UNITED STATES DEPARTMENT OF THE INTERIOR FISH AND WILDLIFE SERVICE FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE VOLUME 57 BULLETINS 107 TO 125 ISSUED BY THE FISH AND WILDLIFE SERVICE 1956-58 UNITED STATES DEPARTMENT OF THE INTERIOR FISH AND WILDLIFE SERVICE Bureau of Commercial Fisheries TITLE PAGE AND INDEX VOLUME 57 ISSUED 1962 U.S. GOVERNMENT PRINTING OFFICE • WASHINGTON, D.C. • 1962 CONTENTS OF VOLUME 57 Bulletin No. P;i(;^' 107. V.vLiDiTV OF .\r,K detp:kmi.\'ation from sc.\les, .\.\d growth ofm.vrkedL.\keMichig.\n L.vKE TRoiT. Bv Louellii E. C'uble. (Issued October 1956.) l-.')!t 108. (_''OMP.\R.VTIVE STUDY OF FOOD OF BIGEYE AND YELLOWFIN TUNA IN THE CENTRAL PaCIFK'. B3" Joseph E. King and Isaac I. Ikehara. (Issued October 1956.) 61-y the fishermen. Initially, the officers removed the fin scar from each fish (in some cases, also a scale sample) and sent them to the Institute for Fisheries Research of the Michi- gan Department of Conservation in Ann .\rl)or for payment of the reward. Later, however, most of the fish weic shi])ped iced, either in the round or (hcssed," to the Institute where the scale sam- ples were taken, measurements recorded, and the deformed or missing fin described in some detail. Sex was not recorded. Vp to July 22, 1952, 1,603 fish had been sent to the Institute for Fisheries Research. Of this number, 96 could not be identified with any one of the three plantings or lacked essential records: ?'. e., record of the missing fin was lacking, the fin or combination of fins rcporteil missing or ab- normal had not been used in the experiment, or fins were reported by the State observer as normal in every resptH't, length measurement was not re- coi-ded, or scale sample was not taken. For the 1 ,.507 fish that, on the basis of fin records alone, could have been marked lake trout, the annual recoveries were as given in table 2. Al- though this group includes individuals with "natinally (h'formed" fins (malformations not resulting from earlier clipping), tlie data of table 2 give a rougii estimate of th(> percentage return from the several [jhnitings. Because it is (loui)tful that the recoveries from area 8 were fish with l)()na fi(h" markings, the percentage of returns arc shown for aieas 1-6 only. Recoveries from the 1945 planting exceech'd those from the 1946 planting almost 4:1, aiul c.Nreeded recovei-ics from tile 1944 " tvills and \ iseera n'lnoveil. planting 11:1. but the I'ccoveries of marked lake trout from all plantings were in exceedingly small percentages of the numbers of fish released. About 0.67 percent of the marked lake trout re- leased in 1945 but only O.Oti percent of the 1944 planting and O.IS percent of those planted in 1946 were recovered. The low |)ercentages of return and abrupt termination of ca|)tiires probably were due to the rapid reduction of the population by the sea lamprey. Xo explanation can be ottered for the higlu'i- percentage of return irom the 1945 than from the 1944 and 1946 plantings. A large majority of the recoveries of marked lake trout in northern Lake Michigan (areas 1-6) were made in the fourth year after planting. The fish had evidently reached a sufficiently large size at that age to be most easily caught in the nets employed in the fishery at the time. The localities and relative numbers of recov- eries are shown in sectional maps of Lake Michigan (figs. 1 and 2). Bouiithiries of tliese sections are superimposed on a map of tiic entire lake (fig. 3) to indicate their position with reference to the houiuhiries of the statistical areas or districts 1-8 regularly employed in analyses of commercial fishery statistics for the State of Michigan waters of Lake Michigan (Van Oosten, Hile. and .lobes 1946; Hile, Eschmeyer, and Lunger 1951). The largest catclies of niaiked lake trout were made out of Manistitiue, Mich., in area 2, and in the vicinity of the islands of areti 3, with the great est concentrati'in about lieaver Ishiiid and tiie shoals to the east of tiiis island. A few s[)ecinieiis were caught in each of areas 1, 5, and ti; 2 trout, taken just across tlie line in tiie northern i)art of area 7 by fisliernieii from I'entwater, are included witii tliose caught in area 6. .\o recoveries were made between Little Saliie Point in the nortltern FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE NAUBINWAY Scale of Miles TRAVERSE CITY Figure 1. — Northern Lake Michigan showing points of release and capture of marked hike trout. Phmting kjcations designated as follows: 1944, square enclosing an X; 1945, circle enclosing a +; 1946, triangle enclosing dot. Re- coveries from the three plantings are indicated as follows: 1944, squares; 1945, circles; 1946, triangles. The sizes of the symbols indicate numbers of fish recaptured at the various points, the smallest symbol of each year class is for 1-4 fish thro\igh the largest for more than 49 fish. part of area 7 and tlie vicinity of South Haven (area 8), more tlian 60 miles distant, where 102 lake trout with deformed or missing; fins were taken. Lake trout witli ahnormal fins, captured on the Wisconsin side of tlie lake, are not shown on the map. Most of tlie 142 fish taken were caught north of Algoma; a few, 1 or 2 off each port, were taken off Two Rivers, Cedar Grove, Milwaukee, and Racine. The records on tliese fisii are not sufficiently detailed for profitable study. Rather than iTJect individual fish arl)itrarily all samjjles, pioporly documented and iiaving "possible" fin markings, were accepted for study. The large size of certain lake trout whose missing fins in<- tions l)ut a summary of tlie basis for the rejection of the samj)les from area S is given at tliis |)()int. GEOGRAPHICAL DISTRIBUTION OF RECOVERIES In areas l-(>, tiie earliest reeoveries were made near the loeality of planting. As the fish grew oldei- and larger the captiiies were more widely disti-ibuted. They scattered to some extent in all directions, hut the principal movement was in a northwest I'lly direction toward Maiustiqiie and thence westerly and southwesterly until some fish were recaptured along tlie Wisconsin shore. Captures of lake trout with deformed fins were fewer and the distribution was discontinuous southward from the localities in which the plant- ings were made. No recoveries at all were made between the extreme northern part of area 7 and the neighborhood of .South Haven.'" If it is as- sumetl that lake trout reported off South Haven were actually marked fish, it is difficult to under- stand why none were caught in the heavily fished 60-mile-long area en route to the more southerly waters. On the other hand, if the lake trout reported from area 8 are considered to be wild- stock lake trout with abnormal fins, the trouble- some question arises as to why no trout of the same category were reported from that 60-mile stretch." The discontinuity of distribution of the recoveries does not provide convincing evidence, but does, nevertheless, give cause to regard with suspicion the genuineness of the mark (deformed fin) on the fish caught at South Haven. FINS ON RECOVERED LAKE TROUT Records of degree of regeneration of the pectoral fins '- in terms of regenerated rays (table 3) and lengths of the abnormal fins (table 4) on recovered lake trout were similar in that they suggested no basis for the separation of marked "> \'an Oosti'ii (19511) dcsciibcd the distribution of rccovprios of these sanu' fish tlirougli 1949. Subsequent captures did not chanKe the general situation greatly, except that the progressive scattering of the growing fish continued. " The answer possibly may lie in the enterprise of a single fi.sherman. Of the 102 recaptures from southern Lake Michigan, 94 were turned in by the same operator. Conceivably fishermen in the waters to the north observed similar al)normalities hut did not believe them to be thi' result of fin-clipping. " The collection of fish with dorsal and adipose fins dijjped is too small to give reliable results, but 43 (75.4 percent I of a total o( 57 specimens were jurtgi'd to have true marks. .lust one lake trout with this mark was caught in area 8. The mark (dorsal and adipose fins removed) proved somewhat confusing becau.se of the presence of fish w itll oni' fin deformi'd anil the other normal. 37832G O — 50 2 lake trout of areas 1-6 from naturally propa- gated individuals of this region, but did indicate rather conclusively that the samples from areas 1-6 and area 8 could not have been drawn from the same population. Despite certain disagree- ments as to detail between data on the right and left pectoral fins of trout frotn areas 1-6 (dis- crepancies which could have been the result of the small number of fisii recaptured with a deformed left pectoral fin), the general situation can be de- scribed satisfactorily from the combined records of the two fins. The extent of regneration of fins on lake trout from areas 1-6 was relatively small. In a total of 1,348 individuals, 57.5 per- cent had no regeneration of the fin rays, and 77.5 percent had fewer than 5 rays regenerated. With respect to length of regeneration, 58.2 percent of the fins were without regeneration, and 75.2 percent were not more than K normal length. In area 8, to the contrary, regeneration of most fins was advanced. Of 74 fish, for which there were records of the number of rays in the de- formed fin, but 1.4 percent had no rays regenerated, and only 4.1 percent had fewer tlian 5 rays re- generated as compared with 77.5 percent in areas 1-6. Of 89 fish, for which the length of the fins was recorded, just 1.1 percent of the fins were without regeneration, and only 13.5 percent were not more than ji normal length as compared with 75.2 percent in areas 1-6. The very small percentage (1.1) of fins showing no regeneration in area 8 is strikingly difl'erent from that (58.2) of fins on fish from areas 1-6. The data of tables 3 and 4 have a usefulness in addition to that of demonstrating that samples from areas 1-6 and area 8 were drawn from stocks that were dissimilar with respect to the character- istics of abnormal fins. If the thesis is accepted that most or all of the lake trout from area 8 were unmarked, it can be anticipated that most of the unmarked lake trout in the samples from areas 1-6 also will be among the fish whose fins exhibit more advanced regeneration. DISCREPANCIES BETWEEN AGES READ FROM SCALES AND INDICATED BY ABNORMAL FINS Agreement between ages indicated by fins and read from scales was high (substantially above 90 percent) in fish from areas 1-6, but in area 8 only 39.2 percent of the scale readings agreed with the ages indicat(>d bv abnormal fins. p]ven 8 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 3. — Extent of regeneration of the pectoral fins,, expressed as number of rays, on lake trout marked in 1945 and 1946 Locality of recovery and mark; year of planting offish Number of rays regenerated >8 Unknown Areas 1-6: Eight pectoral (1945) Percenta?e ' -_ Left pectoral (1946) Percentage Ritiht and left pectorals Percentage Area 8: Right pectoral (1945).,.. Percentage Left iiectora! (1946) Percentage Right and left pectorals. Percentage 271 i.348 679 65.0 96 36.8 774 57.5 91 8.7 29 11.2 120 9.2 46 4.5 9 3.5 55 4.2 43 "69" "in2 1 2.4 1 1.4 31 3.0 10 3.9 41 3.1 1 3.0 2.4 2 2.7 36 3.4 10 3.9 46 3.5 48 4.6 13 5.0 61 4.7 3 9.1 3 4.1 30 2.9 17 6.6 47 3.6 1 3.0 2 4.9 3 4.1 25 2.4 14 5.4 39 3.0 1 3.0 1 1.4 17 1.6 28 10.9 45 3.5 2 6.1 2 4.9 4 5.4 41 3.9 33 12.8 74 5.7 26 7.5.8 35 85.4 60 81.1 I Fish with unknown number of fin rays not included in percentages. Table 4, — Extent of regeneration of the perioral fins, expressed {for most fish) as a fraction of the normal length of the fin, on lake trout marked in 194-5 and 1946 Locality of reco\ery and mark; year of planting Number of fish Extent of regeneration No regener- ation Less than t/ij-inch long normal length normal length normal length H W normal normal length length 45 42 4.2 3.9 27 17 10.2 6.4 72 59 5.4 4.4 13 6 34.2 16.8 10 14 19.6 27.6 23 20 25.9 22.5 Full normal length No record of length Areas 1-6; Right pectoral (194.6) Percentage ' Left pectoral (1946) Percentage Right and left pectorals. Percentage Area S: Right pectoral (1946)..-. Percentage Loft i)ectoral (1946) Percentage Right and left pectorals. Percentage 1.077 27! V.iis 679 63.8 95 36. 0 774 58.2 43 "59' 162" 1 2.0 124 11.6 48 18.2 172 12.9 1 2.6 2 3.9 3 3.4 18 1.7 2,7 26 1.9 1 2.6 4 7.8 6 5.6 21 2.0 8 3.0 29 2.2 2 5.3 1 2.0 3 3.4 128 12.0 60 22.7 188 14. 1 15 39.5 IS 35.3 33 37.1 0.8 2 (1.8 10 U. S 1 2.0 1 1.0 1 Fish with fins of unknown length not included in percentages. though this percentage was somowliat higher than would be expected from an assumption of complete iiulcpeiulcnce of age shown by abnormal fins and by scale markings, it does indicate that if the sample from area 8 contained any authentic marked lake trout, their number was extremely small. GROWTH AS INDICATED BY ABNORMAL FINS Presentation here of details on length frequen- cies and average sizes of various age groups of the different year classes as established by abnormali- ties of the fins and by the examination of the scales would be little to the point as the situation is described adequately by the data of table 5 which shows the mean lengths and ranges of length for the several age groups (year classes combined) as indicated by fins. If these lake trout are taken as bona fide fin-clipped fish, we must accept also tlie conclusion that the trout were largest in the first and second years of life (average lengths of age-groups I and II, 23.8 and 17.1 inches, respectively), were smaller, and, for the most part, without growth in later years (range of 12.5 to 12.7 inches for average lengtlis of age-groups III-VI, and only 13.7 inches for age-group VII)." Despite the consiilerable range of length for each age group of lake trout of known Tablk .5. — Average lengths and ranges in length iif age groups as indicated by the occurrence of abnormal fins '{assumed to be true marks) of lake trout from southern Lake Michigan [See text discussion of the probability that few or none of these fisti c luld have come from the various fin-elipping e.vperiments] Number offish Total length (inches) Average Range I - 4 13 28 22 19 13 3 23.8 17.1 12.7 12.7 12.5 12.6 i:i. 7 22. .5-24. 0 n --- 10 4-24. 0 HI - 7.4-21.0 IV 10. 4-2:!. 0 V 10. 4-1.6. 2 VI 10.0-16.6 VII - - 13. 2-14. 5 '" According to Smith and Van Oosten (1940) lake trout tagged at Port Washington, Wis., that averaged 12.8 inches long at tagging were 19.8 indies long about 2 years later. AGE DETERMINATION FROM SCALES OF LAKE TROTIT 9 age that will be demonstrated later, some of the ranges in table 5 cannot be considered reasonable. These lines of evidence, even though they do not exclude the possibility of the presence of a few marked lake trout in the samples from area 8, demonstrate conclusively that the great majority were unmarked wild stock, and that the occurrence of abnormal fins among these fish was not related to the age of the fish. The sample is, therefore, considered unsuitable for use in the present study. Samples from areas 1-6 undoubtedly also include some unmarked fish with abnormal fins; and con- vincing evidence of their presence will be offered. There is no reason to believe they were sufficiently numerous there to harm seriously the materials for the purposes of this investigation. VALIDITY OF AGE DETERMINATIONS FROM SCALES The study of the scales of lake trout, presum- ably of known age, ofTered the rather perple.xing problem of using the same materials for two pur- poses which, in a sense, are mutually exclusive. It was, of course, imperative to examine carefully the scale characteristics of a large series of fish of known age to establish, as exactly as possible, criteria for the determination of age. It was equally necessary to use the same fish as the basis for an objective estimate of the degree of accuracy to be expected in the reading of the scales of lake trout for which the ages are not known. With a small series of fish, accomplishment of both purposes would be impossible, for the investi- gator would become so well acquainted with the scales of individual specimens as to remember their characteristics, especially their unusual features, and hence would be unable to make ob- jective age determinations. In the present large series of 1,405 fish from northern Lake Michigan (areas 1-6), however, memory of scales of indi- vidual fish probably had no biasing effect on the accuracy of successive readings. Even so, pre- cautions were taken to keep the tests objective. A brief statement of the general procedure follows. In a preliminary examination, designed to estab- lish whether or not the scales of lake trout bear markings that can be interpreted as annuli corre- sponding in number to the supposed age of the fish (as indicated by a deformed or missing fin), the scales of several hundred lake trout were read objectively. They were studied for the occurrence of repetitive irregularities in the sculptured pattern without reference to any information about the fish except the date of its capture. When such markings were found, readings and measurements made from them were compared with tlie full data on the individual fish. Another important aspect of the first series of examinations was the establishment of the time of annulus form- ation and the progress of the season's growth, without knowedge of which it is difficult to make accurate readings from scales of fish caught over much of the growing season. After the characteristics of the annulus and the time of annulus formation were well established, the entire series of scales was read twice. During both readings the only information available was date of capture, and eacli second reading was made without knowledge of the age assigned at the first. After completion of the two readings, a careful study was made of the scales of all lake trout for which the ages assigned were not the same at the first and second examination and a best estimate of the correct age was made. EARLY GROWTH OF SCALES The scales of lake trout are cycloid, oval to egg- shaped. Concentric ridges or circuli, arranged about a focus, roughen the outer surface of the scale. The focus may be central or slightly anterior or posterior to the center of the scale (see figs. 8 and 11). Neither radii nor transverse grooves are present. The inner surface of the scale lacks circuli but is not utterly smooth and characterless. Annuli sometimes are clearly visi- ble on this side. The scales are so small, thin, and deeply embedded in the skin as to be relatively inconspicuous. They are dislodged with such difficulty that few are regenerated. Variation in the number of scales, in series along the lateral line, is large, from 180 to more than 200. Squa- mation of the body is complete. Only the head, which is well supplied with mucus pores, and fins, are unsealed. The size of the scale varies greatly from one location on the fish to another. In general, the larger scales are on the posterolateral surfaces of the body and the smaller scales about the fin bases 10 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE and Oil the aiiteroveiitral and aiiterodorsal surfaces. Samples for study were taken from a mid-point on the body, below the anterior part of the dorsal fin, immediately above the lateral line, and thus did not include either the smallest or the largest scales of the individual fish. However, scales from rather limited areas show considerable variation in size and shape. Scales chosen from the sample for study were tliose which seemed most repre- sentative of the larger symmetrical scales. The scales of lake trout appear during the sum- mer of the first year of development. However, neither the age nor the length of the fish at the time of scale formation lias been determined definitely for the lake trout in I>ake Michigan. Both salt-water and fresh-water fishes that have been studied develop platelets, the beginnings of scales, when the yoinig fish are from about 18 to 50 millimeters in total length (Fish 1932; Hilde- brand and Cable, 1930, 1934, and 1938; Cooper 1951 ; Brown and Bailey, 1952; and others). Fisli (1932) described a lake trout larva 21.5 millimeters long from Cape Vincent Hatchery, but did not mention the development of scales. A yolk sac was still present at this size and the appearance of scales would scarcely be expected before ab- sorption of the yolk. In 1953, young lake trout 26 to 56 millimeters long, were taken in Lake Superior in the middle of June and the middle of August by the Fish and Wildlife Service research vessel ( 'isco. The largest of those caught in August was 56 millimeters or 2?i6 inches long. It had a band of scale pockets containing platelets along the entire length of the lateral line. This band consisted of several rows of platelets on either side of the lateral line. The sizes of the platelets were graduated; the larger ones were adjacent to the lateral line; the others became smaller and farther apart with each suc- cessive row. Only in tiie lateral line did the scale structures take alizarin stain readily. These structures were concave ovoids, two in each pocket, one dorsal to and the other ventral to the lateral-line organ, forming partial sidewalls to it. The platelets, situated in dermal pockets, were protected from immediate <'ontact witli tiie alizarin. Consequently, the scale pockets stood out as clear areas after staining. The largest scale platelets, when teased out of the pockets, measured about 0.2 millimeter long. Some were clear and smooth; the first circulus was formed on others. Although some fish such as biook trout form scales first along the posterior ])art of the lateral line (Cooper 1951), a .young lake trout 53 millimeters long had platelets scattered in one or two interrupted rows and in small groups here and tiiere along the anterior end only of the lateral line. The lateral line itself was not in evidence posteriorly. The largest i)latelets on this lake trout were about 0.1 millimeter long and lacked circuli. Probably scales begin to form on lake trout in Lake Superior when the fish are about 50 millimeters long but no histological sections were made to determine this jxiint. It is not known whether young lake trout growing in Lake Michigan develop scales at the same size as those in Lake Superior. One hundred fingerlings, all of the same age but ranging in length from 35 to 85 millimeters, which were reared in the fish hatchery at Charlevoix, Mich., in 1948 and preserved on Septenil)er 17, were examined. The smallest of these lake trout having scales was 47.5 mm. long. This fish liad scales with as many as 4 circuli the full length of tiu' lateral fine. Other specimens 35 to 43 mm. long were without scales and no evidence of a lateral line was seen. Although these young lake trout grew under artificial conditions, development of the scales began at about the same l)0,\ll scah'.s weri' studied ut the sinii' niagnilication (.Xs:t..1). Ulustra- tions (,f I hi' scales have been reduced Xfifi.S. Sec p. 59 fur siRnificanee o( th.' cheek latuled ■'O." of scales). Here, the annulus often is seen dis- tinctly as a ridge on the scale or as a groove on the impression. The groove is well illustrated by the second and third annuli in figure 10, and the third and fourth annuli in figure 9. Another characteristic pattern in the posterior field results at points where circuli of the preceding growing season end and the first circulus of a new season crosses their paths at angles that bring the pattern to a crude V in which the angle of the V points toward the annulus. These V's are in evidence somewhere on nearly every scale, but on the scale shown in figure 7, it is doul)tful whether the fourth annulus would have been located but for the V on the lefthand side, as the annulus is indistinct elsewhere around the scale. The V's are also clearly represented in figure 5 bv tlie second and third annuli, and in figure 8 bv the first, second, and third annuli. Frequently, part of the posterior area of the scale is almost devoid of .sculpturing. Only ragged bits of crooked, discontinuous circuli are scattered about, but even then, circuli extend farther out into this part of the scale at the annulus than between annuli, pointing it out like a crooked finger. In the anterior and lateral fields, three charac- teristics of the pattern of circuli, usually occurring in combination, indicate the location of the annulus. First is the narrowing of the spacing between circuli at the end of a growing season, mentioned earlier and seen in most figures. Usually, in addition, there is a broken circulus here or there along the annulus with another circulus crossing the ends in a "cutting-over" pattern (as in the V formations of the posterior field). The longer circulus which does the cutting-over is the first circulus of the new growth. It is often continuous through the anterior field from the posterior field on one side to the posterior field on the other side of the scale, and may cross or extend partly across the posterior field itself, as shown by the first annulus in figure 12, and by all annuli in figure 11. The third characteristic pattei-n results from the apj)earance of one oi- two very fine, broken lines '* at the annulus. This feature is illustrated by the scale shown in figure 10. Note especially the second and third annuli. The scales shown in figures 5 to 12, also 15A and 16A, are from fish representative of lake trout '• Thes*' do not appear to t'e true circuli. 