI° OI° 01° 04° 19° 20° 14 IS 16 30 27 26 2316 2318 2320 1353 2346 2344 SECTOR and MONTH LONGITUDE DAY OF MONTH STATION 6 FURCILIA 5 4 3 2 I 3 CALYPTOPIS 2 I DEEP LARVAE Fig. 76. Developmental condition of surface larvae in the Weddell drift, showing evidence of eastward dispersal partly in the surface stream and partly in the cold bottom water. For further explanation and details of construction see text. would be feeding and growing and moulting as they went, as the season advanced the swarms would become progressively and conspicuously older as they travelled towards the east. Contrary, however, to expectation we find that in Weddell Middle, for instance, the older swarms are in the west and the younger, indeed conspicuously younger, in the east, the youngest of all at Station 1144 (repre- sented in Fig. 75 by the large black and white plot almost half-way between 30° W and 0°) being in fact the most easterly. While it seems clear enough that the swarms encountered between 34° and 27° W (Stations ¥14,^ 361, 362, 365, 368 and 369), being conspicuously older than those farther west, did in fact originate in larval risings that took place somewhere to the west of where they were recorded, subsequently being carried eastwards to 30° W and beyond in the surface drift, the swarms still farther east (Stations 1994, 11 42 and 1144) are obviously too young to have had a similar history. On the contrary, it seems they must originally have reached the surface in situ, or if not actually in situ at any rate somewhere not very far to the west of where they were found. In other words, unless these easterly Weddell Middle swarms were derived from some local oceanic spawning, which although possible does not (p. 296) appear very likely, it seems they must have got carried to their easterly situation through the spreading of eggs and nauplii in the bottom water flowing from the west. To invoke deep transport to account in part at least for the eastward spreading of the larvae is not after all unreasonable, for we know that the bottom water is moving to the east and have some reason to think (p. 100) that all along its path from west to east very young larvae are rising out of it, upwards and into the surface stream, at Station 11 44, it would seem, in very large numbers. It is equally clear, turning to the position in March and April, that the April swarms in Weddell East just eastward of 0° (Stations 2316, 2318 and 2320) are much too young to be explained in terms of transport in the surface stream from an original locus of spawning so far away as Weddell West. Clearly they must originally have come to the surface much nearer at hand and having regard to their age they could very well it is suggested have done so, in the First Calyptopis stage, somewhere in the 1 One of the stations worked from the Norwegian whale factory ' Vikingen'. 36 DM 314 DISCOVERY REPORTS neighbourhood of say Station 1 144 (although not necessarily so long before as 5 March), subsequently developing to their present state during their journey from there to 0°. From 18° W to 0° the distance in latitude 60° S is approximately 540 miles and although as yet little is known for certain about the speed of the Weddell current, the observations of Weddell himself and of other early and recent navigators^ suggest it might be of the order of between ^ and 2 knots, or say between 1 2 and 48 nautical miles a day. At 24 miles a day the larvae would be carried from 18° W to 0° in 22 days, long enough it might be for them to develop from the First to the Third Calyptopis stage. While again nothing definite is known of the duration of this particular phase, the developmental condition of the MONTH DATE STATION FURCILIA 6 5 4 3 2 I CALYPTOPIS 3 2 I IB 19 22 N42 N43 VII 23 27 28 1492 823 VI2 Scale per cent o 50 100 I 38 80 2653 34 379 255 1 213 — 24 24 — 15 17 1965 VI4 19 20 25 620 622 361 25 362 ♦♦ DEEP LARVAE | 88 80 2653 34 379 255 | 213 MONTH DATE STATION 6 FURCILIA 5 4 3 2 I 3 CALYPTOPIS 2 I DEEP LARVAE * LARVAE ABOUT TO REACH SURFACE Fig. 77. Developmental condition of surface larvae in Weddell West in January and February. For details of construction see p. 312. surface larvae in Weddell West in January and February (Fig. 77) suggests that it might well be of this order. For in January it will be seen, from the i8th onwards, the vast majority are still in the First Calyptopis stage, it being not until 17 February that in this sector the first substantial numbers of Third Calyptopes are found. In northern waters (p. 119) Lebour (1924) found that laboratory reared specimens of Nyctiphanes coiichii (Bell) took 15 days to grow from First to Third Calyptopes, while the plankton investigations of Einarsson (1945) suggest that the same developmental phase in Thysanopoda acutifrons (Holt and Tattersall) might take about 23 days. The corresponding phase in Meganyctiphanes norvegica (M. Sars) might it seems be shorter, Mauchline (1959) finding that in the laboratory it lasts from 9 to 12 days. The young, for April indeed remarkably young, surface swarm encountered in 19° E at Station 2346 (Fig. 76) clearly could not have come from far west in the surface drift. Like the March swarms at Stations 1994, 1142 and 1144 in Weddell Middle it must have come to the surface in situ, or again if not actually in situ, somewhere not far to the west of where we found it. It must have sprung, too, from an oceanic spawning that took place somewhere in this far eastern sector of the Weddell stream, not necessarily, however (p. 304), near the site of Station 2346. To continue the discussion on the March-April distribution of these surface forms it will be seen that although the region is otherwise virtually barren there is a single although substantial March occurrence of Calyptopes on the South Georgia whaling grounds. Their situation to the south-east of 1 In the track chart illustrating his great southern journey of 1823 Weddell records a maximum speed for the surface drift, a little to the west of 30° W, of 21 nautical miles a day, while farther west, off the east coast of Graham Land, Larsen in December 1893 reported north-easterly and north-north-easterly sets of from i to 2 knots (Murray, 1894). Near the South Orkney Islands {Antarctic Pilot, 1948) easterly sets of | knot are common and settings to the east of as much as i^ knots have been reported. Bruce (1918) records an easterly set of 35 miles in 2 days in Weddell West, Wordie (1921 b) stating that in this part of the current the pack travels towards the South Orkneys 'often as fast as 20 miles a day'. In the straits and narrow passages of the South Orkney Islands themselves the strong, 3J-4 knot, tidal currents reported by Powell (1822) and the tide-rips reported by ' Discovery IT (Marr, 1935) in January 1933 provide, it seems, further evidence of the strength of the Weddell stream as it sweeps eastward through the group. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 31S the island, in the tongue of Weddell water which approaches it from that direction, and our almost total failure to discover there any deep larvae from which they could have sprung, suggest strongly, however, that they could not have been produced in situ but on the contrary got carried there in the surface drift. Some surface Calyptopes now, both in March and April, also appear in the Bransfield Strait, but in view of the pronounced scarcity of Nauplii and Metanauplii in these waters, of which we have ample evidence from November to April (Figs. 69-72), it is doubtful if even they too could have been produced in situ. More likely they drifted into the strait (p. 202) with the Weddell or Bellingshausen Sea water which enters it, the former from the east, the latter from the west. MONTH STATION DATE LONG WEST FURCIUA CALYPTOPIS DEEP LARVAE 2280 8 75° 383 14 63° O 50 100 o SO lio ^"^^'^ P^"" <:«"' t MONTH STATION DATE LONG WEST FURCILIA CALYPTOPIS DEEP LARVAE Fig. 78. Developmental condition of surface larvae at West Wind Stations 2280 and 383. Turning now to the situation in the West Wind drift, it will be noted that in the Bellingshausen Sea-Drake Passage neighbourhood there are two substantial recordings, one in March and one in April, of surface Calyptopis forms. These occurrences at first sight suggest that the larval population of this vast tract of the circumpolar sea is not everywhere so poor as the distributional charts so far presented have suggested it to be. It is distinctly probable, however, that neither is a true West Wind occurrence, much of this West Graham Land region being affected by encroachment from the East Wind drift, a tongue of which, as Deacon (1937, p. 17) has shown, deflected northward near Peter I Island, probably he suggests under the influence of a submarine ridge, penetrates in these parts into the West Wind zone. In both instances, therefore, the larvae very likely sprang from East Wind risings, somewhere possibly on the slope-edge of the extensive region of shallow soundings which between Peter I and Charcot Islands stretches far to the north of the continental land, from such risings later moving to the north and east in the surface drift. As Fig. 78 shows, at both stations the swarms recorded are essentially surface ones, that at Station 383 obviously having been in the surface for some considerable time and almost certainly having come from the west. Moreover, the absence at both stations of deep larvae augmenting them from below and the fact that we have never recorded any deep larvae whatsoever in this particular region considerably strengthen the view that in neither instance could they have been locally derived. If it be assumed that these supposed but probably far from hypothetical intruders from the East Wind zone move northwards in more or less the same longitude, and later eastwards in more or less the same latitude, their separation in time and space would give another hint of how long it may possibly take for E. superba to develop from the First to the Third Calyptopis stage. If the observa- tions had been made in the same year they would in fact have suggested an easterly movement of the larvae of about 360 miles in about 37 days, and that during this time they had developed from being First to become dominantly Third Calyptopis forms. The rate of travel, approximately 10 miles a day or a little under 1 knot, seems about right for the West Wind drift and the developmental rate, ahhough rather slower than our previous estimates, is not altogether out of keeping with them. The speed of 36-2 3i6 DISCOVERY REPORTS the West Wind drift is very variable. It is greatest in the Drake Passage, where it is compressed between Cape Horn and Graham Land, and diminishes towards the east, but even in the Drake Passage {Antarctic Pilot, 1948, p. 55) the majority of the currents experienced by shipping are of the order of i knot or less.^ In April 1842 Ross (1847) found the current there setting to the east at a rate of from 12 to 16 miles a day. South of Australia, although speeds of up to 15 and 18 miles a day have been recorded, the average rate of the easterly movement seems to be of the order of from 6 to 10 miles a day (Mawson, 19400). Another typical example of northerly extension of East Wind influence, as will be shown more emphatically later (p. 325, Fig. 86), is provided in Fig. 75 by the very small black West Wind plot south-east of Kerguelen. Table 61. Vertical distribution of the larvae at Station 64'] Second Calyptopis 6 Depth First {m.) Metanauplii Calyptopis 50-0 — 361 100-50 — 20 250-100 — 94 500-250 — ISO 75c^50o 3.702 282 1000-750 577 7 In the West Wind region of the Scotia Sea, that is, in that part of it between the northern boundary of the Weddell drift and the Antarctic convergence, there is one major concentration and three minor concentrations of surface Calyptopis forms, all four occurrences in March. The major occurrence is represented in Fig. 75 by the black and white plot of the second highest order of magnitude immediately to the east of meridian 60° W. It occurred at Station 647, where as already noted (p. 306), the bottom water, spreading from a secondary locus of its formation between the South Orkney and South Shetland Islands, seems to have been carrying large numbers of eggs and deep larvae, some of which, the vast majority as Metanauplii, had already risen to the 1000-500 m. level. The vertical distribution of the total larval population encountered here is shown in Table 61 from which it is clear that the surface population, being 99% First Calyptopes, represents the fairly recent product of the moulting of the Metanauplii rising from below. The three minor concentrations encountered a little farther east, represented in Fig. 75 by the black and white plots of the second lowest order of magnitude, had no deep larvae below them and all, therefore, may have come from the west, having sprung perhaps from West Wind risings such as that for instance recorded at Station 647. There is little further to add concerning the March and April distribution of the surface Calyp- topes, except perhaps to note again their pronounced abundance in the East Wind- Weddell surface stream and their local abundance in such strictly limited areas of the West Wind drift as are affected by cold surface emanation from the East Wind zone. May-June. The distribution of the surface Calyptopes for May-June, after which they outgrow themselves and are no longer to be found in the plankton,^ is shown in Fig. 79. The Antarctic winter has now begun and the sea become frozen over far to the north of the continental land. The 1 Smith (1920), however, states that through this narrow passage the current often flows at a rate of as much as 50 miles a day, but this probably only occurs when the westerly winds are exceptionally strong. The average speed of the drifting derelict ' Dumbartonshire', abandoned by her crew in August 1894 in 40° S, 40° W, works out over a distance of 1850 miles at around 12 miles a day (Russell, 1895). For 9 days, however, she appears to have drifted at 43 miles a day, probably as Russell suggests owing to a gale of wind. 2 In the plankton at least of the northern or Weddell zone. The Calyptopes may not yet in fact have outgrown themselves (p. 285, note 4) in the plankton of the now largely inaccessible East Wind zone. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 317 i.i.i.i.i..j.,i,i,.i,.i,i.i,i,i,i y. 1,1 , 1, 1 .J ~ I • I ■ ' ■ '^^ ISO" W ISO-E 150° Fig. 79. Distribution of surface larvae. First, Second and Third Calyptopes. May and June. Ice-edge May mean. 3i8 DISCOVERY REPORTS high southern region of larval abundance in the East Wind zone, covered for the most part by a great ice-sheet, is everywhere, or virtually everyw^here, inaccessible^ and from now until January of the following year the facts of the distribution can only be determined with certainty for such northerly and gradually decreasing parts of the circumpolar sea as continue to remain ice free, the situation in the East Wind drift becoming largely a matter for inference or conjecture. Taking first what is apparent, the situation in the ice-free north, it will be seen that although our overall sampling of the circumpolar sea in May-June is uniformly meagre, more meagre in fact by far than at any other time covered by our observations, it nevertheless still emphasises the importance of the Weddell drift as the primary carrier of these now exclusively surface forms and their paramount MONTH WO"E- LONGITUDE DAY STATION FURCILIA 6 5 CAL 3 2 I DEEP LARVAE MAY JUNE -30E ir 17" 20" 24° O0° 00° 08° 17° 18° 18° I 2341 2589 lO 1354 1355 1356 1357 1779 1781 1785 I790 1792 1794 + ^411 ""' ilUo I T MONTH LONGITUDE DAY STATION 6 FURCILIA 5 4 3 DEEP LARVAE Fig. 8o. Developmental condition of surface larvae in Weddell East in May and June. instrument of dispersal in these low latitudes. The situation in fact is even more striking than Fig. 79 reveals, for as will be seen later (p. 328, Fig. 88 and p. 348, Fig. 102), at all but one- of the fifteen stations made in Weddell Middle and Weddell East at this time, surface larvae in one developmental phase or another were encountered, for the most part in very substantial numbers. Compare with this the larval barrenness encountered, for instance, between meridians 1 20° and 1 80° E, where our observational coverage in May and June is actually heavier (in the West Wind drift) than in the Weddell zone. In Weddell East it will be seen the significant occurrences of Calyptopes are all in June, there being in fact only one negligible occurrence of this developmental phase there in May, the majority of the larvae encountered then, as Fig. 80 shows, having already reached a dominantly Second Furcilia stage. Various reasons, however, might be put forward to explain this seemingly unnatural situation. The observations for May-June in this sector were made in different years, the May stations in 1934, the June stations in 1936, and it could well be that the 1934 larvae, as they drifted from the west, encountered better feeding than the larvae of two years later and so, perhaps outgrowing their Calyp- topis state at a somewhat faster rate than usual, arrived in Weddell East in a more advanced condition than the majority of them seem to have done in 1936. In any event, in the investigation of such a complex thing as this, the dynamics of the distribution of a heterogeneous system of larval swarms subject to the vagaries not only of a surface stream, but of a bottom current and crosswise intermediate current as well, it is not surprising that the overall ' Inaccessible from the north or seaward, not necessarily, however, from the continental land. It is now known that in at least some parts of Antarctica, the near coastal waters of the East Wind zone may from time to time be surprisingly open even j in the depth of winter. Fuchs (1958), describing the conditions at Shackleton Base in the winter of 1957, gives a very clear picture of how open they can sometimes be. 'To our surprise', he writes, 'the movement of the pack ice under the influence of the westerly current maintained constantly changing leads of open water. Occasionally a strong blow from the south would drive the pack out of sight, so that all through the winter there were periods when open water extended many miles to east and west of Shackleton'. Similar winter conditions have recently been reported between 45° and 80° E (Mellor, 1959). Priestley (1914), I find, records open water from shore to horizon off the Victoria Land coast in 75° S on midwinter's day. 2 A northerly station (p. 328, Fig. 88) that might not in fact have been located in the Weddell drift at all. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 319 picture that emerges is not always a very tidy one. Variation in the growth-rate alone must produce many seeming anomalies and the vagaries of the bottom and warm deep currents, and above all those of the surface drift, many more. Retarding winds, for instance, would tend to delay the arrival of the swarms in the east in certain years and assisting winds to advance it. Variation in the volume and speed of the cold deep current must also play a part. It may be, for instance, that in exceptionally cold winters more bottom water is formed at the head of the Weddell Sea than usual, and that the flow of the bottom current itself is in consequence stronger after such winters than after mild ones.^ From this it would follow that the colder the winter the more swiftly the deep larvae would sub- sequently be carried away to the north and east, and conversely, the milder the winter the slower. c I OOn 090 oeo 0 70 0 60 0 50- 040 0 30 020- o 10 000 TEMPERATURE DEEP CATCH - 100 200 3005 4 00 500 1928 1929 I930 1931 1932 1933 1934 1935 1937 1938 Fig. 8i. Average annual catch of deep larvae and average temperature at 1500 m. The dynamics are further complicated by the existence of the warm crosswise intermediate current they must traverse before they reach the surface stream. Where it flows directly below the surface drift the major effect of this current would be to divert the deep larvae from their easterly course, tending in general to carry them farther south. The stronger the flow the greater the diversion, involving sometimes, it seems possible (p. 302), their transference from the Weddell to the East Wind zone. In general, however, its most likely effect would be to carry them away from a region where the surface drift was strong into another where it was weak, or vice versa, and so in the end again tend to advance or retard the eastward movement of the resultant surface population. As far as the bottom water is concerned, there does appear to be some measure of correlation between what might be described as the annual volume or activity of this current and the annual abundance of the deep larvae encountered during the spawning season. A graphical representation of this relationship is given in Fig. 81, which shows the average temperature at 1500 m. (taking that as 1 Although Fofonoff 's (1956) work suggests that both temperature and sahnity of the water sinking into the bottom layer are likely to remain more or less constant, Deacon (1959) has pointed out that the volume of the transfer can vary, and that we should expect more sinking at the end of winter than at the end of summer. We should expect, too, as I have suggested, more sinking at the end of a severe winter than at the end of a mild one. Worthington (1959), discussing the formation of 'the 18° water' in the Sargasso Sea, notes that on the basis of existing measurements anomalously cold winters will produce a larger volume of this type of water than usual. 320 DISCOVERY REPORTS a measure of the proximity or influence of the cold deep layer at this level or alternatively as a measure of its volume or activity) plotted against the average catch of deep larvae in the 1000-250 m. layer for the 10 years, between 1928 and 1938, in which we have observations in Weddell West and Weddell Middle. The average temperature is calculated from observations covering the period November to April and the average deep catch based on stations, both positive and negative, restricted to the period mid-January to the end of April in Weddell West, and to the period beginning of March to the end of April in Weddell Middle, the restriction in each instance being imposed with the object of excluding from the calculation all stations operated at times when the deep larvae, though doubtless present in both sectors, were too deep to be sampled by our nets. The correlation seems clear enough, especially in the years when the average temperature at 1500 m. was at its lowest and the activity of the cold deep current presumed to be at its strongest. It is true that the stations were not distributed in the same way in the different years, but at least the association of numbers of larvae with cold deep water is of special interest. Recently Cooper (1955) has suggested that fluctuations in Arctic climate may be connected with changes in the biological productivity of the English Channel. To explain certain large changes in phosphate distribution and in species and abundance of plankton animals that had been observed there between 1920 and 1955 he has erected two interesting hypotheses, (i) that in cold Arctic winters saline surface water is cooled further and made heavier than in relatively mild ones, and (2) that this leads to a greater recruitment of fresh deep water in the North Atlantic after cold Arctic winters. Following this short but necessary digression, the commentary on the May-June distribution of the surface Calyptopes may be resumed with some further remarks on the position in the Weddell drift and the extent of its influence to the east, and concluded with some discussion of the now problematical position in the ice-covered East Wind zone. Over the greater part of the Weddell stream it will be seen we have no observations for May and June, such as we have being confined for all practical purposes to Weddell East. We cannot therefore be sure whether the Calyptopes survive in the region west of 0° or not. It seems fairly clear, however, judging from the advanced condition of the majority of the swarms encountered there in April (Fig. 82), that if any do survive they would do so only into May and then only in rapidly declining numbers. Turning now to the eastward limit of the influence of the Weddell drift in these latitudes we find from the March-April, and more especially from the May-June distribution of these surface forms, the first indication of a somewhat remarkable point to which attention will frequently be drawn in the distributional charts that have yet to be presented. It is that the young surface swarms borne east- wards in the Weddell stream do not, at any rate as larvae,'^ seem to get carried beyond about meridian 30° E except for a short distance as occasional stragglers. In other words it begins to appear that where the Weddell drift ends there too the purely larval distribution ends, and ends it seems quite abruptly, from which it follows that the larval development in these latitudes, whether in the deep or surface layers, from beginning to end takes place strictly within the confines of the region in which this great surface stream is flowing. The evidence for this inference is perhaps best illustrated in Figs, no, 112, 115 and 117 (pp. 359-69) in which I have plotted the seasonal distribu- tion of the massed surface larvae based on the gatherings of our towed stramin nets. The facts of the larval distribution in the East Wind drift are now, as I have said, a matter for inference or conjecture. In view, however, of the lateness of the main spawning and hatching that takes place in these high latitudes (p. 306), and of the slowness of the growth that ensues, there would seem to be good grounds for assuming that throughout May and June, under the ice-sheet that now covers this high southern region of larval abundance, the Calyptopis forms must be surviving in ^ See, however, p. 376. 321 HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL large, possibly very large, numbers. The backward development of the swarms we have repeatedly encountered there earlier in the season strongly supports such a view, for even in late April it will be seen (Fig. 83) they have progressed little beyond the First Calyptopis stage. It is probable, in fact, that in May and June the Calyptopes survive in these high latitudes in far greater abundance than they do, for instance, in the far eastern reaches of the Weddell drift. SECTOR 60W LONGITUDE DATE STATION FURCILIA 54° 47° 44° 44° 41° 2 21 20 4 17 W529W200WI99 1331 WI97 WEDDELL WEST 30 W WEDDELL MIDDLE WO E WEDDELL EAST 30 E SECTOR 23° 22° 14° OB° OI° OI° OI° 04° 19° 20' 23 26 27 28 14 15 16 30 27 26 1346 1350 1351 1352 2316 2318 2320 1353 2346 2344 LONGITUDE DATE STATION 6 FURCILIA 5 Fig. 82. Developmental condition of surface larvae in the Weddell drift in April. MONTH DATE STATION FURCILIA 5 5 4 3 2 I CALYPTOPIS 3 2 DEEP LARVAE 28 1662 2 17 28 1671 NI9 1545 3 4 8 10 22 26 26002603 26IO 1312 1713 1720 Scole per cent o so too .ik„„iil 20 20 22 854 855 2335 13 19 424 227 1455 38 — 9 MONTH DATE STATION 6 FURCILIA 5 4 3 2 I 3 CALYPTOPIS 2 I DEEP LARVAE * LARVAE ABOUT TO REACH THE SURFACE Fig. 83. Developmental condition of surface larvae in the East Wind drift. The principal inferences to be drawn from the distribution of the surface Calyptopis population, a general view of which is given in Fig. 84, are as follows : (i) Although there are a few scattered and on the whole insignificant occurrences in November and December, the majority of them, including the larger, in Weddell West in December, the Calyptopes do not reach the surface on a major scale in this locality until rather late in January. They gradually spread to the eastward, reaching 30° W by the end of February. (2) About a month later they begin to appear in the surface waters of the East Wind drift, but it is probably not until March that the completion of the developmental ascent becomes general there. (3) By March in the Weddell drift they have spread to a position about half-way along Weddell Middle and by April cover the whole of the drift from west to east. (4) Their eastward spreading is due partly to the influence of the surface stream and partly to that of the eastward movement of the deep larvae carried in the bottom water below. It is unlikely, how- ever, that the influence of the bottom water as a larval carrier extends farther east than about 18° W. (5) The spreading may be accelerated or retarded largely owing to the complexity and vagaries of the triple current system in which the larval population spends its early existence. (6) The occasional presence of Calyptopes on the South Georgia whaling grounds and in the Bransfield Strait must be ascribed to the influence of the surface drift and not to any local large-scale hatching that takes place in either region. 37 322 DISCOVERY REPORTS Fig. 84. Gross distribution of the Calyptopes. Ice-edges March and September-October means. Large circles indicate catches of more than 10. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 323 (7) The May-June chart, more than any other in the vertical series, emphasises the primary role of the Weddell stream as the carrier and instrument of dispersal of these surface forms in the Atlantic sector. Its influence, however, seems to cease somewhat abruptly in the neighbourhood of meridian 30° E. (8) The Calyptopes have probably all but outgrown themselves in Weddell West and Weddell Middle by early May, but survive in considerable abundance in Weddell East into June. In the late- spawning, slow-growing region of the East Wind drift, however, they must survive, it seems, in great abundance into June, if not later. (9) The larvae seem to take about a month or more to accomplish the developmental ascent, and from the relative positions and ages of the swarms in the surface stream it would appear that the development from First to Third Calyptopis takes about the same time. (10) In its major aspects the distribution of these stages is essentially a feature of the East Wind- Weddell surface stream, the West Wind drift being barren, or virtually barren, except where it is affected by tongues of cold water emanating from the East Wind zone or by the penetration of bottom water into the Scotia Sea. The shallow living First, Second and Third Furcilias Before proceeding to deal with the distribution of the early Furcilias it might be observed that in the principal region of their abundance in the Weddell drift the distribution of the two groups of larvae so far described has conformed in each instance to a distinct pattern and time schedule, a westerly beginning in the early part of the season followed by a gradual spreading to the east as the season advances. To this pattern and schedule both groups of Furcilias, as they successively appear in the plankton, also conform, except that, as would be expected, their westerly beginnings happen later and later in the season. Jfanuary-February. Furcilias 1-3 then first appear in the plankton of Weddell West and the South Georgia whaling grounds in January. They are everywhere, however, very scarce, and as both Figs. 77 and 85 show it is not until late February that they have been recorded in this region in any measure of abundance. In this developmental phase they are now fast beginning to crowd the near surface stratum both by day and by night^ and in consequence are becoming more and more involved in the movement of the surface drift. Their presence in major concentration in longitude 30° W (Stations 361 and 362 in Fig. 77) can therefore, I think, be ascribed to risings that took place considerably farther west, such as that for instance, disregarding the date, recorded (Fig. 77) at Station 1965 in 47° W. The absence of comparable concentrations west of 30° W at this time (that is, in February) is perhaps a little surprising, but is possibly due to the chance failure of our nets to strike them. In the South Georgia area, where our observations are particularly numerous and closely spaced, there has evidently as yet been no large-scale incursion of early Furcilias borne upon the surface drift. The few young Furcilias recorded there, the majority of them in February, might therefore, it seems, represent the local growth of the scattered and, with one excep- tion, meagre Calyptopis population recorded there in January. As far as the Weddell drift is concerned it is, of course, only to be expected that in the first months of their appearance the early Furcilias should be found concentrated exclusively in the western part of the current, for it is there too that the surface Calyptopes are also concentrated (p. 310, Fig. 74) in the January-February period. The barren, or virtually barren, condition of the East Wind drift is also to be expected, for although there is a solitary occurrence there of a single First Furcilia on the northern outskirts of the Ross Sea (a late February record) it is obvious from Fig. 83 that the early Furcilias cannot normally make their appearance in these high latitudes until long after April. 1 See p. no, Fig. n in the section on diurnal migration. 37-2 324 DISCOVERY REPORTS 30° • • i-io ® • lO-lOO (•) ^ lOO-l.OOO '"^"^Fif/z :® ipoo-io.ooo rrr f r r ■ TT ' I ■ ^ > ■ I ■ I ■ f f ■ -f !■■ I ■ r^^ J 5^ I' I'' \mmM I \'IMM'lA'lMM' t'\' V\' V'T^' V ^' S' \ 'N' S 150° WISO'E 130° Fig. 85. Distribution of surface larvae. First, Second and Third Furcilias. January and February. Ice-edge February mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 325 12013 MAR APR ^^^ ,^ . . 1-10 ^^'f/'-^. ® • 10-100 (•) 0 100-1,000 (•) ^^ 1,000-10,000 ■ ^ '-■''>■' ^ ■ ^ ' /■ ' ' ^ " ' J ' ' ' ' ■ ^ ' I ' iri^i' I'l'i-ht'i'i'i'lnn'i'i'l' Piii---^^* I ' '•'• '' I ' '' I ''' ' ' ' ' ' ' ' ' ' ' M ' VM'\' ^ ' t ' V ■ \'\' V ' V ' \' V ■ V' ^ ■ k I ^ I V ■ V'l ^ ■ k ' ' ■ k v\ t ^-t-V 130° W 180°E 150° Fig. 86. Distribution of surface larvae. First, Second and Third Furcilias. March and April. Ice-edge March mean. 326 DISCOVERY REPORTS March-April. The distribution of the early FurciUas in March and April is shown in Fig. 86. It seems clear from this chart that many of the Calyptopis swarms in Weddell West, probably all in fact of those first encountered there in January, have now outgrown themselves, and, having reached the early Furcilia stage, have begun to spread eastwards in the surface drift reaching in March a point about a third of the way along Weddell Middle. Throughout April the eastward movement continues until, towards the end of the month (Fig. 82), it would appear that the whole Weddell drift from west to east, but most notably Weddell West and Weddell Middle, is carrying these stages in great or very great abundance. There is again, however, evidence (Fig. 86) of a seemingly abrupt termination of the larval drift somewhere in the neighbourhood of meridian 30° E. PETER 1 KEBGUELEN- OUT SIDE EAST PETER I CAPE ADARE- LOCALITY ISLAND -GAUSSBERG WIND DRIFT ISLAND -BALLENY IS, LOCALITY COLD TONGUE COLO TONGUE I30°E COLD TONGUE COLD TONGUE MONTH MAR APR APR MAY JUN JUN MONTH DAY of MONTH 8 14 27 27 28 24 24 DAY of MONTH STATION 2280 383 861 887 888 2843 912 STATION ADOLESCENT 1 ADOLESCENT FURCILIA 6 Scale per cent 1 ■ 6 FURCILIA 5 4 1 1 1 0 50 100 1 ,, J- 1 f 5 4 3 J. J.H 3 2 1 w ■ 2 CALYPTOPIS 3 2 ■t 1 1 — n 0 50 100 1 1 1 0 so 100 1 I 1 0 so 100 3 CALYPTOPIS 2 1 DEEP LARVAE ■i 1 - - - - - 1 DEEP LARVAE Fig. 87. Developmental condition of the larvae in the West Wind drift in localities affected by surface emanation from the East Wind zone. It is not of course to be supposed that the gradual populating of the Weddell stream from end to end with these stages is wholly brought about by the eastward movement of larvae sprung from risings in Weddell West. Clearly if this were so the great distance to be covered would involve them in a development beyond the early Furcilia state long before they could reach Weddell East. In so far at least as the far eastern reaches of the current are concerned it is more likely that the early Furcilias encountered there, spring, like the Calyptopes before them (p. 313), from risings in Weddell Middle, such as that, for instance, recorded in March (Fig. 76, Station 1144) in 18° W. The local growth of the dense concentrations of Calyptopes which in March and April (Fig. 75) are scattered throughout the Weddell stream must also be involved in the gradual west to east populating of this great current with the young Furcilia forms. Elsewhere the East Wind drift continues to be barren, or virtually barren, of these stages and there have evidently as yet been no large-scale influxes on the South Georgia whaling grounds, thoroughly sampled though they have been. Some minor concentra- tions, however, both in March and April, have evidently drifted (p. 315) into the Bransfield Strait. The March-April chart for the early Furcilias provides two instances of how certain isolated and probably strictly circumscribed areas of the otherwise barren West Wind zone may occasionally become populated with substantial or moderately substantial larval communities through the influence of cold tongues of surface water emanating from the East Wind drift. The more conspicuous of these is provided by the April plot south-east of Kerguelen, the other by the April plot in the southern part of the Drake Passage, the northward deflection of the East Wind water responsible for these occur- rences being influenced in the first instance by the Kerguelen-Gaussberg submarine ridge and in the second by the ridge Deacon (p. 315) supposes must exist to the westward of Charcot Island and upon which Peter I Island probably stands. It is to be noted, however, that these Furcilias were from the same two stations and samples as the Calyptopes referred to on p. 315. The developmental condition of the larvae in these localities (Stations 383 and 861) is shown in Fig. 87 and it is somewhat HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 327 surprising to find that in both instances they are considerably more advanced than any that have so far been recorded in April (p. 321, Fig. 83) from the East Wind drift itself. The most likely explanation of this anomaly is that at some earlier stage in their development, probably shortly after they reached the surface as First Calyptopes, they got carried away from the comparative darkness of their ice-infested, diatom-poor southern habitat (see p. 351, Fig. 104) and, reaching a region of higher photosynthetic values and better grazing, outgrew their contemporaries in higher latitudes. These East Wind deflections could in fact be described as miniature Weddell drifts in which the larvae develop at a rate broadly comparable with that in the major current. Apart from these isolated occurrences, it will be seen that the principal concentrations of the early Furcilias in March and April are confined exclusively to the main east-flowing Weddell stream. Outside this current, in the West Wind region of the Scotia Sea south of the Antarctic convergence, there are three minor concentrations to be noted, two in March and one in April. Clearly again (p. 316) these must have come from the west and in view of their age it is possible that they might have had the same East Wind origin as that ascribed to the April concentration in the Drake Passage. May-June. The distribution of the early Furcilias in May and June (Fig. 88 ), like that of the Calyptopis stages for the same period, again emphasises (though from the same stations) the import- ance of the Weddell drift as a primary carrier of the larvae and a major instrument of their dispersal. At the same time it is again worth noting that the pronounced local abundance of young Furcilias revealed in this figure cannot, it seems, be other than real. For although varying observational density in given areas may often provide a false impression of the true relative abundance of a species within such areas, in this instance it obviously cannot, for as Fig. 88 shows our coverage of the circum- polar sea for May and June is not only reasonably uniform (excluding the ice-covered East Wind zone) but is everywhere so uniformly meagre that the heavy concentration of larvae we have recorded in Weddell East, where virtually every surface net haul is positive, postulates a real and not merely an apparent abundance there. Although the Calyptopes in Weddell East (p. 317, Figs. 79 and 80) survive in some measure of abundance into June, the early Furcilias are now the dominant stage, a dominance indeed only to be expected with the former fast outgrowing themselves and shortly about to disappear from the plankton of this region. It seems fairly certain, too, having regard to the overall developmental condition of the Weddell swarms in April (p. 321, Fig. 82), that had we observations to show it, the whole of the Weddell stream from west to east would now be found to be carrying larvae pre- dominantly in the early Furcilia stage. It is probable, too, that these stages have by now arrived in mass on the South Georgia whaling grounds, for although our vertical observations it is true are too few to show this conclusively, in four out of five horizontal towings made off the east side of the island in late May 1927 Hardy and Gunther (1935, Appendix 11) record 'Cyrtopias' (= possibly Furcilias) in moderate to substantial numbers. In the ice-covered East Wind zone on the other hand, it is unlikely as explained on p. 323 that the young Furcilias can yet have appeared on any substantial scale. It is interesting, however, to record that among the material brought back by the * Belgica ' expedition which wintered in the pack south- west of Graham Land in 1898, Hansen (1908, p. 7, PI. i) reports that on 4 May a 'good number' of larval E. superba, which from his excellent figures are clearly Furcilias 1-4, were taken in a plankton haul in 70° 33' S, 89° 22' W, a position located well inside the East Wind drift and very close to the point where it breaks away to form the Peter I Island cold tongue. This record is of great interest, since it provides the only evidence we have as to what the real autumnal condition of the larvae in these high latitudes might conceivably be. It need not, however, mean that these, mainly early, Furcilias were at that time either dominant or even abundant there. A ' good number ' does not suggest 328 DISCOVERY REPORTS .,■'.'■',. N . S . r60 80 120 W 180-E 150° Fig. 88. Distribution of surface larvae. First, Second and Third Furcilias. May and June. Ice-edge May mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 329 Is T^ I /. ■I \- ISO" W 180°E ^' k'V V ' M' \ 'N' ^ ' ^ -'^ . ^ .\---r 150° Fig. 89. Gross distribution of the early Furcilias. Ice-edges March and September-October means. Large circles indicate catches of more than 10. 38 330 DISCOVERY REPORTS a particularly large one and dominance or otherwise could only be established if the relative abundance of the Calyptopis stages, then unknown to Hansen, had itself been established at the same time. In fact all that can be concluded from this isolated yet highly important observation is that in this particular locality, and perhaps elsewhere in the East Wind drift as well, the early Furcilias do begin to appear in some, perhaps small, measure of abundance round about the beginning of May. Since the first substantial gatherings of early Furcilias were recorded at Stations 361 and 362 (p. 314, Fig. 77) on 25 February and the earliest substantial rising of First Calyptopes was recorded at 'Norvegia' Station 42, 11° farther west, on 18 January, there would appear to be some grounds for supposing that in the Weddell drift the growth from the First Calyptopis to the First Furcilia stage takes upwards of 40 days. The principal facts relating to the distribution of the early Furcilia stages, a composite view of which is given in Fig. 89, are then as follows: (i) They first appear in small numbers in Weddell West and on the South Georgia whaling grounds in January, but it is not until late February that in the former locality substantial numbers have been recorded. (2) Now crowding the near surface stratum, they gradually spread eastwards in the surface stream, covering the whole of the Weddell drift from west to east by the end of April. (3) The gradual populating of the Weddell stream with these stages is brought about partly through the eastward movement of larval risings that take place in Weddell West, partly through the eastward movement of similar risings that take place in Weddell Middle and partly through the local growth of the Calyptopis population. (4) The May-June distributional chart again emphasises the role of the Weddell drift as the principal carrier of the larvae and instrument of their dispersal in the Atlantic sector. Its influence is again seen to cease rather suddenly in the neighbourhood of 30° E. (5) In May and June the whole of this surface stream is probably carrying larvae predominantly in the early Furcilia stage. (6) While there is some drifting of Furcilias 1-3 into the Bransfield Strait in March and April it is probably not until May that they begin to reach the South Georgia whaling grounds in any sub- stantial measure of abundance. (7) In the late-spawning, slow-growing region of the East Wind drift there are virtually no early Furcilias to be found from January right through to April and, judging from the backward condition of the swarms encountered there in late April, it seems likely that even throughout May and June they are still far from abundant. (8) There is some evidence that in the Weddell drift the growth from the First Calyptopis to the First Furcilia stage takes upwards of 40 days. In the East Wind zone this development, it seems, must be spread over a much longer period. (9) In its major aspects the distribution of the early Furcilias, like that of the Calyptopes, is essentially a feature of the East Wind- Weddell surface stream, the West Wind drift being barren, or virtually barren, except again where it is affected by tongues of cold water emanating from the East Wind zone. Such larvae as do get carried north in these tongues evidently encounter better feeding than they do in higher latitudes and, developing more rapidly than their contemporaries in the East Wind drift, quickly outgrow them. (10) Finally, it will be noted how both slow growth and ice conditions in the East Wind zone combine to render the existence of the early Furcilias in the higher latitudes of Antarctica virtually impossible to disclose, leaving us with the impression that up to the end of June these stages are confined almost exclusively to the lower latitudes affected by the Weddell stream. 