12 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE FiGt'RE 4. — A scale from a 4-year-olcl lake trout 15.3 inches long, marked in September 1945 and recovered May 7 or 9, 1949, showing the degree to which annuli mark the inner surface of the scales. The outer surface of another scale from the same fish is seen in figure 8. The photograph is a negative of an impression in plastic. AGE DETERMINATION' FROM SCALES OF LAKE TROUT 13 Fkurk f). — Scale (if a lake trout marked in Seiiteiniier I'.Mo and recovered June 11. lOlil. A negative photograph of an impression in plastic. 14 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE FifitiRE G. — Scale of a lake trout 14.8 inches Ions, marked in Seijteniber 1945 and recovered July 8, IIU'.I. The 0- mark and first annulus appear to occur together. A narrow band of new growth is present. presumably of known age. Scales of lake trout whoso age, as rea () — 56 3 16 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Figure 8.— Scale of a lake trout marked in September 1945 and recovered May 7 or 9, 1949. Note that focus of the scale is located posterior to the center. No new growth. AGE DETERMINATION FROM SCALES OF LAKE TROUT 17 FiciRE 9. — Scale of a lake trout 13.5 inches long, marked in September 1945 and recovered August 13 or 16, 1949. Note that the band of new growth is wider than in figure 7 even though this fish was caught 2 weeks earlier. 18 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE NEW C, , GROWTH-^j NEW GROWTH NEW J^^ GROWTH Figure 10. — Scale of a lake trout 15.9 inches long, marked in September 1945 and recovered April 17, 1950. The 0- mark is more conspicuous than the first annulus on this scale. The band of new growth is narrow, but wider laterally than terminally. AGE DETERMIXATIOX FROM SCALES OF LAKE TROUT 19 Fif.iRE 11. — Scale of a lake trout 15.1 inches long, marked in September 1945 and recovered April 25, 1950. The focus of this scale is located anteriorly. The annuli are indistinct. Such a scale is difficult to read. The band of new growth is narrow. TIME OF ANNULUS FORMATION New gi'owth on lake trout scales is first seen as a narrow, clear band outside a darker band of the closely spaced circuli of "winter growth." In the early part of the season, new growth is too narrow to be distinguished from spacing between winter circtili. For this reason, new growth was i(h'iitified and measured only when it had attained a width greater than that of the spacing between |)rece(Hng circuli and an outer circulus had formed at least part way around the scale. Hence, in this study, the scales had grown an undetermined, though short, time before growth was recorded. One lake trout had some new growth on its scales -laiuiary 19, but no others appeared with new growth until the latter part of .\Iarch. Similarly, a single specimen without new growth was caught September 23, more than a month after new growth was started on the scales of all other fish in the sample. The two aberrant specimens are 20 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE FiniiRE 12. — Scale of a lake trout 16.8 inches long, marked in Septenihor 1945 and recovered April 25, 1950. The 0-mark is more conspicuous than the first annulu.s. The band of new growth is wider than is usually found on scales of the fish caught in April. AGE DETERMINATIOX FROM SCALES OF LAKE TROUT 21 2 3 4 5 6 7 GROWTH Fici-RE 13. — Scale of a lake trout 26.2 inches long with left pectoral fin abnormal allhouKh the type of abnormality was not described. Caught June 13, 1947. If thi.s lake trout had been marked by the removal of the left pectoral fin, it would have been 1 year old, but it is too large for that age, and 8 checks were on the scales. (See text for dis- cussion of the central check). Xew growth was not uniform in width. On this section, new growth appears only in the lower right area. The deformed fin was an abnormality. 22 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 987654 32 1 Figure 14. — Scale of a lake trout 31.8 inches long with left pectoral fin missing; no regeneration. Caught September 10, 1947. Lake trout with left pectoral fin removed were released in September 1946. If marked, this fish should have been 1 year old. Because of its large size, it probably was of more advanced age. Ten checks were read on the scale. (See p. 59 for a discussion of the central check). The band of new growth is wide. The missing fin was an abnormalitv. listed in table 6 because a fin of each appeared to have been clipped and the ai\nuli on the scales seemed well defined. However, the dates on which new growth on the scales was begun are sulli- ciently unusual to throw some doubt on the authen- ticity of the fin-clip and the accuracy of the age determination from the scales. The percentage of lake trout witli new growth on tiieir scales increased slowly through April and May, but rose rapidly through Jinie and Jidy, passed the 50-percent level during the last week of June, and reached the 100-percent level the last half of August (table 6; fig. 17). Al- though the season's growth was detectable on the AGE DETERMINATION FROM SCALES OF LAKE TROUT 23 Fkure 15. — (A) Scale of a lake trout 8.6 inches long, marked in September 1945 and recovered October 8, 1947. The band of new growth is wider than the entire growth zone of the previous year. (B) Scale of a lake trout 10.8 inches long with a right pectoral fin missing; no regeneration. Caught November 4, 1948. Only 3 checks were found on the scales. As lake trout with the right pectoral fin removed were released in September 1945, the scales shovild have had 4 checks, 3 antnili, and 0-mark, if the fish were one of those marked. The missing fin was, therefore, abnormal. scales of some lake trout by the latter part of Marcli, it could not be seen on otliers initil August. The period for the start of new growth, therefore, extends tlu'ough .5 months. Possibly, tiie period wotdd he shorter for groups of fish, all caiigiit from a small, localized area. The present collec- tion of marked lake trout came from contiguous but relatively extensive areas in the nortlieastern part of Lake Michigan. A diversity of environ- mental conditions in various localities, about which there is at ijresent very little information, may cause growth on the scales of local groups of :<78:):if! o— ,^i(; —4 lake trout to begin at different times so that when the groups are combiiu-d, as in the present study, tlie semblance of a long period for tiie begimiing of growth would result. Assuming normal distribution, the combinetl data fit, within the confidence limits at the 5- percent level of probability, a normal cimaulative curve witii tlie (7 = 20 days and the .'iO-percent level on .Time 18. The test used for goodness of fit was tiie Kolmogorov-Smirnov test described by Massey (1951). Whereas tlie .")0-percent level of the theoretical normal po|)uhuion falls on 24 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Figure 16. — (A) Scale of a lake trout 16.5 inches long, marked in September 1945 and recovered October 5, 1948. Note partial check between the second and third annuli. There was no evidence of a check on the righthand side of the scale. (B) Scale of a lake trout 13.5 inches long with left pectoral fin consisting of 9 twisted rays one-half normal length. Caught December 18, 1948. Four checks appear on the scales. As lake trout with left pectoral fin removed were released in September 1946, the scales should have had 3 checks, 2 annuli, and 0-mark, if the fish were one of those marked. The fin was, therefore, deformed. AGE detp:rmixation from scales of lake trout 25 Table 6. — Progress of anniilus formalion on the marked take trout [Based on rocovcrii's for the calendar years 1947-51 and age groups II-VI. No consistent differences could be detected among age groups In collections from dilTerent years] Date Number without newgrowth Number with new growth Percentage with new growth Date Number without now growth Number with new growth Percentage with new growth Jan 1-15 29 23 g 13 1 20 25 «7 140 55 71 66 0 1 1 0 0 0 4 3 17 39 24 31 56 0.0 4.2 0.0 0.0 0.0 16.7 10 7 20.2 21.8 30.4 30.4 45.9 July 1-15 48 2 0 0 11 0 0 0 0 104 96 151 fiO 62 47 17 3 3 9 68.4 16-31 93.2 Feb I 15 Aug. 1-15 98.7 16-31 100.0 Mar. 1-15 Sept. 1-15 - 100.0 lfi-31 16-30 97.9 Oct. 1-15. 100.0 10-30 16-31 100. 0 May 1-15 Nov. 1-15 - 100.0 1()-31 16-30 100.0 Dec. 1-15 16-30 16-31 0 16 100.0 I Sec page 19 for comments on these specimens. 100 t/1 UJ _J < u (/I z o I \- $ o a. o LU Z I t- I (/) Ll ti. o LU z lU o a. LU a 6 23 8 22 8 23 8 23 8 23 8 22 8 23 8 23 8 23 JAN FEB MAR APR MAY JUNE JULY AUG SEPT Figure 17. — Perci'iitanc of marked lake trout curve drawn by inspection. (lowing new growth on their scales. Empirical data indicated by dots 26 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE June 18, the date on which 50 percent of the marked fish had started new growth on the scales was June 26. Within the 5-percent confidence limits for samples of the same size, new growth on the scales of lake trout in other years would be expected to reach tlie 50-percent level during the last 3 weeks of June. New growth may be identified, then, on the scales of individual lake trout in northern Lake Michigan any time be- tween the middle of March and the middle of August, and about 50-percent of the lake trout will show new growth on their scales by the latter part of June. Because of the long time interval in which new growth may begin, the numbers of lake trout with narrow spacing between the circuli at the margin of their scales diminish gradually from January through August and the numbers with wide spac- ing between these circuli increase correspondingly. In July and August some scales, that began growth early in the season, already had a wide band of new growth with narrowing spacing between the circuli near the edge of the scales. The age of unmarked fisli would be difficult to interpret from such scales. Whether the band of growth liad been formed during the current or the previous season would be a matter of the reader's judgment. On most scales from fisli caught at this season, the growth of the current season was narrower than the growth of tlie previous year, but tiiere were exceptions which gave difficulty. The end of the growing season for the scales of lake trout could not be determined definitely from the scales themselves. As new growth on the scales of individual fish in the sample began at different times during the spring and summer, they may also have completed growth at different times. In summer and early fall, scales having wide bands of marginal growth with narrowing spacing between the outer circuli had the appear- ance of completed growtli, but it is not known that additional circuli do not form later in the season. It remains uncertain, therefore, whether the scales of lake trout attain the full growth of a season shortly after the begiiming of growtli or continue to increase in size, however slowly, until time for the next anindus to form. SUPERNUMERARY OR 0-MARK During the first examination of the scales, it was a surprise to discover tliat tlie number of annulus- •like markings observed was almost invariably greater, by one, than the number of years of age indicated by the clipped fin. Upon further inves- tigation, the reason for the discrepancy was found in the interpretation of the mark nearest the focus. Comparisons of lengths at capture of lake trout of a known age group (age-groups II to V) with calculated lengths for the same year of life showed the outermost markings to be annuli. Although no lake trout of age-group I were captured, it is logically to be expected that on their scales, also, the outermost mark would be an annulus, hence that the central check is supernumerary. This check or mark appears to have been formed dur- ing the fall of tlie fish's first year when they were only slightly larger than at the time of planting. The innermost marking on the scales, referred to hereafter as the 0-mark, is interpreted to be a line of demarcation between an initial slow rate of growtli and a later sudden increase in the rate as indicated by a change in spacing of the circuli at this point. The circuli within the central mark are more broken and more closely spaced tlian cir- culi laid down later (figs. 5 and 12). The mark is usually fainter tiian the annular rings on the scales and is not present on the scales of all specimens.'^ Rarely, scales show the central marking so closely approximated to the first an- nulus (figs. 7 and 11) as to suggest that on other scales it might coincide with the annulus and thus be lacking altogether as on the scale in figure 6; a few have it very close to the focus, but for most specimens the inner mark is a little over halfway from the focus to the first annulus. Although this mark is typically indistinct (figs. 5 and 9), it some- times is the most conspicuous mark on the scales (figs. 10 and 12). Such outstanding marks might easily be taken to be first annuli on fish of un- known age unless the reader were expecting to find, and looking for, a mark within the true first aiuiulus. The 0-mark can only be surmised, at this time, to record some drastic change in the young fish's enivronment or habits of life. A possible explana- tion is that the check results from handling (an- aesthetization, removal of fin, transportation) at the time of planting and the change from hatchery to lake environment. In support of this view is " A separatp inner marking was not found on the .scales o( 4 (0.3 iwrcent) of the marked specimens and it is believed the inner mark on these scales coincided with llle first annulus. AGE DETERMIXATION FROM SCALES OF LAKE TROUT 27 the fairly close agreement between the average calculated length of 3.7 inches (range 1.5-5.9) at time of formation of the 0-mark (computed from scale measurements of recovered marked fish) and the average measured length (3.2 inches; range 2.1-4.3) of samples of fingcrlings at time of release into the lake. On the other hand, the examination of scales of lake trout that almost surely were not marked fish (lake trout from northern Lake Michigan that were unreasonably large for the ages indicated by their deformed fins and fish from area 8 that included few, if any, marked fish) suggested strongly that naturally hatched lake trout in Lake Michigan also form a 0-mark. Such a mark could arise, for example, from a change in environmental conditions, a change of diet, or a shift by the fish to different grounds upon attainment of a particu- lar length (about 3.7 inches in the northern part of the lake). The scales of lake trout for which there was disagreement between the age, as indicated by scales and fin, consistently exhibited a first check that resembled in every way the 0-mark on the scales of marked specimens. The scales in figures 13 and 14 were from fish turned in as recoveries of marked lake trout but they were unquestionably from fishes of natural origin. Fish marked in 1946, averaging 3.2 inches long, could not have attained lengths of 26.2 and 31.8 inches before they were caught in 1947. Actually, the scales showed 8 and 10 checks, respectively. The central checks resemble closely the 0-marks of the scales from bona fide recoveries. This is brought out forcefully by figures 15 and 16 in which the lefthand scales are presumably from bona fide recoveries (age read from the scales and age indicated by the deformed fin in agreement); and the righthand scales are probably from natur- ally propagated fish (ages from scales in disagree- ment with age indicated by fin). It is readily apparent that the structure and size of the central areas of these scales are similar. That the central check on the scales of wild- stock lake trout was in fact a 0-mark and not the first annulus was strongly supported by the good agreement between the average calculated lengths of the naturally propagated fish and the marked, liatchery-reared fish at each of the first three checks on the scales. A few lake trout captured by large-mesh nets in northern Ijake Michigan during 1947 could be identified, without question, as wild stock because they were too large to have belonged to any group of marked fish. The calculated lengths of these fish at all three first checks were greater than for the marked lake trout caught in all nets over a period of years, 1947-51 (columns 2 and 4, table 7). The differences were no larger, however, than would be expected from the small number of fish in the sample and from the powerful selective influences that bore on the older age groups of the more recent year classes in the collections. The calculated lengths of the wild stock caught in nets of all mesh sizes '* differed little from tlie marked fish caught in similar nets (columns 2 and 6, table 7). Calcu- lated lengths of wild-stock lake trout caught by all nets in the southern part of the lake were 0.8-1.0 inch shorter at each of the first 3 checks tlian those of wild stock caught in more northern waters (columns 6, 8, table 7). This large dif- ference between calculated lengths of lake trout from the 2 sections of the lake is indicative of the racial separation of the 2 pojjulations. '* This group of lake trout incluiles. in addition to those positively Identified as wild stock, other lake trout for wtiich the age read from the scales differed from that indicated by the deformeil fin. Evidence is i)resented later to show tliat most, if not all, of these fisli were also wild stock. Table 7. — Calmlaled total lengths (inches) and increments of growth in length of marked lake trout recaptured in northern Lake Michigan and of naturally propagated fish from northern and southern Lake Michigan year classes combined Unmarked lake trout Northern areas 1-6, from all nets Northern -areas 1-6 Southern-area 8 Check or annulus From large-mesh nets From all nets From all nets Length Increment Length Increment length Increment Length Increment 0 3.7 5.9 8.7 4.0 6.9 10.0 3.5 6.6 8.3 2.7 4.8 7.3 1 2.2 2.8 2.9 3.1 2.i 2.7 2.1 2 2.5 1,318 M6 99 102 .\ge groups in sample n- VI III -IX 11- IX II-'' k'lll ' The marked lake trout averaged 3.2 inches long at time of planting. ' This total includes 9 fish obviously too large for their supposed age (see p. 30) and also 7 that could not be assigned to a particular planting (more thaii one fin deformed or the deformed fin not one used as a mark), but wtiicti were too large to have been from any of the three plantings. .\11 fish weri' caught In 194i . 28 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Discrepancies between increments of growth were also small. At the second check the differ- ence was 0.1 inch between marked and unmarked fish from all nets in areas 1-6, but was nil between unmarked or wild stocks from the northern and southern parts of the lake. At the third check the increment of growth of the unmarked fish in areas 1-6 was 0.1 inch smaller than that of the marked fish from the same areas and 0.2 inch larger than that of the unmarked fish in area 8. With regard to the central check on the scales of the naturally reared lake trout, two assumptions are possible. First, that these fish did in fact form a 0-mark during their first growing season; under this assumption these data exhibit no particular conflict with those for planted lake trout. Second, it may be assumed that naturally reared lake trout do not complete a 0-mark, and hence that the calculated lengths for the first three checks on the scales describe the fish at completion of their first, second, and third growing seasons. A corollary to this thesis, namely, that the average length of the unmarked, wild fish from all nets in areas 1-6, at the end of their first year (3.5 inches), was about the same as that of the marked, hatchery fish at formation of the 0-mark (3.7 inches), might be accepted without mis- givings as the hatchery and naturally propagated lake trout spent much of their first year in different environments. If this corollary is accepted, however, it follows that the increment of growth in length of the wild stock in northern Lake Michigan during their second growing season would be only 2.1 inches which is considerably less than the growth indicated for this group during either the first or third (2.7 inches) years. A growth of 2.1 inches the second year would be 0.7 inch less than the growth made in the same environment by the hatchery fish in their second year and 0.6 inch less than the growth made by the marked liatchery fish between their introduc- tion into the lake in September at a length of 3.2 inches and formation of the first annulus when they were 5.9 inches long. The growth made by the wild stock between formation of tlie first two checks on their scales, nevertheless, was very nearly the same as that made by the marked fish between formation of the 0-mark and the first annulus. It would be expected that the wild stock would grow at about the same rate as the introiluced fish after Sep- tember, but if they did, and the first check on the scales were the first annulus, they could not have grown any the fore part of the season. The length of the wild stock at the end of the second year would be 5.6 inches or 0.3 inch shorter than the marked stock at the end of their firs t year and 3 . 