331 HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL The shallozv living Fourth, Fifth and Sixth Fiircilias It has already been shown (pp. 108-13, Figs. 9-1 1) that once the larvae have accomplished the developmental ascent they show an increasing tendency as they get older to become massed both by night and by day very close to the surface of the sea, Furcilias 4-6 in particular (Fig. 11) evidently being so close that the uppermost (50-0 m.) vertical net, in which these stages are almost exclusively taken, passes it seems only for a second or two through their zone of concentration, capturing in consequence as a rule only small or negligible numbers. In other words, the vertical net can no longer be regarded as an efficient sampler of the larvae at this stage of their development and, therefore, the charts for Furcilias 4-6, while giving a reliable enough picture of their horizontal distribution and its dynamics, provide only an inadequate idea of the actual abundance of these stages during their life in the plankton. A true picture of this abundance can now in fact only be obtained from the stramin nets in which as will be shown presently (p. 366, Figs. 1 15 and 116) the late Furcilias are taken in large, often enormous, numbers. SECTOR 60W WEDOELL WEST LONGITUDE DATE O/^IA/ WEDDELL 1,1^1- 30W^,^;,,, WOE SECTOR 59°5S'S2'5l"49°49°4«'.4tf43°4f4d"38'34"3l°3l"30°* I 27'27°27 26' 23 1 8' 12 30 9 22 8 9 8 3025 21 26 2OI92I0I89 2 89 124 5 647 639 636 637 2298 2300 373 1990 1987 365 369 1142 1328 2293 638 W527 375 374 1138 372 368 1994 1144 LONGITUDE DATE * SOUTH GEORGIA Fig. 90. Developmental condition of surface larvae in the Weddell drift in March. March-April. They first appear, as Figs. 90 and 91 show, in the second half of March, again, like the Calyptopes and early Furcilias in January and February, in Weddell West. Except for one instance, they seem everywhere to be rather scarce and while this might be expected in the first month of their appearance it must in fact be only partly real, our horizontal data showing that at the close of the month the late Furcilias, especially Furcilia 5, may sometimes be present in Weddell West in enormous numbers. In other words this apparent March scarcity must be ascribed in part at least to the low sampling power of the vertical net. In April there is some indication of a spreading from Weddell West into the western part of Weddell Middle, although alternatively the three April occurrences in Weddell Middle might be ascribed to the local growth of a very small percentage of the substantial early Furcilia population (Fig. 90 and p. 325, Fig. 86) recorded there in March. Weddell East is barren except for two very small occurrences, both in late April. Both (p. 321, Fig. 82, Stations 1353 and 2344) can be ascribed to the local growth of the early Furcilias already recorded there. It is somewhat surprising, having regard to the position in March, that there are no even minor occurrences of late Furcilias in Weddell West in April, for as Fig. 82 shows the principal larval concentrations then consist in the main of early Furcilias or younger stages. Two alternatives might be put forward to explain this anomaly : (i) our nets simply through chance failed to strike the principal older concentrations or (2), and perhaps the more likely, our stations fell in a year (or years) when for some reason or other the larval development in Weddell West was more retarded than usual. It has already been shown (p. 319, Fig. 81) that there appears to be a direct correlation between the 38-2 332 DISCOVERY REPORTS 30° 60 9a 120 Fig. 91. Distribution of surface larvae. Fourth, Fifth and Sixth Furcilias. March and April. Ice-edge March mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 333 1^ • I ' .''' I i ■ ' ' ! ' / , I'l'l'.' J ■'■!'! ■/'>■>'"> -l-l'I'I'I'I'KI'I'l'l'in'l' 150° ' 1 ' I ■! M M M ■ I ■ \ M ■ \ ' I ■ t '\ M ■ \ ' \ ■ \ ■ \ ■ ; ' V M -T^ ^ M ■ t I V ' \ ■ ( ■ ^ A ■ ^ ''^v ' V-^-^ W 180*E 150- Fig. 92. Distribution of surface larvae. Fourth, Fifth and Sixth Furcilias. May and June. Ice-edge May mean. 334 DISCOVERY REPORTS annual volume or activity of the bottom water flowing away from the Weddell Sea and the annual abundance of the deep larvae that are encountered during the spawning season. It is interesting, therefore, to record that (see again Fig. 82) Stations WS 197, WS 199 and WS 200 fell in 1928 and Station 133 1 in 1934, both years in which, as Fig. 81 shows, the activity of the bottom water was apparently weak and the average catch of the deep larvae low. It may be, therefore, that the apparent inactivity of the bottom water in these years resulted in the arrival of the deep larvae in Weddell West at a later date than usual and that in consequence the development to the late Furcilia state was correspondingly late. Elsewhere, outside the Weddell zone, the East Wind drift is without trace of a late Furcilia popula- tion. Nor indeed could one be expected, since in these high latitudes, even in late April (p. 321, Fig. 83), the larvae have still, it seems, not outgrown their Calyptopis phase. In the otherwise barren West Wind drift the occurrence of very small numbers of late Furcilias may be noted at two stations (both in April, one of them again at Station 861) south-east of Kerguelen, and the occurrence of a minor (also April) concentration noted in the West Wind region of the Scotia Sea. In view of its relatively great distance from the Weddell drift, to the western part of which the late Furcilias are now principally confined, the Scotia Sea occurrence seems something of an anomaly. It might, how- ever, like the Bellingshausen Sea-Drake Passage occurrences discussed on p. 315, have been of East Wind origin, having sprung perhaps from a rising that took place somewhere to the south of Peter I Island. In the heavily sampled region round South Georgia and in the equally heavily sampled Bransfield Strait there has evidently been some little growth of the few early Furcilias already recorded there, but as yet no large-scale incursion of late Furcilias into either locality. Since the earliest substantial rising of First Calyptopes was recorded on 18 January (p. 314, Fig. 77) and the first substantial numbers of Furcilia 5 recorded between 18 and 26 March (Fig. 90), the development from the First Calyptopis to the Fifth Furcilia stage in the Weddell drift would appear at its fastest to be accomplished in from 60 to 70 days. From the results of the plankton investigations of Heegaard (1948), in the Gullmar Fjord, Sweden, it appears that the corresponding development in Meganyctiphanes norvegica takes from 42 to 49 days, a distinctly faster rate than I seem to find in E. siiperba influenced perhaps by the fact that his specimens were taken in considerably warmer conditions (4° C or higher) than are encountered in summer and autumn (p. 80, Table 6) in the Weddell drift. May-June. The distribution of the late Furcilias in May and June is shown in Fig. 92, the occurrences, all but one of very small numbers, emphasising, like the Calyptopes and early Furcilias from the same stations, the importance of the Weddell stream and the seemingly abrupt termination of its influence eastwards of 30° E. The principal concentrations of larvae in Weddell East, which is virtually the only sector of the current in which we have observations at this time of the year, are still (p. 318, Fig. 80) mainly in the early Furcilia state, or younger, and as yet only a very small per- centage of individuals has moulted and advanced beyond the Third Furcilia stage. We have no May-June observations in the Weddell stream farther west. The developmental condition of the swarms encountered there in March and April, however (p. 321, Fig. 82 and p. 331, Fig. 90), suggests that both in Weddell West and Weddell Middle the late Furcilias would now be found in great abundance. They must also by now I think have arrived in quantity on the South Georgia whaling grounds, for although our vertical nets (for May) reveal only negligible numbers there, Hansen (1913) in June 1901 and again in June 1902, records 'numerous' to 'very large numbers' of E. superba larvae, which Rustad (1930) suggests were very probably late Furcilia forms. In the East Wind drift the occurrence of a single Fourth Furcilia may be noted off Enderby Land. It is likely, however, that under the now extensive ice-sheet there the principal concentrations HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 335 of the larvae must be still (p. 321) mainly in the Calyptopis stages or at most (p. 330) in the early Furcilia state. The West Wind drift, otherwise barren, is again seen to be affected in two localities by cold East Wind tongues, one in the neighbourhood of the Balleny Islands, the other near Peter I Island. In both localities, although the vertical gatherings are very small, our stramin nets (p. 363, Fig. 112) show a moderately substantial late Furcilia population was present, and in both the advanced developmental condition of the swarms recorded (p. 326, Fig. 87), relative to the probable cor- responding condition (p. 327) of their contemporaries in higher latitudes, is again to be noted. We have no May-June observations south-east of Kerguelen. The developmental condition of the swarms recorded there in April, however (p. 325, Figs. 86 and 87), suggests that there too a moderately sub- stantial population of late Furcilias would now be found. July-December. The distribution of the late Furcilias in winter (July-September) and spring (October-December) is shown in Figs. 93 and 94. The low sampling power of the vertical net is now becoming increasingly evident, especially during the last three months of the larval existence, when, as Fig. 94 shows, the scatter of the positive occurrences is wide and the occurrences them- selves only of small or very small numbers. It must, therefore, again be emphasised that during the final phases of the larval life our vertical samples provide primarily a picture of the distribution and its dynamics and only an inadequate conception of the relative abundance of the young krill before they finally outgrow their larval state. The facts of the distribution themselves it will be seen are now much obscured by the vast extent of the winter pack, for now, even in the Weddell drift, formerly wide open, they can only be determined with certainty for a relatively narrow ice-free zone in the north. In fact beyond this northerly strip, all that is left open of a vast area of larval abundance, of the other formerly accessible krill-rich waters, the South Georgia whaling grounds alone remain ice-free and open to investigation, the Bransfield Strait being closed until November and the East Wind drift until January. The periodic covering and uncovering of such an enormous area of the circumpolar sea by ice must raise many questions in Antarctic oceanography to which the answers may for long remain obscure, and perhaps the answer to the overriding question as to what in fact does happen to the vast population of larval krill which now apparently exists, much of it in semi- or near-total darkness, below the sea ice, may in the end only be provided by the establishment of elaborately equipped winter stations on the pack itself. Both America and Russia have already done this in the Arctic by air (Tolstikov, 1957; Armstrong, 1958), but on the highly mobile Antarctic floes^ it might prove (Debenham, 19306) a more hazardous and difficult undertaking. It might, however, be done by freezing in powerful vessels of the ice-breaker type. In any event, as Robin (1959) has said, a wide unexplored field now beckons to Antarctic oceanographers, for as yet we know virtually nothing about life below the polar pack during the southern winter. In the meantime there must inevitably be much speculation and much, and perhaps wrongful, assumption. In so far, however, as the East Wind zone is concerned it now seems likely that wintering vessels, perhaps of the amphibious landing-craft type, could do useful work in the extensive areas of open water that from time to time it seems (p. 318, note i) appear in the coastal waters there. ^ They are particularly active in the Weddell Sea. I quote from Hurley (1925). 'Though sailing west during the day the currents had carried the ice-floes on which we had rested during the night swiftly to the east. Not only had we lost all the distance sailed but the drift had actually gained thirty miles on our efforts! ' Worsley (1931), again, refers to the embarrassing situations that can arise through the movements of bergs travelling faster through the pack than the pack itself is moving, 'ploughing' he says through the floes as though they were so much tissue paper. Or to quote John (1934) on a mile long berg I once watched driving through the pack many years ago, 'in doing so it was closing up and "pressuring" the pack in a wide arc before it, and charging through it, leaving a broad wake behind'. It was travelling at about i\ knots. In the Arctic basin on the other hand the ice is probably not moving faster than from i to i\ sea miles per day (Dunbar, 19556). 336 DISCOVERY REPORTS / . ^ ■ ,/ i-n • • • l-IO ® ® • lO-lOO 100-1,000 1.000 - 10.000 ; I ;■,;.,,, I, ,y,, .,,,,,.,.,,,,,,.,,/,,.,,,,,, ^.,,,i.,,,,/,|,|,,,|, I |,|,, 1.|.|.|.mI M I . I ■ I M ■ I M ' \ ■ 1 ■ 1 ■ I ■ 1 ■ \ ■ ^ ■ V ^ ^^^^^ i.\. t' ^.y. V ■ s A . ^ .\ ■ ISO" W 180"E 130° Fig. 93. Distribution of surface larvae. Fourth, Fifth and Sixth Furcilias. July-September. Ice-edge July-August mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 337 1 "^. -AT^Ci-^i c : OCT NOV DEC ^oPj^P/^QK^S 7 ] ® ® ;1 ® m 11® • ' " o l-IO 10-100 100- 1,000 1.000-10.000 *'!^? ' ^'^ ' ' .' ' .' ■-' ■ / .' '; 'f'T--f- 'i'TT'-i-'i •l-i'fi' fl'i'i'\ 'r-t-n (/, ■I'I'IM'k\ '''-''!'-''■'' ■ ■ \ ■ \ ■ \ ' \ ' \ ■ \ t ■ \ ^ ■ \ ' V ' ■. ' ^ ■ \ ' ISO" W 180°E ISO- Fig. 94. Distribution of surface larvae. Fourth, Fifth and Sixth Furcilias. October-December. Ice-edge late December mean. 39 338 DISCOVERY REPORTS While then the relative abundance of the older Furcilias in this late phase of their development cannot adequately be determined from the gatherings of the vertical nets, it is clear from their distribution that throughout winter and spring, the Weddell drift, and the South Georgia whaling grounds to which it spreads, are the principal carriers of these stages in the ice-free north. The Bransfield Strait, it seems, might also be a carrier, and perhaps an equally important one, the oc- currence of very small numbers there in November (Fig. 94) suggesting that throughout winter at any rate the late Furcilia population in this locality might be fairly substantial. Except for the occurrence of a single Sbcth Furcilia in September, again it is interesting to note in the region south- east of Kerguelen affected by cold surface emanation from the East Wind drift, the July-December vertical nets reveal an absence of late Furcilias throughout the West Wind zone. This, however, in part at least, is only an appearance that springs from the low sampling power of the vertical net when used for gathering these predominantly surface-living forms. Our larger stramin nets (p. 369, Fig. 117), with their far greater sampling power, reveal, in spring at any rate, not absence but great scarcity. Although we can now do little more than guess at what is happening in the East Wind drift, it seems probable that under the ice-sheet which now extends so far to the north of this high southern region of larval abundance, the late Furcilias (p. 334) do not make their appearance on a major scale until some time after June. In any case it must be assumed they develop slowly, attaining the adolescent state considerably later than their contemporaries in the ice-free, better illuminated, conditions of the northern zone. In fact judging from the overwintering of some of them, in the Sixth Furcilia stage, until well into the following year (p. 371), and from the exceptionally small size of the individuals comprising many of the adolescent or mixed larval and adolescent swarms encountered in these high latitudes from January to March (p. 355, Fig. 107), it is possible that in the East Wind zone the Sixth Furcilia in many instances remains dominant, possibly up to the end of December. The only December observation we have in these high latitudes (Fig. 107, Station RS9) does in fact show it so surviving. Finally, turning again to the Weddell drift, it may be noted that while the pronounced scarcity of late Furcilias recorded in September (Fig. 93) must be attributable in some degree to the low sampling power of the vertical net, the equally pronounced scarcity recorded in December (Fig. 94) is only to be expected since it is then (p. 339) that the last Sixth FurciUas surviving in this surface stream moult and become adolescent.^ The developmental condition of the larval, or mixed larval and adolescent, swarms encountered in the ice-free Weddell zone from July to December is shown in Fig. 95. In this presentation of the data the graphs portraying the percentage stage frequency are based in the vast majority of instances on the analyses of the abundant material from the stramin nets, the vertical nets as a rule (p. 331) providing only negligible samples. They are based principally on the larval and adolescent stage frequency determinations of Eraser (1936, Appendix i) but include many similar determinations by other members of the Discovery staff. By mid- July then, it will be seen, the Weddell drift from end to end is carrying larvae mainly in the Fifth and Sixth Furcilia stage, the latter predominating, although the former may also occasionally be encountered (Stations 2863, 2864) as the dominant, if only slightly dominant, stage. A few Fourth Furcilias still survive and in several instances it will be seen the Sixth Furcilia has already moulted and become adolescent although as yet, apart from one instance (Station 2854) only relatively small numbers would appear to have progressed beyond the final larval stage. In the South Georgia area, had we July observations to show it, a similar position would doubtless be disclosed. Throughout August ^ It should be recalled, however, that the 'Terje Viken' whale stomach records (p. 138) point to the occasional, if perhaps exceptional, survival of the Sixth Furcilia in Weddell East in some considerable measure of abundance until the second half of December. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 339 the Sixth FurciHa is clearly dominant and adolescent stages, although proportionately few, are much more frequently encountered than in July. The Fourth and Fifth Furcilias survive but are now becoming scarce. Rarely, however, even in late August, the Fifth Furcilia (Station WS 263) may be encountered as the dominant stage. In September the Sixth Furcilias and adolescents are found in roughly equal propor- tions, the adolescents, however, now slightly preponderating, and the Fifth Furcilia has practically LOCALITY DATE STATION ADOL FURC WW |w m|w e| w m |w e|w m 10 M \2 13 13 IS 16 16 2853 2854 2B57 2365 2859 2863 2372 2864 WEDDELL EAST SOUTH GEORGIA IB 19 21 22 22 24 24 27 28 28 28 28 28 29 29 2394 2398 2405 2406 2408WS256 2412 WS259WS260WS26IWS262WS263WS264WS266WS267 Scale per cent o so too LOCALITY DATE STATION ADOL FURC MONTH SEPTEMBER OCTOBER -First half MONTH LOCALITY SOUTH GEORGIA | W W | WEDDELL EAST W E 1 S G 1 WEDDELL WEST |W M LOCALITY DATE 4 13 17 17 17 17 IB IB 19 25 25 29 30 1 335556668 DATE STATION WS276WS279WS280WS2BIWS2a2WS283WS285WS286WS28B243l 2432 2443 2445 2447WS294WS295WS299WS300WS3OIWS302WS303WS304WS3IO STATION ADOL FURC 6 5 lillfTlflTT^^ --TTITTI*! ADOL 6 FURC 5 3 2 1 III III 0 so 100 o so lOO c 1 II III so 100 0 SO 100 3 2 1 LOCALITY DATE STATION ADOL FURC 6 5 4 3 2 I OCTOBER -Second half WEDDELL EAST WEDDELL MIDDLE 16 17 18 19 20 21 23 25 26 26 27 28 28 28 29 31 453 454 455 459 460 461 462 463 464 465 2465 467 2467 2468 469 2475 ■PTTTTtTTTfTITTT* Scale per cent nn» NOVEMBER- First half W m|S G|W E| S G~ 471 472 2478 4BI 483 TIITT MONTH LOCALITY DATE STATION ADOL 6 FURC 5 4 3 2 I DATE STATION ADOL FURC 6 5 4 3 2 NOVEMBER- Second halt SOUTH GEORGIA SOUTH GEORGIA 15 16 17 17 19 21 22 22 22 23 23 25 26 27 29 7 9 II II 19 I 484 487 488 491 494 500WS487WS488 502 503 507 511 517 518 523 * 2513 *- 527 537 539 STATION S G W E S G LOCALITY TTffTTTAT-T Scale per cent W W . WEDDELL WEST WM WEDDELL MIDDLE WE WEDDELL EAST SG SOUTH GEORGIA BS BRANSFIELD STRAIT * SAMPLE FROM SIGHTED SWARM Fig. 95. Developmental condition of larval or mixed larval and adolescent swarms in the surface waters of the ice-free Weddell zone in winter and spring. disappeared. In Weddell East, however, it appears that the Sixth FurciHa may occasionally be en- countered dominant not only up to the end of September but even into early October (see Stations 2443, 2445 and 2447) and this (see below), it seems, may be a phenomenon peculiar to the far eastern reaches of the current. In October and November the swarms are predominantly adolescent. The Sixth Furcilia survives in a varying measure of abundance throughout both months and may occasionally, although rarely, as for example at Stations 2447 and WS 488, be encountered as the dominant stage. It becomes very scarce, however, shortly after the middle of November and by December the last surviving larval stages have for all practical purposes^ outgrown themselves and reached the adolescent state.^ 1 In Weddell West a single Sixth Furcilia was in fact recorded in the 50-0 m. vertical haul at Station 760 in December 193 1. ^ See, however, p. 338, note i. 39-2 340 DISCOVERY REPORTS With reference to the occasional dominance of FurciHa 6 in Weddell East that persists throughout September and October, it will be seen on closer inspection of Fig. 95 that not infrequently the young swarms there, whether purely larval or both larval and adolescent, are younger, sometimes conspicuously younger,^ than many that are encountered at much the same time, or even up to a month or more earlier, in the more westerly parts of the current. Thus the Weddell East swarm at Station 2406 on 22 August is substantially younger than any of the three swarms recorded in Weddell West and Weddell Middle from 10 to 12 July, and the swarms at Stations 2405, 2406 and 2412 on 21, 22 and 24 August younger than the majority of those recorded on the South Georgia whaling grounds at approximately the same time. Again the Weddell East swarms at Stations 2443 and 2445 in late September and at Station 2447 in early October are conspicuously younger than those recorded from Weddell Middle, Weddell West and the South Georgia whaling grounds a few days later, and substantially younger than those encountered earlier on the South Georgia whaling grounds and in Weddell West from 4 to 19 September. The occasional persistence of the Sixth Furcilia in Weddell East as a substantial or even dominant component of the young winter and spring swarms can be traced up to the end of October and into November and, if the ' Terje Viken ' stomach records (p. 138) be taken into account, until the second half of December. There is a conspicuously young Weddell East swarm, for instance, at Station 2475 on the last day of October and another, less con- spicuous, at Station 2478 in early November. It seems that this persistently recurring phenomenon must be associated in the main with the late risings that take place in Weddell Middle, risings it will be recalled (p. 313, Fig. 76) which result in the arrival of a young Calyptopis population in Weddell East as late as the middle of April. It must be associated, too, with the late risings that occasionally, although apparently rarely (p. 314), take place in Weddell East itself, such as that for instance recorded (see again Fig. 76) at Station 2346 in late April. Finally, although it has already been stated that the Sixth Furcilia becomes very scarce in the latter half of November we have recorded late in that month, at Station WS 488 in the Bransfield Strait, one instance (Fig. 95) where it has in fact persisted as the dominant stage. This is perhaps not surprising, for in November in this southern channel, from which the ice has so recently cleared away, the conditions, in so far as surface temperature goes (p. 82, Table 8), approximate closely to those of the higher latitudes of the East Wind drift, where as already noted (p. 286, note i) the Sixth Furcilia may survive over December into March of the following year.- The principal facts then relating to the distribution of the late Furcilias, a composite view of which is given in Fig. 96,^ may be summarised as follows: (i) The vertical net is not a very efficient sampler oflhese stages because of their intense surface crowding and while it provides useful information as to the limits of their distribution, an adequate idea of their relative abundance can only be provided by the stramin nets. (2) They first appear in the latter part of March in Weddell West, spreading into the western part of Weddell Middle by April. In May and June they are probably spread across the whole of Weddell West and Weddell Middle in large numbers. In Weddell East, however, although small, for the most ^ Perhaps after 'younger' I should add 'or if not younger, conspicuously more backward in development'. ^ See also, however, p. 368. ' In view of the poor sampling power of the vertical net when used for the capture of these stages the insignificant catches from May (excepting May in the East Wind drift) to November have been regarded as having some significance and have been plotted as such in Fig. 96. In May and June, however, in Weddell East they may, as suggested in (2) above, have been given rather more significance than perhaps they deserve. The June West Wind occurrences north of the Balleny Islands and north-east of Peter 1 Island, although meagrely represented (p. 335) in the vertical nets, have been taken from the more substantial stramin net gatherings. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 341 ISO W180-E 150° Fig. 96. Gross distribution of the late Furcilias. Ice-edges March and September-October means. Large circles indicate catches of more than 10 (but see footnote 3). 342 DISCOVERY REPORTS part very small, numbers are encountered from late April to June, they evidently do not appear in any substantial measure of abundance until July. (3) While we have no evidence of any large-scale incursion of these stages into the Bransfield Strait or the South Georgia whaling grounds in March and April, it seems clear, particularly from the evidence of Hansen (1913), that in the latter locality at least they must have arrived in substantial numbers by about May. (4) The East Wind drift is barren of late Furcilias throughout March and April and it is probably not until some time after June that they first begin to appear there on any appreciable scale. (5) From July onwards the facts of the distribution can only be determined with certainty for such northerly areas of larval abundance, that is, the Weddell drift and the South Georgia whaling grounds, as continue to remain unaffected by the encroachment of the winter ice. In both regions the mass of ] the larvae in the plankton now consists exclusively of late Furcilia forms. By mid-July the Sixth Furcilia has become dominant and in general remains so throughout August until about mid- September. From then onwards the early adolescents become increasingly the more abundant, the j Sixth Furcilia becoming very scarce towards the end of November and finally disappearing from the plankton (of the northern ice-free zone) in December. Largely, however, owing to the late arrival in Weddell East of the Calyptopes that spring from the March risings in Weddell Middle the Sixth Furcilia may occasionally be encountered dominant in the far eastern reaches of the current through- out September and into early October. (6) The West Wind drift, as far at least as the vertical nets can show, is barren of late Furcilias j except again for such parts of it as are affected by tongues of cold water emanating from the East Wind zone. In such tongues Furcilias 4-6 are probably to be encountered most frequently and in greatest j abundance from May to September. (7) No late Furcilias have been recorded in the Pacific sector, though here it has been largely! the West Wind drift only that has been sampled in the months in which these stages might be expected. (8) Finally, it will again be noted how both slow growth and ice conditions in the East Wind zone I combine to render the presence of the late Furcilia population in these high latitudes virtually | impossible to disclose. Total eggs and larvae and total surface larvae The bi-monthly distribution and relative abundance of the total eggs and larvae (the latter com- 1 bining both deep and shallow living forms) from November to April is shown in Figs. 97-99, corresponding illustrations for the total surface larvae (the shallow living forms in the top 250 m.) from January to June being shown in Figs. 100-102. The distribution and relative abundance of] the total surface larvae for the periods November-December, July-September and October-December j (p. 309, Fig. 73; p. 336, Fig. 93 and p. 337, Fig. 94) have already been shown. These nine charts serve principally to summarise the more important features of the larval distribu- tion, and the factors controlling it, and in general to reinforce such conclusions as have already been I drawn. The following is a summary of the principal facts they reveal : (i) The early, November-December, beginnings of the larval life-cycle in Weddell Westj (Fig. 97), notably in the far western reaches of that sector of the current. (2) The improbability, clearly suggested by the concentration of the total surface larvae in Weddell j West in January-February (Fig. 100) that there can be simultaneous beginnings occurring all[ round Antarctica under the winter ice-sheet which in November-December still covers an enormous ' area of the circumpolar sea. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 343 30° 120 150° W 180°E 150° Fig. 97. Distribution of total eggs and larvae at all depths. November and December. Ice-edge late December mean. 344 DISCOVERY REPORTS ® • lO-lOO ""Jf^'^- (•) ^ 100-1,000 1.000-10,000 ' ■ t ■ t ■ I ' 1 ' ■ ! ' I '! ■! 'fni'i •' 'xt-r' I' ni'i-f'in-f\'^. ISO" W 180°E ^ ^ ' S ' t • t M ■ ^ \ ' ^ ■ V ' V ■ V ■ 0' V ' ■ ^ . V ■ \ ■ '. . ^ . \ . v^^^^-r-^ 150° Fig. 98. Distribution of total eggs and larvae at all depths. January and February. Ice-edge February mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 345 120; ISO" W 180°E 130° Fig. 99. Distribution of total eggs and larvae at all depths. March and April. Ice-edge March mean. 40 346 DISCOVERY REPORTS 120 130° W 180°E ''' I ''■!''' I ''I r\ MM ■ r t'\' tM ^M'X- VV M'T^ k ■ ^ ■ t ■ V ■'- I I ISO- Fig. 100. Distribution of total surface larvae. January and February. Ice-edge February mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 347 1,000- 10,000 ' f' I > / > > ' i ' t ■ ' ,' w M ; ' > w M I ' / ■ I ' J ■ 1 1 ; ■ f ■ r ■ I M ' I ■ r y»^; ■^j"-::'-^-^''- 60 i'i'\m'.m l'^■l.lM'^^'T^T~ v ^ \' \' \ ' v \ ' V ', ■ \ - ^ ' V 'N ■•v^ 'V 90 120 150° W 180°E 150° Fig. loi. Distribution of total surface larvae. March and April. Ice-edge March mean. 40-2 348 DISCOVERY REPORTS It ■,.,■,■ ■! !M^I•^^1^l^l^l•\■l^■lMM^^M^\'\■v-\'■^^^\'^^^^'^■^\^^'\■ '."^'';^f 'f 'J ' r ■ 150° W 180°E iso° Fig. 102. Distribution of total surface larvae. May and June. Ice-edge May mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 349 (3) The gradual spreading, during the next 4 months, of the larvae to the east, until by April the whole of the Weddell current, both above and below (Figs. 99 and loi), is carrying them in large or enormous numbers. (4) The later, February and March, beginnings of the main larval outburst in the coastal or near coastal waters of the East Wind zone (Figs. 98-101). (5) The seemingly abrupt ending of major eastward transport of larvae in the Weddell stream in about 30° E (Figs. 93, 94, loi and 102). (6) The larval barrenness of a vast area of the circumpolar sea, namely, the West Wind drift, except for certain circumscribed and widely scattered regions where it is affected (Figs. 93, loi and 102) by cold water deflected from the East Wind zone. (7) The virtual absence of eggs and of any larvae whatsoever from an enormous area of the West Wind drift in the Pacific sector (all nine charts). Gross distribution of total eggs and larvae Fig. 103 shows the gross distribution and relative abundance of the eggs and larvae where- soever and whensoever they have been encountered in the circumpolar sea. It uses the vertical net hauls of every southern voyage we have made since 1927,^ the massed observations of these many cruises emphasising again the overriding importance of the Weddell stream as a carrier of the larval krill, throwing into sharp relief the general barrenness of the West Wind drift, a barrenness extending up to particularly high latitudes in the Pacific sector, and demonstrating that all but a relatively insignificant part of the immense total larval population must pass the winter beneath the polar pack. Growth and dispersal in the surface drift The northern or Weddell zone. From the dates of the earliest substantial risings in Weddell West and the dates of the earliest major occurrences of the Sixth Furcilia in Weddell East, the only sector of the current in which we have observations covering every month of the year, some estimate can be formed of the time the young surface swarms take to complete their larval development.^ The first substantial rising recorded in Weddell West took place (p. 314, Fig. 77, Station N 42) on 18 January, the first Weddell East swarms with the last larval stage clearly dominant (p. 339, Fig. 95, Stations 2365 and 2372) being recorded on 13 and 16 July. This suggests that the purely larval developmental phase in the surface takes about 6 months to complete. There is, however, some uncertainty in this estimate. First, it is difficult to reconcile such a long period with the early appearance of the penultimate (Fifth Furcilia) stage which as we have seen (p. 331, Fig. 90) may be encountered already dominant in the surface swarms in the second half of March, only 60-70 days after the earliest Calyptopis risings. Second it is based on an assumption that can only partly be upheld, the assumption that the populating of the far eastern reaches of the Weddell stream is brought about exclusively through the eastward spreading in the surface drift of larvae that spring from risings in Weddell West. Taking these objections in turn, it is obvious that if we are to accept unreservedly such a long period as normal for the surface life-cycle of the larva it must be assumed that once it reaches its penultimate phase there must ensue a long and scarcely credible period, lasting several months, during which it remains dormant as the Fifth Furcilia. That some such period of dormancy or retarded growth might, ^ Augmented (p. 289) by those of the Norwegian and Australasian expeditions between 1927 and 193 1. ^ The earliest appearance of the Sixth Furcilia is in the second part of March (p. 331, Fig. 90) when very small numbers are recorded in the plankton of Weddell West. We are concerned here, however, with the major aspects of the larval development, in other words with the time when the Sixth Furcilia first appears as the dominant stage in the larval, or mixed larval and adolescent, swarms. 350 DISCOVERY REPORTS 30° ISO" W 180'E 130° Fig. 103. Gross distribution of total eggs and larvae. Ice-edge September-October mean. however, exist is not altogether impossible, for in the second half of March, as Fig. 104 shows, the volume of the phytoplankton in the Weddell drift, on which the larvae have hitherto been well nourished, decreasing suddenly, has practically reached its winter minimum, a major decline in food supply that could conceivably lead to a postponement of the final moult until later in the year. Evidence that there occasionally is such postponement is provided in Fig. 95 by three instances (Stations 2863, 2864 and WS 263) where in Weddell Middle and on the South Georgia whaling HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 351 grounds the Fifth Furcilia has remained dominant in the surface swarms until far into the winter months. It might be concluded then, that while there is rapid and continuous growth from January to about the middle of March, the winter slowing up of the larval development, culminating in the long drawn out existence of the Sixth Furcilia in the plankton, happens quite suddenly in the latter part of March, and begins it seems with the Fifth Furcilia stage. 3000 2500 ?; 2000- I500- 1000 O U 500- -WEDDELL DRIFT -E. WIND DRIFT I 1 1 1 1 — Jon Feb Man Mar: Apr May Jun Fig. 104. Phytoplankton values (plant pigments per cubic metre) in the Weddell drift and the corresponding values in the coastal region of larval abundance in the East Wind zone (after Hart, 1942). The second objection is perhaps of greater concern. It has already been shown (p. 313, Fig. 76) that the populating of the far eastern reaches of the Weddell stream, at the outset at least, is brought about partly through the eastward spreading in the surface drift of larvae that spring from March risings (Fig. 76, Stations 1994, 1142 and 1144) in Weddell Middle and partly, though less commonly, through risings that take place (Fig. 76, Station 2346) locally. If the March risings, then, be taken as the starting-point, the life-cycle of the larvae in the surface would appear to take roughly about 4 months to complete, the later April rising at Station 2346 suggesting it could con- ceivably be accomplished in three. Both estimates seem more reasonable than the original one, both being more in keeping with the speed with which the larvae earlier on pass from the First Calyptopis to the Fifth Furcilia stage and both still allowing for a period of dormancy in the latter which it seems might result from the early onset in March of winter or semi-winter phytoplankton conditions. Similarly, the July concentrations of Sixth Furcilias in Weddell West and Weddell Middle (p. 339, Fig. 95) would have their origin in March and April risings in Weddell West (p. 321, Fig. 82 and p. 331, Fig. 90), suggesting again that the surface life-cycle might be accomplished in from 3 to 4 months. Many of the larvae rising in Weddell West from January to April, all, that is to say, that do not get diverted to the South Georgia whaling grounds, but pass onwards in the main east- flowing surface stream, must eventually, of course, turn up in Weddell East, but in view of the distance they have to travel and the relatively short time it seems to take for the majority to reach the last larval stage, they probably do so as Sixth Furcilias of some considerable standing or mainly as young adolescents. The backward development of the young swarms that has repeatedly been recorded in Weddell East (p. 340) must clearly be aggravated by the fact that by the time the Calyptopes reach 0° in mid-April (p. 313, Fig. 76) the phytoplankton values (Fig. 104) have for all practical purposes already reached their winter minimum. In other words, as compared with the successive batches of Calyptopes that come to the surface in Weddell West from January to early March, when the phyto- plankton is at its summer maximum, the larvae in Weddell East, in so far as feeding goes, begin their 352 DISCOVERY REPORTS surface life at a marked disadvantage. However, from July, or at any rate August, onw^ards these indifferently nourished, slowly developing swarms will continually and increasingly be augmented by the arrival of older swarms from the west, swarms dominated by well-established Sixth Furcilias, or by both Sixth Furcilias and young adolescents, stemming no doubt from the early January-February risings that take place in the far western reaches of the current. Throughout winter, therefore, and probably into spring, we should expect to find in Weddell East two distinct larval communities, a younger representing the backward product of the March risings in Weddell Middle and the late April risings in Weddell East, and an older representing the forward product of the January-February risings in Weddell West. Moreover, with the older swarms continually coming in from the west we should expect also to find that in each successive month of their appearance they would be more prominently displayed in the western part of Weddell East, that is, near o°, than in the more easterly reaches of this sector. As Fig. 105 shows both expectations have in fact been substantially realised. Some estimate, too, can now be formed of the rate at which the larvae, developing as they go, traverse the Weddell drift from west to east. If longitude 45° W in 60° S be taken as the mean position around which they are rising in Weddell West from January to ApriV and 15° E in 60° S as the mean position in Weddell East around which the Sixth Furcilias, or Sixth Furcilias and early adolescents, are being encountered 180 days later from July to October, it could be argued that since the distance from 45° W to 15° E in 60° S is approximately 1800 miles, the larvae move eastwards at a rate of about 10 miles a day. A similar although slightly smaller figure is obtained if the March risings in Weddell Middle be considered in the same way. If, for instance, 15° W in 60° S be taken as the mean position for the March risings in this sector, and 15° E in 60° S the mean position of the first major concentrations of Sixth Furcilias recorded in Weddell East 120 days later in July, then since the distance to be covered is now 900 miles the larvae would appear to be moving at a rate of approxi- mately 8 miles a day. Taking a more concrete case from Fig. 76 (p. 313) in which we see a rising in Weddell Middle (Station 1144) in 18° W followed 40-42 days later by the appearance of surface swarms in 01° E dominated by the Second and Third Calyptopis, it would appear that, the distance from 18° W to 01° E in 60° S being approximately (p. 311, Fig. 75) 570 miles, the young krill on their moving belt travel eastwards at about 14 miles a day or say at the rate of about J knot. The average speed of the Weddell drift may in fact be somewhat higher than this. Sets of | knot (p. 314, note I ) it is true are common, but stronger sets of i, i J and even 2 knots have been reported, although the latter possibly may be of no more than local significance having been recorded perhaps only when strong winds were blowing the right way. It need not, however, be supposed that the larvae, planktonic though they are, move with precisely the same speed as the surface stream. It might in fact be expected that they would move more slowly since for all we know the vertical move- ments of the Calyptopis and Furcilia stages, somewhat restricted (p. 108, Figs. 9-1 1) though they seem to be, may lead to some slowing up of the larval drift relative to that of the surface stream. For although we know these movements take place for all practical purposes within the limits of the Antarctic surface layer it is extremely unlikely that the whole 200 m. thickness of this mass of water is moving bodily eastward at a uniform rate, the maximum rate of drift most probably being confined to a relatively narrow surface stratum, the deeper layers moving more sluggishly, the deeper the slower. Clearly then, such descent as the Calyptopes may undertake into the deeper water must lead to some slowing up of their resultant drift to the east. The Furcilia stages on the other hand, with their much more limited or negligible vertical movement (p. no. Fig. 11), possibly travel more swiftly eastward than the Calyptopis forms. ^ The actual mean position is 43° W in 59° S. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 353 There is also the possibiUty, and we cannot ignore it, that the larvae, to a certain degree at least, can resist the surface drift. Bishai (i960) has shown experimentally that 1-5 day old herring are not entirely passive drifters, that they respond positively to current, orientate themselves against it, and try to resist it by swimming upstream, keeping station for at least an hour in face of a current with velocity between 0-58 cm. /sec. (0-013 rnile/hr.) and 1-03 cm./sec. (0-023 rnile/hr.). At higher velocities they drifted with but not so fast as the current. Day old herring (average length 6-5 mm.)^ subjected to a current of 7-60 cm./sec. (about 4 miles a day) drifted along at an average rate of only 4-2 cm./sec. LONGITUDE ADOLESCENT FURCILIA 6 5 1. 4 3 2 I ADOLESCENT FURCILIA 6 5 I. 4 ■I 3 |> 2 I ADOLESCENT FURCILIA 6 5 M 4 " 3 M 2 " I ADOLESCENT FURCILIA 6 ■ ■ 5 4 ADOLESCENT FURCILIA 6 5 4 00° 00° 01° 02° 02° 03° 04° 05° 05° 05° 06° 13° 14° 15° 16° 17° 20° + JULY Scale per cent I I I ^ O 50 100 AUGUST +Tf^f Scale per cent TT Scale per cent "JH- SEPTEMBER Scale per cent TTTTTTT" ' ,L . I Scale per cent OCTOBER NOVEMBER T f LONGITUDE ADOLESCENT 6 FURCILIA 5 4 3 2 ADOLESCENT 5 FURCILIA 5 4 3 2 I ADOLESCENT 6 FURCILIA 5 4 " 3 " 2 " I ADOLESCENT 5 FURCILIA 5 4 ADOLESCENT 6 FURCILIA 5 " 4- Fig. 105. Developmental condition of the young winter and spring swarms in Weddell East, showing the swarms in the west, the younger in the east. older The southern or East Wind zone. Whereas in the Weddell stream the larval growth-rate can be followed with some degree of precision throughout the whole year, in the East Wind drift it can only be followed with certainty from January to April (p. 321, Fig, 83), the onset of winter conditions in May preventing further access, or at any rate further effective access,^ to these high latitudes for the rest of the year. For 8 months, from May to December, they remain largely frozen over and it is not until January of the following year, 1 1 months after the earliest recorded major East Wind rising (p. 310, Fig. 74), that the developmental condition of the i-year old swarms en- countered then permits some estimate to be formed of what the larval growth-rate below the winter ice might be, and of how long the young first year swarms in the extreme cold and prevailing darkness of their wintry environment exist in a purely larval, or principally larval, developmental phase. ' This is about the average length of the Second Furcilia o E. superha. ^ Certain narrow strips of the extreme northerly reaches of the East Wind zone are in fact occasionally accessible in May (p. 348, Fig. 102). Such parts as remain open, however, lie for the most part (p. 346, Figs. 100 and loi) outside the main East Wind region of larval abundance. 