1 inches shorter than the marked fish at the end of their second year. To justify this relationship, it isneces- sary to assume that the wild stock grew erratically during their first or second year. The rates of growth in later years were about the same for the marked and unmarked lake trout. Although it cannot be stated categorically that the central check on the scales of the unmarked fish was not the first annulus, neither does it seem reasonable to assume that it is. The evidence strongly favors the belief that the first check on the scales of the naturally propagated lake trout is a 0-mark formed during the first growing season. This view is supported further by the appearance of the check itself (pattern, and location on the scales). See figures 15 and 16. The contribution of data on lake trout from southern Lake Michigan to the problem of the 0-mark is greatly limited by the lack of recoveries of planted fish from this area for comparison with the wild stock. Nevertheless, the much smaller increment of growth before formation of the first check on the scales of lake trout in southern than in northern waters makes it difficult to assume that the 0-mark of these naturally propagated fish is a first annulus. If this assumption is made, it is necessary to believe that these fish were only 4.8 inches long at the end of two full growing seasons or 1.1 inches shorter than tlie marked fish from northern Lake Michigan at the end of one year (5.9 inches). Alternatively, if it is assumed that the first check is a 0-mark, the calculated length at that point is somewhat smaller than that for the northern fish at formation of this check. vSubsequent growtli is only slightly less for the soutiiern than the northern fish. This growth pattern follows closely that of the marked fish. If the hypothesis, that most or all naturally reared lake trout do form a 0-mark on their scales during their first growing season, is ac- cepted, the question then arises as to the extent of error that this structure might introduce into the work of a competent and careful scale reader who is not aware of its existence. The only objective information on this point comes from AGE DETEKMIXATIOX FROM SCALES OF LAKE TROUT 29 records of culculatccl lengths for 97 lake trout captured in large-mesh gill nets off Montague, Mich., October 1, 1947 (Van Oosten 1950). The scales of these fish were read by Dr. Frank W. Jobes who did not record having observed the 0-mark. The calculated lengths from 82 of the fish in the year classes 1939-43 yielded an average length of 5.1 inches at the end of the first year of life. This average is Ijctwecn (1.5 inches higher and 1.0 inch lower than) the averages 3.6 and 6.1 inches obtained in tlie present study for the lengths at formation of tiie 0-mark and the first annulus, respectively, from 17 lake trout of the same year classes from southern I..ake Michigan (off South Haven in area 8) caught in the same year and in nets of the same mesh size (table 23). These differences suggest that on some scales Dr. Jobes may have measured the first aniuilus to the 0-mark rather than to the first annulus. How- ever, the calculated lengths " for the later years of life of the lake trout from Montague and South Haven were close enougli to indicate good agreement on the assessment of age. From tliese data, it appears tlial without a knowledge of the 0-mark, errors in measuring to the first annulus of lake trout scales, due to mis- interpretation of the central check, might be numerous enough to bias seriously an estimate of the first-year growth of I^ake Michigan lake trout, but errors of age determination would be few. AGREEMENT BETWEEN FIRST AND SECOND READINGS The two readings of lake trout scales, mentioned previously, were made on the scales of all fish in the collection. No samples were discarded, how- ever difficult to read. The second series of read- ings was begun several months after the first was completed and, for each fish, a second scale was read and measured, after comparison with the other mounted scales in the sample. The two readings agreed on age for 96.8 percent of the fish. Errors of interpretation, not involving change in age, reduced agreement to 91.4 percent of the specimens. Because of experience gained during the first reading, and standardization of proce- dures, the second reading disclosed errors in the earlier work as shown in table 8. Many of the disagreements resulted from the omission of a '• Sums of the incrcmfnts of growtfi. Those for the lake trout from .Mon- tapue. Mich., were obtained from the puhlislu'd data. measurement of the central or 0-mark, and the mistaken location of annuli. However, there were also disagreements on the number of annuli. The number of annuli located during the second reading varied from that recorded during the first reading for 45 (3.2 percent) of the fisii as follows: 1 annulus more for 18 fish, 1 annulus less for 24 fish, 2 less for 2 fish, and 3 less for 1 fish. The differences in percentage of such disagreements among the data for tlu' tliree plantings were not large. Disagreements in measurement, not re- sulting in change of age, occurred for scales of 76 (5.4 percent) of the fish. T.\BLE 8. — Comparison of first and second readings by the same person, of scales from the '^marked" lake trout Item Year of planting Totals 1944 1945 1946 57 1,077 271 1,405 Differences from first to second reading Rpsultinp in change of ape: 14 13 2 4 9 18 2 24 2 1 Total 3 5.3 29 2.7 13 4.8 45 Percentage 3.2 Not resulting in change of age; .\ssumption of marginal growth in 4 10 45 4 1 9 8 Current season's marginal growth 1 2 12 Age same but one or more annuli 56 3 5.3 59 5.5 14 5.2 76 5.4 fi 10.5 88 8.2 27 10.0 121 8.6 Disagreements in readings due to omission of the central check at the first reading were recorded, but were not considered to be errors in reading because the importance of measuring the 0-mark was not fully understood at the beginning of the first reading. Measurements of tlie central mark had been taken commonly, liowever, when loca- tion of the first annulus was math' easier In- defi- nitely locating the central check. The scales of some lake trout present such prob- lems of interpretation that readings made at differ- ent times are likely to disagree. Much of this un- certainty is dispelled by long familiarity with scales from fish of known age. Most readers dis- card the more didicult scales (usually about 5 per- cent of the total) as unreadable. U this practice liad been followed in the present study, some of 30 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE the (lisajiroemoiits bctwopii readings miglit liave been eliminated. AGREEMENT BETWEEN AGES READ FROM SCALES AND AGES FIXED BY DEFORMED FINS The final readings of the lake trout scales agreed witli the supposed ages of the fish for 1,319 of the 1,405 or 93.9 pereent of the specimens from nor- tliern I^ake Michigan. The presumed age is fk'- termined as the time between tiie date of capture and the year tlie fish would have been liatclied if the damaged fin were a true mark of identification. Detailed information is given in tal)le 9 for the 86 lake trout for whicii the supposed ages and the ages as read from the scales were in disagreement. Of this numlier, 9 fisii (indicated by asterisks in the talile) were so large in relation to their sup- posed age tiiat it may be assumed with confidence that they were unmarked fisii with malformed fins. No de])en(lal)U' objective standard was found from which to judge wlietlier or not the re- maining 77 lake trout were iiona fide recoveries of marked lake trout. They must accordingly be classed collectively as of "uncertain status." Data presented in a later section, however, give evidence that a large percentage of these fish liad not been marked. A summary of the discrepancies in age with respect to the degree of divergence (including the 9 fish designated in table 9 as too large for their supposed age) is given in table 10. Disagree- ments on age were mostly of 1 year (68.6 percent): but were of 2 years for 18.6 percent and more than 2 years for 12.8 percent. T.\BLE 9. — Inforiiintion on SfS lake trout from northern Lake Michigan for which aqe indicated by the mark did not agree with age read from the scales [Asterisks designate fish that obviously were too large in relation to their supposed ages to have been bona fide recoveries of marked fish] T.\Bi,E 0. — Information on 86 lake trout from northern Lake Michigan for which age indicated hi/ the mark did not agree with age read from the scales — Continued Total length (inches) Sup- posed age Age read from scales ml?ked Condition of an 31.8' 12.3'.. 21.2' 20.0' 14.H* _. 13.3* 12.7 __ 20.0* __ 19.0* 16.5' 12.0 22.0 21.5 20.0 20.0 17.6 14.1 1 1 2 2 2 2 2 2 2 2 2 3 3 3 I 3 9 3 4 5 3 3 5 6 7 r, 3 5 4 6 5 5 1946 1946 1945 1945 1945 1945 1945 1946 1946 1946 1946 1944 1944 1944 1944 1944 1944 No regeneration, or fin missing. ?3 normal length. No informLition. Do. yi normal length, rays twisted. Short stub, 2 twisted rays. ]r'i normal length. .\o information. Small stub. No information. Uo. Do. Do. Adipo.se missing, dorsal normal. Do. Do. Do. Total length (inches) 17.5 1.5.4 14.8 13.3 11.7 11.5 11.2 10.8 10.0 9.5 22.0 lfi.8 15.9 12.7 12.7 12.1 12.1 11.9 11.6 11.3 11.0 14.1 21.8 20.5 20.0 19.3. 18.2. 18.0 16.1 16.0 15.6. 14.7 13.9 12.5. 12.1. 11.7. 9.7 18.0. 14.9. 14.3. 13.4. 13.1. 12.0. 17.3. 14.4 13.7, 13.0. 12.S. 11.8. 11.8. 20.0. 19.4. 19.0. 18.8. 18.7. 18.6. 18.0. 17.0. 17.0. 16.3. 14.0. 13.8. 12.4. 12.2. 11.3. 11.2. 10.8. 13.4- 13.0- Sup- posed age Age read from Year marked' Condition of fin 1945 1945 1945 1945 1945 1945 1945 1945 1945 2 1945 5 1946 4 1946 5 1946 4 1946 4 1946 4 1940 4 1946 4 1946 5 1946 2 1946 4 1946 3 1944 5 1945 7 1945 5 1945 6 1945 6 1945 1945 3 1945 5 1945 5 1945 5 1945 5 1946 5 1945 3 1945 :i 1945 2 1945 5 1946 5 1946 5 1946 5 1946 5 1946 5 1946 6 1944 4 1944 4 1944 4 1944 3 1944 4 1944 4 1944 6 1946 6 1946 6 1945 6 1945 6 1945 6 1945 6 1946 fi 1945 4 1945 6 1945 4 1945 3 1945 0 1945 4 1945 4 1945 4 1945 3 1946 6 1946 4 1945 No regeneration, or fin missinfr. Do. Do. Do. Do. Do. Do. Do. Little regeneration. No information. Fin scar not seen in Ann Arhor. No regeneration, or fin missing. Do, l^ normal length, 8 rays. '/4 normal length, 1 curved ray. ^ii normal length, 10 twisted rays. \i> normal length, rays twisted. !^ normal length, 4 twisted rays. Ml normal length, fi twisted rays. H normal Irntzth. 12 twisted rays. No regt'niTation. or fin missing. H inch long, 1 twisted ray. Adipose missing, 4 rays in dorsal. 34 normal length, some rays fused and curved. 34 normal length, fi twisted rays. S rays. Xo regeneration, or fin missing. H normal length, 4 rays. ^4 normal length, H rays, Xo regeneration, oi' fin missing. Do. Do. Almost normal length, 7 rays. H normal length, 9 twisted rays. \i normal length, ray^ twisted. No regeneration, or fin missing. V-i inch long, 7 rays. \\ inch long, 1 curved ray. X^o regeneration, or fin missing. ^i normal length, H rays. \i normal length, fi rays. X'o regeneration, or fin missing. Do. H inch long. 2 twisted rays. Adipose missing, dorsal normal length but with all rays crooked ^6 distance from back. Adipose torn, dorsal normal. Do. Do. Adipose missing, dorsal normal. Adiposi- small, dorsal normal. Adipose missing, dorsal normal. J.4 inch long. 2 twisted rays. i^ normal length, b rays, Xo regeneration, or fin missing. Do. Do. Do. },-2 inch long, 2 twisted rays, i.^ normal length. ^ rays. No regeneration,- or fin missing H normal length, 2 rays. ^1 normal length, rays broken. Normal length. H norma! length, 4 curved rays. ^4 normal length, 8 normal rays, fi twisted ravs. 34 normal Irngth. 14 iwisted rays. ^4 normal length. 9 rays, ^^ normal l*'ngth. 3 twisted rays. ^^ normal length. 10 rays. 3i norma! length, 9 twisted rays. Table 10. — Smriwary of (he extent of disagreements on lake trout showing discrepancies between supposed ages and those read front the scales Areas 1-fi .\ge discrepancy Number of fish Percentage .59 16 11 68.6 2 years 18.6 >2 years 12.8 AGE DETKRMIXATIOX FROM SCALES OF LAKE TROUT 31 FACTORS OF DISAGREEMENT Disagroempiits hctwocn ajres as read from scales and supposod ages can arise from misinterpreta- tion of the scales from bona fide recovei-ies, and also from the inclusion in tlie sampl(> of lake trout that had not been marked. Both types of errors may be represented in the disagi-eements discussed in the precedino; section. Although the relative importance of these factors cannot be estimated closely, the data do provide some instructive in- formation in tlie matter. Errors of Reading Errors of reading may originate in tlie inter- pretation of scale patterns which, properlv diag- nosed, could lead to a correct determination of the age of the fish. Errors may arise also from defective scales, that is, scales that failed to form certain aimuli, developed accessory checks indis- tinguishable from amudi, or had a pattern so diffuse that any reading is questionable. As was pointed out earlier, the present collection certaiidy contained some lake trout that were not recoveries from plantings of fin-clipped fish. It is impossible, therefore, to attribute any individual disagreement strictly to error on the part of the scah> reader. It is possible, however, to gain a general idea of tlie clarity and dependability of scale patterns from the examination of a large series of scales, a high percentage of whicli must be from bona fide recoveries of planted fish, even thougii the status of an uidividual specimen must be recognized as uncertain. Careful study of the hundreds of scales from which readings agreed with supposed age led to the conclusion that over the age-span represented, the markings were almost always clear, and that failure to form an aiumlus must be rare. Some annuli were extremely faint, espe- cially in the posterior field but faint year-marks usually could be detected in the lateral fields. The presence of an occasional indistinct annidus does, nevertiieless, indicate the possibility of others so weak as to be overlooked. Accessory checks between atniuli, other than the 0-mark discussed in the preceding section, were not connnon and when present caused little trouble because they rarely, if ever, extended completely around the scale. Another factor which may iiave been a source of some error is the int.'rpretation of marginal growth. Dining th(> period of aiuiulus formation it is occasioiuilly didicult to decide whetlu-r the marginal band rei)resents completed growth of the previous year or rapid growth of the current season. Inclusion of Unmarked Fish With Abnormal Fins Overwhelming evidence was |)resented earlier that the "recoveries" from southern Lake Michi- gan (area 8) included few, if any, marked lake trout. Since there is no reason to believe that the development of a])norniaI fins among naturally propagated fish is exclusively a property of tin- stock of lake trout in southern Lake Michigan, it was to be anticipated that the recoveries from northern Lake Michigan, though principally marked fish of Imtchery origin, woulil also iiu'lude some naturally hatciied lake trout (and possibly some unmarked hatchery-reared lake trout that tleveloj)ed abnormal fins).-" Relation of disagreements to appearance of the fin If data on the "extent of regeneration" of the fins of lake trout from area 8 (tal)les 3 aiul 4) are typical for abnormal fins on wild fish, then, in samples from northern Lake Michigan (areas 1-6), the great majority of fish with fewer than 5 rays regenerated or with fins less than y, normal length would be bona fide recoveries of marked speci- mens, whereas most unmarked fisli with abnormal fins would appear in the group sliowitig greater regeneration. If these conchisions are valid and if the collection of lake trout from northern Lake Micliigan contains appreciable numbers of wild fish, a correlation should l)e found between the extent of regeneration and tlie percentage of dis- agreement l)etween supjjosed ages and ages read from scales. Tills expectation is met l)y the data of table 11, for the lowest percentage disagreenu'iit (3.8 per- cent) occurred among fish with fewer tiian .5 fin rays regenerated less than half normal length. For the other three groups in the nuiin body of the table the percentages ranged from (5.3 (trout with fewer than r, fin rays regenerated but iialf normal length or longer) to 10.7 (fish with fins less than half normal length but having 5 or more fin rays regenerated). The value of 6.9 percent ■" Allhough the pvmTitaBi' of wild-stuck hiko Iiout with alinorllKil fins is smiill, llir total niinihiT irport.-il hy fishmiii'ii can he con.siilcrahlc when all catches arc hi'Inc .scnititlizcil for dcfornicil fins. The perccntaRC of hatchery fish with ahnormal fins is also low. Dr. I'aul K.schincycr. who has hcon in charge of fin-clip|iin(! o|)eraIions at the fnitcd Stales Kish and Wildlife Service Fish Hatchery near Charlc'voix, .Mich,, several seiisons. states that an occasional finserlinc lake trout reared in the hatchery has an accrs-sory fin hut very few finperlings have deformed fins. 32 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE for trout with more than 5 fin rays at least half normal length offers a slight inconsistency, since, on the basis of the assumptions made, this per- centage should have been the largest. Table 11. — Relation of extent of the regeneration of pectoral fins, expressed in terms of number of regenerated rays and length {fraction of normal) to percentage disagreement between ages as determined from scales and as indicated by deformed fins Length of fin Less than half normal. Half normal or longer. All lengths Item Xiiiiibcr tif fish |-j\uiiihiT i)f 60 3 8.3 3 100.0 Total or average . . 36 5 13.9 1948: <20 123 61.8 3 2.4 20-40 61 30.7 6 9.8 40-60 15 7.5 1 6.7 >60 0 Total or average. . . 199 10 6.0 1949: <20 -- 91 17.2 3 3.3 20-10 -- 251 47.5 2 0.8 40-fiO— 156 29.6 3 1.9 >60 30 5.7 4 13 3 Total or average . . . 528 12 2.3 19.M): <20 59 23.4 2 3.4 20-tO 48 19.0 4 8 3 40-60 138 54.8 8 .5.8 >60 7 2.8 3 42.8 Total or average- - - 252 17 6.7 \m\: <20 0 0.0 0 0.0 20-40 5 18.6 0 ao 40-60 . . . 9 13 33.3 48.2 1 8 11. 1 >60 61.6 Total or average. . - 27 9 33.3 ' Some lake trout were omitted from this table because the description of the locality of capture was indefinite. 2 All disagreements were on lake trout captured along the north and east shores of Lake Michigan, areas 2, 3, 5, and 6. There were no disagreements on those captured in areas 1 and 4. The sequence of changes through the years fol- lowed the expected pattern. The percentage dis- agreement (between supposed age and age read from scales) was relatively high (13.9 percent) in 1947 when only 36 recoveries were made. As the number of recoveries rose to a maximum of 528 in 1949, the percentage disagreement dechned to AGE DETERMINATION FROM SCALES OF LAKE TROUT 33 a miiiiinuin of 2:.i. Decreases in tlu- luimber of recoveries to 2.')2 fish in 1950 and a mere 27 in 1951 were accompanied hy increases in percentage dis- agreement to 6.7 and 33.3 percent, respectively. The onU-r with respect to the size of the annual total number of lake trout recaptured was practi- cally tlie reverse of the order of the percentage disagreements. The one exception was between ranks 2 and 3 where the difTerences in percentage disagreement were small but sufficient to reverse the order of the ranking as shown: Year Number of fish Rank Percentage disagree- ments Rank 1949 _ 528 252 199 36 27 1 2 3 4 5 2.3 6.7 5.0 13.9 33.3 5 1950 3 1948 - 4 1947 2 1951 1 Still another significant feature of the annual totals is the limited range in the number of dis- agreements (from 5 in 1947 to 17 in 1950). The indicated variability is much below that of total recaptures for corresponding years. For example, from 1947 to 1949 the catch of fish with deformed fins increased 14.7 times but the number of dis- agreements increased only 2.4 times. Thus it ap- pears that the number of disagreements tended to fluctuate about a fairly stable level and to be rela- tively independent of the number of recaptures of marked fish. This relation is precisely the one which should obtain if a liigh percentage of the disagreements were caused by the presence of un- marked fisli. The data on tlie relation between locality of capture and percentage agreement within and between calendar years exhibit certain incoti- sistencies most of which can be attributed to tlic small numbers of fish in some entries. D(>finite trends can l)e detected, nevertheless. It is seen, for example, that tlie percentage disagreement between supposed ages and ages read from scales was invariably nil or small (0.0 to 3.4 percent) for lake trout recaptured within 20 miles of tlie point of release. The percentages were large, on the other hand, foi' trout recaptured more than 60 miles from the locality of planting. Only in 1949, when 13.3 percent of the fish were in disagreement on age, was there evidence of consideiai)le numbers of bona fide marked fish in this aiea. In the re- maining 3 years in which recaptures were re[)orted from distances greater than 60 miles, the per- centages ran from 42.8 to 100.0 (numbers of fish were small but the figures probably are significant because of consistently high values). For lake trout captured at tlie two intermediate distances, the percentage disagreement was nil at 20 to 40 miles in 1947 and 1951, but only 5 fish were reported each year. The remaining records for fish captured at 20 to 40 or 40 to 60 miles indi- cate a general inverse relationship between per- centage disagreement and number of lake trout reported. In the largest sample, 251 fish at 20 to 40 miles in 1949, the percentage disagreement was only 0.8; the two samples in the range of 100 to 200 fish had percentages of 1.9 and 5.8; and the four samples containing fewer than 100 fish had per- centages ranging from 6.7 to 22.2. The data of table 12, taken as a whole, lend strong additional support to tlie belief that a con- siderable part of the disagreements between sup- posed ages and ages read from the scales can be attributed to the presence in the sample of un- marked lake trout with abnormal fins. Relation of disagreements to size of fish It was stated in an earlier section that 9 of the 86 lake trout, for whicli the supposed ages and ages read from scales did not agree, were too large for their supposed age and lience almost certainly were not recoveries of marked fish, but merely had ab- normal fins (these fish are designated by asterisks in table 9). The basis for this conclusion is to be found in the length-frequency distributions of table 13. The 9 fish include 2 members of age-group I (marked lake trout of this age seemingly were still too small to be captured in commercial nets) and the 7 lake trout of age-group II that lay well outside the range of length for lake trout of the same age for which scale reading and supposed age agreed. For the remaining fish, length does not appear to offer a safe criterion for judgment as to whether any particular individual in a "no" column was or was not a marked fish. The frequencies and mean lengths for the paired groupings are so different, however, as to leave no doubt that the lake trout, for which supposed age and age as read did not agree, included considerable numbers of un- marked fish. Despite the wide ranges in length of individual age groups and the extensive overlap between successive age groups, the distribution of 34 FISHERY BITLLETIN OF THE FISH AND WILDLIFE SERVICE Table 13. — Length-frequency distribution of "marked" lake trout at capture, in age groups indicated by deformed or missing fin (all year classes combined) [Fish in "yes" column of cacli ago group are those tor which age read from scales agreed with age indicated by abnormal fin. and fish in the "no" column are those for « hich ages disagreed. Total lengths in inches] Age group Total length I II III IV V 1 No = Yes No Yes No Yes No Yes No 7 0 7 4 1 7 5 7 9 8 0-8 4 4 7 4 2 7 6 3 4 1 1 8 5-8 9 1 1 1 4 2 12 28 33 31 39 31 27 20 8 9 5 2 2 1 "T ! 