41 354 DISCOVERY REPORTS 30° • > ; , f '-rTi-r^T'/' > ' I 'T'J'f-t' 1 ' I ^ yj rT , • i 6 4 Tl @^ ^^ RS9 537 815 813 2547 2561 2562 2560 2004 2606 2603 260O 1154 1297 1361 W540W540 S W540 W538 643 2594 24 23 368 38 1781 I360 MAY JUNE ^ WEDDELL ZONE ■ EAST WIND ZONE ADOLESCENT 6 FURCILIA 5 " o) EAST WIND SWARMS DEVELOPING BELOW THE ICE STATION 52 48 44 40 36 32 28 24 20 16 12 8 4 Fig. 107. Developmental condition of the young first- and second-year swarms in the Weddell and East Wind zones, showing the backward state of the southern swarms following their prolonged 'wintering' below the polar pack. Note on vertical scale : specimens were measured to the nearest millimetre, the diagram showing the percentage length frequencies in 4-mm. groups, the figure in the vertical scale being the fourth in the group. Thus the block opposite ' 8 ' in the scale shows the proportion of specimens measuring 5-8 mm. inclusive (In later text-figures with 2-mm. groups the scale figure is the second of the group, e.g. 14 refers to 13 and 14 mm.). surface existence under such conditions, should from the outset (p. 321, Fig. 83) develop slowly or that II and even 13 months later, in January and March of the following year, the Sixth Furcilia (p. 372, Fig. 120) should still be surviving in these high latitudes, if only in rapidly declining numbers. Under the winter ice-sheet that envelops them from May to December the far southern larvae, it seems, continue to develop slowly, evidently not completing their surface life-cycle until long after this phase has been accomplished and passed in the less rigorous conditions of the northern zone, the backward state we have repeatedly recorded in high latitude 11-15 month old East Wind swarms in January, March and May^ (Fig. 107) providing the clearest evidence of how slow the winter growth-rate must be.^ Fig. 107 shows the developmental condition of typical young Weddell and East Wind swarms, over a period covering the winter and spring of the first year of their growth and extending into autumn of their second, the length frequency being shown in the lower part of the ^ We have no high latitude East Wind data for February and April. ^ As Thomas (1959) remarks, the vast ice-sheet that invests so much of the larval population throughout winter and spring can best be described as 'frozen solid', and where such conditions exist he adds it would appear that production of a basic holophytic food supply cannot take place because ' solar radiation cannot penetrate the solidified media in sufficient quantities per unit time'. Below the ice off Mawson in 64° E Bunt (1959) reports a virtual absence of phytoplankton all through the winter months. It would not, therefore, be at all surprising if it should prove that, in the East Wind zone as a whole, winter growth among the larvae came virtually to a standstill. It is difficult in fact to understand how the young krill, especially the East Wind krill, survive the polar night at all. Are they, as Marshall and Orr suggest (19536), carnivorous then? That they are is at least a possibility, Hart it will be recalled (p. 44) having recorded the frequent occurrence of large Foraminifera (Globigerina sp.) in the stomachs of post-larval specimens. The occasional presence of Foraminifera, Radiolaria and Tintinnids in the stomachs (p. 173) is also reported by Barkley. 41-2 356 DISCOVERY REPORTS diagram with the stage frequency, where determined, above. We have no precise stage frequency data covering this period from the East Wind zone. The Sixth Furciha, however, was recorded surviving there at Stations 2547, 2560, 2561 and 2562 in January, and at Stations 2600 and 2603 in March. It will be seen then that whereas in the Weddell swarms the peak length frequency or modal value in November (Station WS 487) stands at 17-20 mm., the corresponding value in the East Wind zone MONTH SEPTEMBER MONTH DATE STATION ADOLESCENT 5 FURCIUA 5 ,0 " 3 32 0 tr 30 0 28 2 26 > 24 22 20 UJ 0 18 z < 16 a. 14 I t- 12 0 10 DATE STATION ADOLESCENT RJBCIUA 6 32 ^ 30 O 28 O 28 28 29 29 29 W26I W263 W256 W267 W258 13 17 17 17 18 IB 19 19 4 4 4 5 5 5 6 5 W275 W279 W280W282 W283 W285 W286 W287 W288W295 W296 W297 W299 W300 W30I W302W303 IIITTIfllTTTTfTTI 26 5 2* ^ 22 ^ 20 - 18 S 16 5 14 °^ .2 X 10 « 6 i^ b Scale p«r cent O 50 lOO UUtjUtt+itt * * * * . O so 100 ****** MONTH DATE STATION ADOLESCENT FURCILIA 6 " g 30 " a 26 I 24 « 22 Z re " 16 < cr 14 12 J 10 g 8 UJ 6 6 7 7 8 W304 W305 W306 W3I0 Scale per cent o so lOO * * * * 22 23 23 25 26 27 29 472 481 483 484 487 488 491 492 494 500 502 503 507 511 517 518 523 ITTTT1TTTTT"T STATION ADOLESCENT 6 FURCILIA S ul 3 32 O 30 o 28 ? 26 i 24 z 22 20 UJ O IB z < 16 a 14 j- 12 z 10 UJ 8 -I 6 * SWARMS IN OR CLOSE TO ICE Fig. io8. Developmentarcondition of young first-year swarms on the South Georgia whaling grounds, showing the backward state of those encountered in or near the ice (For vertical scale see legend to Fig. 107). a month later (Station RS 9)^ is still only 9-12. It seems obvious, therefore, that since the Sixth Furcilia at Station WS 487 was represented by some 10% of the total larval and adolescent stock, this, the last larval stage, must have been surviving clearly dominant (cf . Stations WS 267 and WS 285) in the young swarm recorded at Station RS 9. And this being so it seems probable that, were it possible to penetrate into high East Wind latitudes to show it, the Sixth Furcilia would be found surviving in the young first year swarms pre-eminently dominant up to the end of November, and possibly still dominant, but surrendering its dominance to the adolescent, in December. It would seem, therefore, that in these high latitudes, taking the mid-February rising off the Princess Martha Coast (p. 310, Fig. 74) as the starting-oif point for the development, the life-cycle of the larva in the surface takes upwards of 9 months or more to complete. ^ This is the only observation we have as to the spring condition of the young swarms in the East Wind zone. It was obtained (p. 255, note i) from a factory ship. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 357 The prevailing presence of the pack and all it implies, extreme cold, low photosynthetic values and inferior grazing, is, I feel sure, the basic factor responsible for the slowing up of the growth-rate in the East Wind zone. It is interesting, therefore, to record that much farther north among the young swarms encountered in, at the edge of, or very close to the pack-ice during our winter and spring surveys of the South Georgia whaling grounds in August, September and October of 1928, a cor- responding slowing up of the growth-rate can readily be detected. The length and stage frequencies of the young swarms encountered then together with those of the November survey of 1930^ are shown in Fig. 108, the pack-ice or near pack-ice swarms being distinguished by an asterisk. From August to November it will be seen that whereas in the open sea there is a steady, if at first slow, upward trend in both growth and development, in or near the pack, there is a pronounced backward trend characterised by a marked lag in developmental condition and a general inferiority in modal length. Our knowledge of the speed of the East Wind drift is meagre. Weddell (1827) records a west by northerly set of 31 miles in 3 days in 72° 38' S, 35° W, McNab (1839) a westerly set of 12 miles a day near the Balleny Islands, Ross (1847) a south-westerly set of 7 miles a day in 68° 34' S, 12° 49' W and speeds of from 9 to 18 miles a day for the cyclonic movement in the Ross Sea. Bruce (1918) mentions north-westerly sets of from 10 to 12 and of from 12 to 18 miles a day near 67° S, 37° W, Fuchs (1958) reporting a west-north-westerly set of between 1 5 and 20 miles a day in approximately the same position. Near the Japanese base, Syowa, in 40° E, the westerly current is estimated to be travelling at some- thing over 9 miles a day (Fukuoka, 1959). In the Australian quadrant of the East Wind zone a strong westerly set of up to 15 miles a day was reported for two days following an easterly gale, the observa- tions of the 'Aurora', spread over three voyages, being 'overwhelmingly in favour of a westerly drift averaging 6 or 7 miles per day with a slight northerly component' (Mawson, 1940a:). It seems, how- ever, that it might be moving faster near the coast, in the narrow belt where the larvae are concen- trated, than farther out to sea. Westerly (or in the east Graham Land region north-westerly) coastwise currents setting at from 24 to as much as 72 miles a day- have in fact been reported (Donald, 1896; Antarctic Pilot, 1930; Deacon, 1937) from which it might be concluded that the coastal waters in these high latitudes are moving with a speed at least comparable with, if not faster than, that of the Weddell drift, and that the larvae they carry, westwards into the Weddell Sea and north-westwards out of it, move at much the same resultant rate, approximately 8-14 miles a day, as they do in the more northerly east-going stream. At 8 miles a day it might be expected that larvae having their origin in risings such as those, for example, recorded in 0° off the Princess Martha Coast at the beginning of March (p. 303, Fig. 71), moving coastwise round the Weddell Sea and developing as they went, would eventually turn up in Weddell West, a journey of approximately 2200 miles, in a matter of 275 days, or some 9 months after their original advent at the surface. During this time, if our estimate for the high southern larval growth-rate is correct, they would have developed to a dominantly Sixth Furcilia stage and it is interesting, therefore, to record that at the time and place expected, in November about half-way along Weddell West, we did in fact encounter two mixed larval and adolescent swarms in 1 Based throughout on the comprehensive measurements and stage frequency determinations of Fraser (1936, Appendix i). 2 Priestley (1923) reports even higher speeds off the Spit at Cape Adare, noting that in summer and autumn 'icebergs of some size come sweeping past the Spit at a rate which at the full of the tide may reach a maximum of 3 or 4 knots'. The current runs particularly strongly along the Ross Ice Barrier face, the 'Terra Nova' (Debenham 1923, end map 11) reporting a westerly set there of as much as 3 knots. Scott (1905) mentions a berg here that had travelled 70 miles to the west in 12 days, adding 'but the surface water had been moving at a greater speed, as we could tell by its effect on the ship'. Recent Japanese observations on the movements of Antarctic icebergs (Kumagori and Yanagawa, 19586) suggest that the westerly set experienced by Scott was probably of the order of | knot. 358 DISCOVERY REPORTS which the Sixth Furcilia if not dominant had, as the length frequencies (Fig. 109) in both instances show, clearly been in a position of dominance some very little time before. In this locality, and so late in the year, it is difficult or impossible to attribute the existence of swarms of such backward develop- mental stature to the January-April risings that take place in Weddell West itself, for the product of such risings would long since have passed on to the east, to the South Georgia whaling grounds and beyond. They must, in fact, have made their original ascent much farther upstream, and having regard to the probable duration of the East Wind surface Ufe-cycle, and the probable rate of transport in the west-going stream, the slope waters of the Princess Martha Coast are indicated as their most likely locus of origin. The young, dominantly Sixth Furcilia, swarm encountered in November in the SECTOR , ■ 0., and 60W WEDDELL WEST ,'o,„ SECTOR 30 W and MONTH NOVEMBER MONTH LONGITUDE 43° 46° LONGITUDE DATE 9 10 DATE STATION 1866 1858 STATION 24 24 uj 22 22 ^ g 20 20 " < 18 1 18 < 16 1 I 16 f 14 1 ff 14 I g ,2 f T ,2 'i y 10 T r-r-i 10 y 8 Scole pjr cent ° " '»» 8 SAMPLE 1 33 475 SAMPLE N9 MEASURED 1 33 475 NS MEASURED Fig. 109. Length frequencies of young first-year swarms in Weddell West in November. Bransfield Strait (p. 339, Fig. 95, Station WS 488) might also it seems have had a similar history, while the underdeveloped mixed larval and adolescent swarms recorded in the East Wind drift between 0° and 30° E in January (p. 355, Fig. 107, Stations 2547, 2560, 2561 and 2562) could well it seems have stemmed from the March risings that take place in the coastal waters of the Australian sector, such as that, for instance, recorded (p. 311, Fig. 75) at Station 1713 in 120° E. The distance coastwise from Station 1 7 1 3 to the mean position where these young January swarms were encountered is approximately 2500 miles, a distance which at a resultant rate of westerly drift of 8 miles a day the larvae would cover in approximately 10 months' time, long enough it would seem for them to have grown to a state in which, although the early adolescents were dominant, the Sixth Furcilia was still substantially repres ented. The principal facts relating to the growth and dispersal of the larvae in the surface stream may be summarised then as follows: (i) In the Weddell drift the development from the First Calyptopis to the Fifth Furcilia stage would appear at its fastest to be accomplished in from 60 to 70 days and the whole surface life-cycle, Calyptopis i to Furcilia 6, in from 3 to 4 months. (2) In the East Wind zone the development is much slower, the surface life-cycle it seems taking upwards of 9 months or more to complete. (3) The resultant rate of larval transport in the surface drift, both in the Weddell and East Wind zones, is probably of the order of from 8 to 14 miles a day. i The larvae, adolescents and adults from the oblique and horizontal samples The massed surface larvae Summer. The distribution and relative abundance of the massed surface larvae in summer (January- March), based on the data from the towed stramin (0-5 m. and loo-o m.) nets, is shown in Fig. no. As Eraser (1936, p. 138) remarks, the stramin net is selective in its fishing, giving a better HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 359 : ® lOO-l.OOO : ^ 1,000-10.000 20 I 18 " ,6 f 14 g 12 1^ lO B 6 N2 EXAM- - INED 75 30 79 47 25 35 24 67 75 40 41 42 56 63 73 41 32 38 28 75 70 62 61 48 69 35 20 21 34 65 26 33 25 60 57 60 44 37 27 36 68 62 36 34 30 38 39 27 31 NS EXAM- -INED Fig. 114. Length frequencies of larval and adolescent components of young winter and spring swarms showing heavy contribution of the last larval stage to the small (11-20 mm.) food of the whales. For vertical scale see legend to Fig. 107. mere separation into larvae and adolescents can be called staging), the length frequency of the sample has been presented graphically in every instance where the latter was determined. Both winter and spring diagrams then show not only the developmental condition of the surviving larvae, but also that of the small whale food, or 11-20 mm. class, to the mass of which the Sixth Furcilias provide a major contribution. The heavy extent of this contribution, especially from September onwards, is illustrated in Fig. 114, which, except for one instance (Station 2478), is based throughout on the measurements and larval and adolescent determinations published by Fraser (1936, Appendix i). In this diagram 49 young winter and spring swarms have been separated into their larval and adolescent components, the length frequencies of each component in 2-mm. groups being shown side by side for comparison. Only stations at which 20 or more individuals of each component were measured are shown. Besides showing the degree of overlap in length frequency of the two components Fig. 114 also illustrates how typically, notably in September and October, the Sixth Furcilia does not moult and become adolescent until it reaches a length of 1 1-12 mm. In fact, as represented here, the transition from last larval stage to adolescent is disclosed as a thing of some abruptness, losing, in the 2-mm. length grouping, some of the indefinite merging of one developmental phase into another upon which Fraser comments. For instance, in August the majority of the young individuals of 10 mm. and under, are larval, and the majority of those of 1 1 mm. and over, adolescent. Again, in September, October and early November there occurs repeatedly an abrupt step from 12 mm. and under, majority larval, to 13 mm. and over, majority adolescent, and in November (in some instances) a jump from 14 mm. and under, majority larval, to 15 mm. and over, majority adolescent. Both winter and spring developmental diagrams then are intended to serve a dual purpose, to be read first in conjunction with the winter and spring massed larval distributional charts and later in conjunction with the 1 1-20 mm. (small whale food) distributional charts which follow presently the massed larval series. 366 DISCOVERY REPORTS 30° o o • I-IOO ® lOO-I.OOO (^ 1,000-10,000 10.000-100,000 ■ - r' ■/■/■: • 'ir I VJ ISO" W180°E 150° Fig. 115. Distribution of the massed surface larvae in winter. Ice-edge July- August mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 367 We measured the young krill to the nearest milUmetre from the anterior margin of the eye^ to the tip of the telson, the vast bulk of the work being carried out daily at sea in the course of the routine examination of the samples from the stramin nets. To eliminate the effect of (i) personal errors intro- duced because the labour of measuring was shared by several (at least six) individuals, (2) errors arising through the variable contraction or distortion of the specimens in the formalin fixative, and (3) multiple errors arising through the very rapidity with which perforce so much of this enormous task had to carried out, the length frequencies are presented in 2-mm. groups. The number of specimens upon which each developmental graph (whether expressed as a length or stage frequency) is based is shown at the bottom of each seasonal diagram, both of which, although based largely on the field observations described above, include the highly accurate measurements (to the nearest ^ mm.) and larval and adolescent determinations of Fraser (1936, Appendices i and 11). JULY AUGUST SEPTEMBER STATION ADOL FURCILIA 18 16 14 12 10 B 5 N? EXAM- -INEO B 9 10 II II 12 12 13 13 15 16 16 [IB 19 21 22 22 24 2B282B 29 29 29 2 4 13 17 17 17 IB IB 19 19 25 25 29 30 2B5I 2853 2855 2B57 2859 2372 23942405240B W26I W264 W267 I403 W279W2B2 W285 W287 2431 2443 2852 2854 2856 2365 2863 2864 2398 2406 2412 W263 W266 W26B W276 W280W283 W2B6 W288 2432 2445 ■i*f%fi|ftkW™"if™f M Scale per cent M 41 103 33 51 58 75 80 47 40 32 no ISO 50 lOO 45 lOO 362 100 lOO lOO 100 62 192 56 100 lOO 186 100 52 59 59 50 lOO 100 lOO 405 90 ADOL 6 FURCILIA 16 14 12 10 N2 EXAM - - INED Fig. 1 1 6. Developmental condition of the massed surface larvae and small whale food in the 1 1-20 mm. range in winter. The distribution and relative abundance of the massed surface larvae in winter (July-September) is shown in Fig. 115. They are now, as Fig. 116 shows, exclusively late Furcilias, predomin- antly the Sixth, and becoming increasingly adolescent. As in autumn they continue to fill the Weddell drift and the South Georgia whaling grounds, for the most part in enormous numbers, and again, judging from our negative August line of observations between 30° and 60° E, there as yet does not appear to have been any major overflow of this teeming population eastwards of 30° E. The West Wind drift is barren except as before where it is affected by cold East Wind penetrations, namely, south-east of Kerguelen, north-east of the Ross Sea near 150° W (both September records) and again (possibly) in the southern part of the Drake Passage. Such observations as we have in the Pacific sector (in the West Wind drift only) are negative except as aforementioned in such small part of it as is affected by north-easterly outflow from the Ross Sea. As to what is taking place in the East Wind drift, now completely encompassed by the winter pack, further comment is unnecessary, the probable position there having already been outlined in the commentary on the winter and spring distribution of the larvae based on the data (pp. 334-8) from the vertical nets.^ The developmental condition of the principal concentrations of the massed surface larvae in winter, based throughout on our observations in the Weddell drift and on the South Georgia whaling grounds, is shown in Fig. 116. The major features of it may be summarised thus: (i) Substantial numbers of purely larval swarms (Furcilias 4-6) persist throughout July and August, giving place in September to older swarms consisting almost exclusively of Sixth Furcilias and adolescents. ^ We found it convenient, and time saving, to use this, rather than the distal end of the rostrum, as the anterior end of the body. In E. superba the rostrum projects forward as far as, but no farther than, the eye. In fixed specimens, however, the eye blankets the rostrum when the animal lies, as it will only do, on its side. " See also pp. 355-6. 368 DISCOVERY REPORTS (2) Whether among purely larval, or mixed larval and adolescent swarms, the Sixth Furcilia remains dominant almost to the end of w^inter, but in September is rapidly surrendering and occasionally losing its dominance to the early adolescent. (3) Exceptionally it seems the Fifth Furcilia may remain dominant until late August, a pheno- menon to which reference has already been made on pp. 349-51. (4) Fraser's conclusion (1936, p. loi) that 'the vast majority of the larvae reach the Furcilia 6 stage before the end of the southern winter', while correct as far as his material went, can now only be considered valid for the northerly region of larval abundance in the Weddell drift and on the South Georgia whaling grounds. In their high latitude region of abundance in the East Wind zone, from which Eraser had no material, clearly, it has now been established, they grow so slowly that vast numbers of them not yet in the Sixth FurciUa stage must persist at least throughout September and possibly (pp. 334-8) until long after. (5) Both larval and mixed larval and adolescent swarms display a heterogeneous pattern of modal values throughout the season. Spring. The distribution and relative abundance of the massed surface larvae in spring (October- December) is shown in Fig. 117. Except for a single instance in Weddell East (Fig. 118, Station 2447) where a few Fourth and Fifth Furcilias were recorded still surviving in early October, the surface larvae, everywhere throughout the ice-free zone, have reached the Sixth Furcilia stage and in this state survive until November by the end of which, in the Weddell drift at least,^ all have become adolescent. During this, the final phase of their existence in the plankton, they continue to be massed in considerable abundance throughout the greater part of the Weddell zone, although nowhere, except] for three instances on the heavily sampled South Georgia whaling grounds, in such enormous numbers as are encountered in winter. There is, however, a curious, largely barren, area in Weddell West, and two alternative explanations are suggested to account for it. Either (i) our nets through chance missed the larger mixed larval and adolescent swarms or, (2) and perhaps the more likely, the region is one from which by springtime the young purely larval swarms encountered earlier in the year have already passed onwards in the surface stream to the South Georgia whaling grounds and beyond. The isolated minor occurrence of the Sixth Furcilia that appears about half-way along this section of the drift, the plot representing the larval component of a young mixed larval and adolescent swarm (Fig. 118, Station 1868) recorded there in November, need not necessarily, therefore, have been of local origin. It could probably in fact be traced to a rising that originally took place in the East Wind drift, possibly, as suggested on p. 357, off the Princess Martha Coast. Similarly, the November swarm with a dominant larval component encountered in the Bransfield Strait (Fig. 118, Station WS 488) could, in view of its backward developmental state and position at the extreme western end of the channel, also have got carried there from the East Wind drift, being probably traceable to the north- ward deflection of East Wind water that occurs near Peter I Island. Between 30° and 60° E, it is evident that there has now been some minor overflow of young spring swarms from the richly populated Weddell drift. Elsewhere the West Wind drift is largely barren of larval stages except again south-east of Kerguelen where there is a conspicuous instance of larval encroachment from the East Wind zone, the major occurrence of the Sixth Furcilia recorded there (Fig. 118, Station 1640), an early December record, suggesting that in isolated regions such as this the young first year swarms diverted from the East Wind zone, although developing considerably more rapidly (p. 327) than their contemporaries in higher latitudes, grow less quickly than, and in consequence do not, in general, become exclusively adolescent until later than, the swarms that are carried in the Weddell stream. All our observations in the West Wind region south-east of Kerguelen 1 See again, however, p. 338, note i. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 369 ® 100-1,000 (^ 1,000-10,000 ifi) 10,000-100,000 • /• > !■! ■fl-I ■ - " ■ / M . I i M I ■ < ■ J i I ' I ■ I I / ' I i I ■ I ' J ' / ' n-M--- ISO" W 180"E 130° Fig. 117. Distribution of the massed surface larvae in spring. Ice-edge late December mean. 43 370 DISCOVERY REPORTS from September to December (Fig. 119) combine in fact to support this view, the young swarms recorded in this locality without exception showing a distinct tendency to grow more slowly than their contemporaries in the Weddell zone and to attain complete adolescence at a somewhat later date. And so it would appear that in the purely larval developmental phase there are three distinct growth- STATION ADOL FURC 6 " 5 ., 4 If) a 34 2" U30 2 28 ^" ~22 ui 020 < IS fM il2 8 N2 EXAM- -INED I 23 3 3444 SSS 566657789 10 MI2I6 17 18 19 2020 2I 23 25 26 26 26 26 28 2828 29 31 I 12 2 2 2447 W294 W295 W297 W299 W30I W302W304W305 W310 1835 1837 454 459 1846 452 454 2455 457 2458 2475 471 I003 2478 1821 1823 W296 1824 W300 1325 W303 1826 W306 1832 1836 453 455 460 461 463 2463 455 2467 469 ^^1 Tft Tir Til in flvff ■ w/Mf ra Scale per cent O so 100 Q so 100 . . 236 24 lOO 71 lOO 61 lOO lOO lOO 55 538 lOO lOO 100 57 100 103 608 97 39 73 58 291 lOO 488 lOO 321 77 224 64 50 lOO 50 lOO 45 50 48 48 52 30 IIO lOO 121 3IO lOO 66 STATION ADOL 6 FURC 5 ■ 4 " u) 34% 32 g 30-^ 28 2 26 ra 20 o 18? 12 i 10 y 8 r«EXAM- -INED STATION ADOL FURC 5 " 5 2 4 5 9 10 1314 15 16 17 17 18 19 19 21 21 21 22 22 22 22 23 23 24 25 25 25 26 26 26 27 27 28 28 29 29 30 2479 I009 1668 483 487 491 494 W4e6 209I W4B8 2093 507 2IOO I039 1628 518 1632 523 2116 1856 1655 481 484 488 492 I029 500W467 502 503 1624 517 I04I 2103 I630 2I09 2113 '■iPPffllPmFTIIPT"""TT"fl'TlfT^ir DECEMBER I 3 4 7 7 8 9 19 19 30 2119 2128 2137 2513 539 I640 2136 2139 537 W496 STATION ADOL 5 FURC 5 • Fig. ii8. Developmental condition of the massed surface larvae and small whale food in the 11-20 mm. range in spring. For vertical scale see legend to Fig. 107. rates to be considered, an extremely slow one in the East Wind drift, a somewhat faster, although still rather slow, one in the north-going outflows from these high latitudes, and a rapid one in the more northerly latitudes affected by the Weddell stream. Farther east there are two further instances of East Wind influence in the West Wind zone, namely, north-east of the Ross Sea (November records) and west of Graham Land (October and November records). The eastern half of the Pacific sector, well covered by our observations, is barren or virtually barren of larval stages, but again all samples are outside the East Wind drift. The developmental condition of the principal concentrations of the massed surface larvae in spring is shown in Fig. 118 which, although based primarily on our pre-eminently abundant material I HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 371 from the Weddell drift and the South Georgia whaling grounds, covers also the region immediately to the east of the Weddell stream between 30° and 65° E (swarms distinguished by a dagger), the West Wind region south-east of Kerguelen (swarms distinguished by an asterisk) and includes several stations in the Bransfield Strait and west of Graham Land (swarms distinguished by two asterisks) where the larvae encountered were probably (p. 371) of East Wind origin. Briefly the major features illustrated are the following: (i) The growing dominance of the early adolescent over the Sixth Furcilia encountered in the Weddell zone throughout October. MONTH DATE STATION ADOL FURC 6 .. 5 40 m36 i32 <28 24 ^20 il6 y 12 8 4 9 10 17 17 19 25 26 26 27 28 29 I 3 3 4 7 7 7 B 9 II 2879 2B80W282W283 !l029 511 517 1039 518 1632 52312119 1640 I540 2128 S 2136 2137 2139 2513 S L .11. . .^/ N2 EXAM -INED SEPTEMBER NOVEMBER DECEMBER T"TTIT=^TT MONTH DATE STATION ADOL 6 FURC 5 " 40 36 UJ 32 i 24 20? 16 i 12 ui 18 100 100 20O 32 lOO 197 74 1570 100 143 1637 86 55 57 87 266 189 99 S = SAMPLE FROM SIGHTED SWARM Fig. 119. Developmental condition of young swarms in the West Wind region south-east of Kerguelen compared with the corresponding condition in the Weddell zone at the same, or approximately the same, time. (2) The absolute dominance of the early adolescent over the Sixth Furcilia encountered in the Weddell zone in November and the virtual disappearance of the latter from the plankton before the end of the month. (3) The survival of the Sixth Furcilia in small numbers between 30° and 65° E in November and in the West Wind region south-east of Kerguelen, in one instance (Station 1640) in large numbers, into early December. (4) The overall backward condition of the November-December swarms of East Wind, or prob- able East Wind, origin as compared with the corresponding condition of many of their contemporaries in the Weddell stream. Summer. The continued survival of the Sixth Furcilia in the higher latitudes of the East Wind drift, but nowhere else, over spring and into summer (January-March), 11-13 months after the first major risings (p. 310, Fig. 74) take place in this coastwise stream, is shown in Fig. 120. The numbers that so survive it will be seen are negligible. The gross distribution of the massed surface larvae is shown by seasonal symbols in Fig. 121. In order to bring the inadequately sampled First Calyptopis (p. 360) into the distributional picture and further to bring together on a single page as much relevant information as possible, the data from the vertical as well as from the stramin net hauls have been used in the construction of this figure throughout. The seasonal symbols employed have been used only where gatherings of major signifi- cance have been recorded, hauls of less than 100 in the stramin nets and of less than 50 in the three uppermost vertical nets being distinguished by small black plots wherever they were encountered throughout the year. The principal features thus illustrated are the following: (i) The great summer outburst in Weddell West which by autumn has involved the whole of the Weddell zone and continues to do so throughout winter into spring. 43-2 372 DISCOVERY REPORTS Fig. 120. Distribution of the Sixth Furcilias that survive from the spawnings of a year or more before. Ice-edge February mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 373 Fig. 121. Gross distribution of the massed surface larvae. Ice-edges February, May and July-August means. 2^4 DISCOVERY REPORTS (2) The absence of any major overflow of this teeming population, as larvae, into the West Wind region east of 30° E. (3) The existence, at all times of the year, of discrete pockets of larval abundance in the otherw^ise virtually barren West Wind drift that springs from the intrusion of cold surface tongues carrying larvae from the East Wind zone. (4) The barrenness, almost absolute, of the heavily sampled West Wind zone of the Pacific sector and of such northerly parts of the East Wind drift as have also been sampled there. (5) The high latitude region of larval abundance in the East Wind drift detectable, owing to the ice conditions there, only in summer and autumn. (6) The vast area of the principal region of larval abundance in the East Wind-Weddell surface stream which in the depth of winter is completely encompassed by the polar pack. The small whale food In the construction of the winter and spring charts illustrating the distribution of the 1 1-20 mm. class, the larger Sixth Furcilias which (p. 365) contribute heavily to the small whale food population in September, October and November, and upon which the whales (p. 138) undoubtedly sometimes feed, have not been included. The charts in consequence deal exclusively with krill, of the 11-20 mm. class, that have definitely been recognised as adolescent. The inclusion of the Sixth Furcilia would have involved overlap and confusion in the presentation of the distributional data, the obscuring (see below) of some important aspects of the distribution itself and in any case, our catch-figures show, have resulted in no more than a stepping-up here and there of some of the lower orders of abundance of the 11-20 mm. class without, however, materially affecting the major facts of its distribution gathered from the adolescent material alone. Winter. The obvious starting-off point for plotting the distribution of the small whale food is winter, for it is winter which marks the beginning of the post-larval life when the Sixth Furcilias surviving from the previous summer's hatchings first begin to moult and become increasingly ado- lescent. Its distribution and relative abundance during this period is shown in Fig. 122. By now, everywhere throughout the ice-free zone, the young winter swarms, consisting essentially of mixtures of larvae and adolescents, and exhibiting considerable variation both in length frequency and range (p. 367, Fig. 116), in most instances fall within an overall maximum range of 8-16 mm. All, or practically all, in addition to their purely larval component, include krill of the 11-20 mm. class, both larval and adolescent, and it is therefore not surprising that in its major aspects the winter distribution of the latter should follow closely that of the younger, purely larval, developmental phase (p. 366, Fig. 115) from which it stems. In fact the only essential point of difference between the winter distribution of the massed larvae and that of the 1 1-20 mm. adolescents, to which the former are giving rise, is that whereas the larvae are spread along the Weddell drift from west to east in more or less uniformly enormous numbers, comparable concentrations of adolescents appear exclusively in the western part of the drift with only moderate, diminishing to minor or insignificant, concentrations in the east. Having regard to the winter distribution of the massed larvae, which are no less abundant in the east than in the west, the relative scarcity of the early adolescent population recorded in Weddell East shows that in their development the young winter swarms as a whole are more advanced in the western than in the far eastern reaches of the drift, ^ a phenomenon that can doubtless be traced to the backward condition of the larval and mixed larval and adolescent swarms we have repeatedly recorded eastward of 0° (p. 339, Fig. 95 and p. 353, Fig. 105) from July onwards. 1 A fact which would have been obscured if the heavy winter concentrations of large Sixth Furcilias in the 11-20 mm. range had been used in the construction of Fig. 122. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 375 ® loo- 1,000 1000-10,000 M) 10,000-100,000 V ' w w :•!•/> Ki'iii't' I'l'i-i ■ f ■ r-^TT-n-^ 150° \V 180°E 150° Fig. 122. Distribution of the small whale food in winter. Ice-edge July-August mean. 376 DISCOVERY REPORTS Except where it is affected locally by East Wind influence, namely, south-east of Kerguelen, north- east of the Ross Sea and again possibly in the southern part of the Drake Passage, the West Wind drift is to all appearances barren of whale food in the 1 1-20 mm. range. In the East Wind drift, although the position is obscure, it seems probable (pp. 355-6) that the young swarms survive the winter in a purely larval state and that the early adolescents in the 1 1-20 mm. range, already well established in the Weddell zone, do not appear until later. Turning again to Fig. 116 (p. 367) in which the developmental condition and length frequencies of the young winter swarms are shown, it will be seen that as might be expected they develop an increasingly dominant 1 1-20 mm. component as the season advances, a component rather infrequently dominant in July, more frequently dominant in August and exclusively dominant in September. Spring. The distribution and relative abundance of the 1 1-20 mm. class in spring is shown in Fig. 123. By October the young swarms, everywhere throughout the ice-free zone, develop length frequencies which for all practical purposes (p. 370, Fig. 118) fall exclusively within the 1 1-20 mm. range, and spring in consequence becomes the season when this group of very small (larval and early adolescent) whale food is at the peak of its abundance. It is the season too par excellence of the early adolescent when the last surviving larval stage, the Sixth Furcilia, becoming scarcer and scarcer, finally disappears from the plankton. ^ Vast numbers of the 11-20 mm. class are now found on the South Georgia whaling grounds, in Weddell Middle and Weddell East. There is again, however (see p. 369, Fig. 117), a conspicuous scarcity of these young swarms in Weddell West, a scarcity that as already suggested could be attributed to the fact that the still younger swarms, so abundant in this region earlier in the year, have already drifted out of it by spring, moving eastwards in the surface stream. The two minor concentrations of the 1 1-20 mm. class recorded in Weddell West (Fig. 118, Stations 1866 and 1868) together with the minor to moderate concentrations recorded in the Bransfield Strait could, therefore, it seems (p. 368), be of East Wind rather than of local origin. In the West Wind drift there is more evidence of the easterly overflow of the young swarms that (p. 368) evidently takes place from the richly populated Weddell drift in November, the three minor adolescent concentrations plotted between 30° and 65° E representing the 11-20 mm. components of three mixed larval and adolescent swarms that it seems must have drifted away from the region of rich larval and post-larval abundance west of 30° E. Elsewhere in the West Wind drift the influence of surface deflection from the East Wind zone is to be seen in the moderate to substantial population of early adolescents that has been recorded (in November and December) in the region south-east of Kerguelen, while farther east, between 180° and 150° W, there is another, although much less substantial, population of these young stages that has evidently been carried north-eastwards out of the Ross Sea. Several very small occurrences appear too, west of Graham Land, in the region affected by the north-going East Wind flow that occurs near Peter I Island. For the rest the barrenness of the heavily sampled eastern half of the Pacific sector is again noteworthy. In the East Wind drift, judging from the position there in January (p. 355, Fig. 107), it seems likely that the small whale food is now beginning to get well established and that towards the end of spring it might be present as a substantial or even dominant component in the young mixed larval and adolescent swarms in these high latitudes. Turning again to p. 370, Fig. 118, it will be seen that the principal features of the develop- mental condition of the young spring swarms, in so far at least as their 11-20 mm. components are concerned, may be summarised as follows: (i) Throughout October, whether they be of purely larval, mixed larval and adolescent, or purely adolescent stock, they fall fairly, and for all practical purposes exclusively, within the 1 1-20 mm. class. 1 That is, from the plankton of the ice-free northern zone, not however (p. 371), from the plankton of the East Wind drift. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 377 oO lOO- 1.000 (^ I.00O-I0.O0O (fi) 10,000-100.000 ' -■ ' ■■■ ■/-■/-/ •/'/■/■/'/'/' 150° W 180°E 150° Fig. 123. Distribution of the small whale food in spring. Ice-edge late December mean. 44 378 DISCOVERY REPORTS 50° 4'J-^i'i.j'i.i>i.i,hJ M. I. y, I, /,,.■..,/./,/,.,/, -^ 30° / , GO 9CF 120 ® loo-ipoo ® I.OOO-IO.OOO ^P) 10,000-100,000 •/ ri-rrt ■ '■■!./.] ,..,.<.,, mm, ■/■ttttt^t^ ISO '■'■■■'•'■\'"' '■'■'■■■'■'■'' W 180*E \ '■''■■ ^ ' ^ '^ ■ ^ ' ^ ' ^ ■'■■■■■' ■ 150° Fig. 124. Distribution of the small whale food in summer. Ice-edge February mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 379 (2) In November, a period of increased feeding and growth, while the majority still fall mainly within the 11-20 mm. range, all but a very few have developed, notably in the second half of the month, a more or less pronounced, and in some instances dominant, over 20 mm. component. (3) In December they tend to fall mainly within the range of the over 20 mm. class, except in water of East Wind origin where they are inclined to be backward and are still, as in November, mainly inside the 11-20 mm. range. (4) Throughout the season they display in their length frequencies a heterogeneous pattern of modal values. Summer. The distribution and relative abundance of the 11-20 mm. class in summer (January- March) is shown in Fig. 124. Taking first the position in the northern or Weddell zone, it is evident that there is such rapid growth among the young, now one year old, swarms in December that by January the small 1 1-20 mm. class, so prominent in spring, virtually everywhere outgrows itself and, except for rare and for the most part negligible instances to the contrary, virtually everywhere disappears from the plankton. In other words, in the northern zone generally — along the whole of the Weddell drift, on the South Georgia whaling grounds and in the Bransfield Strait — there has been almost total transition of these early adolescent forms to a whale food of the staple (over 20 mm.) size. There are, however, two instances in Weddell Middle where transition has not been so complete as else- where. Both were recorded in late January 1931 near a large body of pack which at that time, as Mackintosh and Herdman's ice-chart (1940, PI. Lxxxiii) shows, lay far to the north of its summer mean. It seems possible then, that being located in a region from which the winter ice had evidently but recently cleared away, the young yearling swarms recorded there had developed more slowly (p. 356, Fig. 108) than their contemporaries in warmer ice-free water and so had survived the spring in a condition more commonly encountered (p. 355, Fig. 107) in higher latitudes. It is the East Wind drift, where the growth-rate (p. 358) is so slow, that is now virtually the sole locus of abundance of the 1 1-20 mm. class. In these high latitudes the young first year swarms, having spent the better part of 9 months under the winter ice, are found, now as swarms of a whole year's standing, still in a very backward state (p. 355, Fig. 107), many of them with a strong to dominant 1 1-20 mm. component, a condition in which some of them may persist even right through to the end of summer.^ These backward swarms, as Fig. 124 shows, appear to be particularly abundant in the Atlantic sector, the most conspicuous of them in the highest latitudes near the continental land. Throughout the West Wind drift the scarcity of small whale food is most pronounced, especially in the closely sampled Pacific sector. Autumn. The distribution and relative abundance of the 11-20 mm. class in autumn (April-June) is shown in Fig. 125. By the end of summer, it is clear, the yearling swarms have practically everywhere outgrown their early adolescent state with the result that throughout the ice-free zone small whale food in the 1 1-20 mm. range virtually does not exist in the plankton of the autumn months. Of the four very small occurrences recorded, two significantly are in the cold East Wind drift, both May records, the third, an April record, in the almost equally cold Bransfield Strait and the fourth, also an April record, far south in Weddell East. Autumn, then, is the season when the now 16-18 month old swarms belong almost exclusively to the staple (over 20 mm.) class. Judging, however, from the very backward condition of a young summer East Wind swarm recorded near 70° S in the Atlantic sector on 17 March (p. 396, ^ Since in many instances the young summer yearling swarms, particularly in the East Wind drift, consist essentially of mixtures of small (11-20 mm.) and staple (over 20 mm.) whale food, the detailed description of their developmental condition is deferred until the section dealing with the summer developmental condition of the staple size with which it will be seen the corresponding condition in the 11-20 mm. range is closely linked. Meanwhile see p. 396, Fig. 136. 44-2 38o DISCOVERY REPORTS 30° I Fig. 125. Distribution of the small whale food in autumn. Ice-edge April mean. I HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 381 Fig. 136, Station 2004) it seems probable that under the autumn ice-sheet that now extends over so much of the East Wind zone some whale food in the 1 1-20 mm. range must survive at least into April. The gross distribution of the small whale food is shown by seasonal symbols in Fig. 126, the symbols employed again representing gatherings of not less than 100 in the straminnets, catches of less than that figure being distinguished by small black plots throughout the year. The principal facts of the distribution thus presented are briefly as follows : (i) The conspicuous massing of the 11-20 mm. group that is found in the main east-flowing Weddell stream, on the South Georgia whaling grounds and in the Bransfield Strait throughout winter and spring. (2) Its virtual disappearance from the Weddell zone in summer and its continued survival then in the East Wind drift. (3) Its absence, virtually complete, everywhere from the ice-free parts of the circumpolar sea in autumn. (4) The minor degree of the easterly overflow that takes place from the Weddell drift in spring, an overflow chiefly affecting the West Wind region between 30° and 60° E. (5) The extensive area, curiously barren of this class, that appears in Weddell West in spring, a phenomenon probably attributable to the fact that by this time the majority of the young swarms encountered there earlier in the year have been carried eastwards out of it in the surface stream. (6) The moderate to substantial local populations of this class that in winter and spring have been recorded in the West Wind region south-east of Kerguelen and north-east of the Ross Sea, populations that owe their existence to northerly outflows stemming from the East Wind drift. (7) The barrenness elsewhere of the West Wind zone, especially noticeable in the heavily sampled Pacific sector. (8) The very large area covering the southern part of the Weddell drift from about 30° W to 30° E which appears conspicuously barren simply because the vast majority of the observations that cover it were made, and can in fact only be made, at a time (that is, in summer and autumn) when the early adolescent whale food, in this particular area at least, does not, or virtually does not, exist in the plankton. The whole of this region from July to December would in fact be carrying a dense popula- tion of the 11-20 mm. class which like the larvae (p. 373, Fig. 121) would be under the winter ice. A more realistic representation of the distribution and regional abundance of the small whale food during the pelagic whaling season is obtained if, as in Fig. 127, the summer and autumn condi- tions, using monthly symbols, are shown together on a single chart. This brings into sharp relief the great scarcity of this class that is found in the Weddell zone throughout summer and autumn and the relative abundance in which it persists from January, through February, until March in the East Wind drift. Finally, before proceeding to deal with the third and last section of this series, the staple whale food, this is perhaps the most appropriate juncture at which to present in a more connected, and it is hoped, more realistic manner than has hitherto been possible, the major facts of the distribution and life- history of the young first year swarms in the Weddell zone, the only region of euphausian abundance that remains open to investigation throughout every month of the year. The facts it is true have already been given, although up till now, owing to the nature of the distributional plan, in a somewhat scattered and disconnected form. The chief object of this fresh approach to the problem, which is based essentially on a more comprehensive grouping of the monthly data than has hitherto been adopted, is to show at a glance how the surface life-cycle of the larvae in this northern zone begins and ends in the Weddell drift, how the influence of this great current, as a purely larval carrier, seems 382 DISCOVERY REPORTS oO .; 1/.^ , ,|>,. ,/ J .; I,,,,/,, ,,,,,,, ,/,,,,,,,,,/,,,,,,,,,/,,,,,,,, 150° W 180°E l'l'''IM'\'tMM'l'\MM'l'lAM'l ■^■^'^■^■^' t'T^ ^ ■ V M ■ V ■ \ ' M v ■ \ . k A-^-r-^ ISO" Fig. 126. Gross distribution of the small whale food. Ice-edges February, May and July-August means. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 383 ISO W 180*E 130° Fig. 127. Distribution of the principal concentrations of the small whale food in summer and autumn, showing its marked abundance in the East Wind zone during the pelagic whaling season. Ice-edge February mean. 384 DISCOVERY REPORTS to cease rather abruptly east of 30° E, and how such overflow as takes place beyond this meridian is an overflow primarily of adolescent forms chiefly affecting the region between 30° and 60° E. Thus Fig. 128 shows how the young surface swarms, at first purely as larvae, gradually spread along the Weddell drift from west to east between December and April, Fig. 129 this movement in greater detail and Figs. 130 and 131 how as larvae and early adolescents, the former from May to November, the latter from July to December, they remain heavily massed throughout the entire Weddell surface stream without undergoing any major onward movement to the east except as adolescents in spring. The spring manifestation of the easterly influence of the Weddell drift, which is most conspicuous between 30° and 60° E,^ may be a phenomenon associated with a massive release of cold water resulting from the breaking up and melting of the enormous ice-sheet it carries through- out winter and early spring. The spring break-up, and subsequent melting and dispersal of the Weddell pack from the South Sandwich Islands eastwards takes place with great rapidity. I quote from Hansen (1934): 'Before melting, the zone of pack-ice usually breaks up in several places, with the result that it divides into large drifting islands of pack-ice. The process of melting may take place very rapidly, especially in the eastern [Weddell East] part of the zone. The factory ships may be entirely surrounded by ice one day, to find themselves in ice-free water the next. Strong north and east winds, in particular, help to melt the ice quickly, as the high seas they bring smash up the ice '. The staple whale food Spring. The natural starting-point for the portrayal of the distribution of this, the staple diet of the whales, is spring, for it is then, notably in November (p. 370, Fig. 118), that the young, now 10-12 month old, swarms derived from the previous summer's spawning first begin to develop a substantial over 20 mm. component. It is in spring, too, that the whales first begin to arrive in, quantity on the feeding-grounds. The distribution and relative abundance of the staple feeding-stuff I during this season is shown in Fig. 132. Taking first the position in the ice-free zone, it will be seen from a comparison of Figs. 117, 123 and 132 that the spring distributional patterns, whether of larval, early adolescent or staple whale food, are in all material aspects the same, the principal concentrations of the staple class, like those of the larvae and early adolescents, being massed conspicuously throughout the Weddell drift and very conspicuously round South Georgia. In the West Wind zone there is again evidence between 30° and 60° E of small-scale November overflow from the Weddell stream and again of the effect of East Wind influence south-east of Kerguelen, the very small to minor occurrences plotted in these two localities I representing the over 20 mm. components of the mixed larval and adolescent swarms (p. 370, Fig. 118) already recorded there. The West Wind drift elsewhere is barren of the staple class except again where I it is affected by the current from the Ross Sea, and, west of Graham Land, by the influence of the I northward flow from the East Wind drift that originates somewhere near Peter I Island. Turning again to the Weddell drift it will be seen that in Weddell West the spring scarcity of the j late larval and 11-20 mm. classes, already referred to on pp. 368 and 376, is paralleled by a similar scarcity of the staple class. While again this might simply be due to chance, it has been recorded it | seems on too many occasions, namely, in the springs of eight successive years from 1927 to 19341 inclusive and again in the spring of 1936, for it to be ascribed to chance alone. The region manifestly is 1 one that is subject to periodic invasion by the first and second year swarms that are carried (i) in the East Wind generated current flowing into it from the Weddell Sea, and (2) in the current, partly 1 ^ The major concentrations of small, mainly adolescent, whale food plotted farther east, that is, south and south-east of J Kerguelen, appearing to be mainly, if not all, of East Wind origin. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 385 Fig. 128. Distribution of the massed surface larvae, showing their gradual spreading along the Weddell drift from west to east between December and April (vertical and stramin net hauls). Ice-edge February mean. 45 386 DISCOVERY REPORTS 60" 90 120 ISO" W 180"E 150" Fig. 129. Drift of the Calyptopes and Furcilias in the Weddell surface stream. March and April (vertical net hauls). Ice-edge March mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 387 t.til'l.l . i. I • !.f . I , I , I . : ./. ! . I :, 1 U . ! < .' . ■' , .^ > I > I 'I ' h I ' I' I 'I • I' I' I'll I ' h i< I'm ■!■ I ' i'i^r'--rT ■/''■'■' ISO" W ISCE 150° Fig. 132. Distribution of the staple whale food in spring. Ice-edge late December mean. 390 DISCOVERY REPORTS at least East Wind generated, that enters it via the Bransfield Strait. It seems possible, therefore, that the relatively barren spring condition we have repeatedly recorded here, a condition in which all three classes of the whale food are equally involved, is due partly to the onward passage of the young previous summer's swarms to the east (p. 376) and partly to the fact that the first and second year swarms upstream do not arrive in quantity in this sector until later in the year. The situation in Weddell West in summer (p. 394, Fig. 135) suggests that they may not in fact normally do so until January. The major and nearby minor occurrences of the over 20 mm. class plotted at the eastern end of the Bransfield Strait (p. 370, Fig. 118, Stations 537 and 539) and the single substantial occurrence recorded in about 37° W (Fig. 118, Station 1039) may, therefore, it seems, represent isolated instances of the spring invasion of Weddell West by young first year East Wind swarms each with a strong over 20 mm. component. Although the early adolescents are dominant (p. 376), spring is essentially a time of triple abundance'^ when the surviving larvae and both 1 1-20 mm. and over 20 mm. classes are spread more or less uniformly over every reach of the Weddell drift and in large, sometimes enormous, numbers are present in every part of the northern ice-free zone of euphausian abundance to its farthermost limits. Yet what again of the seemingly impoverished spring condition we have recorded in Weddell West? Can this after all be real, or is it in fact, as I have suggested, merely an appearance that could spring so readily from the chance failure of our nets to strike the surface swarms? It is true no doubt that the swarms encountered there earlier in the year, especially (p. 346, Figs. 100 and loi) the larval swarms, have long since passed on to the east. Yet it is difficult to see how there could ever be interruption in recruitment from the East Wind drift, as it sweeps up from the Weddell Sea, unless it be that periodically this coastwise stream ceases to function, or in other words comes virtually to a standstill. Sverdrup (p. 432) distinctly suggests that some such interruption could happen. In the East Wind zone the position is more obscure. In two instances, however, where there has been major penetration of the ice-belt there, namely, in the Ross Sea by the floating factory ' C. A. Larsen' carrying a Discovery Investigations observer in 1928, and between meridians 45° and 75° E by Mawson's (B.A.N.Z.) expedition in 1929, we have evidence that at least a moderate spring- time over 20 mm. population exists. Having regard, however, to the situation there in summer, throughout which, as we have seen (p. 378, Fig. 124 and p. 383, Fig. 127), substantial numbers of small (11-20 mm.) adolescents still survive, and having regard, too, to the adverse effect of the prolonged winter ice-cover (p. 355, Fig. 107) upon the larval growth-rate in these high latitudes, there can I think be little doubt that spring in the East Wind zone is essentially a time when, the surviving over 11 mm. Sixth Furcilias still contributing heavily (p. 365, Fig. 114), the small whale food vastly predominates over the staple class (but see PI. III). The developmental condition of the principal concentrations of the staple whale food in spring, based (Fig. 132) on our observations in the Weddell drift, the South Georgia area, the Bransfield Strait, the West Wind drift between 30° and 60° E and the East Wind outflow south-east of Kerguelen, is shown in Fig. 133. In this, as in the figures illustrating the corresponding condition in the summer, autumn and winter swarms which follow, the length frequency throughout is expressed in 4-mm. groups, a device rendered necessary because the second year swarms, now representing the main bulk of the staple class, in most instances fall within a length range too wide for the frequency to be expressed adequately by the narrower, 2-mm., grouping adopted for the younger swarms. As Fig. 133 shows, the length range of the second year swarms is often twice or thrice as long as that of the first year swarms, and while the length frequency of the latter can be expressed quite adequately by the narrower grouping (p. 370, Fig. 118) it was found that when that was applied ^ See again the corresponding charts for the massed larvae and small whale food (pp. 369 and 377). HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 391 to the older swarms the resulting graphs in many instances became too attenuated to show clearly where their modal values fell. Since the young first year (10-12 month old) spring swarms (p. 370, Fig. 118) as they grow eventually contribute substantially to the mass of the staple feeding-stuff, especially (p. 379) in the second half of November, their developmental condition is also shown in Fig. 133, first year swarms of East Wind origin being distinguished by an asterisk. DATE STATION 68 64 60 56 52 48 S 44 5 40 2 3 4 4 5 5 8 10 10 10 10 10 10 10 10 10 13 172022 3 6 6 8 13 19 19 22 23 24 25 26 26 25 27 27 29 29 29 W290W295W298 1831 IB35 1835 1835 1835 1839 1844 1854 I009 1875 1029 2094 2IOO 1039 518 1632 2112 f 25IO | 535 537 548 823 624 1825 1835 1835 1835 1835 1835 454 461 1009 1864 494 2093 1624 517 1628 1630 523 757 2139 2513 769 125 539 549 (6)(5) (5)(5) a. 36 I 32 S 28 5 24 -^ 20 16 12 8 4 DECEMBER 4 7 B 8 9 II 12 16 19 19 19 21 21 (?) (3) (3) (5) (4) (4)(5)(4) (5) 10-12 MONTHS OLD SWARMS 22-24 MONTHS OLD SWARMS Scale p I '^f'T'i ';';'/■!' mm;'/' IM III I ■/■Mi-ji] ■/ i ' I ■] ■T^-rn^-rrxy l • l •! ■ l M ■ [ ' 1 A M M ' 1 ' l '^ M ■ l ■ l ■ t A M 150° W 180°E ISO- Fig. 134. Distribution of total available feeding-stuff (small and staple whale food combined) in spring. Ice-edge late December mean. I HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 393 The spring distribution and relative abundance of the small and staple whale food combined is shown in Fig. 134, the data thus marshalled revealing the heavy massing of the total available feeding-stuff in the northern ice-free waters of the Weddell zone and showing that in general it would be the Atlantic stocks of whales that would be the earliest to encounter ample feeding during their southern migration, the Indian and Pacific stocks, the former first in the West Wind region south-east and south of Kerguelen, encountering it progressively later. Summer. The distribution and relative abundance of the staple whale food in summer (January- March) is shown in Fig. 135. The rapid growth, particularly in the Weddell zone, of the first year swarms in December leads by the end of the month to their abrupt^ and almost complete transition from the small to the staple class with the result that in summer, when the modern pelagic whale fishery is in full swing, the mass of the latter everywhere becomes enormously augmented. The effect of this transition is to be seen in the pronounced abundance of the large to enormous concentrations of the over 20 mm. class that is now to be found not only in the northerly Weddell zone, but also (although less conspicuously) in the southerly high latitude region of euphausian abundance in the East Wind drift. In the southern zone, however, the transition is much less abrupt than in the Weddell drift, many of the young, just over one year old, swarms in these high latitudes (p. 379 and p. 397, Fig. 136) persisting until March predominantly as early adolescents in the 1 1-20 mm. range, and this in part at least might account for the relative scarcity of exceptionally large concentrations of the over 20 mm. class apparent there. In the northern zone, it will be seen, there are large to enormous summer concentrations in the Bransfield Strait, in Weddell West, on the South Georgia whaling grounds and in the western half of Weddell Middle. Farther east, in the southern half of the current between 15° W and 20° E, there are minor, large and enormous concentrations, for the most part rather widely scattered, and in the northern half of the current between the same meridians there appears to be an extensive region of scarcity which in view of the rich 11-20 mm. population recorded there in spring (p. 377, Fig. 123) at first seems highly anomalous. It will be seen, however, that over a large part of Weddell Middle, in the northern half of the current between 15° W and 0°, there are virtually no summer observations, whereas the spring observations in the same area are many (Fig. 123) and cover it closely from west to east. In the northern part of Weddell East our summer coverage is also rather meagre with only 27 surface hauls as against 63 in spring. It will be seen, too, that in the same region there is a large area between 4° and 16° E which although well sampled in spring, in one instance with an enormous positive result, was not examined in summer. Turning now to the East Wind drift it would appear from our records that exceptionally large concentrations of the staple class are not of such common occurrence in these high latitudes as they are, for instance, in the western Weddell drift, on the South Georgia whaling grounds and in Brans- field Strait; and it has already been suggested that this apparent scarcity may be due in part at least to the fact that the augmenting of the staple class by the outgrowing of the 11-20 mm. adolescents is a more protracted phenomenon in the southern than in the Weddell zone. The greater frequency of occurrence of very large gatherings in the lower latitudes may be associated too with the fact that our stations there were distributed more or less equally between daylight and darkness, whereas in the East Wind zone the vast majority were made in the continuous summer daylight of these high latitudes — conditions tending to involve our observations in some error, rather difficult to measure, arising through the avoiding action which, in broad daylight in particular, thekrill, even when only half-grown, can readily take of the surface nets. In this coastal region of abundance, with its ^ Abrupt at least in the Weddell zone. 46 °" 394 DISCOVERY REPORTS oO" 30° 120 • lOO-I.OOO 0 1.000-10,000. (§) IO.OOO-IOQOOO •I ■ 1 ■ I • ! • I ' I 'I ' ''I'l'lif': 'I'l'l' I :■!■:':■■ ■|'l"'l ■l'l'l'l'IM'\'IM'ri'\'l'IM'lA'flM' v\' vv k'^ 'Vm' ^' ^' \ 'N ' ^ ' . ' 150° W 180°E ISO" Fig. 135. Distribution of the staple whale food in summer. Ice-edge February mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 395 virtually nightless^ summer, the migrating of the older swarms away from the surface by day must lessen still more the chances of making enormous surface gatherings there. In the Atlantic sector of the East Wind zone attention may be focused on the large area between 20° W and 30° E in which all developmental phases of the surface population, larval, adolescent and adult alike, are absent or conspicuously few.^ The locality has been visited and this scarcity recorded on twelve separate occasions, observations, in aggregate covering every month of the summer, having been made in 1930, 1931, 1932, 1933, 1935 (twice), 1936, 1937 (twice), 1939 (twice) and in the autumn of 1938 in April, the last month in which it remains open to vessels. This repeatedly recorded scarcity, it will be seen, occurs in a region that apparently separates a broad east-flowing zone of abundance in the north from a relatively narrow, coastal, west-flowing zone of abundance in the south, and the region it is suggested might be a backwater, not in fact an actively moving part of the East Wind-Weddell surface stream to which the principal concentrations of the krill are virtually confined. Turning now to the New Zealand side it will be seen that there is another large area, the great expanse of shelf water at the head of the Ross Sea, which to all appearances is barren, again of all three classes of the whale food alike. This phenomenon has already been discussed at some length on pp. 123-6 where, to recall the matter in brief, it was shown that the absence of E. superba from these high latitudes can partly be explained by the fact that the larvae, being carried for part of their existence in the warm south-flowing deep current, cannot, except as occasional stragglers, penetrate on to the shelf because it acts as a barrier to this southward movement. It will be recalled too that in place of the true southern whale food there, there are found vast quantities of another swarming euphausian, E. crystallorophias, which, superficially resembling the krill both in size and behaviour, doubtless is the food of the minke whales so abundant near the Ross Barrier, and, as Mackintosh (1942) also suggests, of the blue and fin whales which, as Villiers (1925) records, that great exploring whaler, C. A. Larsen, with such immense effort,^ hunted in these exceptionally high latitudes in 1923-4. Throughout the West Wind drift the scarcity of the staple class is pronounced, most conspicuously in the heavily sampled Pacific region except, near 90° W, where it is affected by the north-going East Wind flow near Peter I Island. It can be seen, however, that there is now more evidence of the presence of krill in the Pacific sector where the East Wind zone was accessible to sampling here and there. In the West Wind region south-east of Kerguelen only negligible numbers seem to be present, a scarcity of the older krill which appears anomalous in view of the moderate to substantial occur- rences of both early and older adolescents (p. 392, Fig. 134) recorded there in spring. The summer scarcity in this locality, however, could well it seems be put down to inadequate sampling. The minor Pacific concentration recorded in the West Wind region in about 135° W (p. 397, Fig. 136, Station 2244) is probably not a true West Wind occurrence. It was recorded in a position where the wind was from the east, and since the winds reported both north and south of it were also from that direction, it could well it seems have been located in the East Wind drift the northern boundary of which, although plotted a little to the south, must vary considerably. The northern limit of abundance of the staple summer feed then, is manifestly the northern boundary of the East Wind-Weddell surface stream, while its absolute northern limit, it is interesting to note, is not the Antarctic convergence, as has hitherto been supposed, but practically everywhere, except in the neighbourhood of South Georgia, far to the south of it. The situation at South Georgia is to be expected in view of the strong tongue of Weddell water that flows past the island, some- ^ Virtually nightless in the sense that the hours of darkness are so few. ^ See the summer charts for the massed larvae and very young adolescents (Figs, no and 124). ^ The rigours of whaling in the extreme cold of these far southern grounds and the almost insuperable difficulties that Larsen had to contend with are described by Dakin (1934). 46-2 396 DISCOVERY REPORTS times as both spring and summer distributional charts reveal, carrying rare individuals of the over 20 mm. class up to and even over the convergence into the Subantarctic w^ater that lies to the west. The complexity of the water movements responsible for these anomalous occurrences, which it will be seen are confined to the Falkland sector alone, is discussed in greater detail on p. 76. In his analysis of the results of the Discovery Committee's whale-marking programme of 1932-8 Rayner (1940) finds a distinct tendency among fin whales of Area 11 (Brown, 1954, p. 359, Fig. i), apparently during their summer sojourn on the feeding-grounds, to converge on Weddell West, a movement, one is tempted to suggest, that could well be a deliberate seeking out of a region where their food (Fig. 135) is in rich supply. He writes: Although it is fully recognized that whale-marks can only be recovered from those places where whaling is carried out and that the positions and movements of the factory ships have a considerable bearing on the picture rendered by the returned marks, it is also true that the factory ships do not remain long in areas where whales are not to be found. It does, therefore, seem reasonable to suggest that the part of the mouth of the Weddell Sea lying between 35 and 50° W has, for some reason, a strong attraction for Fin whales, which congregate there from a wide area extending to the east and west. In addition there seem to be strong grounds for considering the region between 30° W and 0° to be more sparsely populated with Fin whales. The reasons for this are obscure and a discussion of the possibilities beyond the scope of this paper, but it might be suggested that the influence of the submarine topography along the southern verge of the Scotia Sea on the hydrology of that region sets up conditions favouring the production of rich food stores which attract the whales.^ In his more recent account of the progress of marking Rayner (1948) notes that additional marks recovered since 1939 all corroborate the results arrived at in his earlier paper. Summer, then, is essentially a time when, of the three broadly grouped developmental phases of the surface population, the larval, the early adolescent, and the older adolescent and adult, the third alone, upon which the whales so heavily feed, is at the peak of its abundance, and at that peak is spread throughout the feeding grounds to their farthermost geographical limits. It is a time on the other hand (see again Figs, no and 1 24) when the larvae and early adolescents are limited both in their numbers and geographical range, the main mass of the larvae, just launched on their surface existence, being con- fined to the western Weddell drift but chiefly to Weddell West,^ the main mass of the early adolescents (the small whale food), shortly about to outgrow themselves, surviving only in the East Wind zone. The developmental condition and length frequencies of the summer swarms, that contribute so heavily to the annual fattening of the southern whales, are shown in Fig. 136. Taking in turn (a) the situation in the northern or Weddell zone, and (b) the situation in the southern or East Wind zone, the principal facts presented may be summarised as follows: {a) Northern zone (i) In summer the staple whale food is represented by a heterogeneous assemblage of yearling (13-15 month old) and adult (25-27 month old) swarms, the yearhngs having all but outgrown their early adolescent (11-20 mm.) phase. (2) In the yearling swarms the females are dominantly in stage i in January, stages i and 2 (rarely 3) in February and stages i, 2 and 3 in March. In the adult swarms they are dominantly in stages 5, 6 and 7^ (rarely 8)'* in January, stages 6, 7 and 8 in February and stages 7 and 8 (predominantly the latter) in March, the mounting frequency of stage 8 encountered as the summer advances indicating that spawning is probably at its height towards the middle, rather than at the beginning, of the season. * Some discussion of the questions Rayner leaves unanswered is given on pp. 243-5. * Bearing in mind, however, the far southern region of larval abundance (p. 310, Fig. 74 and p. 311, Fig. 75) in the East Wind zone which cannot be revealed by the stramin nets because the burgeoning surface population there, all through the summer (p. 361, Fig. iii), consists so very largely of the small escaping First Calyptopis stage. ^ Bargmann's stage 7 a or gravid. * Bargmann's stage 7B or spent. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 397 (3) In both young and old swarms the pattern of modal values displayed is heterogeneous throughout the season. (4) In January the yearling swarms tend to fall mainly within a length range of 17-40 mm., the more backward of them, as, for instance, at Stations WS 537, WS 540 and WS 542, having not yet quite outgrown their early adolescent (11-20 mm.) phase. In February the yearling swarms tend to 66 64 S60 O 56 ,■5 52 548 S 44 - 35 UJ a 32 <2e ■^24 ^20 il6 JANUARY FEBRUARY 6 7 8 B ro 1010 10 II r6 17 19 19 I9 202020 2I 21 21 22 22 2425 26 27 27 27 27 27 27 27 27 27 27 27 2B 28 28 303030| 3 57899999 795 575 S 2168 2I70 815 I270 818 2198 B20 W373 312 W538W540W540 824 2SS8 2560W542 2562 2567 1673 I507 351 352 354 797 799 5B0 2168 BI3 602 2195 2197 305 2199 2547 W537 W540W540W540 2558 2559 2561 825 1665 2567 1675 349 351 354 |^~] WEDDELL DRIFT Scale per cent EAST WIND DRIFT (8) (l){3X6){lX2)(8) N 2 MEAS- URED 702 69 325 61 148 48 125 432 174 77 lOO 32 86 522 539 95 19 172 lOO 55 818 185 lOO 55 60 III 51 159 23 87 53 53 167 79 20 467 246 116 588 904 504 35 36 346 171 205 146 82 35 36 82 MONTH DATE STATION 68 64 (/) 603 55 g 52" 48 I 44^ 40z 3*UJ 32 O 28 < 24°= 20f 15 i 12 y 8 N2 MEAS- URED FEBRUARY MARCH 1 17 18 IB 20 21 21 22 22 23 23 23 24 24 26 26 27 28 29 1 I 2 3 4 4 5 5 5 6 8 8 10 10 lO lO 12 14 14 15 15 17 19 21 22 23 23 24 24 25 26 26 2226 222a 624 W152 627 2244 1537 2590 1544 1297 1547 2268 2271 2506 368 540 643 1154 24 2000 38 1713 1715 1717 1720 I 618 522 1531 2BI4 1535 1285 WI56 1543 W9I5 2594 2600 2503 2606 2274 368 643 2614 23 200O2004 1711 2295 1717 1719 558 Scolc per cent Scole per cent WEDDELL DRIFT Scale per cent (2) (2X0 ;-50 144 85 200 71 lOO lOO 153 50 388 lOO 75 42 471 60 342 70 533 50 lOO 100 47 146 ISO 100 118 lOO 630 176 276 415 143 33 21 17 109 269 84 153 lOO 465 57 134 95 lOO 47 75 22 527 91 164 77 241 231 DATE ST 58 54 (/I 50% 55 g 52 O 482 40z 36 ~ u 32 o 28 5 24"^ 20 £ ,6^ 12 Ul e N2MEAS- URED Fig. 136. Developmental condition of the small and staple whale food in summer. For vertical scale see legend to Fig. 107. fall mainly within a length range of 25-44 min., although exceptionally, as for instance at Station 151 1, backward swarms may be encountered with a dominant length range of only 21-28 mm. In March the yearling swarms tend to fall mainly within a length range of 33-48 mm., although again exceptionally, as for instance at Station 658, surprisingly backward swarms may be encountered late in the month in which the females are still in stage i and the length range still dominantly 21-28 mm. as in February. (5) Throughout the season the adult swarms fall dominantly within a length range of 44-60 mm., the majority of them, however, in the 48-56 mm. range. (b) Southern zone (i) As in the northern zone the staple whale food is represented by a heterogeneous assemblage of yearling and adult swarms, the former, however, slowly outgrowing their early adolescent phase and in many instances persisting until well into March with the small whale food as an important or dominant component of the swarm length range. 398 DISCOVERY REPORTS (2) Apart from measurements we have no observations on the developmental condition of the females in the young summer East Wind swarms except that none were recorded carrying spermato- phores. It seems clear, however, judging from the condition of the females in the northern swarms, that in the East Wind drift the majority of them must remain dominantly in stage i and the remainder in stages i and 2, both equally dominant, throughout the season. In the adult swarms they are dominantly in stages 5, 6 and 7 in January, stage 7 (the only observation) in February and stages 7 and 8 in March, our failure to record stage 8 in any measure of dominance until the last month of the season providing perhaps some further evidence of the lateness of the spawning (p. 177) that has been shown to take place in these high latitudes. (3) In both young and old swarms the pattern of modal values displayed is even more heterogeneous than it is in the northern zone, a phenomenon that may be directly associated with the highly variable ice conditions (and the acceleration or deceleration of phytoplankton production that goes with them) in which so many of them spend their summer existence. (4) In January the more backward of the yearling swarms fall mainly within a length range of 12-34 mm., themajorityof them in fact representing whale food predominantly of the small 1 1-20 mm. class. The more forward young January swarms fall mainly within the 21-36 mm. range. Throughout February and March the more backward yearling swarms fall principally within a length range of 17-28 mm., the more forward within a length range of 25-40 mm., the former, even by the end of March, still having not quite outgrown their early adolescent (11-20 mm.) phase. The occurrence as late as 17 March of the very backward swarm (Station 2004) in which the vast majority of the indivi- duals were early (11-20 mm.) adolescents is possibly, however, exceptional. In general it will be seen the overall backward summer condition of the young East Wind swarms as compared with the condition of their northern contemporaries is most conspicuous. (5) Throughout the season the adult swarms fall dominantly within a length range of 36-56 mm., the majority of them, however, in the 40-52 mm. range. It would appear, therefore, that the breeding swarms in the East Wind drift consist of individuals which on the whole are smaller than their con- temporaries in the northern zone. Autumn. The distribution and relative abundance of the staple whale food in autumn (April-June) is shown in Fig. 137. The adolescent and adult swarms that together provide the major sustenance of the whales, now everywhere, except perhaps in the East Wind zone, appear to have undergone a phenomenal decline from their summer magnitude and abundance. The decline in fact is so enormous that one hesitates to believe it can be altogether real. Throughout the circumpolar sea it will be seen our autumn coverage is meagre, more meagre by far than in summer. Indeed in so far as the Weddell current is concerned the density of observation is even lower in autumn than the coverage figures suggest, for they (p. 287) refer not only to the current itself, but also to the West Wind and East Wind zones to the north and south of it. If, however, the observations in the main east-flowing Weddell stream, and in the South Georgia and Bransfield Strait regions affected by it, alone be considered, the coverage figures for the western, middle and eastern reaches of the drift in summer and autumn resolve them- selves as follows: No. of Sector Season observations Weddell West Summer 140 Autumn 29 Weddell Middle Summer 88 Autumn 13 Weddell East Summer 75 Autumn 30 HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 399 30' 60 90 120 150° W 180°E ISO- Fig. 137. Distribution of the staple whale food in autumn. Ice-edge April mean. 400 DISCOVERY REPORTS Thus Weddell West has been sampled nearly five times more thoroughly in summer than in autumn, Weddell Middle nearly seven times and Weddell East two-and-a-half times, the overall summer coverage of the whole current system being more than quadruple the autumn coverage. Disparity in coverage, therefore, especially such pronounced disparity as these figures reveal, might well be held to account for the apparent autumnal decline, although to what extent it might in fact account for it is difficult to gauge. The most that can be said is that it must account for it to some extent since with a discretely swarming organism such as this the probability of striking it in mass in any given area must at least decline as the observations decline. In the final appraisal of our results, therefore, it must be acknowledged that there can hardly fail to be some, perhaps small, measure of unreality in the autumnal scarcity we record. It cannot, however, be altogether unreal, for although our autumn coverage by summer standards is small, it is by no means insignificant. In fact in the western and eastern parts of Weddell East and on the South Georgia whaling grounds it is fairly substantial, or at any rate substantial enough to have provided at least some indication of a pronounced autumnal abundance if any such existed. It must, moreover, be significant that meagre by summer standards although our autumn coverage generally may be it is manifestly adequate to reveal the enormous concentrations of larvae (Fig. 112) that throughout this season are to be found spread along the Weddell stream from west to east. It must be equally significant that our vertical observations in autumn (p. 347, Fig. loi and p. 348, Fig. 102), although fewer and still more widely scattered (especially in May and June) than those of the stramin nets, also reveal the great autumnal abundance of larvae in this surface stream without any suggestion of doubt. In her study on the development and life history of adolescent and adult krill Bargmann (1945) finds that after breeding the adults appear to die oflF, disappearing from the plankton by May. She notes, however, that since spent females actively feeding at the surface are found in April, the absence of third year adults from the very scanty material available to her in autumn and winter might perhaps only be an appearance and not the result of 'a holocaust consequent on exhaustion after breeding'. It seems, however, from the pronounced autumnal scarcity revealed in Fig. 137 that such a holocaust might after all be a reality and that although some of the paired and spent (25-27 month old) swarms may survive for a short time into April (p. 401, Fig. 138, Station 207) the majority do it seems die off" with the result that by May the whaling grounds, formerly teeming with adolescent and adult swarms,^ are probably left largely depleted of the latter and carrying it would appear (Fig. 138) the rising adolescent generation alone. ^ Such a major loss of total available feeding-stuff might well, it seems, in some measure at least, account for the autumnal decline. If, however, it should be that the adults of both sexes having paired and spawned, do not in fact die off and that their virtual absence from our autumnal gatherings is simply due to chance or inadequate sampling, then we should have to suppose that this species, having reached full adult stature, is by autumn already about to embark on yet another, a third, whole year of life. The developmental condition of the oldest swarms we meet in winter, however (p. 405, Fig. 140), shows this to be distinctly improbable. There is evidence it is true (p. 255 and p. 402) that some swarms might in fact live to be fully three years old or more, but it is slender, suggesting that such as do survive for so long are very rare. Finally, I would emphasise, this season of apparent depletion, when the volume of the staple feed is evidently at a very low ebb, falls at the end of a long period, covering, in the northern zone at least, some six months or more, during which a horde of predators, whales, seals, penguins and a multitude 1 That is, staple whale food in the second and third year of their growth. * I am speaking here in terms of total available feeding-stuff. The whaling grounds, in addition to this rising generation, are now of course carrying an enormous larval population, as yet too young, however, to be available as food for the whales. HORIZONTAL DISTRIBUTION. GROWTH AND DYNAMICS OF DISPERSAL 401 of Other animals, has taken a heavy toll of the krill, the mass effect of which would only now be beginning to become apparent. This too would help to account for the autumnal decline. In his recent studies on the diatomaceous skin-film of fin whales and its correlation with fatness Ivashin (1958) finds that the 'maximum state of nutrition', as measured by blubber thickness, is in February, very few whales, whether infected with diatom or not, increasing their fatness in March. He therefore concludes that it is in March that a major decline in the krill takes place from its 'massive summer accumulation'. ^ 68 % 64 O 60 O 56 52 48 44 40 36 32 28 24 20 16 NUMBER MEASURED APRIL MAY JUN I 577777777777 14 15 17 I7 202026 27 27 6 7 7 8 18 2 42 :;07 207 207 207 207 207 2318 2322 855 1351 1359 I360 W434 663 207 207 207 207 207 WI96 665 855 2344 2346 1360 1361 1781 (3)(l)(l)(8)(lX8)Cl)(8)(l)(B)(0(8)(B)(2) (2) (3) (5X2) (2)(3) (I) (2) 91 30 102 69 20 33 39 57 69 53 25 lOO 44 20 41 39 25 50 52 55 50 66 23 370 III 46 lOO 100 68 64 ^ 50 O tr 56 u 52 5 ^« I 44 ^ 35 g 32 < 28 ^ 24 ^r o 20 Z UJ 16 -I NUMBER MEASURED WEDDELL DRIFT EAST WIND DRIFT Fig. 138. Developmental condition of the staple whale food in autumn. For vertical scale see legend to Fig. 107. Taking the autumnal distribution in greater detail and beginning with the northern zone, it will be seen that such minor occurrences of the over 20 mm. class as have been recorded, although widely scat- tered, are confined to the Weddell drift, to the South Georgia whaling grounds, where they show some small measure of crowding possibly overemphasised by the close spacing of our observations there, and to the Bransfield Strait. The minor occurrence, or rather occurrences,^ in the strait may in fact have been of East Wind origin, for when compared with the summer and autumn condition in which numbers of the half-grown swarms occur elsewhere in the Weddell zone (Figs. 136 and 138) the young 16 month old swarms in this southern channel (Fig. 138, Station 207) show a lag in female development and a general inferiority in modal length. The adult swarms, too, recorded at Station 207 exhibit a corresponding inferiority in modal length, a condition more typical of the East Wind than of the northern zone. In the West Wind drift the scarcity or absence of the staple class is everywhere pronounced. In the East Wind zone attention may again be focused on the rather closely sampled region between 0° and 30° E, the so-called backwater referred to on p. 395, in which as in summer there again appears to be a conspicuous scarcity of all classes of the surface population, larval, adolescent and adult alike. Elsewhere, such few autumn observations as we have in these high latitudes suggest that such older stages of the krill as survive the summer are now distinctly more abundant in the southern than in the northern zone. It would appear, too, that by May (Fig. 138, Stations 1359-61) the adult swarms having died off, the staple feeding-stuff in these as in lower latitudes is represented by the adolescent (17-18 month old) generation alone. That the East Wind drift should carry, as it appears to do, a somewhat richer over 20 mm. autumn population than the northern zone is perhaps not altogether ^ In this locality a continuous series of surface nets was towed one after the other along a course extending for approxi- mately II miles. 47 DM 402 DISCOVERY REPORTS surprising, for whereas the greater part of the Weddell system is open to major depredation for 6 months or more, and the remainder, being permanently ice-free, to depredation in some form or another all the year round,^ in the East Wind zone the period of major depredation, in so far as it is effected by whales, seals and penguins,^ must be cut to three months or less because of the prevailing ice conditions there. In other words this high latitude region of euphausian abundance may be con- ceived as being a relatively untapped reserve, in which the krill of all classes, larval, adolescent and adult, stand a greater chance of survival and reproducing their kind than they do in lower latitudes. It is a region above all where the generation of one year's status, benefiting from the long protection it has enjoyed below the winter ice-sheet, survives in relative immunity to breed the following year. Autumn, then, is essentially a time (Fig. 112) when the massed surface larvae everywhere dominate the southern feeding-grounds to their farthermost geographical limits, most conspicuously, but only because it is so readily accessible to vessels, in the Weddell stream. It is a time, on the other hand, when the small whale food in the natural order of things is everywhere absent, or virtually absent, from the plankton of the circumpolar sea, and when the older adolescents and adults of the staple class are practically everywhere, except possibly in the East Wind zone, at a low, perhaps their lowest, level of abundance. The developmental condition and length frequencies of the principal concentrations of the staple whale food encountered in autumn are shown in Fig. 138. Taking in turn {a) the situation in the northern zone, and (b) the situation in the southern zone, the principal facts thus presented may be summarised as follows: (a) Northern zone (i) The staple whale food in autumn is represented principally by a heterogeneous assemblage of 16-18 month old swarms together with (in April only) the few remaining adult (28 month old) swarms that survive after spawning. The younger (16-18 month old) swarms have now completely outgrown their early adolescent (11-20 mm.) phase. (2) In the young swarms the females are dominantly in stages i , 2 and 3 in April, those in stage i at Station 207, however, probably being backward swarms of East Wind rather than of Weddell j origin. In May and June they are dominantly in stages 2 and 3. In the adult (paired and spawned) swarms that survive into April the females are dominantly in stage 8. There is, however (Station 2346), an interesting late April exception from Weddell East. There 41% of the males were in stages 6 and 7 1 and both they and the females had only just begun to pair, the vast majority of the latter, 85%, being! still in stage 5. Clearly, in view of its high modal value, this swarm must be regarded as representing an instance of the attainment of adult (third year) stature with failure to ripen to maturity, and since! Bargmann's elaborately worked out life-history suggests each developmental phase to last 2 months in the male and 2^ months in the female it would appear distinctly possible that in certain, perhaps very rare, instances E. superba does not spawn until it is fully 3 years old or more. (3) The pattern of modal values displayed by both young and old swarms is heterogeneous through- J out the season. (4) From April to June the young 16-18 month old swarms tend to fall mainly within a length] range whose principal limits lie between 25 and 48 mm., the more backward in the 25-36 mm., the] more forward in the 36-48 mm. range. (5) The few remaining adult swarms that survive the spawning into April fall dominantly within! a length range of 44-56 mm. '■ On Elephant Island it will be recalled (p. 43) Gentoo penguins are recorded feeding heavily on krill in the autumn and | winter of 1916. ^ See, also, the section beginning on p. 423 on the coastal topography of Antarctica as an additional factor affecting the j depletion and conservation of the euphausian stock. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 403 (b) Southern zone (i) Although our observations are rather scanty there is evidence that here as in the northern zone the staple whale food is represented principally by a heterogeneous assemblage of young 16-18 month old swarms together with, again apparently only in April, the older adult swarms that survive the spawning. In this region of slow growth, it will be noted, exceptionally backward swarms may be encountered that even as late as May have not yet quite outgrown their early adolescent (11-20 mm.) state. (2) The only observation we have (at Station 1359) as to the condition of the 16-18 month old females in these high latitude swarms suggests that in some, perhaps not infrequently occurring instances, they may persist dominantly in stage i until well into May. There are no observations on the condition of the females in the single adult swarm recorded. (3) The pattern of modal values displayed, by the young swarms at least, is heterogeneous. (4) In April and May the young swarms tend to fall mainly within a length range whose principal limits lie between 25 and 36 mm., the most backward of them, even in May, still mainly in the 21-28 mm. range. Winter. The distribution and relative abundance of the staple whale food in winter (July-September), when only a relatively narrow strip of the northern zone of euphausian abundance in the Weddell stream remains open to investigation, is shown in Fig. 139. Now, as Fig. 140 shows, the older krill are represented exclusively by maturing, although still largely adolescent, 19-21 month old swarms, the previous summer's adult population having evidently completely died off. In the Weddell zone, where the facts of the distribution can alone be determined with certainty, it will be seen that while a moderately substantial population appears to exist in Weddell West and on the South Georgia whaling grounds, the crowding of the positive occurrences in the latter area being doubtless due to the close spacing of our observations there, elsewhere, notably in the rather heavily sampled Weddell East, there is still every indication of the magnitude of the autumnal decline. It would appear, in fact, that the northern decline in staple abundance resulting from the ravages of predators and the dying off of the adult population is a phenomenon which must endure until the rising young genera- tion of mixed larval and adolescent swarms (p. 367, Fig. 116) now, in winter, 7-9 months old and massed in heavy concentration throughout the Weddell zone, eventually begins to outgrow its early adolescent (11-20 mm.) phase. In other words, the depletion of the over 20 mm. class, so evident throughout autumn and winter, cannot, in the northern zone at any rate, be made good (p. 391, Fig. 133) until spring. Although not so numerous as at other times of the year, our winter observations in the West Wind drift are enough to suggest the absence or great scarcity of the staple whale food there. It may be noted, too, that all but one of the four negligible occurrences recorded are located in regions affected by north-easterly outflow from the East Wind zone. The situation in the East Wind drift itself can only be surmised. It seems probable, how- ever, judging from the situation there in autumn (p. 399, Fig. 137), that under the winter ice-sheet in these high latitudes the 19-21 month old adolescent swarms representing the staple class survive in a somewhat greater measure of abundance than they do in the northern zone. Winter then, as Figs. 115 and 122 show, is essentially a time of dual abundance with both larval and early adolescent (11-20 mm.) classes, the former in enormous numbers, dominating the northern zone of euphausian abundance to its farthermost geographical limits. It is a time, in the northern zone at least, of continued scarcity of the staple class. In the East Wind or southern zone, however, the larvae (pp. 338 and 376) are probably the dominant winter class, with the staple whale food of secondary importance and the early adolescents (p. 376) non-existent. 47-2 404 DISCOVERY REPORTS Fig. 139. Distribution of the staple whale food in winter. Ice-edge July-August mean. HORIZONTAL DISTRIBUTION, GROWTH AND DYNAMICS OF DISPERSAL 405 The developmental condition and length frequencies of the principal winter concentrations of the staple whale food recorded in the northern zone are shown in Fig. 140, the major features thus presented being as follows : (i) The staple whale food in winter is represented by a heterogeneous assemblage of swarms about half-way through their adolescence and ranging from 19 to 21 months old. (2) In July, judging from the length frequencies of the swarms encountered, the females are probably dominantly in stage 2.^ In August and September they are dominantly in stages 3 and 4 (rarely 2). MONTH JULY AUGUST SEPTEMBER MONTH DATE 9 10 II 11 16 2852 2853 2854 2855 2864 12 2362 (3) 16 17 19 22 24 27 28 28 2391 2393 2399 24062412 2874 W264W264 (3; (2) (3) (3) (3) (.4) (2) (4) 1 1 1 Scale per cent 0 50 100 2 3 5 5 17 1403 I40S W277 W277 W282 (2) (4) (4) it 27 1813 (4) 1 DATE STATION STATION « 68 %64 i60 ^ 56 |52 i 4B Z 44 68 ,;, 64 - 60 0 56 f 0 Sif V 0 « 00 " 0 • ^^ \ '■A 0 \ 1 i / / ^ ri" -- \ 40° 38' 36* 34° 53 5^ 5(S 40 ••. 40° °o •i-ioo • loo-i.ooo ^1.000-10,000 ^p 10,000-100.000 Fig. 152. Distribution of the small whale food on the South Georgia whaling grounds (stramin net hauls). 40° 38° 36° Si si -^tr- I oa%ov •v °:\ "^"iV i^ \ 40° 38 36 34° °0 •I-IOO •lOO-I.OOO ^1.000-10,000 ^P 10,000-100,000 Fig. 153. Gross distribution of the small whale food on the South Georgia whaling grounds (stramin net hauls). 49-a 420 DISCOVERY REPORTS 40° 38° 36° 34° 53 J 54 • °; JANUARY- MARCH S3' •r" 54' 55' 56' • ■BA^^f^* • 56° o o • ^^ N 40° 38° "*" 34° r 40* 38° 36° 34° < 54 55' 53 O 55° O • A UGUST-SEP rEMBER 1 1- •—•-It V O ^ r\^^ • r/ *x -^^ r / '^ ^ A ^ \ 40° 38° 31? 34° 40° 38° 36° 34° si 54 5i 5<5 • • • J^ — • OCTOBER -DECEMBER 4 . 5^ • O • "" o o o~ • ^^^^ 0 i V • A • • 0 >#" 40° 38r 36° 34° /^s^"^ J • I-IOO •lOO-I.OOO ll.OOO-IO.OOO IIO.OOO-IOO.OOO Fig. 154. Distribution of the staple whale food on the South Georgia whaling grounds (stramin net hauls). ■>o •i-ioo •100-1,000 >i.ooo- 10,000 ^fc 10,000-100,000 Fig. 155. Gross distribution of the staple whale food on the South Georgia whahng grounds (stramin net hauls). LOCAL DISTRIBUTION. THE SOUTH GEORGIA WHALING GROUNDS 421 The massing of the whale food on the north-eastern side of the island under the influence of the branching Weddell stream is emphasised over and over again by the monthly distribution of the large rorquals taken by the South Georgia whalers, the charts of Kemp and Bennett (1932) for the period 1923-31 revealing repeatedly such an enormous preponderance of blue and fin whale kills in the eastern branch of the current that there can be little doubt that they congregate there because of the exceptional richness of their food. As Ruud too (1932) remarks, 'it seems reasonable to suppose that the whales regularly seek out the places where the krill occur in the greatest profusion '. With regard to this local but most striking distributional link-up Kemp and Bennett write as follows : That the principal whaling grounds lie to the north-east of South Georgia is to be explained on hydrological grounds. It is due to the system of currents which prevails in the neighbourhood of the island. One of the currents, which comes from the Pacific, flows through Drake Passage to the south of Cape Horn, and taking a north-easterly direction passes to the west of South Georgia. As it progresses northwards it comes more and more within the influence of the westerly winds and thus tends to take a more easterly course. The second and more important current is of Weddell Sea origin. It passes round the south-eastern end of South Georgia and then sweeps in a north-westerly direction along the coast. Before, however, it reaches the northern end of the land it meets the first current and the prevailing westerly winds and is reflected backwards to take a north-easterly course. The direction of the Weddell current is thus largely due to the presence of South Georgia and to the position which the island occupies athwart the region where the two streams converge. Euphausia superba, the exclusive food of Blue and Fin whales in the Antarctic, is a pelagic prawn which drifts northwards to South Georgia from higher latitudes, the greater part coming from the Weddell area. The whaling grounds lie in the lee [i.e. on the north-east side] of South Georgia, and it is here, in the shelter of the island, that the Euphausians are able to congregate. Another striking example of distributional linkage between krill and predator in this locality is provided, it seems, by the large adult Nototheniid, and prodigious krill-eater, Notothenia rossii, which according to Olsen's findings (1954) is heavily massed throughout the whole of the shelf water along the north-east side of the island, but absent from the wide krill-poor shelf area to the south-west. The monthly developmental condition of the swarms in the main east-flowing Weddell stream compared with the corresponding condition of the swarms in the neighbourhood of South Georgia is shown in Fig. 156. This diagram, constructed on precisely the same lines as Fig. 55 (p. 253), is presented principally to show that there is no mass spawning of the krill in the South Georgia area and that the rich euphausian population there cannot, therefore, be of local origin but must in the main be derived from the swarms that spring from the widespread larval risings that take place in Weddell West and subsequently get carried to the island in the surface drift. The absence of concentrations of eggs coupled with, even more important still, the absence of swarms with a naupliar or metanaupliar mode, provides the principal evidence of the absence of a local spawning, although almost equally conclusive evidence is provided by the extreme rarity with which substantial surface swarms of slightly older denomination (the Calyptopes for instance) have been recorded in these waters throughout the maximum period, November to March, over which spawning is known to be spread. As Figs. 148 and 150 show, the vast majority of our observations between January and March, whether with vertical or stramin nets, record only negligible numbers of young surface forms. Even moderate samples are rare, gatherings of between 100 and 1000 individuals such as we get so regularly at this time in the Weddell stream, and which in the vertical nets at least probably represent the effective sampling of a swarm, having been recorded only twice in the vertical and only thrice in the stramin nets. While it seems fairly clear that the populating of these island waters must be brought about through the incursion of alien swarms from the Weddell Sea, owing to a gap in our observations in this other- wise heavily sampled region we have no precise information as to just when such incursions first begin 422 DISCOVERY REPORTS to take place on a major scale. The gap is spread over the period April to July, there being no effective coverage in April,^ only six observations in May and none in either June or July. Unfortunately this period is critical, for clearly the absence, apart from the rare instances noted above, of young not long hatched swarms from this region from November to March, when the vertical net was used no fewer than 1177^ times (Fig. 156, unbracketed numbers) in South Georgian waters as against only 465 times (Fig. 156, numbers in brackets) in Weddell West, and the widespread occurrence there (Figs. 150 FEB APR NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT 1 1 1 1 1 1 1 1 1 1 1 1 NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT II II NOV DEC JAN MAR 1 1 1 1 0.O A°.°o ~ • •oo ,'.'& «o 0* ooo 00 &'& •• • • 0 o . o S8 ° o o «• o oo o o eo o o 0 • 0 0 oo o«o 0 ee o* • o 9 • «» SJS oo o • oo OO o o*o . ooo 0. • gS8 . FIRST YEAR SWARMS ■> o*o o □ • oo ^ '"^^^^^^'^^rinwi iw^i^ yj •? #^^^ ivi w ^ o*oooo ooo 000 " °*"* 1 SOUTH GEORGIA 1 gjg ooo 000 1 OBSERVATIONAL 1 SSSS o.o i GAP 1 &1 1881 1 ••: Bih (168) 1 - 88 •.'.'se.f. «• SSS ooo cS» 1 ° 0 SSS 1 ::: , eC<~/Miir\ VtAD CUfADkiC ^ THIRD YEAR SWARMS So j * oo j *■ — — — StCJNU YtAH bWAHMb --- — — — - — ----> 231 1 oo o 0 1 304 (119) o , SSS o^ 8S 1 (4^ 0 o. ! .oo 0 1 I80 . «00 %% . o% O™ 1 ::«R SOUTH GEORGIA SWARMS 274 (67) o oo ooo 1 oo 1 88S8S8 WEDDELL SWARMS (69) o. Ill goo^oflSJIfl IS8 1 = 888 88 8SS88 gis 1 »"> 1 1 1 1 1 1 1 1 NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT 1 1 1 1 1 1 1 1 1 1 i 1 NOV DEC JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC JAN MAR FEB APR 60 56 52 48 44 42 40 38 36 34 32 30 28 26 24 22 20 18 16 14 12 lO 8 7 6 5 4 3 2 Fig. 156, Growth of the swarms at South Georgia and in the Weddell drift, showing how the island population is derived from incursions of surface-borne larvae from the Weddell Sea and is not of local origin. and 156) of late Furcilia swarms in August, when observations resume, postulates that it must be some time during the period covered by this 'observational gap' that the first major incursions take place. In four of the samples they analysed from five stations made off the east coast in May 1927, Hardy and Gunther (1935, Appendix 11, Table iii) record moderate to substantial numbers ofE.superba ' Cyrtopias ' ^ ( = almost certainly Furcilias) which suggests that it is possibly in that month (Fig. 150), or even in April, that the young swarms carried in such profusion in the Weddell stream first begin to arrive on a substantial scale. The observations of Hansen (p. 334) also suggest that it is in May that the first large-scale incursions take place — of larvae perhaps in the early Furcihastate. Our own observations round the island (Fig. 150) also point rather strongly to its being in autumn that a major invasion is going on by young swarms sprung from risings in the Weddell Sea. For, although few and rather widely scattered, especially on the east side, they do nevertheless show, at every station falling fairly inside the Weddell stream in April and May, in three instances substantially, the presence of a larval population, and at the same time that the south-western backwater, in striking contrast, is barren. 1 Not effective because it was confined (Fig. 148) principally to the south-western backwater. ^ This figure includes vertical net hauls from which samples were analysed by Hardy and Gunther. ^ Since no more precise developmental data are available for these May samples I could not plot them in Fig. 156. THE COASTAL TOPOGRAPHY OF ANTARCTICA 423 THE COASTAL TOPOGRAPHY OF ANTARCTICA AND ITS INFLUENCE ON THE DEPLETION AND CONSERVATION OF THE SURFACE POPULATION In the northern part of the Falkland Islands Dependencies the ice-free areas of the Trinity Peninsula and its adjacent large islands, with those of the South Shetland, South Orkney, South Sandwich Islands and South Georgia, together provide innumerable coastal sites of suitable gradient where krill-eating penguins. Ringed, Adelie and Gentoo, have established their rookeries on an enormous scale. In the South Orkney group the naturahsts of the 'Scotia' (Bruce, 191 1) estimated that on Ferrier Peninsula alone, one of the many promontories of the heavily sculptured coastline of Laurie Island and a favourite site for Adelies, there were not less than two million birds. Brown (191 5 6) adding that the Ringed population of Graptolite Island might run to several millions. As Murphy (1928) once put it, the incredible numbers in which these social birds are encountered simply ' beggars description '. From these estimates it would follow that over the whole area outlined above the rich euphausian population of the western Weddell drift must suffer annually a major depletion from the depredations of penguins alone. Severe, however, though their ravages m this locality unquestion- ably must be it does not necessarily follow that penguins take a similar toll of the krill throughout its whole geographical range. Long stretches of the coastline of continental Antarctica are occupied by continuous vertical barrier ice cliffs on which these krill-eating birds, here exclusively Adelies,^ even if they could gain a footing, would not choose to build. Suitable rookery sites, therefore, are not only generally less concentrated in the East Wind than they are in the western part of the northern zone but in certain parts of the former, notably throughout a vast section of the continental coast reaching from Enderby Land to Vahsel Bay and beyond, are absent altogether, the barrier along this particular stretch running for approximately 2000 miles not only unbroken by any rock, but free throughout from offshore rocky islets on which the birds might establish their nests.^ Excellent photographs of this ice-bound coast showing its gross unsuitabihty for congregating Adelies, and other rock-frequenting birds, are given by Christensen (1938), Swithinbank (1957a, i957*). Gessner (1942) and van Rooy (1957)- The persistence of fast ice occasionally thrusting far out to sea during the breeding season could also it seems be a serious embarrassment to the nursing birds, their rookeries from time to time (Law and Bechervaise, 1957) being cut off for such enormous distances from open water by undispersed winter ice that the chicks starve to death by the hundred before their parents can return with their food. In terms of consumption of Euphausia this would mean that, bird for bird, the far southern adult population, wherever such adverse conditions occur, has fewer mouths to feed than the corresponding population in lower latitudes. The long frontage of the Ross Barrier presents another obstacle to the nesting birds and I think it extremely likely too that all along the Pacific side, from Alexander Island westwards to the Bay of Whales, suitable breeding sites will prove to be absent or at most exceeedingly rare.=^ Incidentally, these immense ice barriers, pushing their fronts far out to sea (Law, 1952) into deep water, create in many places round Antarctica quast- pelagic coasts along which true coast fishes, such as, for instance, are so numerous and take such heavy 1 Apart from the Emperor, which is essentially a squid-eater, the Adelie is the only penguin frequenting the coasts of the continental land beyond the Graham Land peninsula. It does not extend to the Subantarctic islands. Elton (1927, pi. vii) publishes a photograph, from Mawson (1915), showing a vast congregation of birds with the legend 'Adelie penguin rookery on Macquarie Island'. This is a rookery of Royal penguins. 2 In a map just published Sladen (1958) shows this stretch of the coast without a single AdeUe breeding site. 3 The latest American expedition to visit this sector reports none in the southern parts of the Bellingshausen and Amundsen Seas (between 120° and 80° W), and that the birds themselves are exceedingly scarce, being represented as a rule by lone individuals or at most by two at a time (Anon., 19606). 424 DISCOVERY REPORTS toll of the krill in the neighbourhood of South Georgia, would be unlikely to occur. Thus, it would appear, the krill in these high latitudes, already enjoying at least some measure of immunity (p. 402) from the major depredations of whales and seals, exist in some measure of immunity too from penguins and possibly coast fishes, an immunity it seems that might be particularly high on the Atlantic or Weddell side. And so, enormous as must be the mortality suffered annually by this species through the ravages of its northern predators, in the East Wind zone both pack-ice and barrier contribute something to its protection, combining to create in these high latitudes something of a natural reserve. The impenetrable core of the pack-ice in the heart of the Weddell Sea, if the krill are there (see pp. 71-3), must also contribute to its protection. EUPHAUSIA SUPERBA AND E. TRIACANTHA The gross distribution and relative abundance of (i) E. superba (over 20 mm. group), and (2) the smaller circumpolar E. triacantha (all stages excluding larval), based on the data from the upper (loo-o m.) oblique stramin nets, is shown in Fig. 157, the data for E. triacantha having kindly been provided by my colleague Mr A. de C. Baker, who has recently published a comprehensive report upon it (1959). This figure is presented principally to show the enormous disparity in abundance exhibited by the two species, the larger a highly concentrating swarmer, the smaller, judging from the remarkably low order of abundance in which it persistently occurs in our samples, an organism scattered broadcast more or less uniformly throughout its circumpolar range. The disparity in numbers, enormous even when only the older stages of the krill are considered, would of course be vastly accentuated by including the post-larvae of this species in the construction of the chart. Although other factors as yet obscure must also be involved, it is possible that the great depths at which the krill eggs hatch contribute something to the immense scale on which this species exists. For at these depths other pelagic animals are scarce and the resultant larvae it seems might begin their existence in some measure of immunity from depredation. Moreover, such relative immunity as they may enjoy at the outset of life does not end there, for the ascending larvae it seems in most regions are being carried towards further immunity, the southward movement of the warm layer through which they pass bringing some of them in the end into the high latitudes where major depreda- tion is at a minimum. Or again, if the krill spawn on the bottom, and we have some evidence that they might, it could be supposed that the eggs so laid would enjoy a greater measure of protection than they would if drifting about on the surface at the mercy of the myriads of plankton animals that doubtless prey upon them. Singer (1959), referring to the spawning of sand eels {Ammodytes spp.) in the relative security of the shallow sandy bottoms of the North Sea, suggests that this might contribute at least a little to the immense scale on which these heavily preyed-on fishes exist. In his studies on the reproductive and larval ecology of bottom-living invertebrates, Thorson (1950) emphasises that it is the early part of the life-cycle that is most vulnerable and that the abundance of older individuals will to a large extent be dependent upon the numbers of larvae that survive to settle down to their benthic existence. In so far as it may affect the annual recruitment of the teeming krill population the relative immunity its immensely deep early developmental phase enjoys could well it seems therefore be critical. As Thorson has stressed, a fundamental knowledge of the factors governing the numbers in a population is only to be obtained through a consideration of the limiting values, not the average values, of the ecological factors, and we must, therefore, focus our attention he says on the weakest link in the life-cycle, normally to be found during the breeding period and larval development, 'when the requirements of the organisms from the environment are often much more definite ' than at other times. Thus in a demersal fish such as the haddock the relative numerical EUPHAUSIA SUPERB A AND £•. TRIACANTHA 435 Fig. 157. Distribution and relative abundance of E. superba and E. triacantha. 50 426 DISCOVERY REPORTS Strength of each brood in the first, second and third year of Ufe is estabUshed before it becomes demersal, the factors determining the numbers in each instance being operative during the vulnerable pelagic phase of the haddock's life-history (Parrish, 1956). In E. superba, it would seem, we have an instance where the factors governing the abundance are operating in the reverse direction, the basis from which its fantastic numbers spring being perhaps laid down during its early existence in the deeps long before it reaches the surface stream. Simpson (1956) summarises the present position of our knowledge regarding survival in demersal fishes during the pelagic phase in the following passage : . . . the accumulating evidence and speculation on theoretical grounds point towards the fundamental mortalities being biological ones, which at some stage become density-dependent, that they are not cataclysmic but persistently high while the fish is small and helpless, predation probably being the main cause. Superimposed on this mortality there may be expected to occur more specialized types of mortality differing from species to species, and probably from year to year, currents being important to some species in carrying the larvae to destruction, storms may destroy millions of eggs by violent agitation, food shortages may lead to small survival in other species. While the former type of mortality, predation, may normally be the main cause of death, both annual variations in predation and these other factors may be expected to be responsible for annual fluctuations in survival. During its deep lifetime, it seems clear, as it hatches from the egg and grows to the Metanauplius form, E. superba can be little affected by such biological or physical hazards, and in the prevailing calm of its bathypelagic environment annual variations in such physical hazards as there may be are likely to be small, resulting in little material fluctuation in survival from year to year. For predation must be at a minimum, such currents as there are do not carry the larvae to destruction, but rather the reverse, no storm can shake the eggs, and above all the Nauplii and Metanauplii do not require to feed. They are of course unable to feed and so to them food shortages are of no concern, and density- dependent mortalities, such as from time to time affect the multitudes of a feeding community, are impossible. The swarming habit too, it seems, might contribute something, although at first sight it would seem to place the krill at an enormous disadvantage, presenting as it does a rich and ready reward to major predators such as whales, seals, penguins and fish. In its final phase, however, it must lead to the liberation of astronomically large numbers of highly concentrated eggs, the females of a gravid swarm (p. 250, Fig. 53), as Bargmann (1937, p. 341) has found, producing between 11,000 and 11,500 each. Moreover, having regard to the sex ratio (p. 246, Table 49), in the tightly packed shoals, the prospects of successful mating for all but for perhaps a negligible few of the maturing females would manifestly be extremely high. In fact, as Bargmann (1945, p. 118) remarks, 'faulty implantation hardly ever occurs '. Another, perhaps all-important, factor contributing to the sheer numerical pre-eminence of the southern krill is that they, perhaps alone among Antarctic euphausians, are equipped, and equipped superbly (Hart, 1934; Barkley, 1940; Marshall, 1954), to take the fullest possible advantage of the enormously rich pastures of the circumpolar sea. In this respect their cleaving to a surface habitat, to the optimum zone of phytoplankton luxuriance, might well contribute further both to the stature of these animals and the immense scale on which they have populated the Antarctic seas. A parallel instance of biological success, based on adaptation for algal feeding, is recorded by Fryer (1959) among a group of rock-frequenting cichlid fishes from Lake Nyasa. ' It is significant ', he writes, ' that, basing biological success on numerical abundance of species, the algal eaters are the most successful. This was made abundantly clear during the work at Nkata Bay and elsewhere where algal eaters predominate and where the only really common invertebrate-eater is Labidochromis vellicans'. A parallel too it seems is presented by Calanus finmarchiciis, Lucas (1956^) calling attention to the EUPHAUSIA SUPERBA AND E. TRIACANTHA 427 fine filter feeding of this voracious and extremely successful herbivore as distinct from the coarser, more actively predatory, feeding of other copepods such as Candacia, Temora, Euchaeta and most bathy- pelagic forms, which, having much stronger and less setose mouth parts, triturate their food either before or during feeding. It has been suggested too (Millar, i960) that the large size attained by certain Antarctic ascidians springs indirectly from the rich phytoplankton diet of the plankton animals on w^hich they so largely feed. Dell (1952), referring to the immense quantities of phytoplankton in the Antarctic and the super- abundance of nutrients, phosphates and nitrates, that goes with it, also calls attention to the benefits that accrue from an algal diet, noting that the rich food supply available for the phytoplankton in these southern waters will ' determine the relative abundance of Euphausia which will in turn influence the numbers of the whales and their distribution '. Fisher and Goldie (1959) refer to the particularly high vitamin A reserves present in the Meganyc- tiphanes norvegica population of Loch Fyne and to the relatively great size attained by this species there, noting that these phenomena spring principally from the presence of /?-carotene in the dinoflagellates, diatoms, algae and fern sporangia that contribute to its diet in this western Scottish loch. They find that this diet produces larger specimens of M. norvegica in Loch Fyne than any they have seen from the North Sea, the North Atlantic and the Mediterranean, and that these large animals are much richer in vitamin A than those they have analysed from these localities. The rich, superlatively rich, phyto- plankton diet of E. superba, and the /^'-carotene that goes with it, might contribute much it seems to the relatively enormous size to which it grows. It may be of some significance too that the krill seem to be completely immune to infection by the EUobiopsid Protozoan parasite Amallocystis fagei Boschma, which Boschma (1949) records on Euphausia vallentini, E. frigida, E. recurva, E. lucens, E. hemigibba and Thysanoessa gregaria, and which I have repeatedly noticed is particularly liable to attack E. frigida, an Antarctic species co-existent with the krill in the plankton. It attaches itself to the dorsal surface of the carapace of its host in the region of the genital gland, ' protoplasmatic excrescences ' from the organ of fixation penetrating into the ovary. According to Einarsson (1945), who describes what Boschma thinks is the same species on Thysanoessa inermis, it seems probable that it ' castrates the animals ' it attacks. When one contemplates the hosts of its predators and the vast weight of whale flesh, seal flesh, fish, penguins and other birds that is built up on its wholesale destruction, one is left wondering as to how in its seemingly inexhaustible myriads it continues to survive, and above all at its astonishing capacity year after year for making its staggering losses good. This capacity, by itself, must be one of the overriding factors contributing to its phenomenal biological success. Perhaps the winter ice-sheet, extending as it does over such a vast area of the circumpolar sea, and under which such an immense part of the larval population spends virtually six months of its surface existence, contributes a lot to the ability of this species to maintain its numbers at their existing fantastic level. For above all the winter ice must act as an ecological umbrella under which the overwhelming mass of the krill in its most vulnerable phase is secure from the ravages of its largest predators, enjoying a long period of relative immunity denied to other euphausians that spend their entire existence in the open sea. Above all it seems clear that in the course of evolution this overflowing community has adapted itself superbly both to tolerate and to take the fullest possible advantage of the external metabolites of its neighbours. I quote from Lucas (1947): That microplanktonic organisms do produce substances externally, either during life or in death, which assist some and hinder others of their fellows in their growth, is now confirmed. One point, however, can scarcely be emphasized too strongly: since planktonic organisms do produce such substances, then only those forms which are unharmed 50-2 428 DISCOVERY REPORTS by them can continue to live and develop unaffected in association with the forms producing them. It may be that some benefit from such associations, but if they are not at least tolerant, then they must either migrate or slowly succumb. Only those which are tolerant, too, can succeed after the production of such substances, whilst such tolerant forms must quickly be overtaken in the succession by any forms which have developed favourable adaptations to the conditions. The onslaught of predators upon this species is a phenomenon perhaps without parallel in marine ecology and yet in its untold millions it survives. What, however, if it had no predators, or at any rate suffered vastly less severely from them than it does? It seems inevitable that with its enormous fecundity it would multiply enormously, in the end reaching such fantastic numbers that even the immense wealth of pasture of the southern seas would be unable to support it, and, as happens with fishes (Rounsefell and Everhart, 1953) when their predators have been removed, there would be overcrowding of the polar sea by undersized,^ slow-growing krill, and ultimately a density- dependent mortality would set in and the population would decline. Lack's work (1954, 1955) on animal numbers and the factors that regulate them suggesting it might decline enormously. In the face of such staggering odds the ability of E. superba to maintain itself so successfully passes comprehension. We may well indeed wonder whether in fact it has not already reached a point where in its fantastic numbers it is competing for life on such a massive scale that even now it is bringing about a displacement of its competitors in the same ecological niche and (Hardin, i960) perhaps threatening their very existence. Besides disparity in numbers E. superba and E. triacantha present an interesting contrast in circum- polarity, the larger species, in the major aspects of its distribution, showing a marked asymmetry, the smaller, spread round Antarctica with far less latitudinal variation, a distinctly more symmetrical pattern. The distribution of E. triacantha is marked too by a seeming intolerance of the cold surface current system in which the whale food occurs in such profusion. In fact only in the extreme north, round South Georgia, where hydrological conditions are complex and waters of very mixed origin converge, is it regularly found in this system at all. In short, E. triacantha is essentially a creature of the West Wind drift. Is it right, however, to suggest that it is ' intolerant ' of the somewhat colder, although not all that colder, environment in which we find the vast bulk of the congregating krill ? It may be that it is the way of life of this species, its behaviour, vertical movements and so on, and the behaviour (or dynamics) of the water masses in which it lives from hatching onwards, that in general combine to prevent its co-habitation with its southern neighbour and mixing freely with it. The dynamics] of the environment clearly lead to free mixing near South Georgia, and again it seems, although perhaps in a lesser degree, south-east of Kerguelen and south-east of New Zealand, wherever in factj there is local penetration of the West Wind zone by the branching of the East Wind stream. However, I w^here their major concentrations are concerned, the two species obviously occupy distinct 'ecological] niches', being segregated, as so many other Antarctic plankton animals are, in a seemingly uniform! environment in which as Mackintosh (i960) remarks 'there are few tangible barriers to dispersal'. THE FEEDING MIGRATIONS OF THE BALEEN WHALES. THEIR POSSIBLE ORIGIN AND HISTORY In most baleen whales in both northern and southern hemispheres there is a breeding-feeding rhythml involving them in extensive movements from subtropical and temperate breeding grounds into Arctic] (or Arctic-Boreal) and Antarctic waters to feed. It is worth while speculating as to how this remarkable] rhythm evolved, and how the whales came to undertake such enormous journeys as they accomplish ] today. ^ A parallel and well known phenomenon is the stunting of fish in overstocked lakes and ponds (MacGregor, 1959). THE FEEDING MIGRATIONS OF THE BALEEN WHALES 429 In SO far as the southern stocks are concerned it might be suggested that during a much colder regime, associated with a northerly extension of the polar ice-cap, the cold Antarctic currents carrying the krill penetrated into lower latitudes than they do today, perhaps so far north that at one time both breeding places and feeding places were coincident, or fell not very far apart. It could have happened then that when the ice began to retreat, and the northern limit of euphausian abundance to go with it, the whales learnt to follow their food into the higher and higher latitudes towards which it would imperceptibly have been receding. There is evidence from bottom cores (Deacon, 19606) that the Antarctic convergence has in fact retreated, to where we now find it, from a more northerly position held during the Glacial Period, and that during this same epoch the influence of the Antarctic bottom water must have been more far-reaching than it is today. Similar climatological change, involving a northward shift of feeding stuff from low to high lati- tudes, could also be invoked to explain the extensive migrations of the northern whales. Turning again to Antarctica and its geological history we find in this something that might have been contributing to the evolution of the feeding migration independently of climatological change. There must have been a time I imagine (see Matthews, 1959) when the scattered arcuate group of islands now known as the Scotia Arc was an unbroken mountain chain linking the Andes with Graham Land. Such a barrier, presenting a buffer to the Weddell stream, as it swept up from the south, might it is possible have diverted this strong current far into what are now temperate and subtropical latitudes, carrying the krill into, or very close to, the very regions where the whales now winter, pair and give birth to their calves. It could be imagined then that the gradual disruption or submergence of this mountain chain brought in its train a gradual weakening of the Weddell stream, a slow southerly shift of the krill it carried and again a movement of the whales that over the ages involved them in gradually but imperceptibly increasing southward journeys as they followed their retreating food. Perhaps in Antarctica geological and climatological history are equally linked up with the evolution of this remarkable rhythm. REVIEW OF DYNAMICS OF DISTRIBUTIONAL CONTROL It has been shown repeatedly in this report that the warm deep current carrying the ascending Second Nauplii, Metanauplii and younger First Calyptopes must play an important part in main- taining the krill population within its normal geographical limits and that in the Atlantic sector, where, underneath the Weddell stream, the cold bottom water is presumed to be carrying hatching eggs to the north and east (see pp. 100-102), there seems to be a strong likelihood that the southward movement of the warm layer, through which the resultant larvae are rising, will by some route lead to a replenishment of the coastal population in the East Wind drift that is moving clockwise round the Weddell Sea (see pp. 122-3). It <^^'^> however, be in the Atlantic sector alone that the high latitude East Wind population is being replenished by larval incursions from so far north, for elsewhere, throughout the vast extent of the circumpolar West Wind drift east about from meridian 30° E, there is, it appears (p. 194, Figs. 24 and 25), neither a northerly population of spent or gravid females nor a northerly population of deep south-borne larvae (p. 200, Fig. 28 and p. 307, Fig. 72) from which such replenishment could spring. It has been shown too (p. 169, Table 35), that throughout their circumpolar range the older stages of this species in the staple or over 20 mm. class do not at any time of the year, and least of all (p. 158, Table 28) in the West Wind drift, undergo a mass descent from the north-going Antarctic surface layer into the warm counter-flowing deep current below, so that again the west-going population in the East Wind drift cannot, anywhere it seems, not even 430 DISCOVERY REPORTS in the Atlantic sector, be replenished by incursions of breeders or potential breeders from lower latitudes. Thus while on the Atlantic or Weddell side we can postulate a southward dispersal of the very early larvae as a mechanism of distributional control, we cannot see the same mechanism at work all round Antarctica.^ In the absence of recruitment from the north, that is, from the West Wind drift, the continued existence of the high latitude west-moving population throughout a vast stretch of the East Wind zone, reaching from Graham Land west about to the neighbourhood of Enderby Land, becomes in fact in the present state of our knowledge a very difficult thing to explain. Obviously, however, if there is no recruitment from the north there must be continual recruitment from the east, and this would imply a continuous circulation that could only be maintained it seems if there were transference of East Wind water from the Weddell side to the Pacific side via the narrow channel of the Bransfield Strait. In his study of the hydrology of this enclosed basin Clowes (1934) finds no such transference, although he does find a strong surface movement of East Wind water from the Weddell side which spreads across the north-eastern end of the strait and down the coast of Trinity Peninsula, a movement that seems to be associated (p. 311, Fig. 75) with the drifting of young Calyptopes swarms for some little distance westwards into the channel. Although, however, no through east to west movement has been established it seems possible that it might sometimes happen, Clowes (1934), in a note on the ice conditions of this region, observing that during seasons of prevailing northerly and north-easterly winds the pack-ice will move out of the channel in a south-westerly direction, sometimes, as in March 1924, travelling to the south-west right into the middle of de Ger- lache Strait. The Oceanographic Atlas of the Polar Seas published by the U.S.N. Hydrographic Office in 1957 and Muromtsev (1958, Fig. 80) both show a supposed rotary movement at the surface in the Bellings- hausen Sea immediately to the west of Alexander I Island, a movement involving northward transport of East Wind water into the West Wind zone near Peter I Island and its supposed return, bending southwards and south-westwards, along the Charcot Island-Alexander I Island coast. Such a move- ment, if real, 2 could, it might be supposed, produce a small-scale circulation of euphausians involving the local replenishment of the East Wind drift with either surface-borne larvae (p. 311, Fig. 75 and p. 315, Fig. 78) or older stages returned from the West Wind zone. It is just conceivable, too, that from such an enclosed or locally circulating community the recruitment of the west-going population in the East Wind drift could have its westerly beginnings, without postulating any move- ment of East Wind water from the Weddell to the Pacific side by way of the Bransfield Strait. On the Atlantic or Weddell side of Antarctica Meyer (1923, end chart) shows a cyclonic movement at the surface between 20° and 50° W with its centre in the heart of the Weddell Sea in about 70° S, 35° W, a rotary movement involving the return of part of the east-going Weddell stream to the East Wind drift. Farther east he shows another movement, again involving the return of Weddell water to the East Wind zone, with its centre north of Enderby Land. Hansen (1934, Pis. iv-vii) and the, folder maps in the end-pocket of Atlas over Antarktis og Sydishavet published by Hvalfangernes Assuranceforening (1936) show similar movements, evidently taken directly from Meyer. Kruger (^939> Fig- 68) reproduces Meyer's original map. Liineburg (1940, Fig. 10) shows a supposed current ! moving counter to the Weddell drift ('Weddellgegenstrom') flowing from about 25° E to about 15° W between latitudes 62° and 63° S. Hentschel (1941, Beilage) follows Meyer, Ritscher (1942, Fig- 36), however, showing only his western vortex. Schott (1942, PI. xxii) also shows Meyer's western 1 ' Except in the sense that the deep larvae in the East Wind drift itself (p. 123) are constantly, it seems, being carried by | the warm layer deeper (in the latitudinal sense) into this coastwise current. - The track of the drifting ' Belgica' (de Gerlache, 1943), beset in the pack of the Bellingshausen Sea, does provide some evidence of a rotary movement at the surface here, and according to Rouch (1922) a southward movement probably does exist. 1 REVIEW OF DYNAMICS OF DISTRIBUTIONAL CONTROL 431 cyclone, but places its centre some 5° farther north than Meyer originally placed it. Sverdrup (1942, chart 4), also evidently following Meyer, shows eastern and western cyclones at the surface. Sverdrup, Johnson and Fleming (1946, chart vii) and the Marine Atlas published by the U.S.S.R. Ministry of Defence (1953) again show two rotary surface movements, an eastern and a western, both of them centred about much the same positions as Meyer's. Without figuring it Colman (1950) states there is a clockwise gyral in the Weddell Sea. Leip (1954, p. 21) shows a western vortex only, revolving about 30° W with its centre a little to the north of 70° S, Muromtsev (1956, p. 58) a western cyclone in approximately the same position and an eastern one, again involving the return of the Weddell to the East Wind stream, between 30° and 60° E. Deitrich (1957, Tafel 5), Taljaard (1957, chart i) and Yevgenov (1957, Fig. 18) show the two Meyer cyclonic systems, but Taljaard questions the reality of the eastern one. The 1957 American oceanographic atlas shows only the eastern system, omitting Meyer's rotary circulation in the heart of the Weddell Sea. Bruns (1958, Fig. 105) also omits Meyer's western vortex, but shows the Weddell stream curling back to the East Wind drift between 40° and 80° E. The Times Atlas of the World (Bartholomew, 1958) shows the surface stream moving coastwise round the Weddell Sea, up the east coast of Graham Land and away to the north and east. It omits both eastern and western vortices and in this (see below) probably provides the most reliable interpretation to date of the available hydrological evidence. Ruud (1932) ascribes the abundance and continued existence of the population in the western and eastern parts of the Atlantic sector to Meyer's cyclonic systems, suggesting there may be similar systems operating in other parts of Antarctica, notably, for instance, on the outskirts of the Ross Sea. He suggests that although the bulk of the larvae produced in the western part of the Weddell Sea, probably he supposes in the inaccessible heart of Meyer's western vortex, get carried to the north and east in the surface stream, some get involved in the cyclonic movements and so as adults eventually return to spawn in the higher latitudes from which they sprang. In other words he postulates a mechanism of distributional control involving rotary movements in the surface alone. It will be recalled, however (p. 97), that Ruud's nets did not go deep enough for him to see, as Eraser did, the part the warm deep current, as an early larval carrier, may also play in the return of the krill to the south. Although basically correct, in so far as there is on the Atlantic side an undoubted clockwise move- ment of developing larvae in the East Wind- Weddell surface stream, Ruud's circulation, in so far as its terminal phase (the return of his adults in the surface) is concerned, is based as Eraser observes on very scanty hydrological evidence. Wiist (1933), for instance, suggests that Meyer's two cyclonic movements do not in fact exist, nor do the comprehensive hydrological data of the Discovery Investigations (Deacon, 1937, p. 28) provide any definite indication of a southward movement at the surface linking the east-going Weddell with the west-going East Wind stream. Deacon writes, 'The movement towards the west, the northward current along the east coast of Graham Land, and the current flowing out of the Weddell Sea towards the east, form three parts of a cyclonic movement which extends across the entire width of the Atlantic Ocean. The surface temperature distribution indicates that the cyclonic movement may be completed by a southward movement between 20° and 40° E ; there is, however, very little evidence of such a current at the surface '.