1 ...... 2 7 19 44 84 74 84 HI 90 68 56 26 27 18 .8 3 3 3 1 1 1 10 0-10.4 . . . 2 10 5 10 9 1 11 0-11.4 T 2 1 2 1 2 2 ...... 2 ...... ...... ...... 1 ...... 3 2 6 4 10 5 18 15 24 41 23 31 26 17 14 14 10 5 6 2 11 5-11 9 2 12.0-12.4 --. 12 5-12.9 1 2 1 13.0-13.4 2 13 5-13 9 2 14 0-14.4 2 14.5-14.9 1 15 0-15 4 1 15.5-15.9 16 0-16 4 3 16 5-16 9 1 1 17.0-17.4 3 17 5-17 9 2 1 18 5 18 9 3 19 0-19 4 1 2 20 0-20.4 3 1 2 1 on "i 20 9 1 2 2 1 21 5-21 9 1 2 1 31.5-31.9 1 Number Mean length.-. 2 22 0 39 10.0 10 17.0 255 12.8 27 14.5 732 14.3 22 15.5 280 15.9 25 15.0 I Later age groups not included because the number of fish in each was too small to Yield useful information. ' No fish were captured for which the age read from the scales agreed with this supposed age. the lengtlis and the progressive shift of modes and means of fish in the "Yes" cohimns are much as would be expected. The frequencies in the "no" columns do not exhibit a similarly consistent rela- tionship. They show a random scatter greater than that which can be ascribed to the small num- bers of fish. Modes are lacking, and the means give no indication of the progressive increase in size that should accompany increase in age. The dif- ferences between the two groups with respect to indicated growth is demonstrated by the summary in table 14. Here, as was true for fish from south- ern Lake Michigan, the lake trout for which there was disagreement on age present the ridiculous spectacle of diminishing length with increase in age. A high percentage of tliem obviously could not have been from plantings of marked fingerlings. Another approach to the question of the presence of unmarked lake trout in the samples lies in the comparison of the growth of lake trout for which there was agreement on age with the growth of thosef or which there was not agreement on age. In this comparison it was assumed that none of the lake trout for which there was disagreement were marked fish and that the scales rather than fin abnormalities offer the correct estimate of age. Table 15, which gives the result of this compari- son, is so arranged that the vertical columns give the average lengths at ages indicated by abnormal fins and diagonal rows (from upper left to lower right) contain a series of estimates of the length of lake trout of the same age, as read from the scales. As would be anticipated, if the readings are correct, the lake trout with agreement on age were shorter than those whose ages, read from the scales, were one or more years older than the ages indicated by the deformed fins. Conversely, the lake trout with agreement on age were larger than others whose ages were read one or more years younger than the ages indicated by the fins. In general, the magnitude of this difference in lengths was progressively greater with each in- crease in tiie number of years of disagreement between the supposed age of the fish and the age read from the scales. Despite the considerable variability expected because of the small numbers of fish in some samples and the known large range of lengths within age groups, the means in each diagonal, in the main, fluctuate normally about the average length determined for lake trout for whicli ages from scales and fin marks agreed. Data in summarj- table 16 support the conten- tion that lengths of age groups determined by Table 14. — Comparison of average lengths of lake trout, for which the ages as indicated by scales and fins were the same, with average lengths as indicated by abnormal fins of lake trout for which ages indicated by fins and scales were different [Data from table 13. Number of fish in parentheses] Total length (inches) Age group Scales and fins in agreement Scales and fins not in agreement 22.0 I (2) 17.0 (10) 14.5 (27) 15.5 (22) 15.0 (25) II 1 10.0 \ (39) / 12.8 \ (255) / 14.3 \ (732) / 15.9 \ (280) HI IV V AGE DETKRMIXATIOX FROM SCALES OF L,\KE TROUT 35 Table 15. — Comparison of the average lengths of lake Iroiil whose scale readings disagreed with their supposed age uith the average lengths of lake trout for which the reading agreed with the supposed age ' (In all readings, It was assumed that the central check was a O-mark] Departure of age read from expected age 5 yean; more: Age from scales.. .\verage length.. Number of fish_. 4 years more: .\gi' from scales.. .\vorape length _. Number of fish.. 3 years more: .\ge from scales. .\verage length.. Number of fish., 2 years more; .\ge from scales,, .\verage length.. Number offish.. 1 year more: .\ge from scales. .\verage length. Number of fish. .\s expected: .\ge from scales. .\verage length. Number of fish. 1 year less: Age from scales. .\verage length. Number of fish. 2 years less: .\ge from scales. .\verage length. Number of fish. 3 years less: .^ge from scales. Average length. Number of fish. Age indicated by fln mark reported III 12.3 1 I '5.6 1,319 VII 19.0 1 VI 18.2 2 V 16.4 2 IV 21.2 1 III 13.4 3 II 10.0 39 HI IV VII 20.0 1 VI 20.0 1 V 16.1 6 IV 13.4 15 III 12.8 255 II 10.5 3 VII 19.2 2 VI 18.8 2 V 15.4 13 IV 14.3 732 III 13.5 4 II 9.3 I VII 22.0 1 VI 17.8 11 V 15.8 280 IV 13.0 10 III 12.5 3 II 12.4 1 VI VI 15.6 13 IV 13,0 1 ' In addition to the fish listed in the table, the collection contained 1 lake trout. 31,8 inches long, which, according to the fin. should have belonged to age-group I but the scales indicated it to be a member of age-group IX. • A mean calculated length bii.'^ed on all lake trout for which ages from scales and fin marks agreed. The samples contained no fisli whose scales indicated that they belonged to age-group I. scale readings give a reasonable estimate on growtli of tlie lake trout for which ages from scales and fins disagreed. In tliis table age groups, as established from the scales, have been combined regardless of discrepancies between supposed age and age as read. For age groups II to V, the differences in average lengths between the lake trout with and without agreement on age fell within the range of 0.1 inch (age-group V) to 0.8 inch (age-group IV). The difference was fairly large (2. .5 inches) for age-group VI, but here the average length for trout with agreement (15.6 inches) must be viewed with skepticism as it was 0.2 inch below the mean lengtli for age-group V (15.8 inches). Despite this discrepancy, the data, as a whole, show that the scale readings gave reasonable estimates of the growth of tlie 86 fish with disagreement on age, and hence provide still further evidence of a high percentage of unmarked lake trout among them. Table 16. — Comparison of average lengths of lake trout for which the ages as indicated by scales and fins were the same, with average lengths as indicated by scale readings for the 86 lake trout foi which ages indicated by fins and scales were different [Data from table 15. Number of fish in parentheses.) Total length (Inches) Age group Scales and Sns in agreement Scales and flns not in agreement II f 10.0 \ (39) 10.6 (5) III. 12.8 (255) 14.3 (732) 15.8 (280) 1 15.6 (13) 13.1 (11) 13.5 (27) 15.7 (21) 18.1 (16) ' This low figure is probably due to selective destruction of the lake trout population in Lake Michigan. CONCLUSIONS AS TO THE DEPENDABILITY OF SCALE READINGS The study of tlie scales of lake trout presumably of known age has proved scale readings to be highly dependable over the age span represented in the sample. In the original collection of 1,405 recaptured lake trout from northern Lake Michi- gan, ages as read from scales agreed with ages as indicated by fin marks for l,.'n9 or 9:5.9 percent of the individuals. The actual degree of de- pendability is much greater, however, than this percentage suggests. The evidence is strong that the 86 fish, for which ages were in disagreement, actually included many unmarked individuals on wliich fin development had been abnormal. Nine lake trout could be designated with confidence as "unmarked" because of their unreasonably large size in relation to their supposed age. Criteria were lacking for an objective decision as to whether any one individual among the remaining 77 fish could have been a bona fide recovery, but a series of analyses on the reflation of the disagreements to appearance of the deformed fins, year and locality of recapture, size and growth of fish yielded con- vincing evidence of tlie presence of considerable numbers of unmarked lake trout. Although an exact figure can not be given, it can be stated with confidence that, had the original sample been composed entirely of recaptures from the three plantings of marked fish, the agreement between supposed ages and ages read from the scales would have been well above 95 percent. The O-mark, a check in the field of first-year growtli, was present on the scales of nearly all 36 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE marked lake trout recaptured and, according to the best available evidence, was also a charac- teristic of the scales of most wild fish. It should be emphasized that, dependable as lake trout scales may be as indicators of age, thej- are not read easily. Considerable experience is required before a reader's interpretation of the scale pattern becomes highlj' reliable. Even the experienced reader can do accurate work only if the scale preparations are clear and they are studied carefully with the aid of the best optical equipment. GROWTH OF MARKED LAKE TROUT The study of the growth of marked lake trout is based principally on the 1,319 specimens for which the age as read from the scales agreed with the supposed age. This restriction excludes any bona fide recaptures for which errors were made in scale readings. The 1,319 fish may include a few unmarked fish with abnormal fins that hap- pened to be of the correct age at capture. There is no reason to believe, however, that the number in either of the groups is large; the restricted sample, therefore, may safely be presumed to consist almost entirely of marked fish and also to include nearly all of the true recoveries. Measurements of the marked lake trout were made in Ann Arbor before the fish were preserved but after they had been shipped in ice from the port where they were landed. Although most of the fish were in good condition upon arrival, a few were in advanced stages of decomposition so that length and weight measurements could not be determined accurately. Such fish have been ex- cluded from tables and calculations for which those measurements are requisite. In some tables the total number of fish was further reduced by dropping from consideration the older age groups which were poorly represented. More lengths than weights were obtained because some of the lake trout were dressed (gills and viscera removed) upon arrival in Aim Arbor. LENGTH-WEIGHT RELATION The commonly accepted formula expressing the length-weight relation in fishes is: W=cL" or log M'=log c+n log L where Ii'= weight L = total length and c and « = constants As tlie measurements of length antl weight alike are subject to error, a method developed by Bartlett (1949) was used in fitting a line to the logarithms of individual lengths and weights of 1,197 lake trout ^' from northern Lake Micliigan. The resulting estimate of the relation between woiglit in ounces and total length in inches was: log ir= -2. 4698-1-3.1125 log L The value of 3.1125 for n (which measures the relative rates of increase of weight and length) shows that in these lake trout the weight increased somewhat faster than the cube of the length. In other words, the body form became more robust as the fish grew longer. The departure of the lengtli-weight relation- ship of the lake trout of northern Lake Micliigan from the "cube law" probably was significant. The 5-percent confidence interval of the true slope /3 with ^ = 1.962 for 1,195 degrees of freedom, when calculated by Bartlett's method was 3.13718±0.90129. At the same level of signifi- cance, the least squares method gave 6xy = 3.08414 ±0.04332. Comparisons between empirical weights and tlieoretical weights (as computed from the length- weight equation) are to be found in table 17 and figure 18; the straight line of figure 18 is a graph of the equation. Because talile 17 contains actual and computed values of both length and weight, an explanation of the arrangement may be lielpful. The first row of figures in the left section, for example, states first that the single lake trout 7.2 inches long had a weight of 1.2 ounces at capture (foiu-tli column). In the same row, it is sliown furtlier that the expected weiglit for a 7.2-inch fish was computed to be 1.6 ovuices (fifth column) and that the expected lengtli for a 1.2-ounce lake trout would be 6.6 inches (third column). Agreement between most empirical and calcu- lated weights and lengths can be termed good. Discrepancies usually are small (full agreement at 14 lengths). The larger disagreements occur at 2' This numbor included all the lake trout weighed in the round, l.US presumably marked and 79 for which the ages from scale readings and deformed fins did not agree. 200 100 90 80 70 60 ^50 LU 40 O 2 30 O X 20 1.5 10 9 8 7 6 AGE DETERMINATION FROM SCALES OF LAKE TROUT ■I I , , , , |,,,.|i...|....|....|i.M|....|.i..|..ii|iiii|i.ii|.i.i|...i,..TU^ ■ ■ ' ■ I ■ ' ■ ■ I ■ ■ 37 ' 1 ' ' " I ■' "r log W = -2.470 + 3.112 log L or W= 3.390xI0'l"" 1 I I , I I I I I ,1 ,i,iImi, iiiil ,iil , 1„ In liJ 2 3 456789 10 20 LENGTH (INCHES) 30 40 FioiRE 18. — Length-weight rehitiou of 1,197 lake trout from northern Lake Michigan. (Dots give empirical values line represents values obtained from solution of the length- weight equation]. 38 Table 1/ FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE -Relation between total length (inches) and weight (ounces) of lake trout from northern Lake Michigan, also lengths and toeights calculated with the length-weight equation [Based principally on lake trout recaptured from the 1944-46 plantings of marked fish. See text for details] Num- Total length Weight Num- Total length Weight Num- Total length Weight ber ber of ber of of fish Actual Calculated Actual Calculated fish Actual Calculated Actual Calculated fish Actual Calculated Actual Calculated 1 7.2 6.6 1.2 1.6 24 13.0 13.0 9.0 9.9 10 17.2 17.5 25.2 23 8 1 8.0 7.6 1.9 2.2 15 13. 1 13.1 10. 1 10.2 8.- 17.3 17.5 25.1 24.2 1 8 3 «.3 2.5 2.5 19 13.2 13.1 10, 1 10.4 3 17.4 17. I 23.4 24.6 2 8.4 8.4 2.6 2.6 20 13.3 13.4 10,9 10.7 6 17.5 17. S 26.6 25. 1 2 8.6 8.6 2.7 2.8 32 13.4 13,4 10,8 10.9 6-. 17-6 17,3 24.0 26,6 2 8.7 8.6 2.8 2.8 25 13.5 13,5 11,2 11.2 10 17.7 17 8 26,3 26,0 1 8.8 8.9 3.1 3.0 19 13.6 13.2 10 5 11.4 3. 17.8 17,9 26,8 26,4 1 8.9 8.8 2.9 .11 23 .-.- 13.7 13,7 11,6 11.7 3 - 17.9 18,4 29,4 26.9 1 9.2 9.5 3 8 3 4 30 - 13.8 13.8 12 0 12.0 13 18.0 18,0 27,6 27.4 2 9.3 9.3 3.5 3.5 12 13.9 13,8 12.0 12.2 2 18. 1 18,2 28,2 27.8 2 9.5 11.7 7.2 3,7 27- 14.0 14,0 12.5 12,5 2 18.2 18,2 28,2 28.3 3. 9.7 9.5 3 7 4,0 22 14. 1 14.0 12.5 12,8 4 18.3 18,4 29.4 28,8 4 9.9 9.5 3.8 4,3 30 14.2 14.2 13.0 13, 1 1 18.4 18.4 29.6 29,3 4 10.0 11. 1 6.0 4,4 26 14.3 14.4 13.6 13,4 4 18.5 17.8 26.6 29,8 3 10. 1 10.3 4.8 4,5 27 14.4 14.4 13.7 13,7 5 18 6 19.1 32.7 30.3 4 -- 10.3 10.7 5.4 4.8 22 14.5 14.6 14. 1 14,0 4 18.7 18.8 31.2 30 8 1.- 10.4 10.4 4.9 5.0 24 ...- 14.6 14.6 14,3 14,3 5 18.8 18.8 31.5 31 3 3 10.6 10.6 5.3 5. 1 24 14.7 14,6 14.4 14.6 1 18.9 19.8 36.8 31.9 4 10.6 10.8 6.6 5.3 17 14.8 14,8 14.8 14.9 5 19.0 19.3 33.9 32,4 6 10.7 10.9 5.7 6.4 14 14.9 14.9 15.1 15-2 1 19.1 21,5 47,8 32,9 5 10 8 10.7 5.5 5.6 14 16.0 15.1 15.9 16.6 2 19.2 20,0 38,2 33,6 2 10.9 10.7 5.4 5.7 17 15.1 16.2 16. 1 15.8 4 19.3 19,7 36,2 34,0 10 11.0 11.5 6.7 5.9 17 15.2 15.2 16.3 16.2 2 19.4 19.0 32,4 34,6 8 11. 1 11.2 6 3 6.1 17 16.3 16 3 16.4 16.5 1 19 5 18.9 32.0 36, 1 9 11.2 11.5 6.8 6.2 18 16.4 16.4 16.7 16.8 1 19.6 19.3 34.0 36 7 9 11.3 11,4 6.6 6.4 23 15.5 15.6 17.6 17.2 1 19.7 20,7 42.5 36,2 11 4 11 3 6.6 6.6 23 16.6 15.6 17.5 17.5 1 19.8 19,6 35. 6 36,8 7 11.5 11.3 6.5 6.8 12 16 7 1.5.7 17,9 17.9 12 .--. 20.0 20, 1 38.8 38,0 9 11.6 11.7 7.1 7.0 13 16 8 16.1 19,4 18,2 1 20,1 19,7 36.2 38,6 12 11 7 11 8 7.3 7.2 16 15.9 16.0 18.9 18,6 1 ---- 20,2 19,7 36,5 39,2 8 11.8 11. S 7.4 7.4 7 16.0 15, 6 17.7 19,0 I 20.3 20,6 41,0 39,8 17 11.9 11.9 7.5 7.6 10 16.1 16 2 19.7 19,3 1 20,5 19,7 36,0 41,0 10 12.0 12.3 8.4 7.8 7 16.2 16,2 19.6 19,7 2 21.0 20,5 40,9 44,2 16 12. 1 12.2 8.2 8.0 4 16.3 16,8 21,9 20,1 1 21.2 21,0 48,0 46,6 21 12.2 12.2 8.2 8.2 11 .,-.. 16.4 16,4 20.6 20,6 2 21.5 22.6 65,4 47.6 10 12.3 12.6 8.9 8.4 14 16.5 16,7 21,7 20,9 1-- 21.6 21,9 50,2 48.3 18 12.4 12.3 8-4 8 6 7 16.6 16.5 20,9 21,3 1 21.8 22,0 61,0 49.7 18 12.6 12.6 8.9 8.8 4 16.7 16.5 21,0 21.7 3 22.0 21,3 46.0 51. 1 22 12,6 12.5 8.7 9.0 12 16.8 17.0 23.1 22.1 1 22,3 21,6 47.6 63.3 22 12.7 12.8 9.4 9.2 8 16.9 16.8 22.1 22.5 1 31,0 34,5 208 0 148 6 27 12.8 12.7 9.3 9.5 11 17.0 16.8 21.9 22-9 24 12.9 12.9 9.6 9.7 6 17.1 17.1 23.2 23.3 lengths where averages are based on one or a few fish. Other factors that possibly could have con- tributed to the discrepancies include: annual and seasonal fluctuations in condition; sex and state of gonads (of the larger, mature fish); and gear selec- tion. " It may, accordingly be held valid to use the equation to describe the general length-weiglit relation and thus estimate weight when length only is known or length when weight only is available. The calculated weights (table 17) show that lake trout would be expected to attain the weight of 1 pound at 15.1-15.2 inches, 2 pounds at about 18.9 inches, and 3 pounds at 21.5-21.6 inches. The length corresponding to I/2 pounds, the minimum weight at which lake trout may be taken legally in the State of Michigan, was 17M inches. To test whether the equation representing the length-weight relationship of the lake trout in the sample was also representative of younger fish, " See Farran (1936) and Deasonand Hile (1947) for discussions of the efTects of gill-net selectivity on the estimation of the length-weight relation. calculations of weight were compared with the weights of the control groups reared in ponds at Marquette, Mich. (Shetter 1951). The lengths at which tlie comparisons were made are average lengths at capture of the fish in the control groups. Lengths overlapping those of the lake-reared fish up to 10.7inciies are also inclmled in the tabulation. Length (inches) Number of fish (control groups) Measured weight (ounces) of pond-reared control group Calculated weight (ounces) 2.9 2.007 2.000 946 732 860 2«6 837 289 699 469 262 0.11 .15 .26 1.2 1.3 1.9 1.5 1.9 4.3 4-9 5.3 0.09 13 4 1 .27 (,4 _ 1.1 (5 7 _. 1.3 7.2 1.6 7.3 1.6 76 1.9 97 4.0 10.1 - 4.5 10.7 5.4 AGE DETERMIXATIOX FROM SCALES OF LAKE TROUT 39 The (liffiTcnces hotwec^ii m(>asure(l and calcu- lated weights for each length given did not exceed 0.4 ounce. The average weight of the pond- reared group was only slightly lieavier, 0.08 ounce, than the average calculated weight for all length- intervals represented by the group. GROWTH IN LENGTH The data presented on growth in length of marked lake trout include both lengths at capture and calculated lengths (based on scale measure- ments) at the end of the several years of life and at time of foi-mation of the 0-mark in the first field of growth. All calculations of length were made by direct proportion, that is, on the assump- tion that the ratio of length of fish to diameter of scale is constant at all lengths attained by the fisli after completion of the 0-mark. Although the materials at hand are not suitable for a discrimi- nating test of this assumption (range in lengths is too short and lengths at the ends of the range are represented by inadequate numbers of individ- uals), such data as are available indicate that any systematic errors, from the use of direct propor- tion, must be extremely small. Lengths at Capture The measured lengths of the marked lake trout of each age group, at the time of capture, extended over a wide range which was somewhat greater for the older than for the younger fish (see figure 20). The range within a single age group (year classes combined) varied from 5.4 inches for age-group II to 12.6 inches for age-group III with intermediate ranges for the remaining age groups (table 18). Despite wich- ranges in lengths, the mean lengths for each year of age reached by the three year T.^BLE IS.