^ Model (1958), using all available hydrological data to date, but ignoring data obtained between the surface and 75 m. shows what appears to be evidence of a shallow southward subsurface movement in i8°-25° E at a depth of 75-100 m., and, like Deacon, a strong south-westerly movement in the warm layer between 200 and 400 m. 1 Schott (1924, Fig. 34), Brown (1927, Fig. 10), Debenham (1923, end map A) and Aagaard (1930, 11, end chart) also show the coastwise current moving clockwise round the Weddell Sea, but nowhere any southward return at the surface. 432 DISCOVERY REPORTS I therefore conclude that while the circulation of the krill on the Atlantic side is maintained partly by the clockwise movement of the East Wind-Weddell surface stream, as Ruud originally supposed, at the beginning of the life-cycle the northward and eastward spreading of the bottom water, which there is reason to believe (p. loo) is carrying hatching eggs, and the eastward and southward move- ment of the warm deep water (p. ii8) carrying the ascending larvae, are equally involved. In other words, the only southward movement of the population that the hydrological evidence seems to allow possibly occurs when, as Nauplii, Metanauplii and younger First Calyptopes, the young krill are climbing towards the surface, and even then, I believe, only while they are passing through the warm deep layer. If, however, as has so often been suggested, there is in fact a southward movement at the surface somewhere near 30° E, and this movement does involve the return of Weddell water to the East Wind zone, then the dynamics of control, on the Atlantic side at least, would appear to be simple enough, and the circulation of the krill in this region would in fact follow the broad lines originally postulated by Ruud.^ Such a movement, too, would help to explain the relative poverty of the krill in the West Wind drift east of 30° E. Model's shallow southward subsurface movement between 18° and 25° E, if real, would also help to explain it, in so far as it might lead here (or in 30° E) to the gradual transference of larval or adult swarms from the Weddell to the East Wind zone, if, as seems possible (pp. 105-18 and pp. 268-78), some such swarms undergo a shallow vertical migration by day. If such a trans- ference does in fact take place it would postulate that Model's subsurface current was travelling more quickly than the surface stream and that the migrating swarms, especially the larval swarms, spend more time in it than they do at the surface. We cannot as yet say that they do. The evidence in fact points rather to the contrary (see pp. 105-23, 157-70 and 268-78). It may be too, although Deacon's classic survey of Antarctic hydrology (p. 99) does not suggest it, that all round the continent there is small-scale local cascading of shelf water forming, within the East Wind drift itself, miniature cold bottom currents, which, moving away to the north and east, might possibly give rise to a series of local cold bottom-warm deep-surface circulations like that of the major rotary complex we find in the Weddell Sea. In the absence of transport of East Wind water from the Weddell to the Pacific side by way of the Bransfield Strait a series of such complexes, if real, would help to explain the recruitment and continued existence of the west-bound East Wind popula- tion beyond the Atlantic sector. If, moreover, the cyclonic movement (p. 47) claimed to have been discovered by the Russians off Enderby Land is correct then one could conceive the East Wind popula- tion as being in some measure at least, and in some places, static, part of it perhaps not always travelling to the west. As Sverdrup (1933) remarks, although the principal movement in high latitudes bordering the continental land is towards the west several 'whirls of stationary character appear to exist, and these are present partly because of the prevailing winds and partly because of the bathy- metric features'. Finally, the East Wind zone, enjoying as it does a certain measure of immunity from the major ravages of whales, seals, penguins and possibly other animals, may be regarded in a sense as a relatively untapped reserve of whale food and as such, whatever the factors contributing to the maintenance of its own population may be, must continually be contributing something to the replenishment of the more heavily suffering krill in the Weddell stream. In his recently published Marine Ecology Moore (1958), referring to the Antarctic euphausians as a whole, states that there appears to be ' a regular cycle in which immatures and adults are, in general, ' Recent observations by the Japanese during; the International Geophysical Year suggest rather strongly that such a movement might in fact exist, there being evidence of particularly strong 'recurvature' in 45° E (Kumagori and Yanagawa, iQS^**)- I would repeat, however, that our Discovery observations provide no evidence of this phenomenon. REVIEW OF DYNAMICS OF DISTRIBUTIONAL CONTROL 433 carried into lower latitudes by the shallower water movements, and the eggs are returned to high latitudes by deeper, southward-flowing movements '. In so far, however, as E. superba is concerned, it is to be observed that while some larvae, Calyptopes and Furcilias, are unquestionably carried north- wards (and eastwards) in the surface stream, there is little evidence of any mass return of eggs from northerly spawnings to higher latitudes at a deeper level. On the contrary, it seems, the eggs move north (and east) from southerly spawnings and it is only the very early larvae, the Nauplii, Meta- nauplii and some of the First Calyptopes, that are ever involved in a deep southerly return. As for other Antarctic species little is yet known of their major movements. At least one of them, however, E. crystallorophias, obviously cannot be included in such a simple circulatory scheme, it being strictly neritic, never it seems, either in its larval or adult state, leaving the immediate vicinity of the Antarctic continental shelf. For the rest, John, who has written at large on the distribution of the southern species, makes no mention of a deep return of eggs, neither do Rustad nor Ruud. Fraser does mention a ' deep distribution of E. superba eggs ', but he was referring to a gathering on the Antarctic conti- nental slope in a region where there was no deep southward movement. Moore gives no particular authority for the remarks I have quoted, he does not consult Fraser, nor in this profusely and widely documented book does he refer to either Rustad or Ruud. Ekman (1953), although he does not mention a deep return of eggs, makes much the same sweeping generalisation as Moore. Referring to the ' more important cold-water ' southern species, E. crystal- lorophias, E. superba, E. frigida, E. triacantha and Thysanoessa macrura, he remarks, ' In these species, as in Calanus acutus, the fully grown individuals live only in the surface water and are transported with it towards the Antarctic Convergence, while the younger stages are probably returned by the antarctic return current to the pack-ice where the population has its maximum concentration'. Of these five species, however, again only one, E. crystallorophias, in view of the very high coastal latitudes to which it is confined, could be said to have its maximum concentration in the pack, or at any rate in very cold water, although even in this species, as (p. 75) in E. superba, association with the ice will probably prove to be nothing more than a seasonal phenomenon. Of the others, so far as is known, in E. superba at least, it is the larvae, from the First Calyptopis onwards, rather than the adults, that are carried for vast distances in the surface drift, not necessarily, however, towards the Antarctic convergence, but to the west in the East Wind zone and to the north and above all to the east in the Weddell stream. E. triacantha (Fig. 157) to take another, is certainly not a cold water form, its major concentrations are far from the pack and it does not live (Baker, 1959) only in the surface but in its adult state alternates daily between the surface stream and the warm deep layer. As for E. frigida and T. macrura the major movements of the adults and larvae have not yet been worked out, nor is it yet known that the fully grown individuals of either species are confined to the surface stream. Or to quote Lucas (1956a) on the effect of currents and surface 'swirls' on the distribution of drifting animals, 'The process is also important in the vertical plane; otherwise, for example, the steady drift of cold surface waters from the Antarctic region would tend to depopulate it of plants and surface larvae. Mackintosh (1934) has shown a classic instance in the life history of Euphausia superba, whose surface larvae drift away in their earlier stages, but sink later to be carried by the south-going sub-surface flow back to the Antarctic, where they mature and produce surface eggs'. In his Distribution of the Macroplankton in the Atlantic Sector of the Antarctic, which Dr Lucas cites. Mackintosh makes no such reference to the larval krill, nor does he in any earlier or later work. In a lecture delivered before the Royal Society in January 1950 he refers briefly (Mackintosh, 1950), to what my findings then seemed to show, namely, a massing of the adults at the surface, a sink- ing of the eggs to a depth of 1500 m. or more and a subsequent rising of the newly hatched krill in the course of which they got carried southwards to replenish the stocks in higher latitudes. 51 I'M 434 DISCOVERY REPORTS Ruud and Fraser it is true have shown, or at any rate suggested, that the later larvae drift northwards (and eastwards) in the Antarctic surface layer, or to be more precise in the Weddell stream, but no one has yet shown, or even suggested, that they subsequently sink into the warm deep current to be carried back to the south to mature. On the contrary, our bathymetric data show that perhaps the most striking thing about the whole larval life-cycle is that as the development from Calyptopis One to Furcilia Six proceeds (p. 112) the larvae, far from developing a downward trend, become more and more restricted to the surface drift, never, at any stage in this developmental phase, becoming involved in any mass descent into the warm deep stream. Nor is it certain that when they do mature they produce surface eggs. Obviously, even so far as they are known, the factors controlling the distribution of this species are complex, the problem as a whole presenting many puzzling features that perhaps for long will remain to us obscure. And this must be true of other Antarctic plankton animals as well as of plankton animals elsewhere, that as Dell (1952) has said 'have no obvious method of maintaining themselves in the appropriate ecological niche '. In so far as the currents are concerned in the horizontal dispersal of the krill our observations repeatedly show that their influence, as might be expected, is most marked during the feebly swimming larval and post-larval phase. There is, I believe, horizontal dispersal in the deep bottom water, in the warm intermediate layer and at the surface. But the greatest and most far reaching of these movements, involving the northward and eastward transport of the larvae and their ultimate spreading over the entire face of the Weddell stream, is in the surface drift. In the simplest of terms we can say this movement has its roots in the highest latitudes of the krill's geographical range, being engendered, as it is, by the great current that sweeps coastwise round the continental land, the East Wind stream which Powell (1951), in his account of our Antarctic MoUusca, has described as a 'present' factor that 'must greatly facilitate the lateral distribution of many species'. One last point. It has been shown repeatedly that the krill as they grow tend to become less and less strictly planktonic. A young, only half-grown, swarm for instance (p. 155) has been seen to maintain its position close to a fixed point of reference for hours on end in face of a current of about \ knot, and from this it would appear distinctly possible that the full-grown swarms in similar circum- stances could hold their ground for days, perhaps for many days, on end. Could it not then be that at least the breeders in the East Wind zone, where the average speed of the current is also probably about \ knot, are to a large extent static, not for ever drifting to the west, but keeping station on fixed points of reference for long periods, for all we know perhaps even throughout life} There must be many such points all round the continental coast, rocks, rocky islets, islands, rocky headlands, the ice-foot, the barrier faces, stranded bergs, even the ice-floes themselves, which, although in general moving westwards under the influence of the prevailing wind and current, are often driven due north by violent southerly gales, perhaps even out into the West Wind drift zvhere the current is travelling east not west. If there should in fact prove to be a static population of spawners, or potential spawners, in these high latitudes, and somehow or other we could explain their recruitment when eventually they die off, then we should have a permanent and to some extent untapped nucleus of individuals from which the northern population might spring. I end with a passage from the English translation of Paul Budker's recent book Baleines et Baleiniers. ' Then, the future mothers being filled and the males satisfied — and there is nothing to make us think otherwise — they set out for the south, to find once more the inexhaustible fields of krill, this profusion of provender that is widespread, but not uniformly so, around the Antarctic Continent. And the banquet is resumed, to go on until the next southern autumn. Then always in March to April the I REVIEW OF DYNAMICS OF DISTRIBUTIONAL CONTROL 435 long voyage north will begin again, but it will end in the "happy event" awaited since the preceding season. The whale-calf then conceived will see daylight for the first time in the same warm waters which "cradled" his parents a year previously'. The fields of krill it is true seem inexhaustible, but one wonders how long the even yet not hopelessly depleted stocks of their giant predators will be permitted to enjoy them and for how long they will be allowed, in reasonable security, to conceive, bring forth and nurture their kind. One looks however with hope to the future, for as Bertram (1958) has written, 'The present International Whaling Convention is a remarkable instrument which, if it continues to be applied effectively and even more stringently, may be a monument to the idealistic zoologists of the mid- twentieth century'. SUMMARY 1 . A review of early historical records and recent literature emphasises primarily the abundance and ecological importance of this species in Antarctic waters, its predominantly surface habitat and the circumpolarity and unevenness of its distribution (p. 40). 2. The scope of the observations is described and the limits imposed on them by the polar pack, it being manifest that the periodic freezing and unfreezing of certain large areas of the circumpolar sea and the permanent inaccessibility of others combine to aggravate the difficulty of studying its distribution (p. 48). 3. A broad practical outline is given of the principal features that help to distinguish it in plankton samples, all developmental phases being covered from the egg to the adult state (p. 55). 4. The high latitude coastal current in the East Wind zone and the low latitude oceanic current from the Weddell Sea which the former feeds, together with the Bransfield Strait and South Georgia, to which both East Wind and Weddell water penetrate, are the principal regions of its abundance. Its circumpolarity, therefore, unlike that of many other Antarctic plankton animals, is essentially asymmetrical, the circumpolar West Wind drift being barren or very sparsely populated, except where fed locally from the East Wind drift or by overflow from the Weddell stream (p. 57). 5. Its association with pack-ice throughout every phase of its surface development is seasonal. In winter and spring vast stretches of the principal regions of its abundance are frozen over and it becomes a creature of the ice, but in summer and autumn equally large stretches are unfrozen and it becomes equally a creature of the open sea. Moreover, in certain areas, such as the waters round South Georgia, and perhaps the northern fringe of the Weddell drift, which are neither involved in the winter freeze-up nor affected by the drift of the pack, the krill are never associated with ice at all except in so far as such areas are periodically infested by bergs or are penetrated by water in which ice may have melted previously somewhere 'upstream' (p. 64). 6. It is confined almost exclusively to water south of the Antarctic convergence, water in which the surface temperature range shortens quite appreciably with increasing latitude, the winter minima and summer maxima for the four principal regions of its abundance being: South Georgia —175° and +3-90° C, Bransfield Strait -1-89° and +2-69° C, Weddell drift -1-89° and +2-50° C and East Wind drift —1-89° and +1-45° C. As an older adolescent and adult it encounters its warmest surface conditions (+i-oo° to +3-99° C) on the northerly South Georgia whaling grounds from January to March and its coldest (-2-00° to +1-99° C) in the Weddell and East Wind zones, the latter, however, slightly the colder, during the same period. In the East Wind drift the surface population as a whole, larval, adolescent and adult, spends fully 10 months of the year in temperatures which rarely if ever rise above zero C, the larvae and very early adolescents enduring these severe conditions virtually throughout the year (see p. 409, note i). From June to November the vast majority 51-2 436 DISCOVERY REPORTS of the larvae, everywhere, are developing in sub-zero temperatures, while the very early adolescents, small whale food in the 1 6 — 20 mm. range, owing to the time of year in which they appear in the plank- ton, spend virtually their whole existence in all four regions of euphausian abundance in water that never rises above zero C (p. 75). 7. Study of the vertical distribution of the larvae reveals that in oceanic water the eggs hatch at great depths, possibly far below 2000 m. if the water is deep enough, the resultant larvae rising until they reach the surface as the First Calyptopis stage. The subsequent development, Calyptopis One to Furcilia Six, takes place in the Antarctic surface layer (p. 88 and p. 97). 8. In the course of this ' developmental ascent ' as it has been called the new-born krill pass through their Naupliar and Metanaupliar, and through the early part of their First Calyptopis, existence (P- 97)- 9. In the northern or Weddell zone it is distinctly possible that the young krill are rising out of the northward and eastward-spreading Antarctic bottom water. If so, both this and the surface drift must be involved in their horizontal dispersal to the east (p. 99). ID. Diurnal vertical movement in the larvae, once they reach the surface, is limited, taking place for all practical purposes within the limits of the Antarctic surface layer. Such movement as there is is perhaps most pronounced in the Calyptopis stages, but becomes increasingly less pronounced as the larvae grow, the Furcilias, notably the later Furcilias, tending to be massed more or less per- manently in the surface (50-0 m.) layer. With their limited diurnal movement the young stages, especially the Furcilias, are readily inclined to be carried along in the surface drift (p. 105). 11. As they climb towards the surface the larvae, as Second Nauplii, Metanauplii and First Calyptopes, spend some considerable time, perhaps up to 30 days or more, in the warm deep current which in most parts of the Antarctic has a southerly component, and which, therefore, plays an important part in maintaining the population within the normal limits of its geographical range (p. 118). 12. The absence, or virtual absence, of the krill from extensive areas of shelf water such as exist at the head of the Ross Sea may partly be ascribed to the failure of the warm deep water to effect any major penetration on to the shelf (p. 123). 13. The astronomical abundance in which it exists and the key position it holds in the industry and ecology of the Antarctic seas is revealed by a review of the vast and catholic multitude of its predators (p. 126). 14. The southern baleen whales feed principally on adolescent and adult krill over 20 mm. long (the ' staple whale food '). In the early part of their sojourn on the feeding-grounds, however, notably in spring, they augment their staple diet with large quantities of much smaller euphausians consisting of mixtures of Sixth Furcilias and early adolescents (the 'small whale food') ranging from 11 to 20 mm. long. Whales have also been recorded feeding on the Sixth Furcilia in autumn (early April) when this stage, in some abundance, first appears in the plankton of the western Weddell drift (p. 138). 15. It has been roughly calculated that between 1933 and 1939 the krill, at a conservative estimate, were being grazed down by large baleen whales alone at the rate of about 38 million tons a year, a figure which if expressed in terms of euphausians eaten comes to 47 million million (p. 145). 16. The commencement of the phenomenal burst of growth that occurs in foetal baleen whales during the last two months of pregnancy coincides with the arrival of their mothers at the richly spread table of the Antarctic feeding-grounds (p. 147). 17. The unevenness of its surface distribution, its patchiness or swarming habit, in so far as this is a phenomenon displayed by the adolescents and adults and can therefore readily be observed from the decks of vessels, is described in detail. In the northern zone, notably in the Weddell drift, SUMMARY 437 widespread patchiness has been encountered, sometimes involving areas estimated to be as much as 150 square miles or more in extent. In such areas, as a rule, there are distances of from about a third to a quarter of a mile between patches. Individual patches, although very variable in size, are seldom if ever very large, the smaller covering a few square yards, the larger half an acre or so, or sometimes perhaps a little more. They exhibit an almost infinite variety of shapes largely because, judging from one or two inshore swarms which have been closely observed, they are in a continual state of flux, constantly changing their form, expanding and contracting, elongating this way or that, sometimes appearing as if about to divide or break up, only to re-form, the individuals reversing or changing direction in unison, the whole phenomenon imparting to the swarm the appearance and activity of an amoeba. In colour they vary from pale straw yellow, through ochre, mahogany brown to brick or vivid blood red. They are seen from time to time right on, or very close to, the surface of the sea and appear to be of very shallow draught, in general being disposed in shallow rafts no more than a yard or two thick. The individuals pack tightly, their density, to the eye at least, appearing to be of the order of one euphausian to the cubic inch. Both adolescent and adult swarms are composed of exceedingly active animals reacting either individually or in unison with great rapidity to external stimuli. Where they have been watched in shallow water close to fixed points of reference they have been seen to be capable of maintaining their formation intact for long periods over a given position against a current of considerable force. In general it can be said these vigorous swarming individuals have ceased to be strictly planktonic. The swarm in fact appears to be a unit that, behaving it seems as a single organism, does not break up except when violently disturbed as, for instance, by the passage of a vessel, a unit that even so will quickly re-form (p. 148). 18. Direct observation and the analyses of a very large series of net samples covering a wide bathy- metric range reveal a massive concentration of the whale food at or not far below the surface, and that while its absolute vertical range may extend from the surface down to about 1000 m., in so far as it may be said to exist in concentrations dense enough to satisfy the needs of the whales, it is confined very largely, especially at night, to an extremely narrow zone that does not, it seems, go deeper than 5 or 10 m. below the surface. By day this crowded zone may become somewhat depleted owing to the vertical migration of some of the swarms, particularly the older swarms, to deeper levels. The daytime movement, however, does not seem to extend to depths greatly in excess of 40 or 50 m. It seems, too, to be very erratic and to involve only part of the surface population. The analyses show, too, that there is no mass seasonal descent of either younger or older stages into the warm deep water flowing counter to the surface stream (p. 157 and p. 268). 19. It has been roughly calculated that the density of the krill in the central part of the Weddell drift might be as high as 30 g./m.^ or approximately 50 times the equivalent weight of whale flesh in the inner Antarctic zone (1° isotherm to ice-edge). The same ratio it seems might hold good through- out the East Wind- Weddell surface stream (p. 170). 20. The dietary of E. superba is described in detail. It is clearly and almost exclusively a voracious herbivore, and although at first sight the regular occurrence of certain small spineless diatoms in the stomachs seems to point to some measure of selective feeding, it may be that these forms, being strongly silicified, are not very easily digested, and that many other less strongly silicified species are equally important as food (p. 172). 21. The enormous abundance in some vertical net samples of diatomaceous faecal pellets suggests strongly they are the excreta of the shoaling krill, their more or less uniform distribution from the surface down to 1000 m, indicating they are sinking (p. 176). 22. While much further exploration is required to confirm it, there is some evidence that spawning as a major event is a relatively shallow water phenomenon associated principally with the shelf or slope 438 DISCOVERY REPORTS waters of the high latitudes of the continental land, the discovery there of concentrations of eggs and Nauplii very close to the sea bed suggesting the distinct possibility that in these conditions the eggs are laid on, or sink until they rest on, the bottom itself. A major spawning, lasting from November to March, seems to be associated with the far western (and probably south-western) reaches of the Weddell drift, and another, lasting from late January to March, with the slope waters of continental Antarctica extending from the mouth of the Ross Sea westwards to the Princess Martha Coast, which is as far as our observations go. We have many samples up to very high latitudes on the Pacific side, although relatively few come from the East Wind zone, especially between 90° and 150° W where the ice-cover is so heavy. Such as we have, however, from the high latitude west-going stream, between 80° and 160° W, show little sign of any spawning there. There are many indications too that, although both gravid and spent females are commonly encountered on the South Georgia whaling grounds, there is, paradoxically, virtually no spawning and certainly no successful hatching there. Hatching, in so far as it appears to be a successful event, is essentially a phenomenon of the open sea, taking place in the Weddell zone at great depths, in most instances far away from the continental land. In the East Wind zone there is clear evidence that it takes place on or near the bottom on the continental slope and at greater depths beyond, but notjvery far beyond, the 1000 fathom line. In general hatching in the East Wind zone is much more closely associated with the land than it is in the northern zone. There is no evidence in either the northern or southern zone of large-scale hatching on the shelf itself, nor is there any that the gravid females go to the great depths at which hatching principally takes place in order to lay their eggs, the vertical distribution of the spermatophore-carrying males and females and of the spent and gravid females suggesting that both pairing and spawning are essentially surface phenomena. There is evidence that the eggs themselves, although liberated near the surface, may sink to deep levels, but it is not satisfactory enough to be conclusive. In view of these contradictory phenomena — major spawning in or near coastal waters but hatching in the deep ocean, eggs manifestly concentrated at great depths but gravid females to all appearances at the surface — it is suggested that sinking Weddell shelf water, flowing away to the north and east as the Antarctic bottom current, out of which it seems (9) the new-born krill below the Weddell stream are rising, is perhaps responsible in part at least for carrying the coast eggs away to hatch at much deeper levels farther out to sea. There is an obvious need, however, for systematic exploration of this cold deep stratum, so far scarcely touched even by our deepest vertical nets, and for accurate measurements of its speed, before any reliable picture of the full extent of its influence can be formed. There is much need, too, using large nets, for systematic exploration of the great depths, down to say 3000 m. or more, at which it seems the oceanic eggs are hatched. These are levels far below the deepest observations we have yet made, and it may be that it is there that the gravid females are massed when about to lay their eggs. If, however, they do go to such enormous depths, a journey it seems well within the powers of this large and vigorous species, and one that could happen quite quickly, it is surprising that of the hundreds of observations made below 100 m. during the spawning season not one, with the exception of a single instance on the South Georgia whaling grounds, reveals the slightest evidence of a concentration of them on their way from the surface down. The only alternative explanation of our failure to find evidence of deep oceanic spawning might be that it is undertaken not by massed formations of females, but by single individuals, so widely scattered as virtually to defy capture even by our largest nets, and even this seems wholly inconsistent with the swarming habit (24) to which this species, from hatching right through to the gravid state, seems lifelong prone (p. 176). 23. Although no current measurements have yet been made in the bottom water, historical evidence suggests that in high latitudes below the Weddell Sea, where it seems likely to be carrying hatching eggs, it may be moving fairly quickly (p. 212) I SUMMARY 439 24. Neither animal exclusion nor pack-ice appears to have anything to do with patchiness which seems to spring from the lifelong habit of this species of congregating in discrete swarms in which the individuals are all of the same, or much of the same, age. This phenomenon can be traced from the earliest deep appearance of the Nauplii, throughout the developmental ascent, the whole subsequent life-cycle of the larva in the surface, through early and late adolescence and up to the adult state. With the advance of summer and autumn in the northern zone the age pattern exhibited by the surface swarms becomes increasingly complex as successive batches of eggs hatch out and the existing swarms become augmented by others of later generation. In the southern or East Wind zone, however, largely because of the much compressed spawning season there, the corresponding age pattern is distinctly homogeneous. The rafting or shallow draught of the swarms (17) seems to be a phenomenon persisting from hatching onwards (p. 215). 25. The gaps or 'voids' separating the swarms, so long as they remain untenanted, must be enormously richer in phytoplankton than the water occupied by the congregating krill which in their packed myriads must constitute a grazing unit of immense local destructive power (p. 240). 26. The whales, once they reach the feeding-grounds, probably come upon their food partly by chance, partly by sight and partly, it is suggested, through the sensitivity of the wax plug in the external auditory meatus to high-frequency vibrations produced perhaps by the swarming krill (p. 241). 27. Many of the smaller swarms are probably swallowed by the whales outright, although clearly the larger swarms would not be demolished so quickly. Single whales, or small groups of them, would tend to remain longer in an area of good feeding than large schools, which, by virtue of their greater capacity for local destruction, would shortly have so grossly depleted the swarms on which they preyed that they would be compelled to move on in search of less impoverished fields. The results of whale- marking suggest that the feeding whales do not by any means move entirely at random, but in the main tend to travel against the surface drift in the East Wind-Weddell stream that is carrying their food as it flows (p. 243). 28. The swarms in general are composed of males and females in approximately equal numbers. In adult swarms, however, it is not uncommon to find the gravid females vastly predominating over the males, suggesting that the latter after pairing are the first to die off. Almost throughout the life of the swarm the male is the dominant partner, being slightly larger and always more sexually advanced than the female. After pairing, however, the female appears to undergo a short period of accelerated grovi^th, becoming in the end, both in size and numbers, the dominant partner in the swarm (p. 245). 29. The growth of the swarms in the northern and southern zones is compared, it being apparent that in both stature, in so far as length frequency may be said to express stature, and development, the East Wind swarms, following the late and much curtailed spawning that occurs in these high latitudes and the subsequent slow larval growth-rate there, tend persistently to lag behind their northern contemporaries virtually throughout their existence in the plankton (p. 251). 30. The occurrence of cast skins in the plankton coincides (i) with the period when both yearling and two year old krill are growing at their maximum intensity, and (2) it seems with the spawning season (p. 256). 3 1 . The reaction of the krill to ship and nets in varying conditions of light intensity and turbulence of the sea is examined in detail, it being apparent that the samples obtained by conventional stern nets provide little indication of the natural density of these animals in the sea (i) because the ship in its passage may scatter their surface concentrations, and (2) because of the comparative ease with which the krill, both as individuals and swarms are able to dodge the nets. Our samples, however, manifestly provide a basis for an estimate of the relative density or abundance of this species in one region or another and equally for estimates of the limits of its optimum abundance and absolute 440 DISCOVERY REPORTS geographical range. The krill as they grow develop a mounting capacity to avoid the surface nets, a capacity most conspicuous in full daylight, but decreasing as the daylight wanes. In view of this capacity and because with a swarming, predominantly surface living, population such as this, a net towed on the surface (especially at night) is manifestly more likely to strike a swarm than a net hauled open to the surface obliquely from below, certain rather arbitrary corrections have to be applied to the catch-figures of the stramin nets, corrections affecting principally the samples from the open oblique nets fished in the loo-o m. layer (p. 258 and p. 278). 32. A monthly survey of the distribution and relative abundance of the larval stages based on the data from the vertical nets shows that in the oceanic water of the northern or Weddell zone they first appear as Nauplii and Metanauplii deep down in the far western part of the Weddell drift in December, reaching the surface there on a major scale as First Calyptopes by about mid-January. Under the dual influence of the bottom water and the surface drift, and here to some extent also of the warm deep water, they gradually spread to the east, being encountered in great abundance about half-way along the Weddell drift in March and throughout the whole of this surface stream by the end of April. From then until the last surviving Sixth Furcilias moult and become adolescent in December they continue to fill the Weddell stream, the influence of which as a major instrument of their dispersal appears to cease somewhere near 30° E. The whole of the development from Calyp- topis I to Furcilia 6 takes place in the surface waters of this current and in the surface waters of the South Georgia whaling grounds and Bransfield Strait affected by it, the first surface swarms appearing to drift into the strait in March and to reach the South Georgia area on a major scale at the latest by May. In the southern or East Wind zone the first deep larvae are encountered in oceanic water in late January and early February, the first surface swarms appearing about the middle of the latter month. It is not until March, however, that the completion of the developmental ascent becomes general there (p. 284). 33. In the northern zone the larvae grow rapidly, the first Sixth Furcilias appearing towards the end of March, the whole surface life-cycle, Calyptopis i to Furcilia 6, probably being completed on a major scale in from about 90 to 120 days. In the extreme cold of their southern environment they grow more slowly, the surface life-cycle taking upwards of from 9 to 10 months to complete. Their resultant rate of westerly drift in the East Wind zone and easterly drift in the Weddell is probably of the order of from 8 to 14 miles a day (p. 349). 34. A seasonal survey of the distribution and relative abundance of the massed surface larvae, the adolescents and the adults, based on the material from the stramin nets, reveals, in an even more striking manner than the vertical nets are able to convey, the magnitude of the larval outburst that takes place in the western reaches of the Weddell stream in summer and the subsequent spreading of these stages, developing as they go, to the east. Summer may be described as essentially a time of single abundance when the staple whale food is at its peak and at that peak is spread throughout the feeding-grounds to their farthermost geographical limits. It is a time, however, when the great mass of the larvae, just launched on their surface existence, are confined to the western reaches of the Weddell drift and when the small adolescent krill (the small whale food), shortly about to outgrow themselves, survive principally in the higher latitudes of the East Wind zone. Autumn is essentially a time of purely larval abundance, the young stages vastly outnumbering the older especially in the Weddell zone. It is a time when, the surviving early adolescents having outgrown themselves in the southern zone, the small whale food has virtually everywhere disappeared from the plankton and a time too when the staple class, the paired and spawned adults evidently having died off, is practically everywhere, except possibly in the East Wind drift, at a low, perhaps its lowest level of abundance. The autumnal scarcity of the staple class is very probably associated too with the magnitude of the SUMMARY 441 depletion it suffers through the ravages of spring and summer. Winter is essentially a time of dual abundance with both larvae and the early adolescents to which they are giving rise, the former in enormous numbers, dominating the northern zone again to its maximum limits. It is a time, in the northern zone at least, of continued scarcity of the staple class. In the East Wind or southern zone, however, the larvae are probably the dominant winter class, with the staple whale food of secondary importance and the small (i 1-20 mm.) group non-existent. Although the small whale food is dominant, especially in the East Wind drift, spring alone is a time of triple abundance when the surviving larvae, and both small and staple whale food are spread throughout the whole vast extent of the feeding- grounds to their farthermost limits (p. 358). 35. The circumpolar distribution of surface discoloration, supposedly attributable to krill swarms, follows very closely that of the gross distribution of the total euphausian surface population as revealed by the stramin nets (p. 412). 36. The great summer larval outburst in the western Weddell drift and the subsequent spreading of the larvae to the north and east in the surface stream are shown to be annual events (p. 412). 37. The local distribution of the krill round South Georgia reveals a conspicuous massing of these animals, in every phase of their surface development and at all times of the year, on the north-eastern side of the island where the influence of the surface drift from the Weddell Sea is itself most con- spicuous (p. 413). 38. In view of the relatively short period it remains open to the major depredations of whales and seals and the relative immunity vast stretches of it enjoy from the ravages of penguins, and for all we can tell of a multitude of other animals as well, the East Wind drift may be conceived as being a relatively untapped reserve of whale food and as such must contribute not a little to such other factors as may be involved in the conservation of the krill population (p. 423). 39. The distribution and relative abundance of E. superba and the smaller E. triacantha are com- pared, it being suggested that among other factors as yet obscure the great depths at which the krill eggs hatch, the lifelong swarming that ensues, the enormously rich pastures among which it finds itself and is so admirably equipped to feed, its immunity to EUobiopsid infection, and the protection from surface predators the overwhelming mass of its larvae enjoys throughout the polar winter, all contribute something to the immense scale on which E. superba has populated the Antarctic seas (p. 424). 40. The evolution of the feeding migrations of the baleen whales is considered in the light of the geological and climatological history of Antarctica (p. 428). 41. On the Atlantic or Weddell side of Antarctica the only mechanism we can clearly see working to maintain the krill population within its normal geographical limits is the eastward and southward dispersal of the very early larvae as they rise towards the surface through the warm deep layer, deep larval dispersal from the lower latitudes in which the Weddell stream is flowing resulting in this region in a recruitment of the high latitude west-going population in the East Wind drift moving clockwise round the Weddell Sea. There is no evidence, however, that the East Wind popula- tion elsewhere can similarly be replenished from so far north, the vast extent of the West Wind drift east about from 30° E carrying it appears neither an adult female population nor a deep south-borne population of larvae from which such replenishment could spring. Since at the same time there appears to be no deep return of breeders or potential breeders from low to high latitudes anywhere throughout the circumpolar sea the factors responsible for the continued existence of the west-moving population throughout the greater part of the East Wind drift remain for the present obscure. There is, however, in the East Wind zone itself a southward movement of the ascending larvae carried in the warm deep layer, a movement operating continually to the advantage of the coastal population in these high latitudes (p. 429). ... DISCOVERY REPORTS 442 APPENDIX. TABLES 62-4 I have presented the regional distribution and relative abundance of the krill, as it grows and spreads round Antarctica, on a long series of maps in which increasing orders of abundance are represented by circles of increasing diameter. Such a system, although widely used and on the whole perhaps the most satisfactory we have for portraying distribution in plankton animals, lacks exactness, for orders of abundance are very different from catch-figures, the higher orders, especially, often including gatherings covering an enormous range of magnitude. The following tables are therefore presented to show that the broad differences in regional abundance we repeatedly encounter throughout the circumpolar sea, the richness of the East Wind-Weddell stream and the poverty of the West Wind drift providing the most striking example, are revealed as equally striking phenomena when the figures representing the harvest of our nets are treated more exactly. Using the gatherings of the towed stramin nets, the horizontal (0-5 m.) net on the surface, the obUque (loo-o m.) net in the surface layer (the catch-figures for the latter corrected as requisite as described on pp. 59 and 282-3), this final presentation of the data is based throughout on observations confined to water where the krill might be expected to occur; in other words it refers not to the circumpolar sea as a whole but only to the region south of the 4° C isotherm (of the month of sampling), the thermal boundary which our temperature data seem to show (see p. 76 and especially Tables 4, 5 and 9) is for all practical pur- poses the absolute northern limit of the krill's geographical range. It deals mainly with the regional distribution and relative abundance as a circumpolar phenomenon, showing (Table 62), sector by sector, our average monthly and annual gatherings of the surface population in the Weddell Current and in the West Wind drift, and again, sector by sector (Table 63), the corresponding gatherings in the East Wind zone. It shows too (Table 64) something of the local distribution, presenting our average monthly and annual gatherings round South Georgia and in the Bransfield Strait, where so much of the early whaling took place. The numbers of net hauls from which the averages are derived are shown throughout by italic figures. As in earlier presentations of the data (p. 65) the surface population is divided into three broadly grouped developmental phases, larval (under 16 mm.), early adolescent (16-20 mm.) and older adolescent and adult (over 20 mm.). Attention, however, is again called to the apparent absence or great scarcity of the larvae in the Australasian, Indian Ocean and Atlantic sectors of the East Wind zone, throughout a long coastal belt known (p. 302) to be an important breeding ground of the krill. This is an anomaly that springs (p. 355) from the slow growth- J rate of the euphausians in these high latitudes, where from February right through to April (p. 361, M Fig. Ill) we find the larval population in the surface layer represented dominantly and almost^ exclusively by the First Calyptopis, which, being very small, escapes as a rule through the meshes of the stramin net. The complete absence of very young stages recorded in the Pacific sector between ^ 60° and 150° W seems on the contrary largely real, for there is little evidence that any substantial spawning is going on there (p. 360). In the West Wind zone, although the overall poverty of the surface population is everywhere pronounced, it will be seen that the scarcity of the three developmental phases, larval, adolescent and adult, is not so clearly marked between 30° and 120° E as elsewhere, and that the larvae, although not the older forms, seem to be gathered in some little strength in the Australasian sector and again in the Scotia Sea. In these localities minor anomalies, or slight departures from the rule, are to be expected, the Indian Ocean sector for instance being affected (p. 384) by spring overflow from the Weddell stream, and, south and south-east of Kerguelen, by the branching of the East Wind stream, the Australasian sector by another East Wind branching affecting the region north of the Balleny Islands. In the Scotia Sea there are three factors to be considered, and all, either separately or in combination, I .8 s s S 1 O -Si fe ' en Over 0 20 mm. r2 ? S Co ^ s •S - K. o ■^ . Unde Is mm. -fe ^2 •*-, = o K f^ g o 5 APPENDIX :^ : M I :5 Q000t^O\O^ 443 to I I O 0^ tri -t "^ V 5 •S s ~ to ^1 Si) « N o l>3 CO i o g o . f I lil ►2 '^"= > o" a o -3; :2 - s feJ 12; "N g '-' N ri 'O O 00 fn O CO O O^ O "i" fN. M mo N r*' M N O lO fO o N u-i m v£> ^ N lO r* ^ tN 2 ^ ■^ O so "^ "* O o O "- CO in ro 00 " ■<»• O M- =0 0\ V 0) ■5 J3 •ii O " S" 5 J ^ M O Z Q "S ^ S' n. S < II 3 >. an '^ 3 52-2 DISCOVERY REPORTS Table 63. Average gatherings in the East Wind drift. {Samples from surface and oblique i-m. nets. Corrected figures) 60' Pacific Sector A istralasian Sector Indian Ocean Sector A Atlantic Sector A — "^ W 150' W 120 °E 30° E Jo»H Under Over Under Over Under Over Under Ova No. of 16 16-20 20 No. of 16 16-20 20 No. of 16 16-20 20 No. of 16 16-20 20 Month hauls mm. mm. mm. hauls mm. mm. mm. hauls mm. mm. mm. hauls mm. mm. mm. December 3 — 29 69 4 9 3 294 1 — — 216 — January H — 9 139 30 - 13 334 11 — — 410 27 32 186 124 February 17 — — 128 24 — 5 214 19 — 2 3434 27 — 7 18 March 6 — — 1779 7 — — 12 13 6 34 1016 40 I 64 183 April-May — — 6 381 15 1482 17 7 — 8 Annual averages 40 — S 375 65 2 8 253 50 47 II 1842 111 9 70 102 Table 64. Average gatherings at South Georgia and in the Bransfield Strait. {Samples from surface and oblique i-m. nets. Corrected figures) South Georgia A Bransfield A Strait No. of Under 16-20 Over No. of Under 16-20 Over Month hauls 16 mm. mm. 20 mm. hauls 16 mm. mm. 20 mm. September 21 467 66 1363 — , October 16 53 204 43 — November 53 1185 3135 1 103 19 44 415 520 December 44 27 — 1439 14 I 3 69 January 74 4 4 982 2 — — — February 61 — — 2921 22 — — 2252 March 68 26 2 909 — April 17 215 — 50 22 329 — 95 May 10 331 — 56 — June — — July — — August 21 367 • — 214 — Annual averages 385 237 445 1220 79 102 lOI 791 might be responsible for the slight measure of larval abundance we have recorded there. They are (i) the influence of surface-borne incursions from the Bellingshausen Sea (p. 360), (2) the influence of the Antarctic bottom v^^ater (pp. 305-6), and (3) simply because this is a region in which, hydro- logical boundaries being what they are, it is often rather difficult to separate true West Wind occurrences from Weddell ones. REFERENCES Aagaard, B., 1930. Fangst og Forskning i Sydishavet (Oslo). Ahlstrom, E. H., 1943. Studies on the Pacific pilchard or sardine (Sardinops caerulea). 4. Influence of temperature on the rate of development of pilchard eggs in nature. Spec. Sci. Rep. U.S. Fish Wildl. Serv. no. 23, pp. 1-26. Allen, G. M., 1916. The whalebone lohales of New England. Mem. Boston See. Nat. Hist, viii, no. 2, pp. 107-322. 1942. Extinct and Vanishing Mammals of the Western Hemisphere. American Committee for International Wild Life Protection, Special Publication no. 1 1 (Lancaster, Pa.). Allen, J. A., 1899. Fur-seal hunting in the Southern Hemisphere. Paper no. 12 in D. S. Jordan's The Fur Seals and Fur-Seal Islands of the North Pacific Ocean, pt. 3 (Washington). REFERENCES 44S Andersson, K. a., 1905. Das hohere Tierleben im Antarktischen Gebiete. Wiss. Ergebn. schwed. Sudpolarexped. v, Lief 2, PP- 1-57- Andrews, R. C, 1909. Observations on the habits of the finback and humpback whales of the eastern north Pacific. Bull. Amer. Nat. Hist. XXVI, pp. 213-26. 1916. Monographs of the Pacific Cetacea. 11,— The set whale (Balaenoptera borealis Lesson). Mem. Amer. Mus. Nat. Hist. n.s. I, pt. VI, pp. 291-388. Anon., 1959. Whaling operations in the Antarctic, season 1958-59. Norsk Hvalfangsttid. no. 9, pp. 452-76. 1960a. The structure of Antarctica. Sci. Amer. ccii, no. 3, pp. 86-8. 19606. United States activities in the Antarctic, 1959-60. Polar Rec. x, no. 66, pp. 284-9. Apstein, C, 1909. Die Bestimmung des Alters pelagisch lebender Fischeier. Mitt, dtsch. SeefischVer. xxv, no. 12, pp. 364-73. Ardley, R. a. B., 1936. The birds of the South Orkney Islands. Discovery Rep. xii, pp. 349-76, pis. x-xii. Armstrong, T., 1958. The Russians in the Arctic (London). Aron, W., 1958. The use of a large capacity portable pump for plankton sampling, with notes on plankton patchiness. J. Mar. Res. XVI, no. 2, pp. 158-73. Aksenev, V. A., 1958a. Distribution of whales in the Atlantic sector of the Antarctic. Trans. Inst. Mar. Fish. U.S.S.R. xxxiii, pp. 76-95. 19586. Distribution of feeding grounds and the accumulation of baleen whales in the Antarctic. Trans. Inst. Mar. Fish. U.S.S.R. xxxiii, pp. 96-100. Baalsrud, K., 1955. Utilization of plankton. Norsk Hvalfangsttid. no. 3, pp. 125-33. Bagshawe, T. W., 1938. Notes on the habits of the Gentoo and Ringed or Antarctic penguins. Trans. Zool. Soc. Lond. xxiv, pt. 3, no. I, pp. 185-307. 1939. Two Men in the Antarctic (Cambridge). Bainbridge, R., 1953. Studies on the interrelationships of zooplankton and phytoplankton. J. Mar. Biol. Ass. U.K. xxxii, no. 2, pp. 385-447- 1957. The size, shape and density of marine phytoplankton concentrations. Biol. Rev. xxxii, no. i, pp. 91-115. Baker, A. de C, 1954. The circumpolar continuity of Antarctic plankton species. Discovery Rep. xxvii, pp. 201-18. 1959. The distribution of Euphausia triacantha Holt and Tattersall. Discovery Rep. xxix, pp. 309-39. i960. Observations of squid at the surface in the N.E. Atlantic. Deep-sea Res. vi, no. 3, pp. 206-10. Balleny, J., 1839. Discoveries in the Antarctic Ocean, in February, 1839. [Extract from the Journal of the Schooner 'Eliza Scott', commanded by Mr John Balleny]. The Antarctic Manual for the use of the Expedition of 1901 (London). Banner, A. H., 1954. Some 'schizopod' crustaceans from the deeper water off California. Occ. Pap. Allan Hancock Fdn. no. 13, pp. 1-45. Bargmann, H. E., 1937. The reproductive system of Euphausia superba. Discovery Rep. xiv, pp. 325-50, pis. i-v. 1945. The development and life-history of adolescent and adult krill, Euphausia superba. Discovery Rep. xxili, pp. 103-76. Barham, E. G., 1957. The ecology of sonic scattering layers in the Monterey Bay area. Tech. Rep. no. i, Hopkins Marine Station, California. Barkley, E., 1940. Nahrung und Filterapparat des Walkrebschens Euphausia superba Dana. Z. Fisch. Beiheft i, pp. 65-156. Barnes, H., 1957. Nutrient elements. Chapter 11 in Treatise on Marine Ecology and Paleoecology, I, Ecology, edited by Joel W. Hedgpeth (New York). 1959- Oceanography and Marine Biology (London). Bahrett-Hamilton, G. E. H., 1901a. Seals. Resuh. Voy. Belgica, Rapp. Sci., Zool., pp. 1-19. 19016. Seals. The Antarctic Manual for the use of the Expedition of 1901 (London). 1902. Mammalia. Report on the Collections of Natural History made in the Antarctic Regions during the Voyage of the 'Southern Cross' (London). Bartholomew, J., 1930. South Polar Regions. Folder map in The Polar Book (London). igjS. The Times Atlas of the World. Mid-century Edition (London). Batham, E., Fisher, L. R., Henry, K. M., Kon, S. K. and Thompson, S. Y., 195 i. Preformed vitamin A in marine Crustacea. Biochem. J. xlviii, p. x. B^ A. W. H., Ewing, M. and Linton, L. W., 1959. A quantitative multiple opening-and-closing plankton sampler for vertical towing. J. Cons. int. Explor. Mer, xxv, no. i, pp. 36-46. Beebe, W., 1932. Nonsuch. Land of Water (London and New^ York). Beklemishev, K. v., 1958a. The dependence of the phytoplankton distribution on the hydrological conditions in the Indian sector of the Antarctic. C.R. Acad. Sci. U.S.S.R. cxix, no. 4, pp. 694-7. 19586. The interrelations betzveen phyto- and zooplankton in the Bering and Okhotsk Seas. Abstract from Proc. Ninth Pacif. Sci. Congr. 1957, xvi. 1958c. The plankton stops a ship. Priroda, no. 11, pp. 105-6. 1959. Antarctic divergence and whaling grounds. Bull. Acad. Sci. U.S.S.R., geogr. ser. no. 6, pp. 90-3. 446 DISCOVERY REPORTS Beklemishev, K. v., i960. Southern atmospheric cyclones and the whale feeding grounds in the Antarctic. Nature, Lond. CLXXXVii, no. 4736, pp. 530-1. Beklemishev, K. V. and Korotkevitch, V. S., 1958. Zooplankton of the Indian sector of the Antarctic. C.R. Acad. Sci. U.S.S.R. cxxi, no. 6, pp. 1009-11. Bellingshausen, T., 1945. The Voyage of Captain Bellingshausen to the Antarctic Seas i8ig~i82i (London). Bennett, A. G., 1920. On the occurrence of diatoms on the skin of whales. Proc. Roy. Soc. B, xci, pp. 352-7. 1931- Whaling in the Antarctic (Edinburgh and London). Bentley, C. R. and Ostenso, N. A., 1961. Glacial and subglacial topography of West Antarctica. J. Glaciol. ill, no. 29, pp. 882-910. Bernacchi, L., 1901. To the South Polar Regions (London). Bertalanffy, L. von, 1950. The theory of open systems in physics and biology. Science, cxi, no. 2872, pp. 23-9. Bertram, G. C. L., 1940. The biology of the Weddell and Crabeater seals. Sci. Rep. Brit. Grahamld. Exped. i, no. i, pp. 1-139- 1958. Arctic and Antarctic. A Prospect of the Polar Regions (Cambridge). Best, P. B., i960. Further information on Bryde's whale (Balaenoptera edeni Anderson) from Saldanha Bay, South Africa. Norsk Hvalfangsttid. no. 5, pp. 201-15. Betesheva, E. 1., 1954. Data on the feeding of baleen whales in the Kurile Region. Trans. Inst. Oceanol. Acad. Sci. U.S.S.R. XI, pp. 238-45. Beverton, R. J. H. and Holt, S. J., 1956. The theory of fishing. Contribution no. ix to Sea Fisheries. Their Investigation in the United Kingdom, edited by Michael Graham (London). BiERMAN, W. H. and Voous, K. H., 1950. Birds observed and collected during the whaling expeditions of the ' William Barendsz' in the Antarctic, ig46-ig4y and ig4y-ig48 (Leiden). BiGELOW, H. B., 1909. Cruise of the U.S. fisheries schooner 'Grampus' in the Gulf Stream during July, igo8, with description of a new medusa (Bythotiaridae). Bull. Mus. Comp. Zool. Harv. lii, no. 12, pp. 195-210. 1926. Plankton of the offshore waters of the Gulf of Maine. Bull. U.S. Bur. Fish, xl, 1924, pt. 11, Document no. 968, pp. 1-486. BiSCOE, J., 1831. From the Journal of a Voyage towards the South Pole on Board the Brig ' Tula\ under the Command of John Biscoe, with the Cutter 'Lively' in Company. The Antarctic Manual for the use of the Expedition of 1901 (London). BiSHAi, H. M., i960. The effect of water currents on the survival and distribution offish larvae. J. Cons. int. Explor. Mer, xxv, no. 2, pp. 134-45. BoDEN, B. P., 1950. Plankton organisms in the deep scattering layer. U.S. Navy Electronics Laboratory Rep. 186, 29 pp. 1955- Euphausiacea of the Benguela Current, First Survey, R.R.S. 'William Scoresby', March igso. Discovery Rep. XXVII, pp. 337-76. BoEHM, G. A. W., 1959. The Exploration of 'Inner Space'. Reprinted from Fortune (Ottawa). BoGOROV, B. G., 1938. Biological seasons of the Arctic Sea. C.R. Acad. Sci. U.S.S.R. xix, no. 8, pp. 641-4. 1957- Regularities of plankton distribution in North-west Pacific. Proceedings of the Unesco Symposium on Physical Oceanography (Tokyo). 1958. Perspectives in the study of seasonal changes of plankton and of the number of generations at different latitudes. From Perspectives in Marine Biology, edited by A. A. Buzzati-Traverso (University of California Press). BoRCHGREViNK, C. E., 1895. The 'Antarctic's' voyage to the Antarctic. Geogr. J. v, no. 6, pp. 583-9. 1901. First on the Antarctic Continent (London). BoscHMA, H., 1949. Ellobiopsidae. Discovery Rep. xxv, pp. 281-314, pis. xxxviii-xli. Brachi, R. M., 1953. Examination of some components of cephalopod and sperm-whale liver oils by the chromatographic method. Biochem. J. liv, no. 3, pp. 459-66. Braekkan, O. R., 1948. Vitamins in whale liver. Hvalrad. Skr. no. 32, pp. 1-25. Breathnach, a. S., i960. The cetacean central nervous system. Biol. Rev. xxxv, no. 2, pp. 187-230. Brennecke, W., 1921. Die Ozeanographischen Arbeiten der Deutschen Antarktischen Expedition, igii-igi2. Aus d. Arch. dtsch. Seew. xxxix, no. i, pp. 1-216. Broch, H., 1951. Stylasteridae (Hydrocorals) from southern seas. Discovery Rep. xxvi, pp. 33-46, pis. ii-iv. Brodsky, K. A. and Vinogradov, M. E., 1957. On the distribution of plankton in the Indian sector of the Antarctic region. C.R. Acad. Sci. U.S.S.R. cxii, no. 5, pp. 957-60. Brongersma-Sanders, M., 1948. The importance of upwelling water to vertebrate palaeontology and oil geology. Verh. Akad. Wet. Amst. 2nd sect, xlv, no. 4, pp. 1-112. — ■ — 1957. Mass mortality in the sea. Chapter 29 in Treatise on Marine Ecology and Paleoecology, I, Ecology, edited by Joel W. Hedgpeth (New York). Brown, R. N. Rudmose, 1915^. The seals of the Weddell Sea: notes on their habits and distribution. Rep. Scot. Nat. Antarct. Exped., Zool. IV, pt. 13, pp. 185-97. 19156. The life and habits of penguins. Rep. Scot. Nat. Antarct. Exped., Zool. iv, pt. 14, pp. 249-53. REFERENCES 447 Brown, R. N. Rudmose, 1920. Report of the Interdepartmental Committee on Research and Development in the Dependencies of the Falkland Islands. Appendix iv, pp. 45-7 (London, H.M. Stationery Office [Cmd. 657]). 1923. A Naturalist at the Poles (London). 1927. The Polar Regions. A Physical and Economic Geography of the Arctic and Antarctic (London). Brown, S. G., 1954. Dispersal in blue and fin whales. Discovery Rep. xxvi, pp. 355-84. 1956. Whale marks recently recovered. Norsk Hvalfangsttid. no. 12, pp. 661-4. 1957- Whale marks recovered during the Antarctic whaling season ig^d-^y. Norsk Hvalfangsttid. no. 10, pp. 555-9. 1958. Whales observed in the Atlantic Ocean. Notes on their distribution. Mar. Obs. xxviii, nos. 181 and 182, pp. 142-6 and 209-16. 1959- Whale marks recovered in the Antarctic seasons 1955-36, 1958-59, and in South Africa 1958 and 1959. Norsk Hvalfangsttid. no. 12, pp. 609-16. Bruce, W. S., 1894. No title. Last Chapter in W. G. Burn Murdoch's From Edinburgh to the Antarctic (London). 1905- Bathymetric survey of the South Atlantic Ocean and Weddell Sea. Scot. Geogr. Mag. xxi, no. 8, pp. 402-12. 1906. The Voyage of the 'Scotia' [as recounted by Brown, R. N. Rudmose, Mossman, R. C. and Pirie, J. H. Harvey] (Edinburgh and London). 191 1. Polar Exploration (London). 1913- Measurements and weights of Antarctic seals taken by the Scottish National Antarctic Expedition. Trans. Roy. See. Edinb. XLix, pp. 567-77. 1915- Some observations on Antarctic Cetacea. Rep. Scot. Nat. Antarct. Exped., Zool. iv, pt. 22, pp. 491-505. 1918. Station log of the S.Y. 'Scotia', 1902-1904. Rep. Scot. Nat. Antarct. Exped. i, pt. i, pp. 1-28. 1920. Report of the Interdepartmental Committee on Research and Development in the Dependencies of the Falkland Islands. Appendices i and 11, pp. 33-41 (London, H.M. Stationery Office [Cmd. 657]). Bruns, E., 1958. Ozeanologie (Berlin). Bruun, a. F., 1957a. The ecological zonation of the deep-sea. Proceedings of the Unesco Symposium on Physical Oceanography (Tokyo). 1957^- Deep sea and abyssal depths. Chapter 22 in Treatise on Marine Ecology and Paleoecology, I, Ecology, edited by Joel W. Hedgpeth (New York). Buchet, G., 1895. Quelques Observations sur les Balenopteres des eaux islandaises. Bull. Soc. Zool. Fr. xx, pp. 30-1. BuDKER, P., 1957. Baleines et Baleiniers (Paris). 1958- Whales and Whaling (London). Bull, H. J., 1896. The Cruise of the 'Antarctic' to the South Polar Regions (London). BuLLiVANT, J. S., 1961. Photographs of Antarctic bottom fauna. Polar Rec. x, no. 68, pp. 505-8. Bullock, T. H., 1958. In the Discussion following A. C. Hardy's Toward Prediction in the Sea (see Hardy, A. C, 195812). Bunt, J. S., 1959. Marine studies in inshore waters of Antarctica near Mawson, 1956. Nature, Lond. clxxxiii, no. 4674, pp. 1541-2. Burn Murdoch, W. G., 1894. From Edinburgh to the Antarctic (London). Buzzati-Tra VERSO, A. A., 1958. Perspective in Marine Biology. A Summary of the Proposals put forward by the Symposium which under this title was held at the Scripps Institution of Oceanography in 1956 (University of California Press). Capurro, L. R. a., 1955. Expedicion Argentina al Mar de Weddell (Buenos Aires). Carrick, R., 1959. The contribution of banding to Australian bird ecology. Ch2ipter 22 in Biogeography and Ecology in Australia (The Hague). Carrington, R., i960. A Biography of the Sea (London). Carthy, J. D., 1957. An Introduction to the Behaviour of Invertebrates (London). Cassie, R. M., 1957. The sampling problem, with particular reference to marine organisms. Reprinted from Proc. N.Z. Ecol. Soc. no. 4, pp. 37-9. 1959- An experimental study of factors inducing aggregation in marine plankton. N.Z. J. Sci. 11, no. 3, pp. 339-65. Cendron, J., 1952. line Visite Hivernale a une Rookerie de Manchots Empereurs. Note Ornithologique no. 4, Expeditions Antarctiques en Terre Adehe, 1949-1952, Terre et la Vie, no. 2, pp. 101-8. Chaplin, J. M., 1932. Narrative of hydrographic survey operations in South Georgia and the South Shetland Islands, 1926-1930. Discovery Rep. iii, pp. 297-344. Charnock, H., i960. Recent advances in science: physical oceanography. Sci. Progr. XLViii, no. 190, pp. 257-70. Cherry-Garrard, A., 1922. The Worst Journey in the World (London). Chilton, C, 1909. The Crustacea of the Subantarctic Islands of New Zealand (Wellington). Chittleborough, R. G., 1953. Aerial observations on the humpback whale, Megaptera nodosa (Bonnaterre), with notes on other species. Aust. J. Mar. Freshw. Res. iv, no. 2, pp. 219-26. 1958. An analysts of recent catches of humpback whales from the stocks in groups iv and v. Norsk Hvalfangsttid. no. 3, pp. 109-37. 448 DISCOVERY REPORTS Chittleborough, R. G., 1959a- Australian marking of humpback whales. Norsk Hvalfangsttid. no. 2, pp. 47-55. 1959^- Intermingling of two populations of humpback whales. Norsk Hvalfangsttid. no. 10, pp. 510-21. I959<^- Determination of age in the humpback whale, Megaptera nodosa (Bonnaterre). Aust. J. Mar. Freshw. Res. X, no. 2, pp. 125-43. i960. Determination of age in the humpback whale, Megaptera nodosa (Bonnaterre). Norsk Hvalfangsttid. no. i, pp. 12-38. Christensen, L., 1938. My last expedition to the Antarctic. A lecture delivered before the Norwegian Geographical Society, 22 September, 1937. Chun, C, 1903. Aus den Tiefen des Weltmeeres (Jena). Clark, R. S., 1919. South Atlantic whales and whaling. Appendix to E. H. Shackleton's South (London). Clarke, G. L., 1954. Elements of Ecology (New York and London). Clarke, G. L. and Breslau, L. R., i960. Studies of luminescent flashing in Phosphorescent Bay, Puerto Rico, and in the Gulf of Naples using a portable bathy photometer. Bull. Inst. Oceanogr. Monaco, no. 1171, pp. 1-32. Clarke, R., 1950. The bathypelagic angler fish Ceratias holbolH Krbyer. Discovery Rep. xxvi, pp. 1-32. 1954- Whales and seals as resources of the sea. Norsk Hvalfangsttid. no. 9, pp. 489-510. 1956. Sperm whales of the Azores. Discovery Rep. xxviil, pp. 237-98. • 1957- Migrations of marine mammals. Norsk Hvalfangsttid. no. 11, pp. 609-30. Clarke, R. and Brown, S. G., 1957. International co-operation in Antarctic whale marking from ig45 to 1957. Norsk Hval- fangsttid. no. 9, pp. 513-17- Clarke, R. and Ruud, J. T., 1954. The voyage of the 'Enern' to the Antarctic 1953. Norsk Hvalfangsttid. no. 3, pp. 128-46. Clowes, A. J., 1934. Hydrology of the Bransfield Strait. Discovery Rep. ix, pp. 1-64. 1938. Phosphate and silicate in the Southern Ocean. Discovery Rep. xix, pp. 1-120. CoLBECK, W., 1905. Observations on the Antarctic sea-ice. Geogr. J. xxv, no. 4, pp. 401-5. Colebrook, J. M., 19600. Plankton and tvater movements in Windermere. J- Anim. Ecol. xxix, no. 2, pp. 217-40. 19606. Some observations of zooplankton swarms in Windermere. J. Anim. Ecol. xxix, no. 2, pp. 241-2. CoLLETT, R., 1886. On the external characters ofRudolphi's Rorqual (Balaenoptera borealis). Proc. Zool. Soc. Lond. pp. 243-65. 1912. Norges Pattedyr (Oslo). Collins, J. W., 1886. Notes on an investigation of the great fishing banks of the western Atlantic. Bull. U.S. Fish Comm. VI, no. 24, pp. 369-81. CoLMAN, J. S., 1950. The Sea and its Mysteries (London). CoLTON, J. B., 1959. The multiplane kite-otter as a depressor for high-speed plankton samplers. J. Cons. int. Explor. Mer, xxv, no. I, pp. 29-35. Comfort, A., 1956. The Biology of Senescence (London). Cook, F. A., 1900. Through the First Antarctic Night (London). Cook, J., 1777. A Voyage towards the South Pole and round the World (London). Cooper, L. H. N., 1952. Factors affecting the distribution of silicate in the North Atlantic Ocean and the formation of North Atlantic deep water. J. Mar. Biol. Ass. U.K. xxx, no. 3, pp. 511-26. 1955- Hypotheses connecting fluctuations in arctic climate with biological productivity of the English Channel. Pap. Mar. Biol, and Oceanogr., Deep-sea Res., Suppl. to vol. in, pp. 212-13. Copenhagen, W. J., 1953. The periodic mortality of fish in the Walvis region. A phenomenon within the Benguela Current. Invest. Rep. Fish. Mar. Biol. Surv. S. Afr. no. 14, pp. 1-35. CouTiERE, H., 1906. Crustacis Schizopodes et Decapodes. Expedition Antarctique Fran9aise (1903-1905), pp. 1-9, pis. i-ii. CouTifeRE, M., 1917. Crustach Schizopodes et Decapodes. Deuxieme Expedition Antarctique Fran^aise (1908-1910), pp. 1-8. Crantz, D., 1765. Historic von Gronland (Leipzig). Crisp, D. J., 1959. The influence of climatic changes on animals and plants. Geogr. J. cxxv, pt. i, pp. 1-19. Crisp, F., 1954. The Adventure of Whaling (London). CuMPSTON, J. S., 1939. Antarctica. In Handbook and Index to accompany Commonwealth Map of Antarctica (Canberra). CusHiNG, D. H., 1951- The vertical migration of planktonic Crustacea. Biol. Rev. xxvi, no. 2, pp. 158-92. — — 1955- Some experiments on the vertical migration of zooplankton. J. Anim. Ecol. xxi, no. i, pp. 137-66. CusHiNG, D. H. and Richardson, I. D., 1956. A record of plankton on the echo-sounder. J. Mar. Biol. Ass. U.K. xxxv, no. i, pp. 231-40. Czihak, G., 1959. Vorkommen und Lebensweise der Ophiothrix quinquemaculata in der Nordlichen Adria bet Rovinj. Thalassia Jugoslavica, I, no. 6, pp. 19-24. Dakin, W. J., 1934. Whalemen Adventurers (Sydney). 1953. Australian Seashores (London). Dakin, W. J. and Colefax, A. N., 1940. Theplankton of the Australian coastal waters off New South Wales. Part I. Publ. Univ. Sydney, Zool., Monograph no. i. I REFERENCES 449 Dall, W. and Dunstan, D., 1957. Euphausia superba Dana from a humpback whale, Megaptera nodosa (Bonnaterre), caught off southern Queensland. Norsk Hvalfangsttid. no. i, pp. 6-9. Dallmann, E., 1873-4. MS. log of the steam whaler ' Gronland '. Library of Justus Perthes' Geographische Anstalt, Gotha. Dana, J. D., 1850. Sytiopsis Generum Crustaceorum Ordinis ' Schizopoda' J. D. Dana elaboratus {Pars I). Amer. J. Sci. II, no. 9, pp. 129-33. 1852. Crustacea. U.S. Exploring Expedition under the Command of Charles Wilkes, U.S.N., xiii, pt. i, 685 pp. 1855. Folio Atlas of 96 Plates. Dansereau, p., 1955. Biogeography of the land and the inland zvaters. Chapter 5 in Geography of the Northlands (New York). Darwin, C, 1901. The Origin of Species (London). David, P. M., 1955. The distribution o/Sagitta gazellae Ritter Zahony. Discovery Rep. xxvii, pp. 235-78. Davis, C. C, 1955. The Marine and Fresh-Water Platikton (Michigan). Davis, J. K., 1919. With the 'Aurora' in the Antarctic, igii-igi4 (London). Davtoin, W. H., 1952. Whales and Whaling in the Southern Ocean. Chapter 6 in The Antarctic Today, edited by Frank A. Simpson (Sydney). 1959. New Zealand and South Pacific whale marking and recoveries to the end of 1958. Norsk Hvalfangsttid. no. 5, pp. 213-38. 1960a. An analysis of the Nezv Zealand catches of humpback whales from ig4y to 1958. Norsk Hvalfangsttid. no. 2, pp. 61-74. 19606. The composition of the New Zealand whale catch in 1959. Norsk Hvalfangsttid. no. 9, pp. 401-9. Deacon, G. E. R., 1933. A getieral account of the hydrology of the South Atlantic Ocean. Discovery Rep. vii, pp. 171-238. 1937- The hydrology of the Southern Ocean. Discovery Rep. xv, pp. 1-124. 1939. The Antarctic Voyages of R.R.S. 'Discovery //' and R.R.S. ' William Scoresby', 1935-37. Geogr. J. xciii, no. 3, pp. 185-209. 1 957 «■ The deep-water circulation in the Pacific Ocean. Proceedings of the Unesco Symposium on Physical Oceanography (Tokyo). i<)S7b- The oceans. Chapter 5 in The Planet Earth, edited by D. R. Bates (London). 1959- The Antarctic Ocean. Sci. Progr. xlvii, no. 188, pp. 647-60. 1960^. The sea and its problems. Essay no. 5 in The World around Us, edited by Sir Graham Sutton (London). 19606. The southern cold temperate zone. Proc. Roy. Soc. B, clii, pp. 441-7. Debenham, F., 1923. Report on the maps and surveys. Brit. Antarct. Terra Nova Exped. 1910, pp. 1-94. 1930a. The Polar Regions (London). 19306. The glaciology of the polar regions. Sixth Contribution to The Polar Book (London). 1959- Antarctica. The Story of a Continent (London). Dell, R. K., 1952. Marine biology. Chapter 5 in The Antarctic Today, edited by Frank A. Simpson (Sydney). 1959- Cephalopoda. B.A.N.Z. Antarct. Res. Exped., 1929-1931, ser. B (Zool. and Bot.), viii, pt. 4, pp. 91-105. Dickson, W., 1850. The Antarctic voyage of Her Majesty's hired barque, Pagoda. The United Service Magazine of June and July. Dietrich, G. and Kalle, K., 1957. Allgemeine Meereskunde (Berlin). Dietz, R. S., 1948. Deep scattering layer in the Pacific and Antarctic Oceans. J. Mar. Res. vii, no. 3, pp. 430-42. DoLLO, L., 1908. Notolepis coatsi, Poisson pelagique nouveau recueilli par l' Expedition Antarctique Nationale Ecossaise. Proc. Roy. Soc. Edinb. xxviii, pp. 58-65. Donald, C. W., 1896. Cruise of the ' Balaena' and the 'Active' in the Antarctic Seas, 1892-93. Geogr. J. vii, no. 6, pp. 625-43. Donaldson, H. H., 1924. The Rat (Philadelphia). Douglas, J. S., 1953. The Story of the Oceans (London). Douguet, M., 1950. Rapport sur les Observations faites par le ' Commandant Charcot ' au Cours de sa Croisiere dans V Antarctique durant T&e Austral 1948-1949. Service Central Hydrographique (Paris). Drygalski, E. von, 1928. The Oceanographical problems of the Antarctic. From Problems of Polar Research, American Geographical Society special Publication no. 7 (New York). Dudley, P., 1725. An Essay upon the Natural History of Whales, with a particular Account of the Ambergris found in the Sperma Ceti Whale. Phil. Trans, xxxiii, pp. 78-84. Dunbar, M. J., 1955 a. Marine life. Chapter 6 in Geography of the Northlands (New York). 1955^- Water masses, circulation, ice cover. Chapter 3 in Geography of the Northlands (New York). D'Urville, J. S. C. DuMONT, 1842. Voyage au P6le Sud et dans I'Oceanie (Paris). Einarsson, H., 1945. Euphausiacea. I. Northern Atlantic species. Dana Rep. no. 27, pp. 1-191. Ekman, S., 1953. Zoogeography of the Sea (London). Elliot, H. F. L, 1953. The fauna of Tristan da Cunha. Oryx, 11, no. i, pp. 41-53. Ellsworth, L., 1938. Beyond Horizons (London). 53 DM 4SO DISCOVERY REPORTS Elton, C, 1927. Animal Ecology (London). Emery, K. O., i960. The Sea off Southern California. A Modern Habitat of Petroleum (New York and London). Falla, R. a., 1937. Birds of the British, Australian, and New Zealand Antarctic Research Expedition {ig2g-ig3i), and of the Australasian Antarctic Expedition {1911-1914). B.A.N.Z. Antarct. Res. Exped., 1929-1931, ser. B (Zool. and Hot.), 11, pp. 1-304. 1952. Antarctic birds. Chapter 8 in The Antarctic Today, edited by Frank A. Simpson (Sydney). Feltmann, C. F. and Vervoort, W., 1949. Walvisvaart. Biologische en Technische Grondslagen van die moderne, Antarctische Walvisvaart. Noorduijn's Wetenschappelijke Reeks, no. 35 (Gornichem). Filchner, W., 1922. Zum Sechsten Erdteil (Berlin). Fish, C. J., 1925. Seasonal distribution of the plankton of the Woods Hole Region. Bull. U.S. Bur. Fish, xli, pp. 91-179. Fisher, J. and Lockley, R. M., 1954. Sea-Birds (London). Fisher, L. R., 1958. Whales, krill and vitamin A. New Scientist, in, no. 76, pp. 24-7. Fisher, L. R. and Goldie, E. H., 1959. The food of Meganyctiphanes norvegica (M. Sars), with an assessment of the contributions of its components to the vitamin A reserves of the animal. J. Mar. Biol. Ass. U.K. xxxviii, no. 2, pp. 291-312. Fisher, L. R. and Kon, S. K., 1959. Vitamin A in the invertebrates. Biol. Rev. xxxiv, no. i, pp. 1-36. Fisher, L. R., Kon, S. K. and Thompson, S. Y., 1953. Vitamin A and carotinoids in some Mediterranean Crustacea with a note on the swarming of Meganyctiphanes. Bull. Inst. Oceanogr. Monaco, L, no. 1021, pp. 1-19. Fleming, C. A., 1952. The seas between. Chapter 4 in The Antarctic Today, edited by Frank A. Simpson (Sydney). Fofonoff, N. p., 1956. Some properties of sea water influencing the formation of antarctic bottom water. Deep-sea Res. iv, no. I, pp. 32-5. Forster, G., 1777. A Voyage round the World (London). Fox, D. L., 1957. Particulate organic detritus. Chapter 14 in Treatise on Marine Ecology and Paleoecology, I, Ecology, edited by Joel W. Hedgpeth (New York). Foxton, p., 1956. The distribution of the standing crop of zooplankton in the Southern Ocean. Discovery Rep. xxviii, pp. 191-236. 1961. Salpa fusiformis Cuvier and related species. Discovery Rep. xxxii, pp. 1-32. France, R. H., 1894. Zur Biologic des Planktons. Biol. Zbl. xiv, no. 2, pp. 33-8. Franz, V., 1910. Phototaxis und Wanderung. Nach Versuchen mit Jungfischen und Fischlarven. Int. Rev. Hydrobiol. iii, heft. 3-4, pp. 306-34. Eraser, F. C, 1936. On the development and distribution of the young stages of krill (Euphausia superba). Discovery Rep. xiv, pp. 1-192. 1937. Early larval development in the Euphausiacea. Proc. Linn. Soc. Lond. 149th Session, pp. 89-94. Eraser, F. C. and Purves, P. E., 1954. Hearing in Cetaceans. Bull. Brit. Mus. (Nat. Hist.), Zool. 11, pp. 101-14. 1959- Hearing in whales. Endeavour, xviii, no. 70, pp. 93-8. Fricker, K., 1900. The Antarctic Regions (London). Frisch, K. von, 1948. Geloste und ungeloste Rdtsel der Bienensprache. Naturwissenschaften, xxxv, heft 2, pp. 38-43. Fryer, G., 1959. The trophic interrelationships and ecology of some littoral communities of Lake Nyasa with especial reference to the fishes, and a discussion of the evolution of a group of rock-frequenting Cichlidae. Proc. Zool. Soc. Lond. cxxxii, pt. 2, pp. 153-281. Fuchs, V. and Hillary, E., 1958. The Crossing of Antarctica (London). Fujino, K., 1955. On the body weight of the set whale located in the adjacent waters of Japan. Sci. Rep. Whales Res. Inst., Tokyo, no. 10, pp. 133-41. FuKUOKA, J., 1959. A note on the ice and ocean conditions in the sea near Syowa Base in Antarctica. Rec. Oceanogr. Wks. Jap., n.s. V, no. I, pp. 1-5. Gain, L., 1914. Oiseaux Antarctiques. Deuxieme Expedition Antarctique Fran9aise (1908-1910), pp. 1-194. Garner, D. M., 1958. The Antarctic Convergence south of New Zealand. N.Z. J. Geol. Geophys. i, no. 3, pp. 577-94. Gauld, D. T., 1959. Swimming and feeding in crustacean larvae: The nauplius larva. Proc. Zool. Soc. Lond. cxxxii, pt. i, PP- 31-50- Gerlache De Gomery, A. de, 1943. Quinze Mois dans V Antarctique (Bruxelles). Gessner, W., 1942. Die Deutsche Antarktische Expedition 1938-39. Die deutsche luftfoto-grammetrischen Arbeiten in Neu- Schwabenland (Leipzig). Gough, L. H., 1905. Report on the plankton of the English Channel in 1903. Mar. Biol. Ass. U.K., International Fishery Investigations, First Report (Southern Area), pp. 325-77 (London, H.M. Stationery Office [Cmd. 2670]). Gould, R. T., 1924. The Ross Deep. Geogr. J. lxiii, no. 3, pp. 237-41. 1929. New South Greenland. From Enigmas. Another Book of Unexplained Facts (London). — 1941. The charting of the South Shetlands, 1819-28. Mariner's Mirror, xxvii, no. 3, pp. 206-42. Graham, M., 1956. Synopsis and commentary. Contribution no. x to Sea Fisheries. Their Investigation in the United Kingdom, edited by Michael Graham (London). REFERENCES 4Si Grainger, E. H., 1959. The annual oceanographic cycle at Iglooik in the Canadian Arctic. I. The zooplankton and physical and chemical observations. J. Fish. Res. Bd. Can. xvi, no. 4, pp. 453-501. Gran, H. H., 1912. The plankton production in the North European waters in the spring of igi2. Rapp. Cons. Explor. Mer, Bulletin Planktonique pour I'Annee 1912, pp. 7-142. Gray, J., 1953. How Animals Move (Cambridge). i960. The ascent of falls by salmon. New Scientist, VII, no. 178, pp. 927-8. Griffin, D. R., 1955. Hearing and acoustic orientation in marine animals. Pap. Mar. Biol, and Oceanogr., Deep-sea Res., Suppl. to vol. Ill, pp. 406-17. Groom, T. T. and Loeb, J., 1890. Der Heliotropismus der Nauplien von Balanus perforatus und die periodischen Tiefen- wanderungen pelagischer Tiere. Biol. Zbl. x, nos. 5 and 6, pp. 160-77. Guldberg, G. a., 1887. Zur Biologie der Nordatlantischen Finwalarten. Zoal. Jb. 11, pp. 127-74. GuNTER, G., 1957. Temperature. Chapter 8 in Treatise on Marine Ecology and Paleoecology , I, Ecology, edited by Joel W. Hedgpeth (New York). GuNTHER, E. R., 1928. Notes and sketches made during two years on the 'Discovery' Expedition ig2§-ig2j. Reprinted from the 'Draconian' for 1928 (Oxford). 1936. A report on oceanographical investigations in the Peru coastal current. Discovery Rep. xiii, pp. 107-276. 1949- The habits of fin whales. Discovery Rep. xxv, pp. 113-42. GuozDEN, H., 1959. La Campana Antarctica ig^S-^y. Contribucion del Instituto Antarctico Argentine, no. 19, pp. 3-31. GuRNEY, R., 1924. Crustacea. Part IX. — Decapod larvae. Brit. Antarct. Terra Nova Exped. 1910, Nat. Hist. Rep., Zool. viii, no. 2, pp. 37-202. Haden-Guest, S., 1955. The wealth of the Northlands. I. Renewable resources. Chapter 10 in Geography of the Northlands (New York). Hamilton, J. E., 1934a. The Southern Sea Lion, Otaria byronia [De Blainville). Discovery Rep. viii, pp. 269-318. 1934^- The Sub-Antarctic forms of the Great Skua (Catharacta skua skua). Discovery Rep. IX, pp. 161-74. 1939- The leopard seal Hydrurga leptonyx (De Blainville). Discovery Rep. xviii, pp. 239-64. Hansen, H. E., 1934. Limits of the pack-ice in the Antarctic in the area between 40° W and iiQ° E. Hvalrad. Skr. no. 9, pp. 39-41, pis. iv-vii. Hansen, H. J., 1908. Schizopoda and Cumacea. Result. Voy. Belgica, Rapp. Sci., Zool., pp. 1-20, pis. i-iii. 1913- Crustacea Schizopoda. Swed. Antarct. Exped., 1901-1903, pp. 1-56, pis. i-vi. 1915- The Crustacea Euphausiacea of the United States National Museum. Proc. U.S. Nat. Mus. XLVIII, no. 2065, pp. 59-114. 1921. On some Malacostracous Crustacea {Mysidacea, Euphausiacea, and Stomatopoda) collected by Swedish Antarctic Expeditions. Reprinted from Ark. Zool. xiii, no. 20, 7 pp. Hanson, N., 1902. Extracts from the private diary of the late Nicolai Hanson. Report on the Collections of Natural History made in the Antarctic Regions during the Voyage of the 'Southern Cross' (London). Hardin, G., i960. The competitive exclusion principle. Science, cxxxi, no. 3409, pp. 1292-7. Hardy, A. C, 1928. The work of the royal research ship 'Discovery' in the dependencies of the Falkland Islands. Geogr. J. Lxxii, no. 3, pp. 209-34. 1930. Biology of the polar seas. Seventh Contribution to The Polar Book (London). 1936 a. Observations on the uneven distribution of oceanic plankton. Discovery Rep. xi, pp. 511-38. 1936 i. The continuous plankton recorder. Discovery Rep. xi, pp. 457-510. 1936c. Plankton ecology and the hypothesis of animal exclusion. Proc. Linn. Soc. Lond. 148th Session, pt. 2, pp. 64-70. 1953- Sofne problems of pelagic life. Sixth Contribution to Essays in Marine Biology. The Richard Elmhirst Memorial Lectures (Edinburgh and London). 1955- A further example of the patchiness of plankton distribution. Pap. Mar. Biol, and Oceanogr., Deep-sea Res., Suppl. to vol. Ill, pp. 7-1 1- 1956. The Open Sea, its Natural History, the World of Plankton (London). 1958 a. Toward prediction in the sea. From Perspectives in Marine Biology, edited by A. A. Buzzati-Traverso (University of California Press). 1958 &. In the Discussion following A. C. Redfield's The Inadequacy of Experiment in Marine Biology (see Redfield, A.C., 1958). 1959. The Open Sea, its Natural History, Part II, Fish and Fisheries (London). Hardy, A. C. and Bainbridge, R., 1951a. Effect of pressure on the behaviour of decapod larvae (Crustacea). Nature, Lond. CLXvii, no. 4244, pp. 3S4-5- 195 1 &• Vertical migration of plankton animals. Nature, Lond. CLXViii, no. 4269, pp. 327-8. 1954- Experimental observations on the vertical migrations of plankton animals. J. Mar. Biol. Ass. U.K. xxxii, no. 2, pp. 409-48. 53-2 452 DISCOVERY REPORTS Hardy, A. C. and Gunther, E. R., 1935. The plankton of the South Georgia whaling grounds and adjacent waters, ig26-ig2j. Discovery Rep. xi, pp. 1-456. Harmer, S. F., 193 1. Southern whaling. Proc. Linn. Soc. Lond. 142nd Session, pp. 85-163. Harmer, S. F. and Lillie, D. G., 1914. List of collecting stations. Brit. Antarct. Terra Nova Exped. 1910, Nat. Hist. Rep., Zool. II, no. i, pp. 1-12. Hart, T. J., 1934. On the phytoplankton of the south-west Atlantic and the Bellingshausen Sea, 1^29-31 . Discovery Rep. viii, pp. 1-268. 1935. On the diatoms of the skin film of whales, and their possible bearing on problems of whale movements. Discovery Rep. x, pp. 247-82. 1937. Rhizosolenia curvata Zacharias, an indicator species in the Southern Ocean. Discovery Rep. xiv, pp. 413-46. 1942. Phytoplankton periodicity in Antarctic surface waters. Discovery Rep. xxi, pp. 261-356. 1946. Report on trawling surveys on the Patagonian continental shelf. Discovery Rep. xxiii, pp. 223-408. 1959. A possible correlation between weather conditions and discoloured water phenomena in the eastern central basin of the Caribbean Sea. Mar. Obs. xxix, no. 186, pp. 191-5. Hart, T. J. and Currie, R. I., i960. The Benguela Current. Discovery Rep. xxxi, pp. 123-298. Hartley, C. H., 1934. Sea-birds of West Spitsbergen — their food supply. Bull. Brit. Orn. CI. liv, pp. 164-7. Haynes, F. and Laurie, A. H., 1937. On the histological structure of Cetacean lungs. Discovery Rep. xvii, pp. 1-6, pi. i. Hedgpeth, J. W., 1957. Concepts of marine ecology. Chapter 3 in Treatise on Marine Ecology and Paleoecology, I, Ecology, edited by Joel W. Hedgpeth (New York). Heegaard, p., 1948. Larval stages 0/ Meganyctiphanes (Euphausiacea) and some general phylogenetic remarks. Medd. Komm. Havunders0g., Kbh., ser. Plankton, v, no. i, pp. 1-27. Heezen, B. C, 1959. Dynamic processes of abyssal sedimentation: erosion, transportation, and redeposition on the deep-sea floor. Geophys. J. 11, no. 2, pp. 142-63. Heezen, B. C. and Ewing, M., 1952. Turbidity currents and submarine slumps, and the ig2g Grand Banks earthquake. Amer. J. Sci. CCL, pp. 849-73. Hendey, N. I., 1937. The plankton diatoms of southern seas. Discovery Rep. xvi, pp. 151-364. Hentschel, E., 1941. Das Netzplankton des siidatlantischen Ozeans. Wiss. Ergebn. dtsch. atlant. Exped. 'Meteor', xiii, no. 4, pp. 245-308. Herdman, H. F. p., 1948a. Soundings taken during the Discovery Investigations, ig32-3g. Discovery Rep. xxv, pp. 39-106, pis. xxiii-xxxi. 19486. The Antarctic pack ice. J. Glaciol. i, no. 4, pp. 156-66, 172-3. 1952. The voyage of 'Discovery //', igso-si. Geogr. J. cxviii, pt. 4, pp. 429-40. 1953. The Antarctic pack ice in winter. J. Glaciol. 11, no. 13, pp. 184-93. 1959. Some notes on sea ice observed by Captain James Cook, R.N., during his circumnavigation of Antarctica, 1772^5. J. Glaciol. Ill, no. 26, pp. 534-41. Hersey, J. B. and Moore, H. B., 1948. Progress report on scattering layer observations in the Atlantic Ocean. Trans. Amer. Geophys. Un. xxix, no. 3, pp. 341-53- Heyerdahl, E. F., 1932. Hvalindustrien (Oslo). Hiatt, R. W., 1948. The biology of the lined shore crab, Pachygrapsus crassipes Randall. Pacif. Sci. n, no. 3, pp. 135-213. HiCKLiNG, C. F., 1927. The natural history of the hake. Fish. Invest., Lond. ser. ii, x, no. 2, pp. 1-99. Hinton, M. a. C, 1925. Reports on Papers left by the late Major G. E. H. Barrett-Hamilton relating to the Whales of South Georgia. British Museum (Natural History) London. Hjort, J., 1931. The Emperor's New Clothes (London). 1933. Whales and whaling. Hvalrad. Skr. no. 7, pp. 7-29. Hjort, J., Jahn, G. and Ottestad, P., 1933. The optimum catch. Hvalrad. Skr. no. 7, pp. 92-127. Hjort, J., Lie, J. and Ruud, J. T., 1932. Norwegian pelagic whaling in the Antarctic. /. Whaling grounds in ig2g-ig30 and ig30-ig3i. Hvalrad. Skr. no. 3, pp. 1-37, charts 1-6. Hjort, J. and Ruud, J. T., 1929. Whaling and fishing in the North Atlantic. Rapp. Cons. Explor. Mer, lvi, pp. 1-123. HoBBS, W. H., 1939. The discoveries of Antarctica within the American sector, as revealed by maps and documents. Trans. Amer. Phil. Soc. n.s. xxi, pt. i, pp. 1-71. Hodgson, T. V., 1902. Crustacea. 2. Schizopoda. Report on the Collections of Natural History made in the Antarctic Regions during the Voyage of the ' Southern Cross ' (London). 1505. Preliminary report of the biological collections of the 'Discovery'. Geogr. J. XXV, no. 4, pp. 396-400. 1907. On collecting in Antarctic seas. Nat. Antarct. Exped. 1901-1904, Nat. Hist, ill, pp. i-io. HoLGERSEN, H., 1957. Ornithology of the 'Brategg' Expedition. Sci. Res. Norweg. 'Brategg' Exped. no. 4, pp. 1-80. Holt, E. W. L. and Tattersall, W. M., 1906. Preliminary notice of the Schizopoda collected by H.M.S. 'Discovery' in the Antarctic region. Ann. Mag. Nat. Hist. (7), xvii, pp. i-ii. REFERENCES 453 Hough, J. L., 1956. Sediment distribution in the southern oceans around Antarctica. J. Sediment. Petrol, xxvi, no. 4, pp. 301-6. HOUOT, G. and Willm, P., 1955. Two Thousand Fathoms Dozen (London). Howell, G. C. L., 1921. Ocean Research and the Great Fisheries (Oxford). Hurley, D. E., 1959. Some features of the benthic environment in Cook Strait. N.Z. J. Sci. 11, no. i, pp. 137-47. Hurley, F., 1925. Argonauts of the South (New York and London). Hustedt, F., 1958. Diatomeen aus der Antarktis und dem SUdatlantik. Dtsch. Antarkt. Exped. 1938-39, Wiss. Ergebn. 11, Lief. 2, pp. 103-91. Huxley, J. S., 1912. The Individual in the Animal Kingdom (Cambridge). Hvalfangernes Assuranceforening, 1936. Atlas over Antarktis og Sydishavet. Charts i and 2. Hydrogr.\phic Department, Admiralty, London, 1930. The Antarctic Pilot. 1948- The Antarctic Pilot. Second edition. Ichihara, T., 1961. Blue whales in the waters around Kerguelen Island. Norsk Hvalfangsttid. no. i, pp. 1-20. Igarashi, S., 1957. On the filtering-rate of plankton net. Bull. Fac. Fish. Univ. Hokkaido, viii, no. i, pp. 48-50. Illig, G., 1930. Die Schizopoden der Deutschen Tief see-Expedition. Wiss. Ergebn. ' Valdivia', xxii, Teil 6, pp. 400-625. Ingebrigtsen, a., 1929. Whales caught in the North Atlantic and other seas. Rapp. Cons. Explor. Mer, lvi, pp. 1-26. Inoue, N., Nishizawa, S. and Fukuda, M., 1957. The perfection of a turbidity meter and the photographic study of suspended matter and plankton in the sea using an undersea observation chamber. Proceedings of the Unesco Symposium on Physical Oceanography (Tokyo). International Commission on Whaling, 1959. Tenth Report of the Commission (London). Ivanov, a. and Waterman, T. H., 1958. Elliptical polarization of submarine illutnination. J. Mar. Res. xvi, no. 3, pp. 255-82. Ivanov, Yu. a. and Tareev, B. A., 1959. The question of the structure of the Antarctic divergence zone. Bull. Acad. Sci. U.S.S.R., geogr. ser. no. 6, pp. 82-9. IvASHiN, M. v., 1958. The skin film of diatomaceous algae on Antarctic finwhales. Trans. Inst. Mar. Fish. U.S.S.R. xxxiii, pp. 186-98. Jackson, P., 1954. Engineering and economic aspects of marine plankton harvesting. J. Cons. int. Explor. Mer, xx, no. 2, pp. 167-74. Jacquinot, H., 1842. In the Notes accompanying Dumont D'Urville's Voyage au P6le Sud et dans VOceanie (Paris). Jameson, Professor, 1823. Memoranda by Professor Jameson. Appendix iii to William Scoieshy's Journal of a Voyage to the Northern Whale-Fishery (Edinburgh). JoHANSEN, A. C, 1925. On the diurnal vertical movements of young of some fishes in Danish waters. Medd. Komm. Havun- dersog., Kbh., ser. Fiskeri, viii, no. 2, pp. 1-27. John, D. Dilwyn, 1934. The second Antarctic Commission of the R.R.S. 'Discovery IF. Geogr. J. lxxxiii, no. 5, pp. 381-94. 1936. The southern species of the genus Euphausia. Discovery Rep. xiv, pp. 193-324. Johnson, M. W., 1957. Plankton. Chapter 16 in Treatise on Marine Ecology and Paleoecology, /, Ecology, edited by Joel W. Hedgpeth (New York). Johnston, T. Harvey, 1937. Biological Organization and Station List. B.A.N. Z. Antarct. Res. Exped. 1929-193 1, ser. B (Zool. and Bot.), I, pp. 1-48. Jonsgard, a., 1957. Some comments on W. Dall and D. Dunstan's article on the find o/Euphausia superba in a humpback whale caught near South Queensland. Norsk Hvalfangsttid. no. i, pp. 10-12. 1959- Recent investigations concerning sound production in Cetaceans. Norsk Hvalfangsttid. no. 10, pp. 501-9. Joyce, E. M., 1929. The South Polar Trail (London). Karcher, F. H., 1940. Vber den Algenbewugs auf siidlichen Walen. Z. Fisch. Beiheft i, pp. 13-36. Karsten, G., 1905. Das Phytoplankton des Antarktischen Meeres nach dem Material der deutschen Tief see-Expedition, i8g8-i8gg. Wiss. Ergebn. 