— Mean length (inches) and ranges of length at time of capture, of the year classes of marked lake trout' by age group ' Item Age group II III IV V VI 1944 year class: .Mean length 13.4 9.9-20.0 16 12.9 9. 7-22. 3 190 12.2 9. 9-16. 4 49 12.8 9. 7-22. 3 255 15.2 12.8-21.0 10 14.2 10.5-21.0 555 14.5 11. 1-20.0 167 14.3 10. 5-21. 0 732 15.1 13. 4-19. 7 17 15.8 10. 0-22. 0 240 16.6 13.0-21.6 23 15.8 10.0-22.0 280 Range Number of fish 1945 year class: Mean length Range N'umberof fish...' 194(5 year class: Mean length Kange .Number offish ... Combined year classes: .Mean length. Range Number offish.... 10.0 7. 2-12. 6 31 9.9 8.0-12.0 8 10.0 7. 2-12. 6 39 15.6 12. 2-20. 2 13 I.V6 12. 2-20. 2 13 classes of marked lake tiout were remarkably close together. \o represtnitatives of age-group II of the 1944 year class were taken by the fishermen, but the mean lengths of the 2-year-olds of the 1945 and 1946 year classes differed by only 0.1 inch. Tile mean lengtlis for age-groups III, IV, and V in all three year classes had maximum differences of 1.2, 1.0, and 1.5 inches, respectively. Tlie year classes of marked lake trout planted in Lake Michigan not only grew at similar rates but, regardless of environmental difi"ercnces, they also grew at about the same rate as control groups reared in ponds at tiie State Fish Hatchery, Marquette, Mich. The pond-reared lake trout of the 1944 year class had grown 16.6 inches in lengtli by October 1948 (age-group IV). None of the 1944 year class of marked, lake-reared lake trout were captured in October 1948, but the average length of trout in age-group IV caught from April through September was 15.2 inches which, as would be expected, was somewhat below tlie average for the fish taken only in October. The pond-reared lake trout of the 1946 year class were 10.1 inches long when they were measured in October 1948 (age-group II). Although no re- coveries from the lake-reared fish of the 1946 year class were made in October 1948, the average length of 9.9 inches for fish in age-group II caught from May tlirough September is not far below that for the pond-reared lake trout of the same year class. The best comparison of lake- and pond-reared lake trout comes from the more plentiful samples of the 1945 year class which were measured in May 1948 when they were members of age-group III. At this time the pond- reared fish were 11.7 inches long and the marked, lake-reared fish averaged 11.9 inclies long (table 19). Table 19. — Comparison of total lengths {inches) of take- reared, marked lake trout with those of the pond-reared control groups (.Number of fish in parentheses] Year of planting ' 1944 1945 1946 Pond-reared trout: .Average length- 16.6(200) Oct. 1948 15.2(10) Apr.-Sept.,1948 11.7(378).. May 1948.. 11.9(20). . May 1948 10.1 (196) Oct. 1948 9.9 (8) May-Sept.. 1948 Time of measurement- - I,ake-reare of cajjture as the following tabulation demon- strates: The marked fish recaptured as members of age-groups II, III, and IV measured 10.0, 12.8, and 14.6 inches long. Calculated lengths for the same years of life were 8.7, 11.2, and 13.7 inches. Whereas the calculated lengths give the size of the fish at the beginning of the growing season, the fish were caught somewhat later in the year at various times during the growing season, hence, were expected to be longer. Lengths, obtained by adding increments of growth, for fish in their fifth and sixth years of life show that in those years the lake trout actually continued to grow at rates only slightly lower than those during the earlier years of life (excepting the firet year). The relation of the calculated lengths to tlie empirical data is shown in figure 20. The mean annual increments of growth gradually decreased as tlie lisli became older from 5.9 inches the first vear to 2.8 inches the second 42 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE MEAN LENGTH AND RANGE IN LENGTH OF MARKED LAKE TROUT AT TIME OF CAPTURE COMPARED WITH MEAN CALCULATED LENGTHS T — [ — 1 — I — I — I — r • MEAN LENGTH AT TIME OF CAPTURE X MEAN CALCULATED LENGTH Ql I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 0 SONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJAS FIRST YEAR SECOND YEAR THIRD YEAR FOURTH YEAR FIFTH YEAR SIXTH YEAR Figure 20. — Mean length and range in length of marked lake trout, at time of capture compared with mean calculated lengths obtained by adding annual increments of growth (assuming January 1 the date growth is completed). Vertical broken lines give range of lengths, and dots the mean lengths at capture. Calculated lengths are shown along the solid, diagonal line. year, 2.5 inches tlie third, fourth, and fifth jears, and 2.4 inches the sixth year. Factors of Discrepancies in Estimates of Growth Several factors were considered as possible causes of the discrepancies in the estimates of growth made from the different age groups of marked lake trout: (1) condition of the fish; (2) sex differential in growth; (3) selectivity of nets employed by the fishery; (4) selectivity of lamprey predation. The effects of the first two were not considered important. Nearly all fish captured were taken during the summer months, thus seasonal changes in condition were not a factor. Combining data on the three year classes masked the annual differences Sexual differences had not developed on these fish, most of which were still immature; none cauglit was in gravid condi- tion. The other two factors affecting estimates of growth are discussed later. Selectivity of gill nets is an important factor which would have a tendency to cause discrepancies in estimates of the growth rate. Some marked lake trout (43.4 percent) were caught in the 4/2- inch-mesh gill nets of the whitefish and lake trout fisheries and others (56.6 percent) were taken in the 2j2-inch-mesh gill nets of the chub fishery. The percentage of the total catch of lake trout taken by the 4)2-inch-mesh nets decreased from 91.3 in 1947 to 17.5 in 1950. At the same time the percentage caught in tlie 2}2-inch-mesh nets increased from 8.7 to 82.5. During this period lake trout were becoming so scarce that fishermen were AGE DETERMINATIOX FROM SCALES OF LAKE TROUT 43 turning more and more to cliul) fishing with tho small-mesh nets. A gill net made of a single size of mesh tends to catch the larger fish of the younger age groups but, as the fish grow larger in later years this relation between size of fish and size of mesh in the nets is reversed and the net then catches the smaller individuals of the older age groups. This reversal takes place when the fish are at an earlier age if small-mesh nets are used than if the fishing is done with larger-mesh nets. The marked lake trout of age-groups III and IV, caught in 2,'2-inch-mesh nets, were 1.3 and 0.2 inches longer than tlie mean calculated length for the age group, and those of age-groups V and VI were 0.4 and 4.2 inches shorter than the calculated lengtlis. Fish of all age groups, caught in the 4/2-inch-mesh nets were longer than those cauglit in the 2!2-inch-mesli nets and also longer than the mean calculated lengths for the age groups represented. The discrepancies for age groups II to V fluctuated between 1.7 and 1.2 inches without clear trend. For age-group VI, the difference (0.5 inch) was less than the other differences, but the reduction mav not indicate that the reversal to capture of the smaller fish of a year class was approaching for this net. Prob- ably, larger fish were no longer available for capture. Am- group Mean cal- culated length Length at capture offish caught in nets of: 2>4-inch- niesh 4H-lnch- mesh n 8.7 U.2 13.7 16.2 18.6 10.0 Ill 12.5 13.9 15.8 14.4 12.9 IV 14.9 V 17.5 VI . ... 19.1 Even though the large-mesli nets consistently caught the larger fish, the average size of lake trout taken in them and in the small-mesh nets increased as the fish became larger. Nets of each mesh size were static measures of a segment of a changing range of lengths within the population as the fish of each year class became older, hence, the mode of the lengths of lake trout caught in each net shifted from the lower toward the upper limits of its segment as the average size of the fish increased (table 21). T.\BLE 21. — Calculated lengths (in inches) of marked lake trout (year classes combined) caught in large- and small-mesh gill nets IDifferences are shown below the lengths of flsh caught in each pair of nets) .\ge group Mesh of net (inches) Number ' offish caught Average total length and range of length Lit cap- ture 2 (inches) Calculated lengths at end of year of life 0 1 2 3 4 5 6 / 4H \ r.i ( Mi I 2,4 / 44 \ 2'i 1 Mi \ 24 f 44 \ 24 39 0 187 64 272 449 64 215 3 10 10.0 (7.2-11.9). _ 3.5 5.4 8.5 12.9 (9.7-22.3) 3.8 3.8 0.0 3.9 3.7 .2 3.8 3.7 .1 3.2 2.6 .6 5.9 6.1 -.2 5 9 5.6 .3 0.0 5.5 .5 5.1 4.2 .9 9.2 9.0 .2 9.1 8.7 .4 S.8 8.2 .6 8.1 6.0 2.1 11.7 11.6 .1 11.8 11.0 .8 11.4 10.6 .8 10.8 8.0 2.8 12.5(10.7-15.5)... 14.9 (10.3-21.0) 13.9(10.5-19.1) 17.5(13.0-22.0). 15.8 (10.0-20.1) __. 19.1 (17.7-2n.2)_ .. 14.4(12.2-18.0) 14.3 13.4 .9 14.0 12.9 1.1 13.6 9.9 3.7 v.. VI..._ _. _ 16.7 15.4 1.3 16.0 11.8 4.2 "ii'Q 13.9 5.0 4H 3.8 3.8 3.7 3.7 2.1 5.9 1.9 5.6 3.2 9.1 3.0 8.6 2.6 11.7 2.3 10.9 2.5 14.2 2.4 13.3 2.7 16.9 2.5 15.8 2.9 Length from summation of increments 19.8 m 2.0 17.8 1 Size of mesh in net not recorded for 16 fish. * Fish caught :it clifTerent times during the growing season. Their total lengths :ire not comparable with llu- oalculatcd lengths. Calculated lengths of the marked lake trout emphasize the differences in length between fish caught in the 4}^- and 2j2-inch-mesh nets. The differences increase in size with each year of life (table 21, fig. 21). Undoubtedly, "the small (average length at capture, 10.0 inches), slender lake trout of age-group II captured in large-mesh nets were caught by their teeth or by other en- tanglement in the twine. The size of the mesh ill the net could scarcely have been the determining factor in their capture. In fact, the small repre- sentation from age-group II in the sample (that from age-group III was six times as large) in- dicates that the fish in this age group were too small to be caught systematically in commercial nets of anv mesh size used. Evidently, too, these 44 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 3 4 5 6 YEAR OF LIFE Figure 21. — Calculated lengths of marked lake trout (year classes combined) caught in large- and small-mesh gill nets For each age group the calculated lengths of the fish caught in 4!j-inch-mesh nets are connected by a solid line and those of fish caught in 2H-inch-mesh nets by a broken line. As all age groups represented in the sample are shown in the same graph, the curves do not have a common base, hence none is shown. Consult table 22 for values (inches of length) of points on the curves. The numbers of fish taken by each net are shown in parenthetical boxes. lake trout were the smaller individuals of age- group II. Their mean calculated lengths were all smaller than those for the same years of life of the fish in age-groups III, IV, or V caught in either type of net (with the exception of the calculated length for the second year of age-group V caught in the small-mesh nets which was just 0.3 inch shorter than the one for age-group II). Nets of both sizes of mesh took fish of approxi- mately the same size from age-group III (larger- mesh nets captured only slightly larger fish). The difference in the calculated lengths of fish caught by large- and small-mesh nets increased gradually, as the fish advanced in age, from 0.1 inch in the third year of life of age-group III to 0.9 inch in the fourth year of age-group IV, 1.3 inches in the fifth year of age-group V, and 5.0 inches in the sixth year of age-group VI. The large discrepancies in the older age groups between the calculated growth histories of fish caught in 2^^- /2- and 4j2-inch-mesh nets leave some uncertainty as to the true rate of growth. Pos- sibly the samples from small-mesh nets give better estimates of the growth rates for the younger age groups and the fish from large-mesh nets may provide better estimates for the older age groups. Because of the different selectivities shown by the gill nets of these two mesh sizes, the marked and unmarked lake trout caught in nets of each mesh size were studied separately. The growth rates of marked and unmarked lake trout of the same year classes (1944-46) caught by 4}2-inch-mesh nets in northern Lake Michigan were closely similar. However, with but one exception, sizes ecjual at formation of the first annulus, the calculated lengths of the unmarked fish were somewhat lower, ranging from 0.2 inch at formation of the 0-mark to 1.2 inches at the sixth annulus. The average annual increment of growth in length after the first year was 2.8 inches for the marked and 2.5 inches for the unmarked fish. The calculated lengths of the unmarked AGE DETERMINATION FROM SCALES OF LAKE TROUT 45 4 5 6 7 8 9 YEAR OF LIFE Figure 22. — Calculated lengths (sums of mean increments of growth in inches) of marked and unmarked lake trout of year classes 1944-46, and of the older year classes (1938-43) from the wild stock, caught by 4'2-inch-mesh nets in north- ernLake Michigan, areas 1-6. lake tiout (year classes 1944-46) caught by 2?^- iiich-mesh nets in northern Lake Michigan were also lower than those of marked lake trout caught in the same nets. In fact, the differences between their calculated lengths ranged from 0.4 inch at formation of the first annulus to 1.1 inches at the fifth annulus. The average difference was 0.2 inch greater than the average difference between the groups of marked and unmarked lake trout caught in large-mesh nets. The average annual increment of growth in length for the fish from small-mesh nets was 2.4 inches for both marked and unmarked lake trout but the marked fish were already 0.4 inch longer than tiie luumirked fish at formation of the first annulus (table 22, fig. 22). Although marked and unmarked lake trout of the same year classes caught by small-mesli nets were somewhat smaller than those caught in the large-mesh nets, the calculated lengths of the unmarked fish retained about the same relative position below those of the marked fish that the unmarked fish had to the marked fish caught in large-mesh nets. Table 22. — Calculated lettylhs (sums of mean increments of growth in inches) of marked and unmarked lake trout of year classes 1944-46 caught in Lake Michigan [Increments of growth in parentheses] Locality ot capture and group o( lake trout Mesh of nets (inches) Number offish Calculated lengths at year of life .\verage incre- 0 1 2 3 4 5 6 ment of growth .\reas 1-6: Marked 1 Hi 565 3.8 5.9 9. 1 (3.2) 8.7 (2,8) 0.4 8,6 (3 0) 11.7 (2.6) 11.3 14.2 (2. 5) I.t 6 16,9 (2.7) 16.2 (2.6) 0.7 15.8 (2,5) 14.7 (2.1) 1.1 13.2 (2.1) 1.5 19.8 (2,9) 18.6 (2.4) 1.2 17,8 (2.0) 17.0 (2.3) 0.8 14.8 (1,6) 2,2 Unmarked 1 4K 29 3.6 5.9 (2. 6) (2 .11 Diflerences in calculated lengths 0.2 3.7 0.0 5.6 0,4 10,9 <■> :il 0.6 13.3 (2.4) 12.6 (2.4) 0.7 11.1 (2 01 .\reas 1-6: Marked _ f 2H 738 / 2M 38 3.2 5.2 7. 8 10. 2 (2 6) '2 41 (2. 4) Differences in calculated lengths 0.5 2.5 0.4 4.6 0.8 6,9 (2.3) 0.9 0.7 9.1 (2,2) Area 8: 1 2H 76 (2. 6) Differences in calculated lengths of unmarked fish: Areas 1-6 and area S 2H 0.7 0.6 11 IS 46 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE The discrepancies miglit have been explained on the basis of annual fluctuations of growth, had not the calculated lengths from fish of all the age groups within each year class of unmarked lake trout varied consistently about a lower mean than those of the marked fish. Evidently, the larger size of marked lake trout over unmarked fish of the same year class is a real rather than apparent difference, which suggests that the marked fish may have derived a certain advantage from the hatchery environment during their first summer that carried over into later life. The lake trout of year classes 1944-46 from area 8, caught in small-mesh nets, were decidely smaller than the northern wild stock caught in these nets. The average calculated length of the southern fish at the first annidus was only 4.6 inches and the average annual increase in length to the sixth year of life was 2.0 inches (table 22; fig. 23) compared with a calculated length of 5.2 inches at the first annulus and an annual increase of 2.4 inches for the northern fish. A major difference between samples of marked lake trout from 2}^- and 4/2-inch-mesh nets was the near absence of Lee's phenomenon in the data for the fish taken by the latter gear (fig. 24). These fish were subject to little or no selectivity from the nets, for few of tlie marked lake trout grew large enough to exceed the catching potential of the large-mesh nets. Another factor in bringing about apparent decline, even cessation, of the growth of marked lake trout with increase in age is believed to be destruction by sea lampreys of the most rapidly growing fish. Lengths, at capture, of marked lake trout in age-group V were little greater than those of fish a year younger; and lengths of fish in age- group VI were actually smaller than those in age-group V. A high percentage of the larger specimens in age-groups V and VI (28 percent of those caught in 1951) bore scars or open wounds made by lampreys. Smaller fish were unscarred; hence it is thought that lamprey predation is most severe among larger lake trout 14 or more inches long. It is possible, nevertheless, that small lake trout which have been attacked by lampreys die immediately so they do not come into the nets with wounds as do the larger fish. Hall and Eliott (1954) found also an increase of scarring with increase in length of the fish for the white sucker (Catostoinus commersoni) . They showed that incidence of scarring was consistently greater among suckers more than 10 inches long than among smaller fish and near 100 percent for fish 19 to 20 inches long. Thus the larger fish of the younger age groups and nearly all in the older age groups were being eliminated leaving only small, slow growing individuals. Wild and hatchery lake trout of the same year classes were subject to the same selectivity by the nets and the same predation by lampreys. The marked lake trout and the wild stock of year classes 1944-46 were comparatively free from atta(!ks by lampreys until they were about 14 Figure 23. — (.'alculated lengths (sums of mean increments of growtli in inclies) of marked and unmarlsed lake trout (year classes 1944-46) caught in 2'i-inch-mesh nets. AGE DETERMINATION FROM SCALES OF LAKE TROUT 47 19 - ■ 18 - 17 - * 16 - ■ (^ 15 _ UJ I • 1 14 _ * ■ X 13 — ■J) ii_ 12 ' 8 o A X " ■ (- ta z 10 — lij —1 M ^ 9 ~ • < ♦ t- O 8 _ ■ 1- '=' 7 UJ 1 _ t < 4 6 — A. (_J _/ % — ■ 4 - ♦ ■ 3 2 1 - 1 1 1 1 " I 2 3 4 5 6 YEAR OF LIFE FiGiRE 24. — Calcuhited lengths of age groups of marked lake trout caught in -iH-inch-mesh nets (year classes combined). Symbols: diamonds, age-grovip II; dumb-bells, age-grovip III; dots, age-group IV: triangles, age-group V; squares age-group VI. inclies long, during tlieir fourth year of life. Fish from earlier year classes were suhject to (lie selectivity of largc-mesh nets for a longer period of time than the marked fish, but to a lower level of lamprey infestation because they were caught before the lampreys had made apprecial)le inroads into the lake-trout population in their areas of tlie lake. The best estimates available of the growth of lake trout in the lake before sea lampreys enteied it in large numbers are from data provided by the wild stock from the earlier year classes. In the northern part of the lake, areas 4, .5, and (1, sixteen individuals of year classes 1939-48 were caught by large-mesh nets in 1947. These fish were considerably larger at each year of life than the surviving fish of year classes 1944-46 caught in the same nets from 1947 to 1951. The average calculated length of the lake trout in the earlier year classes at formation of the first annulus was ().9 inches and the average annual increase in lengtli to the sixth annulus was ,S.O inclies com- pared witli .5.9 inches at the first annulus and an annual gain of 2.8 incites for the marked fish of year classes 1944-46 (tables 22 and 2.3, fig. 22). The early year classes of lake trout tliat lived in southern I^ake Michigan were rei)resented by two groups, both captured in 1947 i)y large-mesh nets. The larger sample of 97 fish (82 of which were of year classes 1939-43) was taken in area 7 off Montague, Michigan (Van Oosten 19.")0). The otiier group contained 17 lake trout of tlie same year classes from the collections of fish witli deformed fins caught in area 8 off South Haven, Mich. The average calculated lengtlis of the Iwo groups differed little from the second to the 48 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 23. — Calculated lengths {sums of mean increments of groivth in inches) of unmarked lake trout caught by 4V2-inch-mesh gill nets in 1947 in Lake Michigan [Year classes 1939-43 combined] Number offish Calculated lengths at years of age Average in- crement 01 Locality of capture 0 • 2 3 4 5 6 7 8 9 growth to sixth annulus Areas 4-6-. --- f 16 4.0 6.9 10.0 (3.1) 8.7 (3.6) 9.0 (2.9) 8.8 (2.7) 12.8 (2.8) 12.1 (3.4) 12.1 (3.1) 12.1 (3.3) 16.1 (3.3) l.'i.S (3.2) 15.4 (3.3) 16.3 (3.2) 19.0 (2.9) 18.3 (3. 0) 18.7 (3.3) 18.4 (3.1) 21.8 (2.8) 20.9 (2.6) 21.2 (2. 6) 21.0 (2.6) 24.2 26.4 28.7 } 3.0 Area 7: ) Off Montague, Mich i 82 6.1 23.1 24.7 1 3. 