'Valdivia', 11, Teil 2, pp. 1-136. Kellogg, R. and Whitmore, F. C, 1957. Marine mammals. In Treatise on Marine Ecology and Paleoecology, I, Ecology, edited by Joel W. Hedgpeth (New York). Kellogg, W. N., 1958. Echo ranging in the porpoise. Science, cxxviii, no. 3330, pp. 982-8. Kemp, S., 1934. In the Discussion following D. Dilwyn John's The Second Antarctic Commission of the R.R.S. 'Discovery IP (see John, D. Dilwyn, 1934). Kemp, S. and Bennett, A. G., 1932. On the distribution and movements of whales on the South Georgia and South Shetland whaling grounds. Discovery Rep. vi, pp. 165-90, pis. vii-xlii. Kemp, S. and Hardy, A. C, 1929. Discovery investigations, objects, equipment and methods. Discovery Rep. i, pp. 141-232, text-figs. 1-33, pis. vii-xviii. Kemp, S. and Nelson, A. L., 1931. The South Sandwich Islands. Discovery Rep. in, pp. 133-98. Kling, a., 1922. Fine Schlittenfahrt nach Morell-Land. Chapter 27 in W. Filchner's Zum Sechsten Erdteil (Berlin). Klumov, S. K., 1961. Plankton and the feeding of the whalebone whales (Mystacoceti). Trans. Inst. Oceanol. U.S.S.R. li, pp. 142-56. 454 DISCOVERY REPORTS Knauss, J. A. and King, J. E., 1958. Observations of the Pacific equatorial undercurrent. Nature, Lond. clxxxii, no. 4635, pp. 601-2. Knight-Jones, E. W. and Qasim, S. Z., 1955. Responses of some marine plankton animals to changes in hydrostatic pressure. Nature, Lond. CLXXVii, no. 4465, pp. 941-2. Kohl, L., 1926. Zur Grossen Eismauer des Siidpols (Stuttgart). KoN, S. K., 1958. Some thoughts on biochemical perspectives in marine biology. From Perspectives in Marine Biology, edited by A. A. Buzzati-Traverso (University of California Press). KoROTKEViTCH, V. S., 1958. The distribution of plankton in the Indian sector of the Antarctic. C.R. Acad. Sci. U.S.S.R. cxxii, no. 4, pp. 578-81. KoRRiNGA, P., 1957. Lunar periodicity. Chapter 27 in Treatise on Marine Ecology and Paleoecology, I, Ecology, edited by Joel W. Hedgpeth (New York). Kramp, P. L., 1959. The hydromedusae of the Atlantic Ocean and adjacent waters. Dana Rep. no. 46, pp. 1-283. Kritzler, H., 1952. Observations on the pilot whale in captivity. J. Mammal, xxxiii, pp. 321-34. Krogh, a., 1934. Physiology of the blue whale. Nature, Lond. cxxxiii, no. 3365, pp. 635-7. Kruger, H., 1939. Die Thaliaceen der 'Meteor' Expedition. Wiss. Ergebn. dtsch. atlant. Exped. 'Meteor', xiii, no. 2, pp. II 1-52. Kumagori, T. and Yanagawa, S., 1958a. Recurvature current in the Southern Ocean current. J. Tokyo Univ. Fish. (Special Edition), I, no. 3, pp. 231-40. 19586. Ice berg in the Antarctic Ocean. J. Tokyo Univ. Fish. (Special Edition), i, no, 3, pp. 285-94. Kuroda, N., i960. Analysis of sea bird distribution in the Northivest Pacific Ocean. Pacif. Sci. xiv, no. i, pp. 55-67. Lachambre, H. and Machuron, A., 1898. Andree and his Balloon (London). Lack, D., 1954. The Natural Regulation of Animal Numbers (Oxford). 1955- The mortality factors affecting adult numbers. From The Numbers of Man and Animals, edited by J. B. Cragg and N. W. Pirie (Edinburgh and London). LaFond, E. C, i960. Arctic oceanography by submarines. Reprinted from U.S. Naval Institute Proceedings for September, i960, pp. 90-6. Lakteonov, a. C, 1957. The International Geophysical Year in Antarctica (Leningrad). Larsen, C. a., 1894. The voyage of the 'Jason' to the Antarctic regions. Geogr. J. iv, pp. 333-44, 466-7. Lasker, R., i960. Utilization of organic carbon by a marine crustacean: Analysis with carbon-14. Science, cxxxi, no. 3407, pp. 1098-1X00. Laughton, a. S., 1959. The sea floor. Sci. Progr. xlvii, no. 186, pp. 230-49. Laurie, A. H., 1933. Some aspects of respiration in blue and fin whales. Discovery Rep. vil, pp. 363-406. 1937- The age of female blue whales and the effect of whaling on the stock. Discovery Rep. xv, pp. 223-84. Law, P. G., 1952. Through the pack-ice. Chapter 9 in The Antarctic Today, edited by Frank A. Simpson (Sydney). Law, P. G. and Bechervaise, J., 1957. Anare. Australia's Antarctic Outposts (Melbourne). Laws, R. M., 1958. Growth rates and ages of crabeater seals, Lobodon carcinophagus Jacquinot & Pucheran. Proc. Zool. Soc. Lond. cxxx, pt. 2, pp. 275-88. i959<'- The foetal growth rates of whales with special reference to the fin whale, Balaenoptera physalus Linn. Discovery Rep. xxix, pp. 281-308. 19596- On the breeding season of Southern Hemisphere fin whales, Balaenoptera physalus {Linn.). Norsk Hvalfangsttid. no. 7, pp. 329-51. 1960a. Problems of whale conservation. Trans. 25th N. Amer. Wildl. Conf., pp. 304-19. 19606. Researches on the period of conception, duration of gestation and growth of the foetus in the fin whale, based on data from international whaling statistics. Norsk Hvalfangsttid. no. 5, pp. 216-20. Lebour, M. v., 1924. The Euphausiidae in the neighbourhood of Plymouth and their importance as herring food. J. Mar. Biol. Ass. U.K. n.s. XIII, no. 2, pp. 402-32. 1926. A general survey of larval Euphausiids, with a scheme for their identification. J. Mar. Biol. Ass. U.K. n.s. xiv, no. 2, pp. 519-27. Leip, H., 1954. Der Grosse Fluss im Meer (Miinchen). Levick, G. Murray, 1915a. Natural history of the Add lie penguin. Brit. Antarct. Terra Nova Exped. 1910, Nat. Hist. Rep., Zool. I, no. 2, pp. 55-84, pis. i-xxi. • 191 5 6. Antarctic Penguins (London). LiLLiE, D. G., 1910. Observations on the anatomy and general biology of some members of the larger Cetacea. Proc. Zool. Soc. Lond. pp. 769-92. Lilljeborg, W., 1866. Synopsis of the cetaceous mammalia of Scandinavia (Sweden and Norway). Ray Soc. Publ., 1886, pp. 221-312. LiNDSEY, A. A., 1937. The Weddell Seal in the Bay of Whales. J. Mammal, xviii, no. 2, pp. 127-44. REFERENCES 455 LiNDSEY, A. A., 1938. Notes on the crab-eater seal. J. Mammal, xix, no. 4, pp. 456-61. LiOUVlLLE, J., 1913. Citacis de VAntartique. Deuxieme Expedition Antarctique Francjaise (1908-1910), pp. 1-248, pis. i-xv. LisiTZiN, A. P., i960. Bottom sediments of the eastern Antarctic Ocean. Deep-sea Res. vii, no. 2, pp. 89-99. LisiTziN, A. P. and Zhivago, A. V., i960. Marine geological work of the Soviet Antarctic Expedition, 1955-1957. Deep-sea Res. VI, no. 2, pp. 77-87. LoNNBERG, A. J. E., 1906. Contributions to the fauna of South Georgia. I. Taxonomic and biological notes on vertebrates. K. svenska VetenskAkad. Handl. XL, no. 5, pp. 1-104. Lucas, C. E., 1947. The ecological effects of external metabolites. Biol. Rev. xxii, no. 3, pp. 270-95. 1956 a. Plankton and basic production. Contribution no. Ill to Sea Fisheries. Their Investigation in the United Kingdom, edited by Michael Graham (London). 19566. Plankton and fisheries biology. Contribution no. iv to Sea Fisheries. Their Investigation in the United Kingdom, edited by Michael Graham (London). LuNEBERG, H., 1940. Vber die hydrographischen Untersuchungen siidlich Bouvet in der Walfangsaison 1938-39. Z. Fisch. Beiheft i, pp. 37-63. Macdonald, R., 1927. Food and habits of Meganyctiphanes norvegica. J. Mar. Biol. Ass. U.K. n.s. xiv, no. 3, pp. 753-84. MacGinitie, G. E., 1934. The egg-laying activities of the sea hare, Tethys californicus {Cooper). Biol. Bull., Woods Hole, Lxvii, no. 2, pp. 300-3. 1935- Ecological aspects of a Californian marine estuary. Amer. Midi. Nat. xvi, no. 5, pp. 629-765. MacGinitie, G. E. and MacGinitie, N., 1949. Natural History of Marine Animals (New York). MacGregor, J. S., 1959. Relation between fish condition and population size in the sardine, Sardinops caerulea. Fish. Bull., U.S. Lx, no. 166, pp. 215-30. Mackintosh, N. A., 1934. Distribution of the macroplankton in the Atlantic sector of the Antarctic. Discovery Rep. ix, pp. 65-160. 1937- The seasonal circulation of the Antarctic macroplankton. Discovery Rep. xvi, pp. 365-412. 1942a. The southern stocks of whalebone whales. Discovery Rep. xxii, pp. 197-300. 19426. The natural history of whales. Polar Rec. in, no. 24, pp. 553-63. 1946. The Antarctic convergence and the distribution of surface temperature in Antarctic waters. Discovery Rep. xxiii, pp. 177-212, pis. i-xiv. 1950. The work of the Discovery Committee. Proc. Roy. Soc. A, ecu, pp. 1-16. I959«- Biological problems in the regulation of whaling. New Scientist, v, no. 133, pp. 1229-31. 1959^- Biological problems in the regulation of whaling. Norsk Hvalfangsttid. no. 8, pp. 395-404. i960. The pattern of distribution of the Antarctic fauna. Proc. Roy. Soc. B, CLii, pp. 624-31. Mackintosh, N. A. and Ardley, R. A. B., 1936. The Royal Research Ship 'Discovery IP. Discovery Rep. xiii, pp. 77-106, text-figs. 1-5. Mackintosh, N. A. and Brow^, S. G., 1956. Preliminary estimates of the southern populations of the larger baleen whales. Norsk Hvalfangsttid. no. 9, pp. 469-80. Mackintosh, N. A. and Herdman, H. F. P., 1940. Distribution of the pack-ice in the Southern Ocean. Discovery Rep. xix, pp. 285-96, pis. Ixix-xcv. Mackintosh, N. A. and Wheeler, J. F. G., 1929. Southern blue and fin whales. Discovery Rep. i, pp. 257-540. McIntosh, W. C, 1887. On the pelagic fauna of our shores in its relation to the nourishment of the young food-fishes. Ann. Mag. Nat. Hist. XIX, ser. 5, pp. 137-45. Malkus, W. V. R., 1953. A recording bathypitotmeter. J. Mar. Res. xii, no. i, pp. 51-9. Mann, F. G., 1946. Ojo y Vision de las Ballenas. Biologica, Santiago, iv, pp. 23-71. Marr, J. W. S., 1927. A short zoological report. Appendix to F. A. Worsley's Under Sail in the Frozen North (London). 1933. Exploring the Antarctic in the 'Discovery IP. Discovery, xiv, no. 163, pp. 212-15. 1935. The South Orkney Islands. Discovery Rep. x, pp. 283-382. 1937. Notes on Antarctic Islands. Field Records of the National Institute of Oceanography (unpublished). 1938. On the operation of large plankton nets. Discovery Rep. xviii, pp. 105-20, text-figs. 1-4, pi. ii. I95S- Krill, the whale's food. Zoo Life, x, no. 2, pp. 56-8. 1956. Euphausia superba and the Antarctic surface currents. An advance note on the distribution of the whale food. Norsk Hvalfangsttid. no. 3, pp. 127-34. 1957- Further comments on the occurrence of Euphausia superba in a humpback whale caught off Southern Queensland. Norsk Hvalfangsttid. no. 4, pp. 181-2. Marshall, E. H., 1930. Report on a visit to the Ross Dependency. Geogr. J. lxxv, no. 3, pp. 244-51. Marshall, N. B., 1951. Bathypelagic fishes as sound scatterers in the ocean. J. Mar. Res. x, no. i, pp. 1-17. 1954- Aspects of Deep Sea Biology (London). 456 DISCOVERY REPORTS Marshall, N. B., 1955. Studies of Alepisauroid fishes. Discovery Rep. xxvii, pp. 303-36. i960. Sviimbladder structure of deep-sea fishes in relation to their systematics and biology. Discovery Rep. xxxi, pp. 1-122. Marshall, S. M. and Orr, A. P., 1953(2. Calanus finmarchicus. Egg production and egg development in Tromso Sound in spring. Acta Borealia, A. Scientia, no. 5 (Tromso Museum). 1953^- The production of animal plankton in the sea. From Essays in Marine Biology, the Richard Elmhirst Memorial Lectures (Edinburgh and London). 1955- The Biology of a Marine Copepod Calanus finmarchicus (Gunnerus) (Edinburgh and London). Martinez, A., 1951. Un Botanico en la Antartida Argentina. Idia, no. 46, pp. 14-22. Massy, A. L., 1932. Mollusca: Gastropoda Thecosomata and Gymnosomata. Discovery Rep. in, pp. 267-96. Matthews, D. H., 1959. Aspects of the geology of the Scotia Arc. Geol. Mag. xcvi, no. 6, pp. 425-41. Matthews, L. H., 1929a. The natural history of the elephant seal with notes on other seals found at South Georgia. Discovery Rep. I, pp. 233-56. 1929^. The birds of South Georgia. Discovery Rep. i, pp. 561-92. 1931- South Georgia the British Empire's Subantarctic Outpost (Bristol and London). 1932. Lobster-krill. Anomuran Crustacea that are the food of zvhales. Discovery Rep. v, pp. 467-84. 1937- The humpback whale, Megaptera nodosa. Discovery Rep. xvii, pp. 7-92. 1938a. Notes on the Southern right whale, Eubalaena australis. Discovery Rep. xvii, pp. 169-82. 19386. The sei whale, Balaenoptera borealis. Discovery Rep. xvii, pp. 183-290. 1951- Wandering Albatross. Adventures among the Albatrosses and Petrels in the Southern Ocean (London). Mauchline, J., 1959. The Development of the Euphausiacea [Crustacea) especially that o/Meganyctiphanes norvegica (M. Sars). Proc. Zool. Soc. Lond. cxxxii, pt. 4, pp. 627-39. Maury, M. F., 1855. The Physical Geography of the Sea (London). Mawson, D., 1915. The Home of the Blizzard (London). 1928. Unsolved problems of Antarctic exploration and research. Yrora. Problems of Polar Research, American Geographical Society Special Publication no. 7 (New York). 1940a. Hydrological observations. Sci. Rep. Aust. Antarct. Exped., ser. A, 11, pt. 4, pp. 103-25. 19406. Marine biological programme. Sci. Rep. Aust. Antarct. Exped., ser. A, 11, pt. 5, pp. 127-67. 1942. Narrative of the Voyages of the S.Y. 'Aurora'. Sci. Rep. Aust. Antarct. Exped., ser. A, i, pt. i, pp. 1-104. McNab, J., 1839. Extract from the Log of the Schooner 'Eliza Scott', Captain John Balleny, while S. of 55 S. Latitude, iSjg, kept by John McNab, Second Mate. The Antarctic Manual for the use of the Expedition of 1901 (London). Mellor, M., 1959. Mass balance studies in Antarctica. J. Glaciol. in, no. 26, pp. 522-33. Menegaux, a., 1907. Oiseaux. Expedition Antarctique Fran9aise (1903-1905), pp. 1-75. Meyer, H. H. F., 1923. Die Oberfldchenstromungen des Atlantischen Ozeans im Februar. Veroflf. Inst. Meeresk. Univ. Berl., Neue Folge, Reihe A, Heft 11, pp. 1-32, Chart. Mill, H. R., 1903. The Realm of Nature (London). - 1905. The Siege of the South Pole (London). Millais, J. G., 1906. The Mammals of Great Britain and Ireland (London). Millar, R. H., i960. Ascidiacea. Discovery Rep. xxx, pp. 1-160. Model, F., 1958. Ein Beitrag zur regionalen Ozeanographie der Weddellsee. Dtsch. Antarkt. Exped. 1938-39, Wiss. Ergebn. II, Lief. 2, pp. 63-96. MoiSEEV, P. A., 1957. Influence of oceanologic regimen of the Far-Eastern seas on commercial fish population. From Proceedings of the Unesco Symposium on Physical Oceanography (Tokyo). Moore, H. B., 1931. The specific identification of faecal pellets. J. Mar. Biol. Ass. U.K. n.s. xvii, no. 2, pp. 359-65. 1933- Faecal pellets from marine deposits. Discovery Rep. vii, pp. 17-26. 1950. The relation between the scattering layer and the Euphausiacea. Biol. Bull., Woods Hole, XCIX, no. 2, pp. 181-212. • 1958. Marine Ecology (New York). Moore, H. B. and Bauer, J. C, 1960. An analysis of the relation of the vertical distribution of three copepods to environmental conditions. Bull. Mar. Sci. of the Gulf and Caribbean, x, no. 4, pp. 430-43. Morgan, R., 1956. World Sea Fisheries (London). Morrell, B., 1832. A Narrative of Four Voyages, to the South Sea, North and South Pacific Ocean, Chinese Sea, Ethiopic and Southern Atlantic Ocean, Indian and Antarctic Ocean. From the Year 1822-1831 (New York). Motoda, S., 1953. New plankton samplers. Bull. Fac. Fish. Univ. Hokkaido, in, no. 3, pp. 181-6. 1959- Devices of simple plankton apparatus. Mem. Fac. Fish. Univ. Hokkaido, vii, nos. i and 2, pp. 73-94, pis. i-viii. Motoda, S. and Anraku, M., 1955. The variability of catches in vertical plankton hauls. Bull. Fac. Fish. Univ. Hokkaido, vi, no. 2, pp. 152-75. Muromtsev, a. M., 1956. The Pacific Ocean (Leningrad). 1958. General Account of the Hydrology of the Pacific Ocean (Leningrad). I REFERENCES 457 Murphy, R. C, 1928. Antarctic zoogeography and some of its problems. From Problems 0/ Polar Research, American Geographical Society Special Publication no. 7 (New York). 1936. Oceanic Birds of South America (New York). 1948. Logbook for Grace (London). Murray, James, 1909. Biology. Appendix I to E. H. Shackleton's The Heart of the Antarctic (London). ■ 1910. On collecting at Cape Royds. Brit. Antarct. Exped. 1907-9, Rep. Sci. Invest. I, Biol. pt. i, pp. 1-15. Murray, John, 1894. Notes on an important geographical discovery in the Antarctic regions. Reprinted from Scot. Geogr. Mag. for April, pp. 195-9. 1895. A summary of the scientific results obtained at the sounding, dredging, and trawling stations of H. M.S. Challenger. The Voyage of H.M.S. 'Challenger' (Edinburgh). Murray, John and Philippi, E., 1908. Die Grundproben der 'Deutschen Tiefsee Expedition'. Wiss. Ergebn. 'Valdivia', x, pp. 80-206. Naaktgeboren, C, Slijper, E. J. and Utrecht, W. L. Van, i960. Researches on the period of conception, duration of gestation and growth of the foetus in the fin whale, based on data from international whaling statistics. Norsk Hvalfangsttid. no. 3, pp. 1 13-19. N.\NSEN, P., 1928. The oceanographic problems of the still unknown Arctic regions. From Problems of Polar Research, American Geographical Society Special Publication no. 7 (New York). Nasu, K., 1959. Surface water condition in the Antarctic whaling Pacific Area in 1956-57. Sci. Rep. Whales Res. Inst., Tokyo, no. 14, pp. 137-43. Neaverson, E., 1934. The sea-floor deposits. I. General characters and distribution. Discovery Rep. ix, pp. 295-350, pis. xvii-xxii. Nemoto, T., 1957. Foods of baleen whales in the Northern Pacific. Sci. Rep. Whales Res. Inst., Tokyo, no. 12, pp. 33-89. 1958. Cocconeis diatoms infected on whales in the Antarctic. Sci. Rep. Whales Res. Inst., Tokyo, no. 13, pp. 185-91. • 1959- Food of baleen whales with reference to whale movements. Sci. Rep. Whales Res. Inst., Tokyo, no. 14, pp. 149-290. Nemoto, T. and Nasu, K., 1958. Thysanoessa macrura as a food of baleen whales in the Antarctic. Sci. Rep. Whales Res. Inst., Tokyo, no. 13, pp. 193-9. NiCOL, J. A. C, i960. The Biology of Marine Animals (London). NiSHiWAKi, M., 1950. On the body weight of whales. Sci. Rep. Whales Res. Inst., Tokyo, no. 4, pp. 184-209. Nishiwaki, M. and Hayashi, K., 1950. Biological survey of fin and blue whales taken in the Antarctic season 1947-48 by the Japanese fleet. Sci. Rep. Whales Res. Inst., Tokyo, no. 3, pp. 132-89. Nishiwaki, M. and Oye, T., 195 i. Biological investigation on blue whales (Balaenoptera muscuhis) and fin whales (Balaenoptera physalus) caught by the Japanese whaling fleets. Sci. Rep. Whales Res. Inst., Tokyo, no. 5, pp. 91-167. Nordenskjold, O., 1905. Antarctica (London). Nybelin, O., 1947. Antarctic fishes. Sci. Res. Norweg. Antarct. Exped. no. 26, pp. 1-76. 1951. Subantarctic and Antarctic fishes. Sci. Res. Norweg. 'Brategg' Exped. 1947-48, no. 2, pp. 1-32. Olsen, 0., 1913. On the external characters and biology of Bryde's whale (Balaenoptera brydei), a new rorqual from the coast of South Africa. Proc. Zool. Soc. Lond. lxxii, pp. 1073-89. Olsen, S., 1954. South Georgian cod (Notothenia rossii marmorata Fischer). Norsk Hvalfangsttid. no. 7, pp. 373-82. 1955. A contribution to the systematics and biology of Chaenichthyid fishes from South Georgia. Nytt. Mag. Zool. in, PP- 79-93- Ommanney, F. D., 1933. Whaling in the Dominion of New Zealand. Discovery Rep. vii, pp. 239-52. 1936. Rhincalanus gigas [Brady) a copepod of the southern macroplankton. Discovery Rep. Xiii, pp. 277-384. 1938. South Latitude (London). 1949- The Ocean (London). Omura, H., 1957. Report on two right whales caught off Japan for scientific purposes under Article VIII of the International Convention for the regulation of whaling. Norsk Hvalfangsttid. no. 7, pp. 374-90. 1958. North Pacific right whale. Sci. Rep. Whales Res. Inst., Tokyo, no. 13, pp. 1-52. Ottestad, P., 1956. On the size of the stock of Antarctic fin whales relative to the size of the catch. Norsk Hvalfangsttid. no. 6, pp. 298-308. Parker, G. H., 1902. The reactions of copepods to various stimuli and the bearing of this on daily depth-migrations. Bull. U.S. Fish. Comm. 1901, xxi, pp. 103-23. Parkes, a. S., 1955. Some biological aspects of population problems. From The Numbers of Man and Animals, edited by J. B. Cragg and N. W. Pirie (Edinburgh and London). Parrish, B. B., 1956. The cod, haddock and hake. Contribution no. vii to Sea Fisheries. Their Investigation in the United Kingdom, edited by Michael Graham (London). Parrish, B. B., Saville, A., Cr.\ig, R. E., Baxter, I. G. and Priestley, R., 1959. Observations on herring spawning and larval distribution in the Firth of Clyde in 1958. J. Mar. Biol. Ass. U.K. xxxviii, no. 3, pp. 445-53. 54 ^58 DISCOVERY REPORTS Paulsen, O., 1906. Studies on the biology o/Calanus finmarchicus in the waters round Iceland. Medd. Komm. Havundersog., Kbh., ser. Plankton, i, no. 4, pp. 1-2 1. iQog. Plankton investigations in the "waters round Iceland and in the North Atlantic in igo4. Medd. Komm. Havundersog., Kbh., ser. Plankton, i, no. 8, pp. 1-57. Peres, J.-M., 1958. Remarques generales sur un ensemble de quinze plongees effectuees avec le Bathyscaphe F.N.R.S. III. Ann. Inst, oceanogr. Monaco, n.s. xxxv, pp. 259-85, pis. xix-xxii. Perkins, J. E., 1945. Biology at Little America III, the west base of the United States Antarctic Service Expedition, ig3g-ig4i. Proc. Amer. Phil. Soc. lxxxix, no. i, pp. 270-84. Peters, H., 1955. Vber das Vorkommen des Walkrebschens Euphausia superba Dana und seine Bedeutung fUr die Ernahrung der siidlichen Bartenwale. Archiv. fiir Fischereiwissenschaft, 6, pp. 288-304. Peters, N., 1938. Aus der Naturgeschichte der Wale. From Der netie Deutsche Walfang (Hamburg). Petersen, C. G. J., 1915. Appendix to Report XXI from the Danish Biological Station. igi4. Rep. Danish Biol. Sta. xxii, 7 pp. 1918. The sea bottom and its production of fish-food. Rep. Danish Biol. Sta. xxv, pp. 1-62. PiccARD, A., 1956. In Balloon and Bathyscaphe (London). Pirie, J. H. Harvey, 1905 a. Summer work in the South Orkneys. Scot. Geogr. Mag. xxi, no. i, pp. 34-7. 19056. Deep-sea deposits of the South Atlantic Ocean and Weddell Sea. Scot. Geogr. Mag. xxi, no. 8, pp. 413-17. 1908. Zoological log of the Summer Station, Omond House. Rep. Scot. Nat. Antarct. Exped., Zool. iv, pt. i, pp. 85-101. Pirie, J. H. Harvey and Brown, R. N. Rudmose, 1905. The Scottish National Antarctic Expedition. Second Antarctic Voyage of the 'Scotia'. Scot. Geogr. Mag. xxi, pp. 24-33. PoNOMAREVA, L. A., 1954. Weight characteristics o/ Euphausia in the Japanese Sea. C.R. Acad. Sci. U.S.S.R. xcix, no. i, pp. 169-71. 1959. Reproduction of Euphausiidae of the Sea of Japan and development of their early larval stages. Zool. Zh. xxxviii, no. II, pp. 1649-62. PoNTiNG, H. G., 1923. The Great White South (London). PoRSiLD, M. P., 1918. On 'Savssats': A crowding of Arctic animals at holes in the sea ice. Geogr. Rev. vi, no. 3, pp. 215-28. Powell, A. W. B., 1951. Antarctic and Subantarctic Mollusca: Pelecypoda and Gastropoda. Discovery Rep. xxvi, pp. 47-196. Powell, G., 1822. Notes on South- Shetland, etc. Memoir printed to accompany R. H. Laurie's Chart of South Shetland, including Coronation Island, etc. (London). Priestley, R. E., 1914. Antarctic Adventure: Scott's Northern Party (London). 1923. Physiography (Robertson Bay and Terra Nova Bay Regions). Brit. Antarct. Terra Nova Exped. 1910, pp. 1-87. PuRVES, P. E., 1955. The wax plug in the external auditory meatus of the Mysticeti. Discovery Rep. xxvii, pp. 293-302, pis. xiv-xviii. Raab, F., 191 5. Beitrag zur Anatomic und Histologic der Euphausiiden. Arb. zool. Inst. Univ. Wien, xx, pp. 125-58, pis. i-ii. Racovitza, E. G., 1900a. Resultats generaux de I'Expedition Antarctique Beige. La Geographic, i, pp. 81-92. 19006. General results of the Belgian Antarctic Expedition. Appendix i to F. A. Cook's Through the First Antarctic Night (London). 1903. Citaces. Result. Voy. Belgica, Rapp. Sci., Zool. pp. 1-142. Rae, K. M., 1958. Parameters of the marine environment. From Perspectives in Marine Biology, edited by A. A. Buzzati-I Traverso (University of California Press). Rammner, W., 1956. Sdugetiere. In vol. iv of Brahms Tierleben (Leipzig and Jena). Ravninger, R., 1955. Report on the whale marking voyage of the 'Enern' to the Antarctic ig54. Norsk Hvalfangsttid. no. 6, ^ pp. 310-15. Rayner, G. W., 1935. The Falkland species of the crustacean genus Munida. Discovery Rep. X, pp. 209-45. 1940. Whale marking, progress and results to December iQjg. Discovery Rep. xix, pp. 245-84, pis. xliii-lxviii. 1948. Whale marking II. Distribution of blue, fin and humpback whales marked from ig32 to igjS. Discovery Rep. xxv, ] pp. 31-8. Redfield, a. C, 1958. The inadequacy of experiment in marine biology. From Perspectives in Marine Biology, edited by I A. A. Buzzati-Traverso (University of California Press). Regnard, p., 1 89 1. Reserches expirimentales sur les Conditions Physiques de la Vie dans les Eaux (Paris). Richard, J., 1907. L'Oceanographie (Paris). Riedel, W. R., 1958. Radiolaria in Antarctic sediments. B.A.N.Z. Antarct. Res. Exped., 1929-1931, ser. B (Zool. and Bot), j VI, pt. 10, pp. 219-47. Riedl, R., 1958. An attempt to test the efficiency of ecological field methods and the validity of their results. From Perspectives^ in Marine Biology, edited by A. A. Buzzati-Traverso (University of California Press). Risting, S., 1912. Knolvalen. Norsk FiskTid. xxxi, pp. 437-49. 1922. Av Hvalfangstens Historic (Oslo). REFERENCES 459 RiSTiNG, S., 1927. Whales and uliale foetuses. Statistics 0/ catch and measurements collected from the Nonvegian Whalers' Association ig22-ig2j. Rapp. Cons. Explor. Mer, l, pp. 1-122. RiTSCHER, A., 1942. Die Deutsche Antarktische Expedition ig38-3g. Organisation, Vorbereitung und Verlauf der Expedition (Leipzig). Robert, H., 1922. L'Emploi du Filet et de la Pompe dans les Peches de Plancton. Notes critiques de Mithodologie. Ann. Biol. lacust. XI, Fasc. i, pp. 208-36. Roberts, B., 1940a. The life cycle of Wilson's Petrel Oceanites oceanicus (Kuhl). Sci. Rep. Brit. Grahamld. Exped. i, no. 2, pp. 141-94. 1940ft. The breeding behaviour of penguins with special reference to Pygoscelis papua {Forster). Sci. Rep. Brit. Grahamld. Exped. I, no. 3, pp. 195-254. 1941- A bibliography of Antarctic ornithology. Sci. Rep. Brit. Grahamld. Exped. I, no. 9, pp. 337-67. 1958. Chronological list of Antarctic expeditions. Polar Rec. ix, no. 59, pp. 97-134. Robertson, D., 1892. Jottings from my note-book. Trans. Nat. Hist. Soc. Glasg. iii (n.s.), pp. 22-4. Robin, G. de Q., 1959. Future Antarctic research. New Scientist, v, no. 120, pp. 504-5. i960. The Antarctic. Essay no. 6 in The World around Us, edited by Sir Graham Sutton (London). RoMANOVSKY, V., 1955. L'Homme Poisson et ses Conquetes (Paris). RONNE, F., 1952. Recent expeditions and institutional activities. Contribution by Commander Finn Ronne, U.S.N.R., to chapter 16 in The Antarctic Today, edited by Frank A. Simpson (Sydney). ROOY, M. P. VAN, 1957. Meteorology of the Antarctic (Pretoria). Ross, J. C, 1847. A Voyage of Discovery and Research in the Southern and Antarctic Regions (London). RoucH, J., 1922. Manuel d'Oceanographie Physique (Paris). RouNSEFELL, G. A. and Everhart, W. H., 1953. Fishery Science: Its Methods and Applications (New York). RouTH, M., 1949. Ornithological observations in the Antarctic Seas, ig46-4y. Ibis, xci, no. 4, pp. 577-606. Russell, F. S., 1927a. The vertical distribution of marine macroplankton. V. The distribution of animals caught in the ring-trawl in the daytime in the Plymouth area. J. Mar. Biol. Ass. LT.K. n.s. xiv, no. 3, pp. 557-608. 1927ft. The vertical distribution of plankton in the sea. Biol. Rev. 11, no. 3, pp. 213-56. 1928. The vertical distribution of marine macroplankton. VIII. Further observations on the diurnal behaviour of the pelagic young of teleostean fishes in the Plymouth area. J. Mar. Biol. Ass. U.K. n.s. xv, no. 3, pp. 829-50. Russell, H. C, 1895. Icebergs in the Southern Ocean. Reprinted from J. Roy. Soc. N.S.W., pp. 1-31. RuSTAD, D., 1930. Euphausiacea with notes on their biogeography and development. Sci. Res. Nor\veg. Antarct. Exped. no. 5, pp. 1-83. 1934. On the Antarctic Euphausiids from the ' Norwegia' expeditions ig2g-30 and igjo-ji. Sci. Res. Norweg. Antarct. Exped. no. 12, pp. 1-53. RuUD, J. T., 1927. Zooplancton of the coast of More, ig26. Rapp. Cons. Explor. Mer, Appendix A, pp. 120-4. 1932. On the biology of Southern Euphausiidae. Hvalrad. Skr. no. 2, pp. 1-105. 1936. Euphausiacea. Rep. Danish Oceanogr. Exped. Medit. 11 (Biology), D. 6, pp. 1-86. 1955. The mortality rates of Antarctic fin whale stocks. Pap. Mar. Biol, and Oceanogr., Deep-sea Res., Suppl. to vol. in, pp. 257-60. 1956. Do we knozv anything about the stocks of fin whales in the Antarctic} Norsk Hvalfangsttid. no. 11, pp. 625-30. RuuD, J. T., JONSGARD, A. and Ottestad, P., 1950. Age studies on blue whales. Hvalrad. Skr. no. 33, pp. 1-72. RuuD, J. T. and 0ynes, P., 1959. Whale marking carried out by Norwegian catchers in the Antarctic, seasons ig56-5y and ig§y-8. Norsk Hvalfangsttid. no. 2, pp. 56-62. Saito, L and Mishima, S., 1953. Studies on detection of plankton by echo-sounder {or fish-finder) and selection of fishing station in a fishing ground (preliminary report) detection of D.S.L. by echo-sounder. Bull. Fac. Fish. Univ. Hokkaido, iii, no. 4, pp. 269-72. Sakiura, H., Ozaki, K. and Fujino, K., 1953. Biological investigation of the Northern Pacific baleen whales caught by the Japanese whaling fleet in ig52. Fisheries Agency of Japanese .Government (Tokyo). Sars, G. O., 1875. Om 'Blaahvalen'. (Balaenoptera Sibbaldii, Gray). Forh. VidenskSelsk. Krist. 1874, pp. 227-41. 1883. Preliminary notices on the Schizopoda of H. M.S. 'Challenger' Expedition. Forh. VidenskSelsk. Krist. no. 7, pp. 1-43. 1885. Report on the Schizopoda collected by H.M.S. 'Challenger' during the years 1873-76. The Voyage of H.M.S. 'Challenger', Zool. xiii, pp. 1-228. Bars, M., 1829. Bidrag til Soedyrenes Naturhistorie (Bergen). Schaefer, M. B. and Revelle, R., 1959. Marine resources. Contribution no. 4 to Natural Resources, edited by M. R. Huberty and W. L. Flock (New York). ScHEFFER, V. B., 1958. Seals, Sea Lions and Walruses (Stanford University Press, California). ScHOTT, G., 1924. Physische Meereskunde (Berlin and Leipzig). 54-2 46o DISCOVERY REPORTS ScHOTT, G., 1942. Geographie des Atlantischen Ozeans (Hamburg). ScHOUTEDEN, H., 1902. Le Phototropisme de Daphnia magna Straus. Ann. Soc. Ent. Belg. xlvi, no. 7, pp. 352-62. Schubert, K., 1955. Der Walfang der Gegenwart (Stuttgart). ScHULZ, W., 1931. Die Orientierung des Ruckenschwimmers zum Licht und zur Stromung. Z. vergl. Physiol, xiv, pp. 392-404. ScoRESBY, W. , 1 820. An Account of the Arctic Regions with a History and Description of the Northern Whale-Fishery (Edinburgh). 1823. Journal of a Voyage to the Northern Whale-Fishery (Edinburgh). Scott, R. F., 1905. Results of the National Antarctic Expedition. I. Geographical. Geogr. J. xxv, no. 4, pp. 353-70. Sewell, R. B. Seymour, 1948. The free-swimming planktonic Copepoda. Geographical distribution. Sci. Rep. Murray Exped. VIII, no. 3, pp. 317-579. Shackleton, E. H., 1909. The Heart of the Antarctic (London). 1919- South (London). Sheard, K., 1953. Taxonomy, distribution and development of the Euphausiacea (Crustacea). B.A.N.Z. Antarct. Res. Exped., 1929-193 1, sen B (Zool. and Bot), viii, pt. i, pp. 1-72. Sherratt, R., 1821. Article in the Imperial Magazine, London, the text of which, with an introductory note by Brian Roberts, was pubhshed in the Polar Record, vi, no. 43, in January 1952. Shumskiy, p. a., 1959. Is Antarctica a continent or an archipelago} J. Glaciol. in, no. 26, pp. 455-7. Simpson, A. C, 1956. The pelagic phase. Contribution no. vi to Sea Fisheries. Their Investigation in the United Kingdom, edited by Michael Graham (London). Singer, B., 1959. The food fish eat (2). World Fishing, viii, no. 9, p. 59. Sladen, W. J. L., 1958. The Pygoscelid penguins. Sci. Rep. Falkland Islands Dependencies Survey, no. 17, pp. 1-97. Slijper, E. J., 1958. Walvissen (Amsterdam). 1961. The still unexplained mystery of the whales. Norsk Hvalfangsttid. no. 2, pp. 41-54. Smith, C. E., 1922. From the Deep of the Sea. An Epic of the Arctic. Being the diary of the late Charles Edward Smith, M.R.C.S., surgeon of the whaleship 'Diana' of Hull, edited by his son, Charles Edward Smith Harris, M.B., Ch.B., surgeon, Cunard Steamship Company, late temporary Surgeon-Lieutenant, R.N. (London). Smith, C. V., 1920. Memorandum on the currents of the South Atlantic Ocean. Appendix xxi to Report of the Interdepartmental Committee on Research and Development in the Dependencies of the Falkland Islands, pp. 135-6 (London, H.M. Stationery Office [Cmd. 657]). Sgrlle, T., 1924. Vekten av en Hval. Norsk Hvalfangsttid. no. 7, p. 74. Southern, R. and Gardiner, A. C, 1926. Reports from the Limnological Laboratory. I. The seasonal distribution of the Crustacea of the plankton in Lough Derg and the River Shannon. Sci. Invest. Min. Fish. Irish Fr. St. 1926, i, pp. 1-170. Southwell, T., 1881. The Seals and Whales of the British Seas (London). Station List, 1929. Discovery Investigations Station List ig25-ig2y. Discovery Rep. i, pp. 1-140, pis. i-vi. 1930. Discovery Investigations Station List ig2j-ig2g. Discovery Rep. in, pp. 1-132, pis. i-x. 1932. Discovery Investigations Station List ig2g-ig3i. Discovery Rep. iv, pp. 1-232, pis. i-v. 1941. Discovery Investigations Station List igji-igjj. Discovery Rep. xxi, pp. 1-226, pis. i-iv. 1942. Discovery Investigations Station List ig33-ig35. Discovery Rep. xxii, pp. 1-196, pis. i-iv. 1944. Discovery Investigations Station List ig35-ig37. Discovery Rep. xxiv, pp. 1-196, pis. i-iii. 1947. Discovery Investigations Station List ig3y-ig3g. Discovery Rep. xxiv, pp. 197-422, pis. iv-vi. 1949. Discovery Investigations Station List ig3i-ig38. Discovery Rep. xxv, pp. 143-280, pis. xxxiv-xxxvii. 1957. Discovery Investigations Station List ig^o-igsi. Discovery Rep. xxviii, pp. 299-398, pis. iii-v. Stephansson, v., 1921. The Friendly Arctic (New York). 1928. The resources of the Arctic and the problem of their utilization. From Problems of Polar Research, American Geographical Society Special Publication no. 7 (New York). Steuer, a., 1910. Planktonkunde (Leipzig and Berlin). 19 II. Leitfaden der Planktonkunde (Leipzig and Berlin). Stommel, H., 1958. The Gulf Stream. A Physical and Dynamical Description (London). Stonehouse, B., 1953. The Emperor penguin Aptenodytes forsteri Gray. I. Breeding behaviour and development. Sci. Rep. Falkland Islands Dependencies Survey, no. 6, pp. 1-3 1. 1956. The Brown Skua Catharacta skua lonnbergi {Mathews) of South Georgia. Sci. Rep. Falkland Islands Dependencies Survey, no. 14, pp. 1-24. i960. The King penguin Aptenodytes patagonica of South Georgia. I. Breeding behaviour and development. Sci. Rep. Falkland Islands Dependencies Survey, no. 23, pp. 1-81. Stott, F. C, 1936. The marine foods of birds in an island fjord region in West Spitsbergen. Part I. Plankton and in shore benthos. J. Anim. Ecol. v, no. 2, pp. 356-69. S VERDRUP, H . U . , 1 93 1 . The origin of the deep water of the Pacific Ocean as indicated by the oceanographic work of the ' Carnegie'. Beitr. Gcophys. xxix, pp. 95-105. REFERENCES 46« SvERDRUP, H. U., 1933. On vertical circulation in the ocean due to the action of the wind with application to conditions zvithin the Antarctic circumpolar current. Discovery Rep. vii, pp. 139-70. 1942. Oceanography for Meteorologists (New York). 1954. Oceanography. Chapter 5 in The Earth as a Planet, edited by Gerard P. Kuiper (Chicago). SvERDRUP, H. U., Johnson, H. M. and Fleming, R. H., 1946. The Oceans. Their Physics, Chemistry, and General Biology (New York). Swallow, J. C, 1955. A neutral-buoyancy float for measuring deep currents. Deep-sea Res. iii, no. i, pp. 74-81. Swallow, J. C. and Worthington, L. V., 1957. Measurements of deep currents in the western North Atlantic. Nature, Lond. CLxxix, no. 4571, pp. 1 183-4. 1961. An observation of a deep count ercurrent in the western North Atlantic. Deep-sea Res. viii, no. i, pp. 1-19. Swithinbank, C, 1957 a. The morphology of the ice shelves of Western Dronning Maud Land. Sci. Res. Norweg. Brit. Swed. Antarct. Exped., 1949-52, in, pp. 1-37. 19576- The regime of the ice shelf at Maudheim as shozvn by stake measurements. Sci. Res. Norweg. Brit. Swed. Antarct. Exped., 1949-52, III, pp. 43-75. Sysoiev, N. N., 1957. Developments and future in the oceanographic instrument construction in the U.S.S.R. Proceedings of the Unesco Symposium on Physical Oceanography (Tokyo). Taljaard, J. J., 1957. Geographical and hydrological features of the Antarctic. Chapter I in Meteorology of the Antarctic, edited by M. P. van Rooy (Pretoria). TIning, a. v., 1955. On the breeding areas of the swordfish (Xiphias). Pap. Mar. Biol, and Oceanogr., Deep-sea Res., Suppl. to vol. 3, pp. 438-50. Tattersall, W. M., 1908. Crustacea. VII. Schizopoda. Nat. Antarct. Exped. 1901-1904, Nat. Hist, iv, pp. 1-42, pis. i-viii. 1913. The Schizopoda, Stomatopoda, and non- Antarctic Isopoda of the Scottish National Antarctic Expedition. Trans. Roy. Soc. Edinb. xlix, pp. 865-94. 1918. Euphausiacea and Mysidacea. Sci. Rep. Aust. Antarct. Exped., ser. C (Zool. and Bot.), v, pt. 5, pp. 1-15. 1924. Crustacea. Part vm.— Euphausiacea. Brit. Antarct. Terra Nova Exped. 1910, Nat. Hist. Rep., Zool. viii, no. I, pp. 1-36. Taube, E., 1915. Beitrdge zur Entwicklungsgeschichte der Euphausiiden. II. Von der Gastrula bis zum Furciliastadium. Z. Wiss. Zool. cxiv, pp. 577-656. Taylor, G., 1916. With Scott: The Silver Lining (London). 1930. Antarctic Adventure and Research (New York and London). Taylor, R. J. F., 1957. An unusual record of three species of whale being restricted to pools in Antarctic sea-ice. Proc. Zool. Soc. Lond. cxxix, pt. 3, pp. 325-30. Tchernia, p., 1950. Observations sur la D.S.L. faites a bord du ' Commandant Charcot ' {Campagne ig48-4g). Comite Central d'Oceanographique et d'fitude des Cotes (C.O.E.C.). Bulletin d' Information, no. i, pp. 7-20 (Service Hydrographique de la Marine, Paris). Tebble, N., i960. The distribution of pelagic polychaetes in the South Atlantic Ocean. Discovery Rep. xxx, pp. 161-300. Thomas, C. W., 1959. Lithology and zoology of an Antarctic Ocean bottom core. Deep-sea Res. vi, no. i, pp. 5-15. Thompson, S. Y., Ganguly, J. and Kon, S. K., 1949. The conversion of j^ -carotene to vitamin A in the intestine. Brit. J. Nutr. Ill, no. I, pp. 77-8. Thorpe, W. H., 1956. Learning and Instinct in Animals (London). 1958. Ethology as a new branch of biology. From Perspectives in Marine Biology, edited by A. A. Buzzati-Traverso (University of California Press). Thorson, G., 1946. Reproduction and larval development of Danish maritie bottom invertebrates, with special reference to the planktonic larvae in the sound (Oresund). Medd. Komm. Havundersog., Kbh., ser. Plankton, iv, no. i, pp. 1-523. 1950- Reproductive and larval ecology of marine bottom invertebrates. Biol. Rev. xxv, no. i, pp. i-45- 1957. Bottom communities [sublittoral or shallow shelf). Chapter 17 in Treatise on Marine Ecology and Paleoecology, I, Ecology, edited by Joel W. Hedgpeth (New York). 1958. Parallel level-bottom communities, their temperature adaptation, and their 'balance' between predators and food animals. From Perspectives in Marine Biology, edited by A. A. Buzzati-Traverso (University of California Press). Tinbergen, N., 1953. Social Behaviour in Animals (London). TOLSTIKOV, E. L, 1957. The Arctic and Antarctic programme of the International Geophysical Year. From Proceedings of the xith General Assembly of the International Union of Geodesy and Geophysics (Toronto). ToMiLiN, A. G., 1954. Adaptive types in the order Cetacea {the problem of an ecological classification of Cetacea). Zoologicheskii Zh. XXXIII, no. 3, pp. 677-92. Tonner, F., 1935. Schwimmreflex und Zentrenfunktion bei Aeschna Larven. Z. vergl. Physiol, xxii, pp. 517-23- ToNOLLi, V. and Tonolli, L., 1958. Irregularities of distribution of plankton communities: considerations and methods. From Perspectives in Marine Biology, edited by A. A. Buzzati-Traverso (University of California Press). 462 DISCOVERY REPORTS ToTTON, A. K., 1954. Siphonophora of the Indian Ocean together with systematic and biological notes on related specimens from other oceans. Discovery Rep. xxvii, pp. 1-162. Trouessart, E. L., 1907. Mammiferes Pinnipedes. Expedition Antarctique Fran9aise (1903-1905), pp. 1-27. True, F. W., 1904. The whalebone whales of the western North Atlantic, compared with those occurring in European waters; with some observations on the species of the North Pacific. Smithson. Contr. Knowl. xxxiii, no. 1414, pp. 1-332. Tucker, D. W., 19590. A new solution to the Atlantic eel problem. Nature, Lond. clxxxiii, no. 4660, pp. 495-501. 19596. Eel migration. Nature, Lond. clxx.xiv, no. 4695, pp. 1281-3. U.S.N. Hydrographic Office, 1955. Antarctica. Map compiled from available sources to 1955 (Washington). 1956. Field Report Oceanographic Observations U.S. Navy Antarctic Expedition ig54-ig55. Misc. Rep. 16331 (Washington). i957«- Operation Deep Freeze II, ig^S-igsy. Oceanographic survey results. Tech. Rep. TR-29 (Washington). • 19576- Oceanographic Atlas of the Polar Seas. Part I. Antarctic. H.O. Publ. no. 705 (Washington). U.S.S.R. Ministry of Defence, 1953. Marine Atlas, vol. 11, Physico-Geographic (Moscow). Vangstein, E., 1956. Estimates of the southern populations of the larger baleen whales. Norsk Hvalfangsttid. no. 9, pp. 483-4. Vanhoffen, E., 1896. Schivarmbildung im Meere. Zool. Anz. xix, no. 520, pp. 523-6. 1902. Biologische Beobachtungen. Dtsch. SiidpolExped., Veroff. Inst. Meeresk. Univ. Berl. i, pp. 55-72. 1903. Biologische Bericht. Dtsch. SiidpolExped., Veroff. Inst. Meeresk. Univ. Berl. v, pp. 143-54. ViLLiERS, A. J., 1924. To the Frozen South (Hobart). 1925. Whaling in the Frozen South (London). Vinogradov, M. E., 1955. Vertical migrations of zooplankton and their role in the nutrition of deep-water pelagic fauna. Trans. Inst. Oceanol. Acad. Sci. U.S.S.R. xiii, pp. 71-6. Waite, E. R., 1916. Fishes. Sci. Rep. Aust. Antarct. Exped., ser. C (Zool. and Bot.), in, pt. i, pp. 1-92. Walford, L. a., 1955. New directions in fishery research. Pap. Mar. Biol, and Oceanogr., Deep-sea Res., Suppl. to vol. 3, PP- 471-3- 1958. Living Resources of the Sea (New York). Waterman, T. H., 1955. Polarization of scattered sunlight in deep water. Pap. Mar. Biol, and Oceanogr., Deep-sea Res., Suppl. to vol. 3, pp. 426-34. 1Q58. Polarized light and plankton navigation. From Perspectives in Marine Biology, edited by A. A. Buzzati-Traverso (University of California Press). Webster, W. H. B., 1834. Narrative of a Voyage to the Southern Atlantic Ocean (London). Weddell, J., 1827. A Voyage towards the South Pole (London). Wexler, H., 1959. The Antarctic Convergence — or Divergence^ From The Atmosphere and the Sea in Motion, edited by Bert Bolin (New York). Wheeler, J. F. G., 1934. On the stock of whales at South Georgia. Discovery Rep. ix, pp. 351-72. Wild, F., 1923. Shackleton's Last Voyage (London). Wilkes, C, 1852. Narrative of the United States Exploring Expedition (London). Williamson, G. R., 1959- Three unusual rorqual whales from the Antarctic. Proc. Zool. Soc. Lond. cxxxiii, pt. i, pp. 135-44. Wilson, D. P., 1958. Some problems in larval ecology related to the localized distribution of bottom animals. From Perspectives in Marine Biology, edited by A. A. Buzzati-Traverso (University of California Press). Wilson, E. A., 1902. Notes on Antarctic seals. Report on the Collections of Natural History made in the Antarctic Regions during the Voyage of the ' Southern Cross ' (London). 1905. The distribution of Antarctic seals and birds. Geogr. J. xxv, no. 4, pp. 392-6. . 1907a. Mammalia. Cetacea and Pinnipedia. Nat. Antarct. Exped. 1901-1904, Nat. Hist. 11, pp. 1-66. 19076. Aves. Nat. Antarct. Exped. 1901-1904, Nat. Hist. 11, pp. 1-116. Wilton, D. W., 1908. Zoological log of 'Scotia'. Rep. Scot. Nat. Antarct. Exped., Zool. iv, pt. i, pp. 1-60. WiMPENNY, R. S., i960. Young plaice hauls off the English east coast. Fish. Invest. Lond., ser. 2, xxiil, no. i, pp. 1-20. Wolff, T., i960. The hadal community, an introduction. Deep-sea Res. vi, no. 2, pp. 95-124. Wordie, J. M., 1921a. Depths and deposits of the Weddell Sea. Trans. Roy. Soc. Edinb. lii, pt. iv, no. 30, pp. 781-93. 19216. The natural history of pack-ice as observed in the Weddell Sea. Trans. Roy. Soc. Edinb. lii, pt. iv, no. 31, PP- 795-829. Wordie, J. M. and Kemp, S., 1933. Observations on certain Antarctic icebergs. Geogr. J. lxxxi, no. 5, pp. 428-34. WoRSLEY, F. A., 1927. Under Sail in the Frozen North (London). 193 1. Endurance. An Epic of Polar Adventure (New York). 1959. Shackleton's Boat Journey (London). WoRTHiNGTON, L. V., 1959. The 18° zvater in the Sargasso Sea. Deep-sea Res. v, no. 4, pp. 297-305. Wright, C. S. and Priestley, R. E., 1922. Glaciology. Brit. Antarct. Terra Nova Exped. 1910, pp. 1-485. REFERENCES 463 WiJST, G., 1933. Das Bodenwasser und die GUederung der Atlantischen Tie/see. Wiss. Ergebn. dtsch. atlant. Exped. 'Meteor', VI, pt. I, pp. 1-107. 1955. Stromgeschmindigkeiten im Tiefen- und Bodenwasser des Atlantischen Ozeans atif Grand dynamischer Berechnung der Meteor— Profile der Deutschen Atlantischen Expedition 1925-27. Pap. Mar. Biol, and Oceanogr., Deep-sea Res., Suppl. to vol. 3, pp. 373-97. Wynne-Ed\v.\rds, C. J. C, i960. A hydrographic survey unit in British Antarctica. Geogr. J. cxxvi, pt. 3, pp. 310-17. Yevgenov, N. I., 1957. Marine Currents (Leningrad). Young, J. Z., 1954. Memory, heredity and information. From Evolution as a Process, edited by Julian Huxley, A. C. Hardy and E. B. Ford (London). Zelikman, E. a., 1958. On gonad maturation and female productivity in species of euphausians abundant in the Barents Sea. C.R. Acad. Sci. U.S.S.R. c.wiii, no. i, p. 201. i960. Some aspects of the behaviour of Barents Sea euphausiids, and possible reasons for their seasonal vertical migration. C.R. Acad. Sci. U.S.S.R. x, edition 4, pp. 62-7. Zenkevich, L. a., 1957. The distribution of sea floor fauna in the North-west Pacific. Proceedings of the Unesco Symposium on Physical Oceanography (Tokyo). 1958. Immediate problems in the development of marine biology. From Perspectives in Marine Biology, edited by A. A. Buzzati-Traverso (University of California Press). Zenkovich, B. a., 1937. Food of far eastern whales. C.R. Acad. Sci. U.S.S.R. xvi, no. 4, pp. 239-42. ZiMMER, C, 1913. Untersuchungen iiber den inneren Bau von Euphausia superba Dana. Zoologica, Stuttgart, Lxvii, pp. 65-128, text-figs. 1-5, pis. viii-xiv. 1914. Die Schizopoden der Deutschen Siidpolar-Expedition, igoi-j. Dtsch. SiidpolExped. xv, pp. 377-445. 1956. Euphausiacea. Bronn's Klassen, v, Abt. I, Buch 6, Teil iii, Lief, i and 2, pp. 1-286. ZoBell, C. E., 1953. The occurrence of bacteria in the deep sea and their significance for animal life. lUBS Deep Sea Colloquium, XIV Int. Zool. Congr., Copenhagen. ZoBell, C. E. and Morita, R. Y., 1956. Bacteria in the deep sea. From The Galathea Deep Sea Expedition 1950-1952 (London). art} v/ Snuov ".'r[ • III aTAJ4 ^£ s;c) .u ,; arjasQ' 8s no anos b/uW JaB^ ,riy, PLATE III Stomach of a small 72 foot male blue whale, cut open to show the enormous quantity of krill that can be swallowed, even by a young animal. The handle of the flensing knife is 5 ft. long, and there may be between i and 2 tons of krill present. The two fish, swallowed fortuitously, each about 1 2 in. long, are evidently Notolepis coatsi, itself a prodigious krill-eater. This whale was killed in the southern, krill- rich, East Wind zone on 28 December 1947, in 62° 39' S, 93° 01' E. {Photo: Robert Clarke) DISCOVERY REPORTS, VOL. XXXII PLATE III