1 Area 8: f Ofl South Haven, Mich l 17 3.6 6.1 23.6 23.2 26.3 '25.0 I 3.0 Areas 7 and 8 combined (from 3rd to 8th years){ 99 j 3.0 seventh aiimilus. Differences at the eighth annu- lus were due to tiie small number of measurements (6 for area 7 and only 1 for area 8). As explained earlier, the calculated length at the first annulus for lake trout caught in area 7 (as published), was not based on the same criteria as the data on the O-mark and the first annulus treated in this paper. For this reason, the calculated lengths of the fish from areas 7 and 8 were combined only from the second to the eighth annulus. The mean calculated length at the first annulus of the fish from area 8 was 6.1 inches and the average annual increase in length of the combined groups was 3.0 inches (table 23). Comparison cannot be made of these figures with like figures from lake trout of year classes 1944-46 from the southern areas of the lake as none of those fish were caught in large-mesh nets. The calculated lengths for the early year classes, however, were very much larger than those for the fish of year classes 1944-46 caught in small- mesh nets (table 22). Selective destruction of the more rapidly grow- ing individuals by sea lampreys and by nets of the commercial fishery leads to a decrease of gi'owth rate with increase of age which would not exist within a stock not subject to such selective mor- tality. It is a natural consequence of continued selective destruction of large fish, that each older age group should be composed of slower-growing fish than the younger age groups. Because the combined eft'ects of biased samplitig and selective destruction of the marked lake trout by lampreys cannot be measured, it must be recognized that tiie "normal" growtli of lake trout in Lake Michigan probably was not determined precisely. However, tlie use of summations of the mean increments of growth in length to describe general growth tends to lessen the effects of selective mortalitv and thus to vield curves more representative of the true rate of growth than otherwise could be obtained from these data. A third cause for discrepancies in estimates of growth of lake trout in Lake Michigan, not, however, affecting area estimates, is geographic differences in size and growth. Lake trout in- habiting the northern part of the lake were larger at each year of life than those in the southern part of the lake. Tliis difference in size is ap- parent in comparisons of fish in the same year classes caught in nets of the same mesh size. Examples: the early year classes (1939-43) caught in 4}^inch-mesh nets (table 23, fig. 25), and the later year classes (1944-46) caught in 2)^inch- mesh nets (table 22, fig. 23). For tliese and other groups of lake trout from the two parts of the lake, the differences appear to stem principally from a slower growth of the southern fish during their first summer to formation of the first annulus. The southern fish of the early year classes caught in the large-mesh nets, at formation of the first annulus, were 0.8 inch shorter than a similar group of the more northern fish, but the average annual increases in length in later years were identical. Those of the more recent year classes caught in 2K-inch-mesh nets, at formation of the first annulus were 0.6 inch shorter and the average annual increases in length were 0.4 inch less than the annual gains of the unmarked northern fish of the same year classes. The consistency of the discrepancies between the calculated lengths of lake trout from southern and northern Lake Michigan indicates that they represent a true geograpliical difl'erence of growth between the two populations. GROWTH IN WEIGHT Weights were available for only 1,118 of the 1,319 marked lake trout, but these were sufficient AGE DETERMINATION FROM SCALES OF LAKE TROUT 49 4 5 6 YEAR OF LIFE Figure 25. — Calculated lengths (sums of mean increments of growth in inches) of unmarked lake trout caught in Lake Michigan by 452-inch-mesh nets in 1947. Year classes 1939-43 of the fish from areas 4-6 were combined as were those from areas 7 and 8. for determination of mean weights at capture of the fisli in each age group represented. Further information on growth in weight was obtained by calcuhitiiig weights corresponding to calcuhited lengths at the end of the several years of life and and at tlie time the 0-mark was formed. These calculated weights were computed by the length- weight ecjuation. Weights at Capture The range of weight in all age groups of the marked lake trout was large, as would be expected from fish that differed so greatly in length. Both the average weights and the ranges of weight of the different age groups are presented in table 24. In 9 of the 12 age groups for which data are given in the body of the table, the weight of the heaviest fish was more than 5 times that of the lightest (the advantage was more than 10-fold in age- group IV of the 1945 year class). In the remain- ing 3 age groups the heaviest trout weighed 2.0 to 4.5 times as much as the liglitest. Despite the great variability in weight, the mean weights of certain age groups of tlH> different year classes were similar. The average weight ranged from 4.3 to 4.4 ounces in age-group II, from 6.6 to 9.7 ounces in age-group III, and from 11.2 lo 15.2 ounces in age-group IV. The range of the mean weights was somewhat larger in age-group V (15.2 to 27.4 ounces). Comparabl(> data on the weiglits of the fish in age-gi'oup VI were not available. Table 24. — Mean weight (ounces), at time of capture, of the year classes of marked lake trout, by age groups Item Age group II III IV V VI 1944 year class: 7.0 3.5-32.0 10 9.7 3. 9-22. 4 174 6.6 3. 7-21. 5 38 9.0 3.5-,12.0 222 15.2 9. 1-46. 0 9 11.2 4. 7-47. 8 523 14.0 6.0-41.0 116 11.8 4. 7-47. 8 648 15.2 10.5-21.2 11 20.5 5. 1-48. 8 184 27.4 8.2-50.2 12 20.6 5. l-.5fl. 2 207 1945 yeur class: Sle:in weight Range Xumber offish 1946 year cl;iss: .Mean weight Range Xumber offish Tomhined year class: Mean weight 4.4 1.2-8.3 28 4.3 1.9-8.5 8 4.4 1. 2-8. 5 3fi 26.9 11.5-36.5 5 26.9 11,5-36.5 Xumber offish 5 Calculated Weights The growth in weight of the marked lake trout (as determined by the length-weight equation from the calculated lengths shown in table 20) was slower in the earlier than in the later years of life. Wiiereas the most rapid growtli in lengtli occurred during the first year, growtli in weight proceeded slowly through the second year. Tiie weights calculated for the first year of life were typically less than 1 ounce and averaged only 3.0 ounces at the end of tlie second vear. The annual 50 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 25. — Weights of the marked lake troui at capture and as ralriilated for the end of each year of life ' [Weight in ounces] Year planted Number o specimens Weight at capture Calculated weight at end of year of life Age group measured weighed 0 1 2 3 4 5 6 31 8 28 8 4.4 4 3 4.4 7.0 9.7 6.6 9.0 15.2 11.2 14.0 11. S 15.2 20.5 27.4 20.6 26.9 0.15 .18 .17 .18 .23 .20 .22 .18 .22 .20 .20 .17 .20 .15 .20 .07 0.57 1.10 .66 1.16 .85 .85 .90 .85 .94 .85 .90 .76 .72 .76 .72 .34 2.6 3.0 2.6 3.6 3.4 3.0 3.4 3.2 3.0 2.8 2.8 2.2 2.6 2.3 2.5 1.2 jj . 1946 16 190 49 10 174 38 8.2 7.4 5.7 7.2 7.6 6.6 6.1 6.4 5.0 5.6 4.7 5.4 2.8 1945 III 1944 10 555 167 9 523 116 13.7 11.7 12.5 12,0 9.2 10,4 10.4 10,2 5.4 IV 1944 17 240 23 U 184 12 14.9 17.9 19.0 17.5 9.5 V --- VI .. 1945 13 5 16.2 Number of specimens. Mean increment of gro\ Weight from summatio Mean calculated weigh Increments of mean we 1,319 0.21 .21 .21 .21 1,319 0.65 .86 .86 .65 1.319 2 1 3.0 2.9 2.0 1,280 3.4 6.4 6.4 3.5 1,025 5.2 11.6 11.3 4.9 293 7.3 18.9 17 4 6.1 13 6.7 25.6 16,2 ght -1.2 I Weights calculated with length-weight formula from calculated lengths shown in table 20. 3, ' ^ I UJ 5 I 14 2 3 YEARS OF AGE Figure 26. — Calculated growth in length and weight of marked lake trout [summation of ann\ial increments of growth]. addition of weight increased sharply from 2.1 ounces in the second year to 3.4, 5.2, 7.3, and 6.7 ounces in succeeding years. The calculated incre- ments of weight of fish in the older age groups (especially age-group VI) would have been larger except for selective mortality of the more rapidly growing lake trout which resulted in reduction of the average length increment (table 25, and fig. 26). Calculated weights obtained by summation of the mean increments of growth in weight were slightly smaller than weights of the fish at capture for tlie same reason that the calculated lengths were smaller than the measured lengtlis. The differences in weight ranged between 0.2 and 2.6 ounces as shown in the following tabulation: Year of life 2 3 4 5 6 Weight from summa- tion of calculated increments 3.0 6.4 11.6 18.9 25.6 Age group II III IV V VI Weight at capture 4.4 9.0 11.8 20.6 26.9 PROGRESS OF SEASON'S GROWTH As a first approach to the estimation of the progress of the growth of lake trout during the growing season, tabulations were prepared of the sizes attained by the age groups of the marked fish at capture in each month of the year. AGE DETERMINATION FROM SCALES OF LAKE TROUT 51 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I MJ JASONDJFMAMJJ ASONDJ FMAMJ JASONDJ FMAMJJASO n m EZ ¥ AGE GROUP Fir.iRE 27. -Mean length? and mean weights of tlie marked lake trout at time of capture. Year classes 1944-4fi com- bined. [Curves drawn by inspection.) The average lengths of lake trout of the 1944-46 year classes of the same age group were originally tahulated by semimonthly periods, but as division of the data into shorter time intervals did not provide additional information, the averages of table 26 (see also fig. 27) were based on monthly groupings. Altliough the month-to-month changes in the average lengths of the age groups were de- cidedly irregular, the figures do give the general impression that much of the increase in length took place in the late summer and fall. In other words, rapid growth seems to have started about the end of June and to have continued at least through October, possibly longer. The records of average weight of the age groups at capture support a similar interpretation (table 27, fig. 27). T.\Bi.E 26. — Average lengths (inches) of marked lake trout at time of capture (Data for 1944, 1945, and 1940 year classes combined.' Number of specimens in parentheses] .lanuary February March April May June.- , July .\ugust .. September- October November. December-. Mean length.. Age group 8.0 (1) 8.5 (3) 9.8 C4) 9.7 (5) 9.8 (12) 9.9 (7) III 11.5 (7) 10.0 (39) 15.4 (1) 9.9 (1) 12.0 (2) 11.6 (8) 11.8 (22) 11.9 (36) 12.7 (30) 12.9 (64) 13.5 (72) 11.9 (7) 12.4 (8) 14.1 (4) IV 13.8 (27) 14.6 (13) 14.1 (17) 14.9 (80) 13.9 (156) 13.9 (120) 14.0 (183) 14.8 (109) 1.^3 (13) 15.9 (6) 15.6 (4) 15.0 (4) 12.8 (255) 14.3 (732) 15.0 (24) 15.2 (6) 16.3 (5) 15.7 (24) 15.8 (81) 15.8 (6(1) 16.7 (34) 17.6 (33) 16.0 (12) 13.1 (1) 15.8 (280) I Lengths of the 13 fish in age-group VI omitted because the data arc too scattered to be of value in this table. 52 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 27. — Average weight (ounces) of marked lake trout at time of capture (Data for 1944, 1945, anil 1946 year clas,ses combined in parentheses] ' Number of specimens Age group II III IV V 6.9 (1) 3.7 (1) 8.4 (2) 7.7 (1) 7.4 (20) 8.2 (28) 9.9 (26) 9.6 (57) 11.3 (69) 9.6 (5) 11.9 (9) 12.6 (4) 11.0 (19) 15.2 (13) 13.1 (16) 15.0 (66) 12.7 (1.65) 13.0 (80) 12.9 (172) 16.5 (102) 14.8 (12) 19.4 (6) 14.7 (4) 19.6 (4) 18.1 (18) 17.1 (4) March 17.6 (4) April 18.9 (19) May i.9 (i) 2.6 (3) 4.5 (4) 4.1 (3) 4.3 (12) 4. 2 (6) 19.7 (65) June July \ugust 22.8 (38) 23.6 (28) 30.2 (24) September 25.0 (7) 6,8 (7) Mean weight . _ - 4.6 (36) 9.9 (222) 13.6 (648) 21.9 (207) I Weights of the 13 fish in age-group VI omitted because only 5 fish were weighed and the data are too scattered to he of value in this table. More dependable data on the progress of the season's growth may be obtained by romputation of growtli from scale measurements. Examples of the distribution of these increments are contained in the records for the 555 lake trout of age-group IV from the 1945 3'ear class, the largest year class in the collections. Their increments of growth in length were computed by semimonthly periods (table 28). The amount of growth attained by individual lake trout in any stated time varied widely. By the end of April, the range in the amount of seasonal increment of growth in length was from nil to 1.4 inches. This range continued nearly constant and the mean advanced only slightly (0.14 to 0.40 inch) until the middle of July." In the latter part of July the range in lengtlis of the increments began to broaden and by the end of August the spread was 2.8 inches. In the fore part of August, some lake trout were still just beginning to grow whereas others had been grow- ing since the middle of March or possibly even longer. It was largely because of this wide spread in the time of the onset of growth that the average increment was still only 0.22 inch in the first half of June. Subsequent more rapid in- crease carried the average to 1.7 inches in the first half of September. Returns of lake trout were so sparse during the remainder of the year that dependable estimates of growth cannot be » iMp,v growth cannot be recognized on the scales until the first circulus haS been formed, a circumstance which probably accounts for the small propor- tion, at any time, of fish having as little as 0.2 inch caclulated growth. The smallest calculated length increment is more often 0.4 inch. Hence the fish usually had grown nearly \i inch by the time the annulus could be read with confidence. made from them. It is especially difficult to form a judgment as to the time the season's growth ends. It appears from the data in table 29 that the growth of the fish in age-group IV had not been completed by the end of December, when the average increment (4 fish) was 1.95 inches or 0.64 inch below the figure of 2.59 inches computed for the full season from age-group V of the same year class. (The fish in age-group V that had not yet completed the fifth annulus gave nearly the same estimate of growtli in the fourth year, 2.60 inches, as did those on whose scales the fifth annulus was visible, 2.58 inches.) Records of tlie percentage of the season's growth completed by age groups of the 1945 year class up to various dates of capture, despite gaps in the data and the siaall numbers of fish on which certain percentages were based, give evidence of annual differences in tlie progress of growth and of irregular growtli in some years (table 29). These points are well illustrated by the curves in figure 28 which were fitted by inspection to the empirical data. The data were scanty for the lake trout of the 1945 year class in age-group II. The single trout captured in the first half of June had made no growtli. Percentages of growth completed by fish caught later in the season rose quickly to 51 in early August but fluctuated erratically thereafter. Seven fish recovered in December had grown more (percentage, 115) than the "expected" increment for the full season calculated from measurements of the fish in age-group 111. The 4 lake trout of age-group 111 caught in late April and early May 1948 exhibited no new growth, but those captured during the last half of May had completed 7 percent of the expected growth for the season. The percentage dropped ill early June, but thereafter it increased steadily (except in the first half of September) to 94 per- cent in early October. The single trout caught in December had gained only 79 percent of the expected total increase. Age-group IV, captured in 1949, seems to have started growing early in the season. Possibly the single lake trout with new growth in January could be dismissed as aberrant, but all semimonthly collections from the latter half of March onward contained some fish that had begun to grow. The advantage of this early start was later lost, how- ever, for the percentage of new growth remained AGE DETERMINATION FROM SCALES OF LAKE TROUT 53 ■PS p a n li is £| o ■^re^/;x^l^|:-^^c §3! 1 I I I l.-HNr-accoiCtr^t^r-ociiJC*':— " \ ;— ■ E S 1 :^-H ;« : ; .^ ^g « :::;:;::;:; 1 J2 B O 1 -" i'""^ i ; ; ; "S 1 ! ! : i I ; i i ; i"' ; i i x: o O 1 : i : : ; : i i'" ; : i^ ; 1 ';;::;; ^ '> CO '•'<'•'• 'i 'fe B a m ; i'^ i : i : : i i i 1 ; ! M! i i|-- : ; i : i T"^-^^ i i i M i : i ; ^^ s ■< 1 1 i^N^^ eo«'^N I i • ■ ; i" ; ; : i i i i ; ss iC _< itCO-*^ t*40 Tj-r-.-. i 1 . ;;;;;;!;!:; 1 (O> ;;;;;;;;;;; S2 1 I ; 1 1 ! I 1 j ! j d 1 g i- i 1 I • d lO N ■ ■ ■ ' NO d >. C 03 •-9 1 »o .—1 . . d 1 "* I 1 I ' d at _^ o c E £ O c O N -* *D 00 C o o d o" o - 00 d N ■* «D CO q e* W « ao "5 ; 1 : E ; > "CO J5S 1-^ ^E & . § X o (U F ox: s TJ a F > a a R c Sr * o £ •o a S i" J= = » F te a V) £ 3-S e S L i" - u. 54 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 29. — Average percentage of season's growth completed at various dates during the season hy age groups of the marked lake trout from the 1945 year class [The bases for the percentages are the increments of growth for full seasons computed from measurements of the scales of fish of the next higher ape groups of the same year class] Age-group II Age-group III Age-group IV Age-group V Unweighted Date Number offish Average percentage Number of fish Average percentage Number offish Average percentage Number offish Average percentage average ' percentage 4 6 2 9 0.0 2.3 .0 .0 11 11 3 0.0 .0 .0 (1.0 16-31 - -- 1,2 Feb 1 15 .0 16-28 Mar 1-15 -- 16-31 16 21 40 79 31 36 47 100 57 71 17 4 1 4 2 1 3 2.3 1.2 5.4 5.0 5.0 8.5 13.5 15.4 29.0 40.6 53 7 66.8 63.7 60.6 80.7 83.4 62.6 Apr 1-15 1 18 48 22 32 27 14 11 24 5 8 4 .0 9.4 8.5 13.4 fi.7 8.0 30.8 54.9 65.2 88.9 70.5 92.4 .6 16-30 i 3 17 14 14 9 12 21 25 42 29 2 0.0 .0 7,2 3 8 19. 1 25.0 43.2 53.4 65.7 64.8 72.9 94.1 4.9 May 1-15 -- 4,5 16-31 8.5 June 1-15 --- 1 1 2 0.0 13.3 17.2 4,8 16-30 ... ._ 13,5 July 1-15 - 22.1 16-31 42.4 Aug 1-15 4 1 1 7 7 50.8 41.8 16.4 50.4 54.7 52.5 16-31 62.5 Sept. 1-15 54.6 16-30 69.8 Oct 1-15 69.8 16-31 Nov 1 15 16-30 Dec 1 15 7 114.9 i 79.3 4 75.3 1 58.6 81.9 2.56 2 36 2.59 2 24 31 190 555 240 ' In order that age-group IV. which was represented in nearly all semimonthly periods, would not exert undue influence on the trend oidy those periods which were represented hy one or more other age groups are included. 2 Based on the 13 fish in age-group VI. at 5 from mid-May through June. Beginning with the first lialf of July, the percentages were consistently smaller than those for age-group V and, with one e.xception (early September), were also below the percentages for age-group III. The erratic variation of the percentage of com- pleted growth for age-group IV during September- December can be attributed partially to the small numbers of fish in the samples, but the generally low level (61 to 83 percent; 75 percent for 4 lake trout caught in late December) is further evidence tliat the seasonal growtli was not completed at the end of the calendar year. As the average increment of growth of the 4 lake trout caught the last part of December was only 2.0 inches, the actual amount of growtli between December 31 (ages change on January 1) and the completed growth of 2.6 inches at formation of the fifth annulus was 0.6 inch. It was pointed out earlier that the average estimate of the growth of age- group IV for the entire season, calculated from measurements of the scales of lake trout in age- group V taken in 1950, was the same for fish without the fifth annulus as it was for those that had that annulus visible. However, the average increment for 11 lake trout caught the first half of January, which did not have new gi-owth on their scales, was only 2.3 inches (88.5 percent of the total increment), whereas the average incre- ment for an ecjual number of lake trout caught the last part of the month was 2.59 inches (100 percent of total increment). Increments for 4 fish taken between the first of February and the 15th of April were low (2.13 inches), but the average increment for 12 fish taken the first half of May (2.63 inches) showed a slight rise over that for January. These few fish, caught January to May, do not furnish definitely reliable information on the end of tlie growing season, however they do indicate that in certain years lake trout may continue to grow tlirough the winter months. The 25 lake trout of age group V taken in Jan- uary and February and a single fish caught in early April had not started to grow. The incre- ment of new growth on the scales of lak-e trout captured in the last half of April amounted to 9 percent of the expected total increment, but this percentage showed no clear tendency to increase during May and June. A sharp upturn, beginning in July, however, carried the percentage to 92 in late September (with a single exception to the trend in the first half of the month). The single AGE DETERMIXATIOX FROM SCALES OF LAKE TROUT 00 100 - 90 80 5 O 70 a. 60 50 40 30 20 I 0 -o--o-^ JAN FEB MAR APR MAY JUNE V -^ /^ >-^- A _, _^..-'" jC^ in ^^ 12 o ■'■^ / / /> ° /■ / / /x / /x / ^ / / < / • / y A / / • / X / ^ X / k A 4 .'' / / ,-" / X / X / o/' O // A > trout of ajio-tiroup V caught in Doccmhor, however, liati compUUed only 58 percent of the cxi)ecte(l total growth. A start of growth followed by a stoppage or near-stoppage as demonstrated for age-group I\' in the last half of April through May, and for age-groii]) V in May and June, might he exiiecled to produce irregularities in the scale structure. Nevertheless, examination of the scales of lake trout of age-groups IV and V captured late in the growing season revealed no checks or marks that could be attributed to this stoppage. Some of the irregtdarities in the data of table 20 can be attributed to the inadequacy of the samples, but the majority give evidence that the coin-se of the season's growth v'aries considerably from one calendar year to another. (There is no evidence of a ju'ogressive change witli age), '{"his year-to- year \!iri(itioii iind the uncerlaiiit v as lo the tinu' growth ends (data were conflicting even among the best represented age groups) prohibit a general description of seasonal growth of the marked lake trout. Growth may start as early as March or as late as June. Once started, growth may follow a regular course; but in sonu' years it may stop, completely or nearly so, for a period of several weeks. The end of the season as well as the start probably varies from year to year. In some seasons growth may continue into the next cal- endar year. Because of the variation in the start and finish of the growing season, growth of lake trout in Lake Michigan is likely to occiu- in at least 9 or 10 months of the year, i)ossibly in even more. The most rapid growth, nevertheless, ap- l)ears normally to take place in Jtdy and August. The percentages at the right of table 29 indicate that nearly half of the total season's growth occurs in these months. The same set of figtn-es shows 56 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE that, in general, the lake trout gained about 30 percent of their growth sometime after the middle of September. The growing season for lake trout in Canadian waters is shorter. Kennedy (1954) found that the lake trout in Great Slave Lake "grow only between late May and the middle of September, with no growth at any other time." Of the seasonal growth of the lake trout in South Bay, Lake Huron, Fry (1953) stated, "The lake trout . . . add about 1 inch to their total growth increment for the year by mid-September. The total for the major year class represented in 1949 (the 1944 year class) . . . was estimated at 1.8 inches. This increment would indicate the rapid growth observed from June to September probably continued at least until mid-October." SUMMARY From 1 to 1)2 million liatchery -reared lake trout (average length 3.2 inches) were liberated into northeastern Lake Michigan in September of each of the years 1944-46. About 10 percent of these fingerlings were marked by the removal of fins. In the years subsequent to the plantings, 1947-52, fishermen captured 1,747 lake trout with abnormal fins of whicli only 1,507 were adequately documented. Of the latter group, 102 caught off South Haven, Mich., difTered so much from those caught in the northern part of the lake that all, or nearly all, were considered to be unmarked wild lake trout with abnormal fins; hence, they were excluded from the main sample. The scales of the remaining 1,405 fisli were studied to determine the validity of age readings from scales and the rate of growth of lake trout in Lake Michigan. Lake trout scales are small and have concentric circuli. They develop first as platelets adjacent to the anterior end of the lateral line when the fish are about 2 inches long and rapidly cover all the body except the head. Probably joung lake trout in Lake Michigan are fully scaled before the end of their first summer. Even though the scales were rather difficult to interpret, simple criteria for recognition of the annulus were determined. The annulus is gcn- erallj' indicated by wider spacing between circuli outside closely spaced circuli, but this arrange- ment, usually most clearly seen in the lateral fields, is seldom definite enough to be followed entirely around the scale. Other indications of an annulus are: a V-shaped pattern in the circuli of the lateral fields, a ridge across the posterior field, also such irregularities as broken or crooked circuli and fine accessory lines. An annidus is usually located by a combination of these criteria. The annulus was formed on the scales of some lake trout as early as the middle of March, of the majority during June and July, and of a few as late as the middle of August. In addition to the expected number of annuli for the marked fisli, a central check was found within the first annulus which has been designated the "0-mark." The scales of the unmarked, wild-stock lake trout from Lake Miciiigan exam- ined during this study also carried the central check (0-mark). Two readings were made of the markings on the scales. The ages read agreed on 96.8 percent of the specimens. The number of annuli read from the scales agreed with the age of the fish indicated by the deformed fin for 93.9 percent of the lake trout in the sample of presumably marked fish. Most of the disagreements were of 1 year but some were of 2 or more years. The principal difficulty in the way of determin- ing the accuracy of age readings from the scales of the lake trout from northern Lake Michigan resulted from the presence in the collections of a small percentage of unmarked fish. The exact number of these fish could not be comited but evidence from several lines of investigation led to the conclusion that nearly all the 86 fish, for which the age read from the scales disagreed with that indicated by the deformed fin, were unmarked lake trout. The average lengths of the age groups indicated by the deformed fins of the 86 "uimiarked" fish were very different from those of the age groups of the 1,319 "marked" fish (those with agreement between age indicated by the fin and that read from the scales); furthermore, the average length of the 86 fish decreased with increase of age. On the other hand, at ages read from the scales, the growth curve for these 86 fish was similar to that of the 1,319 "l)ona fide" recoveries. It was concluded, therefore, that the age read from the scales rather than tlu' age indicated by the deformed fin was correct for most fish. Tlu» evidence strongly indicates a liigli depend- AGE DETERMINATION FROM SCALES OF LAKE TROUT 57 ability of ago readings from lake trout scales. The reader does, nevertheless, need considerable experience with scales from fish of known age to become proficient in recognition of the 0-mark and annuli. The estimate obtained of the relation between weight in ounces and total length of the fish is expressed by the formula: log Tr= -2.4698+3.1 125 log L The range of total lengths at capture of fish within an age group of marked lake trout was wide. The average length for an age group of one year class, however, was close to those for the same age group of the other two year classes. Lake- and pond-reared fish had attained about the same lengths at 2, 3, and 4 years of age. The calculated lengths of the fish at various ages prior to capture were computed by direct proportion from the diameters of the annuli. The calculations from 2 scales were averaged. The calculated lengths (sums of the mean increments of growth in length) being lengths of the fish at the end of growing seasons were, as would be expected, somewhat smaller than the mean lengths of the fish of the same age groups at time of capture which was, in most cases, after the beginning of a new growing season. The lengths calculated from the fish in age- groups III-VI exhibited Lee's phenomenon of gradually decreasing values with increasing age. Most of the discrepancies are explained by selec- tive destruction of the most rapidly growing fish by nets and sea lampreys. Scars and open wounds made Vjy lampreys were found more often on large than on small lake trout. The destruction of the large, fast-growing fish could account for the small size of the fish remain- ing in the older age groups which were caught after the population had been materially reduced. Gill nets of the two sizes of mesh most com- moidy used in Lake Michigan caught lake trout of greatly different sizes. During tlie years marked lake trout were caught, the fishermen gradually shifted from use of large- to small-mesh nets. The large-mesh nets caught larger fish than the small- mesh nets and the difference became greater as the fish grew older. It is questionable, therefore, whether a general average gives a true estimate of the growth of these lake trout. The fish caught in the small-mesh nets may give the better esti- mate of the growth of the younger age groups, whereas those caught in the large-mesh nets may be more representative of the older age groups. Lee's phenomenon, prominent iti measurements of the first group, is almost lacking from the meas- urements of the fish in the latter group. Summing the increments of growth in length minimizes the efl'ects of biased sampling and selec- tive destruction of the fish. The weights of the marked lake trout were similar to the lengths in that the weights of in- dividual fish at capture varied greatly within age groups and the mean weights for the age groups at capture were slightly larger than the calculated weights. Although the most rapid gain in length occurred during the first year of life, the gain in weight was least in this year and much greater in later years. Seasonal growth of the marked lake trout re- flected the long period of annulus formation. The growing season was extended and variable. Growth for the three year classes indicated a long period of slow growth in the spring, rapid growth from the end of June through October, and slower growth again on into December. Monthly distribution of the increments of growth in length of the 1945 year class suggested that lake trout may occasionally have a somewhat longer season of growth. The average percentage of growth completed at semimonthly intervals for the sepa- rate age groups showed that the growing season varied consideraVjly from one year to the next. Not only the time of the beginning but also of the end of the growing season may vary several weeks, even months. Because of this lack of uniformity in the time of start and finish, growth of lake trout in Lake Michigan may be expected to take place in 9 or 10 months of the year. As large- and small-mesh nets caught fish of different sizes and the destructiveness of the sea lampreys increased during the years the marked lake trout were in the lake, it was necessary for estimation of the growth in length, to select fish of the same year classes caught in the same calen- dar years by nets with mesh of the same size. The marked fish (year classes 1944-46) caught in large-mesh nets were slightly larger than the un- marked fish also caught in the northern part of the lake, which suggests that the marked lake trout gained some small advantage from early care in the hatchery. Lake trout caught in large-mesh 58 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE nets in northern and southern areas of the lake could not be compared because no fish of year classes 1944-46 were caught in large-mesh nets. Those caught in small-mesh nets were consider- ably smaller than both marked and unmarked lake trout caught in these nets in northern waters. Lake trout that had lived in Lake Michigan before the sea lampreys became numerous were larger and had grown at a faster rate than the marked fish. Two samples of lake trout of these early year classes from the southern part of the lake, caught in 1947 by large-mesh nets were so similar they are believed to have been drawn from the same population, but one that differed from the northern population by an important char- acteristic. Growth during the first summer to formation of the 0-mark was much less than for fish in the more northern waters. Subsequent annual growth in length was the same in both areas. LITERATURE CITED Applegate, Vernon C. 1951. The sea lamprey in the Great Lakes. Scien- tific Monthly, vol. 72, pp. 275-281. Applegate, Vernon C, Bernard R. Smith, and Willis L. Neilsen. 1952. Use of electricity in the control of sea lam- preys: Electromechanical weirs and traps and electrical barriers. U. S. Fish and Wildlife Service, Spec. Sci. Rept.: Fish. No. 92, 52 pp. Applegate, Vernon C, and James W. Moffett. 1955. The Sea Lamprey. Sci. Amer., vol. 192, No. 4, pp. 36-41. Bahtlett, M. S. 1949. Fitting a straight line when both variables are subject to error. Biometrics, vol. 5, pp. 207-212. Brown, C. L D., and Jack E. Bailey. 1952. Time and pattern of scale formation in Yel- lowstone cutthroat trout Salmo clarkii Lewisii. Trans. Amer. Micro. Soc, vol. 71, No. 2, April, pp. 120-124. Butler, Robert L., and Lloyd L. Smith, Jr. 1953. A method for cellulose acetate impressions of fish scales with a measurement of its relia- bility. Prog. Fish-Culture, vol. 15, pp. 175- 178. Cole, Leon J. 1905. The German carp in the United States. Rept. U. S. Comm. Fish. 1904, pp. 523-641. Cooper, Edwin L. 1951. Validation of the use of scales of brook trout, Salvelinus fontinalis, for age determination. Copeia, 1951, No. 2, pp. 141-148. Cooper, Gerald P., and John L. Fuller. 1945. A biological survey of Moosehead Lake and Haymock Lake, Maine. Maine Dept. In- land Fisheries and Game, Fish. Survey. Rept. No. 6, 160 pp. Deason, Hilary J., and Ralph Hile. 1947. Age and growth of the kiyi, Leiirirhthi/f; kiyi (Koelz), in Lake Michigan. Trans, .\nier. Fish. Soc, vol. 74 (1944). Eschmeyer, Paul H., Russell Daly, and Leo F. Erkkila. 1953. The movement of tagged lake trout in Lake Superior. Trans. Amer. Fish. Soc, vol. 82, pp. 68-77. Farran, G. p. 1936. On the mesh of herring drift net.s in relation to the condition factor of the fish. Jour, du Cons., vol. U, pp. 43-52. Fish, Marie Poland. 1932. Contributions to the early life histories of sixty-two species of fishes from Lake Erie and its tributary waters. Bull. U. S. Bur. Fish., vol. 47, pp. 293-398. Ford, E. 1933. An account of the herring investigations con- ducted at Plymouth during the years from 1924 to 1933. Jour. Marine Biol. Assoc vol. 19, No. 1, pp. 305-384. Fry, F. E. J. 1949. Statistics of a lake trout fishery. Biometrics, vol. 5, pp. 27-67. 1953. The 1944 year class of lake trout in South Bay, Lake Huron. Trans. Amer. Fish. Soc, vol. 82 (1952), pp. 178-192. Fry, F. E. J., and W. A. Kennedy. 1937. Report on the 1936 lake trout investigation. Lake Opeongo, Ontario. University To- ronto Stud., Pub. Ont. Fish. Res. Lab., No. 42, pp. 3-20. Greeley, John R. 1934. Fishes of the Raquette Watershed, with annotated list. Suppl. 23rd -Ann. Rept. N. Y. Cons. Dept. In: A Biological Survey of the Raquette Watershed, pj). 109-135. 1936. Fishes of the area (Delaware-Susquehanna Watershed) with annotated list. Suppl. 25th Ann. Rept. N. Y. Cons. Dept. 1935, pp. 45-88. Hall, A. E., Jr., and Oliver R. Elliott. 1954. Relationship of length of fish to incidence of sea lamprey scars on white suckers, Calosto- mus commersoni, in Lake Huron. Copeia, No. 1, pp. 73-74. Hildebrand, Samuel F., and Louella E. Cable. 1930. Development and life history of fourteen teleostean fishes at Beaufort, N. C. Bull. U. S. Bur. Fish., vol. 46, pp. 383-488. AGE DETERMINATION FROM SCALES OF LAKE TROUT 59 Hii.DEBRAND, Samiel F., and LovELLA E. Cable. — Con. 1934. Reproduction and development of whitings or kingfishe.s, dnim.s, spot, croaker, and wcakfi.-ihes or sea trouts, family Sciaenidae, of the Atlantic Coast of the United States. Bull. U. S. Bur. Fish., vol. 48, pp. 41-117. 1938. Further notes on the development and life history of some teleosts at Beaufort, X. C. Bull. U. S. Bur. Fish., vol. 48, pp. 505-642. HiLE, Ralph. 1949. Trends in the lake trout fishery of Lake Huron through 1946. Trans. Amer. Fish. Soc, vol. 76 (1946), pp. 121-147. HiLE, Ralph, and Howard J. Buettner. 1954. Statistics of the lake troiit fishery of Lakes Huron, Michigan, and Superior, 1949-53. Great Lakes Fishery Committee. Minutes of Annual Meeting, St. Louis, Mo., pp. 36- 40. HiLE, Ralph, Pail H. Eschmeyer, and George F. LlNCER. 1951. Decline of the lake trout fishery in Lake Michigan. Fish. Bull. U. S. Fish and Wildlife Service, vol. 52, pp. 77-95. Hodgson, William D. 1929. Investigations into age, length, and maturity of the herring of the southern North Sea. Part in. The composition of the catches from 1923 to 1928. Min. Agric, and Fish., Fish., Invest. Ser. II, vol. 11, No. 7, 71 pp. Kennedy, W. A. 1954. Growth, maturity and mortality in the rela- tively unexploited lake trout, Cristivomer namaycush, of Great Slave Lake. Jour. Fish. Res. Bd. Canada, vol. 11, pp. 827-852. M ASSET, Frank J. 1951. The Kolmogorov-Smirnov test for goodness of fit. Jour. Amer. Stat. Assoc, vol. 46, pp. 68-78. Miller, R. B., and W. A. Kennedy. 1948. Observations on the lake trout of Great Bear Lake. Jour. Fish. Res. Bd. Canada, vol. 7, pp. 176-189. MiLNER. Ja.MES W. 1874. Report on the fishes of the Great Lakes; the result of inquiries prosecuted in 1871 and 1872. Rept. U. S. Comm. Fish.. 1872-1873, pp. 1-78. Moffett, James W. 1952. The study and interpretation of fish scales. The Science Counsellor, vol. 15, No. 2, pp. 40-42. Reibisch, Johannes. 1899. Ueber die Eizahl bei Plenronectes plalessa und die Altersbestimniung dieser Form aus den Otolithen. Wiss. Meers., Abt. Kiel, X. F., Bd. 4, s. 231-249. SCHNEBERGER, EdWARD. 1936. Tagged trout reveal new facts and figures. The Fisherman, vol. 5, No. 7, p. 1. Shetter, David S. 1951. The effect of fin removal on fingerling lake trout {Cristivomer namaycush). Trans. Amer. Fish. Soc, vol. 80 (1950), pp. 260-277. S.viith, Oliver H.. and John Van Oosten. 1940. Tagging experiments with lake trout, white- fish, and other species of fish from Lake Michigan. Trans. Amer. Fish. Soc. vol. 69 (1939), pp. 63-84. Van Oosten, John 1949a. The sea lamprey — a threat to Great Lakes fisheries. State Government, vol. 22, pp. 283-284, and 289. 1949b. Progress report on the sea lamprey study. The Fisherman, vol. 17, Xo. 3, pp. 6, 9, and 10. 1950. Progress report on the study of Great Lakes trout. The Fisherman, vol. 18, Xo. 5, pp. 5, and 8-10. and Xo. 6, pp. 5 and 8. Van Oosten, John, Ralph Hile, and Frank W. Jubes. 1946. The Whitefish fishery of Lakes Huron and Michigan with special reference to the deep- trap-net fishery. Fish. Bull. V. S. Fish and Wildlife Service, vol. 50, 1950, pp. 297-394. U. S. GOVERNMENT PRINTING OFFICE 1956 O — 378326 COMPARATIVE STUDY OF FOOD OF BIGEYE AND YELLOWFIN TUNA IN THE CENTRAL PACIFIC By JOSEPH E. King and Isaac I. Ikehara, Fishery Research Biologists The predominant species of tuna captured on longline-fisliing surveys of the Fish and Wildlife Service's Pacific Oceanic Fishery Investigations (POFI) are the yellowfin, Neothunnus macropterus (Temminck and Schlegel), and the bigeye, Para- thun?}us sibi (Temminck and Schlegel), with a catch ratio of about 5 to 1 in favor of the yellow- fin. These are large tanas, the yellowfin oc* casionally reaching a weight of 200 pounds and the bigeye a weight of 300 pounds in the tropical Pacific. The two species have a marked super- ficial resemblance in general body shape and coloration and arc not always differentiated in the commercial catch. Murphy and Shomura (1953a, 1953b), in dis- cussing results of experimental longline fishing conducted by POFI, point out interesting differ- ences in the distribution of these two species. In the tropical Pacific, the bigeye have been taken in greatest numbers north of latitude 5° N. The best catches of yellowfin, on the other hand, have been made in the general region of the Equator, sometimes to the north when the area is under the influence of southeast tradewinds, and some- times to the south when the northeast trades are dominant. Tliis shift in abundance that appears to be related to changes in the prevailing winds can now be explained, at least partially, from our knowledge of the ocean currents and their effect on the basic food supply (Cromwell 1953).' Al- though the peaks in abundance do not correspond exactly, tlie general area of high yellowfin catch is also the area of greatest zooplankton abund- ance (King 1954). The horizontal distribution of the bigeye, however, does not seem to conform to the general pattern that the most fish are found where food is most abundant. There is also some evidence of difference in the vertical distribution of yellowfin and bigeye. While the results are rather variable, there have ' .\Iso a manuscript by O. E. Sette: Nourishment of central Pacific stocks of tuna by the equatorial current system (Proceedings of the 8th Pacific Science Congress). been indications on certain POFI cruises to the equatorial area that the best catches of bigeye came from greater depths than those of the yellowfin (Murphy and Shomura 1953b). In Ha- waiian waters the bigeye occurs in greatest num- bers during the winter months from Octol)er to May, whereas the yellowfin is most abundant from May to September (Otsu 1954). Brock (1949) points out that the Hawaiian longline fishermen try to increase the catch of bigeye after the yellow- fin season by lengthening the hook lines in order to fish deeper. Also, unlike the yellowfin, the bigeye — at least the adults — are rarely taken by surface-fishing methods. Nakamura (1949) states that the bigeye is thought to occur at the deepest levels of any of the tunas. It appears that the bigeye prefers somewhat colder water than does the yellowfin, or perhaps the two species have different feeding habits or food preferences which influence their distribution. The purposes of this study are to describe the food of bigeye tuna in the central Pacific, to com- pare the foods of bigeye and yellowfin tuna - captured at about the same time and place, to determine whether differences occur which are associated with the horizontal and vertical distri- bution of these fish, and to obtain information on food preferences of each fish which maj^ be useful to the commercial fishery. The experimental fishing carried out by POFI has provided collec- tions of bigeye and yellowfin stomachs which are essentially alike in respect of time and area and which were obtained with standardized fishing methods. Therefore, we believe the resulting data should provide reliable comparisons of the food of these fish because these several variables have been controlled. There is an extensive literature, reviewed pre- viously by Reintjes and King (1953), dealing with the food of yellowfin, whereas there are only a very ' The food of yellowfln was previously described by Reintjes and King (1953). 61 62 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE few references pertaining specifically to the food of bigeye. Suyehiro (1942) desciibes this species as a very voracious fish and lists the following items as appearing in its food: amphipods, shrimp, cuttlefish, squid, sardines, sauries, bonitos, needle fish, and a viper fish. In the report of the South Seas Tuna Fishery Investigations for 1950 (,Kan- agawa et al. 1951), the food of 27 bigeye is shown to include the following: 7 squid, 1 octopod, 3 deca- pod Crustacea, 1 fan fish, 15 needle fish, 1 file fish, 6 pomfret, and 4 lantern fish. This suggests that the bigeye, like many other tunas, has a varied diet. A large number of the tuna stomachs reported on here were examined by John W. Reintjes, Sueto Murai, and T. J. Roseberry, former em- ployees of POFI. We appreciate their services in a difficult and generally disagreeable task. We are grateful to other staff members of POFI for assistance in handling these large fish aboard the vessels and in removing and preserving the stomachs. SOURCE OF MATERIALS This report is based on examination of 439 yellowfin and 166 bigeye stomachs collected on 11 cruises of Fish and Wildlife Service vessels during the years 1950 to 1953 (table 1). The yellowfin data include 125 stomachs collected in 1950 and 1951 and previously reported on by Reintjes and King (1953). These collections and the additional 314 yellow^n stomachs obtained in 1952 and 1953 were obtained at the same stations, or near the same stations, as furnished the bigeye stomachs included in this report. The sampling area (fig^ 1), extended along the Equator between Maufim^ 119° W. and 180° and approximately from lat- itude 17° N. to latitude 14° S. at its greatest width. The tuna were captured by longline at depths of about 150 to 500 feet. This method of fishing, as practiced by POFI, has been reviewed by Murphy and Shomura (1953a); the design and construction of the gear was described by Niska (1953). Only fish caught 25 miles or more from land are considered in this study ; therefore local differences due to reef faunas should be reduced to a mini- mum. The sampled fish varied widely in size, from 87 to 172 cm. fork length for the yellowfin, and from 77 to 196 cm. fork length for the bigeye (fig. 2). Weights of fish given in this paper were 180° 170° 160° 150° 140° 130° I2C ° 20'' ' o "t> O YELLOWFIN ONLY • • BIGEYE ONLY a YELLOWFIN 8 BIGEYE • i ► 10° 0° • 1 c c )0 o c o o o o * 3 9 • o* *f • ° q? o o » • 9 • C oc 8 • • 9 c c c o )0 o >o < o o 0 » > 1 c ( t a » > k t a a c o o 10° (* o 20° — 10° 10° 180° 170° 160° 150° 140° 130° 120° Figure 1. — Locations of the stomach collections of yellowfin and bigeye tuna captured by experimental longline fishing in the central Pacific, 1950-53. FOOD OF BIGEYE AND YELLOWFIN TUNA 63 Table 1. — Distribution of yellowfin and bigeye stomachs collected from the central Pacific, identified by vessel, cruise, time of year, and locality Sampling area Yellowfin Bigeye Vessel Cruise No.— Period Range or longitude (W.) Range of latitude Xumber captured Number of stomachs exam- ined Percent of catch sampled Number captured Xumber of stomachs exam- ined Percent of catch sampled Hugh M. Smith Hugh M. Smith- _ 7 11 11 1 1 2 12 13 18 14 15 Oct.-Xov. 1950 Aug -Sept. 1951 157°-167'' ISOMse" 155°-180° IW-ISO" 140° ll°X-0° 15''X-»''S.... 5°X-7'>S 9°N-1°S e^N-S'N 6''N.-2''N---- 7°X.-5'>S..... 17°N.-5°S.... 9°N.-10°S-... 4°N.-14''S.... 10°N.-6° S.... 132 457 210 72 42 720 146 135 60 106 197 1 106 M53 59 44 19 1 55 40 69 20 80 34 28 61 17 3 1 41 67 65 10 22 93 30 43 11 60 28 29 50 19 46 14 36 6 17 2 13 5 10 17 19 27 64 John R Manning ,. Jan.-Mar, 1952 . 20 Charles H. QUbert Cavalieri May-June 1952 June-July 1952 Aug.-Sept. 1952 Aug.-Sept. 1952 Oct.-Nov. 1952 Oct.-Xov. 1952 Jan.-Mar. 1953 Apr.-June 1953 40 Cavalieri . . .- 140°-142'' 140°-150° isr-no" 120°-131° 140°-150° isflo-no" 22 John R. Manning John R. Manning Hugh M. Smith- -. John R. Manning John R. Manning 18 34 34 100 59 I Of this number, only 38 (29 percent of the catch) were considered comparable in respect to time and place to the bigeye collections and were included in this report. ' Of this number, only 87 {19 percent of the catch) were included in this report for the same reason as above. 1 — i — I — I — r YELLOWFIN 1 — I — \ — I — I — I — I — I — 1 — \ — r I r^^rT^^rrr. FORK LENGTH (CM ) BODY WEIGHT (LBS ) -1 — 1 — t — I — ' — I — I — I — I — I — r- BIGEYE 100 47 120 110 160 FORK LENGTH (CM ) 81 127 187 BODY WEIGHT (LBS ) 200 360 FiGiRE 2. — LeiiKth-frequeiicy distribution of yoUowfin and bigeye tuna from which stomach.s were collected. derived from length measurements converted by means of length-weight tables provided in the POFI Scientific Field Manual (unpublished). METHODS At sea, the stomacii was removed as soon as possible after the fish was captured, placed with any iTgui-gitaled material in an unbh'aclied-inuslin bag. and preserved in lO-peicent formalin. A label bearing date, species najme, fork length, fish- ing method, hook number, bait used, name of observer, vessel, and cruise number was placed with each stomach. Tuna landed with their stomachs everted were not sampled. The stomach was removed by one of the follow- ing methods: (1) The abdominal cavity was opened by a longitudinal midventral incision, the small intestine was severed posterior to the pyloric valve, and the stomach was freed by cutting through the esophagus; or (2) the gill membrane was slit along the line of attachment with the cleithrum posterior to the fourth gill arch, the viscera was pulled out, and the stomach was removed by cutting througli the small intestine and esopliagus. In the laboratory, the stomachs were soaked in fresh water for a period of 16 to 24 hours to remove excess formalin. Each stomach was then slit open, and the contents were carefully removed and separated into groups according to kind of organism. Identifications were made as com- pletely as was practicable, and the number of each species or group of organisms present was recorded. Each species or group was measured volumetrically by the displacement of water in a graduated cylinder of appropriate size. Bait used to capture tlie tuna was omitted from this analysis. The methods and literature used in the identification of the food organisms were essentially the same as that employed bv Reintjes and King (195.S) and will not be I'eviewed here. Berg's (1947) system of classification and nomen- clature was primarily useii foi' the family names of tlie forage fishes. 64 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE A detailed list of the food organisms found in the tuna stomachs is presented in the appendix (table 11). For each kind or group of organisms there are shown (1) the total number of sucli organisms, (2) the number of stomachs in which they occurred, (3) the percentage of occurrence, (4) the total aggregate volume of each food element, and (5) the percentage of total volume. Regardless of the method or methods of anal- ysis used, there are many uncontrollable variables inherent in food studies which detract from the precision of the results. It is our belief, however, that for a fish with a generalized diet, such as that of the tuna, any of the commonly used meth- ods of evaluation will give substantially the same results if a sufficiently large number of specimens are examined. In reporting the results of our studies on tuna food we use both the percentage- of-occurrence and the percentage-of-volume meas- urements (as described by Reintjes and King 1953) and the average volume of food per stom- ach. The food items that rank liigh in number, volume, and frequcncv of occurrence are most likely to be important foods. No attempt has been made to apply statistical tests of significance to the data. It is likely tliat the variates used — volume of food per stomach, percentage of occurrence, and percentage of total volume of the organism — are not distributed nor- mally and that the means are correlated with the variances or standard deviations. To apply meaningful tests of significance, transformation of the data would be necessary. Moreover, several of the comparisons that will be made involve two-way or three-way classification of the data. Even if suitable transformations were derived, the application of advanced analysis of variance techniques would be hampered by unequal subclass numbers. Furthermore, it appears that in both yellowfin and bigeye there is an increase in mean volume of food per stomach with increase in size of fish. ^ \\^ ^ » * f ■* i 9"' ^■■ h Fir.URE 3. — Exainple.s of type.s of food commonly found in .stomach.s of yellowfin and bigeyf tunas: Left to right: pom- fret (1), truncated .sunfi.sh (1), snake mackerel (1), lancet fish (1), shrimps (3), viper fish (15), hatchet fi.sh (3), euphausids (8), juvenile stomatopods (3), crab megalopa (12), squid (3), and paper nautilus (1). 10 9 - I C3 e UJ » 7 >- o o 6 m ii o b m _] 4 CC Ul 0. 3 ,^ (> o d -I o > FOOD OF BIGEYE AND YELLOWFIN TUNA YELLOWFIN BIGEYE 65 1200 ' 1 1 1 I 1 - MOO - (A) - 1000 - ■ - 900 - - STOMACH 8 § — • - IT 600 UJ a. — - 500 o o ~ 400 _i o - • - ^ 300 - — 200 - j^L-^ _____ - 100 A 'l :-:l:v-.J,'V1^S:;;^i.-: • .■(•■ ■ 1 T T T T (A) ^ :■■■■■■ -f • • -i-:- (B) "1 r "1 r (B) - 40 280 320 360 80 120 160 200 240 0 40 80 120 160 200 240 BODY WEIGHT (LBS) BODY WEIGHT (LBS) Figure 4. — Regressions of (A) food volume per stoniacli and (B) food volume per unit body weight on total body weight for 439 yellowfin and 166 bigeye captured on longlines. and tliere is also a decrease in average stomaeli content per unit of body weight (cc./lb.) with increase in size of fish. Tlie least-s(itiares trend lines shown in figure 4 (tliere is no a priori reason for assuming rectilinearity) indicate tlH> need for covaiiance nu'thods of statistical analysis, again after stiitahie liansformations. Finally we must point out the great variability of the data as illustrated by the wide scatter of points about tiie trend lines. This gi-eat variability reduces the opportunity- of denionstrMting stat ist icidly signifi- cant differences, particularly when the data are analyzed in subgroups which contain few speci- mens in each. Because of the difficidties outlined above, in the following sections we have tabulated average values and have discussed difi'crences and trends without attempting to ajjpraisc their statistical significance. Consequently, the inferences that we make must be regarded as suggestions oidy. They may form the bases for hypotheses which can be tested more stringently in the futuic. 66 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE RESULTS The food of both the yellowfin and the bigeye was primarily fish, of great variety, and squid (table 11, Appendix). Other mollusks, such as the argonauts and octopods, and crustaceans were of minor importance.^ Figure 5 ilhistrates the percentage of occurrence of the major food items. Figure 6 shows the percentage of aggre- gate total volume of each major food element, which indicates its relative importance by bulk. Representatives of 48 fisli families and 1 1 invertebrate orders were found among the stomach contents of the yellowfin, as compared with 36 fish families and 9 invertebrate orders for the big- eye* Despite this great variety in the food, only OTHER FOOD Figure 5. — Percentage of occurrence of the major food elements. s Among the results of this study, not referred to elsewhere in the report but perhaps worthy of mention, were observations on the number of stomach parasites. Among the bigeye, 26 percent of the stomachs examined were Infested with nematodes and 32 percent with trematodes. The infestation was somewhat less among the yellowfin, being 16 percent for nematodes and 26 percent for trematodes. * The greater variety in the food of the yellowfin as compared with the big- eye is due, we believe, simply to the fact that more than twice as many yellow- fin stomachs were examined. a few items were of primary importance to either species. For both the yellowfin and the bigeye, those food elements ranking high in number, volume, and frequency of occurrence were squid, of the families Ommastrephidae and Loliginidae, and among the fish the pomfret {CoUyhus drachme) and snake mackerel {Gempylus serpens) were important. Certain fishes, such as the tunas (Thunnidae) and the sun fishes (Molidae), were relatively important in volume but ranked low in number and frequency of occurrence, indicating that they are only occa- sionally utilized. Crustacea of the order Stomato- poda, prominant in number in the food of yellow- fin, were completely lacking from the bigeye stomachs. The young of other tunas, especially skipjack, formed a much more important part of the yellowfin diet than that of the bigeye. In the following sections of this report we shall try to describe the major differences and similarities in the foods of these two species of tuna as related to such factors as size of the tuna, area and depth of capture, season, and features of the equatorial- current system. Variation in Food with Size of Tuna In general, for both yellowfin and bigeye, there was an increase in food volume per stomach with an increase in size of the tuna (fig. 4). With the hope of minimizing the effects of this factor, in our examination of differences in the food specif- ically related to size of tuna we have split the data for both species into two size groups, (1) those less than 140 cm.^ and (2) those 140 cm. and over, in fork length (table 2). This provided for each species two groups of fish roughly equal in number. In the yellowfin the larger size gfoup contained 29 percent more food per stomach, and in the bigeye it contained 16 percent more. The ratios of stomach content to body weight are almost identical for the two species (table 2). Although Crustacea make up a very small per- centage of the food of these large, deep-swimming fish, in both species tlie smaller specimens con- sumed greater amounts of such organisms as crab larvae, shrimp, and ampliipods. In both species, the larger specimens consumed less fish and more mollusks — as percentage of total volume — than did the smaller size group; this was particularly true for the bigeye. The per- ' A 140-cm. yellowfin from the equatorial Pacific weighs approximately 118 pounds, while a 140-cm. bigeye weighs approximately 127 pounds. FOOD OF BIGEYE AND YELLOWFIN TUNA 67 ^fi-'f-'-"'^ S- ^^ YELLOWFIN BIGEYE FifiiRE 0. — Comparative importance, in volume, of the major food elements. centage by occurrence and percentage by volume for the fish famihes prominent in the diet exhibited little variation with the size of the tuna. Variation in Food with Depth of Capture Figure 7 is a diagram of one unit (a basket) of POFI longline gear, showing the arrangement of hook-bearing dropper lines and the general lay of the line with respect to the surface. Although an attempt is made to set the line at each station in a standard fashion, with an average distance be- tween buoys of about 900 feet, the actual depth of fishing is quite variable depending upon the amount of sag in the main line, which is gi-eatlv influenced by wind and current conditions. FiriiRK 7. — .Arrangement of a unit (ba.sket) of POFI standarfi loMKlinc gear showing the float lities, main line, hook-bearing dropper lines, and the general lay of the line with respect to the surface. :{887:)4 ( ) — Sft 2 Murphy and Shomura (1953a) have calculated that the ma.ximuin possible depth for hooks 1 and 6, with a 900-foot buoy interval, is 310 feet; for hooks 2 and 5, it is 450 feet; and for hooks 3 and 4, it is 530 feet. These maximum depths are seldom achieved, however, because of the rela- tively strong surface currents generally prevailing in this region. The miniminn depths are even more uncertain; therefore it is difficult to define a depth range for each of the hooks. We postulate that liooks 1 and 6 may fish at depths of 1.50 to 300 feet, hooks 2 and 5 at depths of 250 to 400 feet, and hooks 3 and 4 at depths of 300 to 500 feet. Despite this variation and the imcertain- ties involved, it is worthwhile, without attempt- ing to designate actual fishing depths, to make comparisons between tlie shallow (hooks 1 and 6), intermediate (hooks 2 and 5), and deep (hooks 3 and 4) levels of capture with respect to dift'erences in stomach contents. Because of tlie rather slight differences in composition of tlie food associatetl with the size of the tuna, the two size groups (140 cm.) were combined for liiis study. Table 3 shows the variation in composition of stomach contents with igeye stomachs were collected at 0-24 miles, and few (eigiit) were collected in the 25-99 mile interval; tlierefore, the data do not provide the desired information on difterences in the food related to this feature. There was some indica- tion that the consumption of squid and pomferts by the bigeye increased in an offshore direction, as compared with tlieir uniform utilization by the FOOD OF BIGEYE AND YELLOWFIN TUNA 69 SQUID FISH 100 90 80 - w 70- UJ 2 I- 50 z UJ " 40 o 30 ID 3 O 20 10 O 100 YELLOWF (109) (120) (165) i- A 90 - 80 ■ 70 60 UJ Z I- 50 z UJ <-> 40 u 5 30 o 20 10 0 BIGEYE (18) (45) SHALLOW INTERMEDIATE DEEP FinuRE 8. — Variation in the major food categories (total fish and squid) as related to the depth at which the tuna were captured. Number of stomachs is shown in parentheses. yellowfin (tahlp 4), and in the bigeye the average food vohune increased with greater distance from land while in the yellowfin the volume varied irregularly. Variation with Season and Longitude To examine differences related to time of sampl- ing, the various cruises were grouped into four seasonal periods, as indicated in table 5. For botii species the largest average volume of food occurred in the April-July period, with ()ctob(>i- Xovember averaging the lowest in the yellowfin and August-September the lowest in the bigeye. If we consider the average volinnc per stomach of the major food elements, we find tlial, in our samples of both tunas, fish were consumed in greatest amount during April-.Tuly and in least amount during August-September (fig. 9 and table 5). The average volumes of squid and the major fisli families represented in the food did not vary in parallel fashion for thebigeyeandyellowfin. When the data from the various cruises are combined with regard to longitude but witliout regard to time of year, we obtain the results pre- sented in table 6, with the variation in availability ^ SQUID TOTAL FISH I80'>-I55"' 1 50<^ 140° WEST LONGITUDE 1 30°- 1 20° /<-■ FiciRE * — Variation in the major foods as related to time of year that the tuna were captured. Number of stomachs is shown in pareiithe.ses. 70 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 3. — Variation in volume and composition of stomach contents with depth of capture of yellowfin and bigeye tuna Food organisms Average volume (cc.) per stomach Percentage of occurrence Percentage of total volume Shallow Inter- mediate Deep Shallow Inter- mediate Deep Shallow Inter- mediate Deep Crustaceans: 1.8 0.7 26.6 47.2 11.8 0.8 47.4 43.2 9.0 5.3 5.3 4.4 10.6 0.0 1.0 5.0 1.8 0.0 0.6 0.0 88.1 91. 9 0.7 1.4 30.7 63.9 5.4 0.9 54.9 86 7 10.8 22.5 6.7 22.7 14.4 0.5 0.4 6.0 4.4 3.4 0.4 0.3 92.2 153.3 1. 1 3.2 33.7 36.0 7.4 6.9 78.6 83.7 13.4 21.1 9.0 14.5 22.2 2.5 0.3 3.6 12.8 1.4 0.9 0.1 121.7 129.9 39.4 33.3 84.4 72 2 47.7 33.3 91.7 72.2 53.2 27.8 33.9 33.3 9.2 0.0 11.9 11.1 1.8 0.0 27.5 5.5 44.2 35.6 90.0 84.4 49.2 17.8 91.7 86.7 64.2 26.7 35.0 42.2 9.2 2.2 4.2 17.8 1.7 4.4 23.3 8.8 46.7 46.9 90.3 83.9 37.0 40.7 93.9 86.2 66.7 35.8 33.9 39.6 9.1 2.6 7.3 22.2 2.4 1.2 27.3 7.4 2.0 0.7 30.2 51.4 13.5 0.8 53.8 47.0 10.2 5.8 6.0 4.8 12.0 0.0 1.2 5.5 2,0 0.0 0.6 0.0 0.8 0.9 33.3 41.7 6.8 0.6 59. 5 56.5 11.7 14.7 7.3 14.8 15.6 0.3 0,4 3.9 4,8 2.2 0,6 0.2 0 9 Bigeve 2 5 Squids: Yellowfin 27 7 Bigeye _ ... 27 7 Other mollusks: Yellowfin 6 1 Fish (total): Yellowfin . . . 64 6 Bramidae: Yellowfin ....... . 11 0 Bigeye Oempylidae; Yellowfin .. . _. 16.3 7 4 11 2 Thunnidae: Yellowfin 18 2 1.9 Sudidae: Yellowfin 0 2 ■» 7 Molidae: 10 6 1 1 Other foods: OS 0,0 All foods: Yellowfin Number of stomachs examined- Yellowfin 109 18 141 152 120 162 0.7 0.6 120 45 140 148 118 148 0.8 1.0 166 81 142 142 122 133 1.0 1.0 Bigeye Average fork length (cm,): Bigeye __ Average weight (lbs.): Yellowfin . Average volume (cc.) food per pound of body weight: Yellowfin FOOD OF BIGEYE AND YELLOWFIN TUNA 71 Table 4. — Variation in volume and composition of sto/nach contents with distance of place of capture from nearest emergent land Food organisms Average volume (cc.) per stomach Percentage by occurrence Percenuge of total volume 25-99 miles 100-399 miles 400 mi. and over 25-99 miles 100-399 miles 400 mi. and over 25-99 miles 100-399 miles 400 mi. and over Crustaeeans: Vellowfin 1.2 3.7 30.7 9.8 1.2 3.0 79.2 28.0 14.7 1.5 4.8 21.9 1.6 0.0 0.7 3.7 38.2 CO 1.1 0.7 11.3. 4 46.1 1.8 1.9 24.2 17.9 4.4 2.2 52 1 94.0 9.5 15.0 5.8 18.5 5.9 10.8 0.6 5.1 9.6 4.2 0.8 0.2 83.9 116.2 0.4 2.5 34.8 64.2 11.1 5.5 71.1 81.5 12.4 30.7 8.0 12.2 25.5 2.4 0.3 3.1 6.0 0.0 0.3 0.1 117.9 153.8 53.3 75.0 90.0 75.0 13.3 25.0 96.7 62.5 43.3 25.0 36.7 12 5 10.0 0.0 10.0 12.5 6.7 0.0 26.7 12.5 52.7 42 8 89.6 76.2 36.8 34.9 94.5 85.7 61.7 25.4 31.8 38. 1 .VD 3.2 10.4 17.5 3.0 4.8 32.8 6.3 7.7 35.8 84.1 87.4 48.1 30.5 90.9 85.3 62 0 3S. 9 31.2 38.9 11. 1 3.2 2.9 16.8 1.9 0.0 15.4 6.3 1.1 8.2 27.1 21.7 1. 1 6.6 69.8 62.0 13.0 3.4 4.2 48.6 1.4 0.0 0.6 8.2 33.7 0.0 1.0 1.5 2.2 1.7 29.0 15.4 5.3 1.9 62.5 80.9 11.4 12 9 6.9 15.9 7.0 9.3 0.7 4.4 11.5 3.6 1. 1] U. 1 0.4 1.6 Squids: 29.6 41.8 Other mollusks: 9.5 3.6 Fi5h (total): Yellowfin - 60.3 53.0 Bmmidae: Vellowfin 10.5 20.0 (Jeinpylidae: 6.8 7.9 Thunnidae: Vellowfin - - --- - - 21.7 1.5 Sudidae: 0.3 2 0 Molidae: 5.1 0.0 Other foods; 0.3 0.0 All foods Number of stomachs: 30 8 138 153 112 166 1.0 0.3 201 63 140 139 118 125 0.7 0.9 208 95 142 148 123 149 1.0 1.0 Bigeve .\veraBe fork length (cm.): Yellowfin .Average weight Obs.): Average volume (cc.) food per pound of body weight: Bigeye ... 1 ■■ 1 72 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 5. — Variation in volume and composition of stomach contents with time of year at which the tunas were captured Food organisms Average volume (cc.) per stom- ach Percentage by occurrence Percentage of total vo ume Jan- March April- July Aug.- Sept. Oct.- Nov. Jan.- March April- July Aug.- Sept. Oct.- Nov. Jan.- March April- July Aug.- Sept. Oet.- Nov. Crustaceans: Vellowfin 1.1 1.9 27.0 21.8 7.9 3.2 56.9 109.5 8.0 29.1 8.1 11.5 10.4 0.2 1.4 34.4 72.4 14.5 3.9 105.7 116.4 14.9 51.8 8.0 16.9 25.7 4.9 0.2 2.6 32.9 3.0 0.3 0.03 155.2 194.1 0.6 1.1 40.3 40.3 4.5 5.6 52.0 51.4 14.1 4.5 1.9 8.0 25.1 0.0 2.2 0.2 0.3 0.2 97.7 98.7 2.1 5.3 21.1 30.9 5.4 2.9 54.8 76.2 10.8 12.8 8.6 24.3 5.2 16.6 0.5 1.8 14.7 3.0 0.8 0.1 84.1 115.3 60.1 48.0 94.5 80.0 46.9 48.0 96.1 92.0 60.1 24.0 45.3 48.0 7.8 13.3 36.0 0.0 43.7 12.0 26.8 37.0 76.0 78.3 47.9 30.4 91.5 89.1 50.7 43.5 23.9 37.0 16.9 6.5 7.0 19.6 12.6 2.2 9.9 6.5 29.9 25.9 96.3 86.1 39.3 13.0 90.7 70.3 69.2 33.3 24.3 28.0 5.6 0.9 5.6 0.9 7.5 3.7 43.6 48.8 91.0 92.7 21.0 26.8 97.0 95.2 60.9 31.7 28.6 39.0 5.3 4.9 5.3 12.2 2.3 4.9 25.6 9.8 1.2 1.4 28.8 16.0 8.4 2.3 60.6 80.2 8.5 21.3 8.7 8.4 11.1 1.0 9.5 0.0 0.9 0.2 0.1 0.7 22.2 37.3 9.3 2.0 68.1 60.0 9.6 26.7 5.2 8.7 16.6 2.5 0.1 1.3 * 21.2 1.6 0.2 0.02 0.7 1.1 41.2 40.9 4.7 5.6 53.2 52.1 14.4 4.6 2.0 8.1 25.7 0.0 2.2 0.2 0.3 0.2 2.4 Bigeve . _ 4.6 Squids: Yellowfin 25.1 Bigeye — Other mollusks: Yellowfin . .. 26.8 6.4 2.5 Fish (total) : 65.1 Bigeye 66.0 Bramidae: 12.9 Bigeye 11. 1 Gempylidae: Yellowfin 10.2 21.4 Thunnidae: Yellowfin . 6.2 14 4 Subidae: 0.9 13.0 0.0 0.6 1.5 Molidae: Yellowfin - 17.5 2.6 Other foods: 0.9 0.2 93.8 136.6 0.9 0.08 All foods: Number of stomach examined: Yellowfin 128 25 140 148 119 148 0.8 0.9 71 46 141 141 121 129 1.3 1.5 107 54 140 149 116 151 0.8 0.7 123 41 141 146 120 142 0.7 0.8 Bigeye Average fork length (cm.): Vellowfin Bigeye . - . Average weight (lbs.): Yellowfin Average volume (cc.) food per pound of body weight: Bigeye FOOD OF BIGEYE AND YELLOWFIN TUNA 73 V7?i SQUID I TOTAL FISH olOO S o (- f 75 a. UJ a. o 50 0 150 5 100 < z o YELLOWFIN (71) (128) (107) (123) ^ 1 I ^ i 25 BIGEYE (25) (46) i i (41) (54) I I -JUL AUG- SEP SEASON OCT -NOV Figure -KT — Variation in the major foods as related to the general longitude of capture of the tunas. Number of stomachs is shown in parentheses. of the major food items shown in figure 10. The chief simihirity between the two species lies in the lower volume of total fish in the food of tunas captured in the region of 140°-150° W. longitude. The utilization of squid, Bramidae, Gempylidae, and TluHinidae does not vary in any regular pattern for the two species. A majority of the Thunnidae appearing in the food of yellowfin captured in the area of 120°-130° W. were Auxis ^Aojo/y/, which was not prominent in the food in the more western regions and which in the l)igeye was represented by only one specimen, also from the 120°-13n° \V. region. P^or both bigeye and yellowhn, the largest specimens were captured in the eastern region (120-1:50° W.) and the smallest in the western region (155° W-180°). When the variation in volume of stomach contents is considered in terms of imit volume per iniit of body weight, we find 388734 O- 56 3 only slight regional diflferences for the yellowfin but a rather large variation for the bigeye (table ()). \w the bigeye, specimens from the western region contained 1.5 cc. of food per pound of body weight, as compared with 0.6 cc. for specimens from the central region and 1.0 cc. for specimens from the eastern region. These three values closely parallel the corresponding average volumes of total fish per stomach (115.3, 55.4, and 95.8 cc). Variation with the Current System The general pattern of the Pacific equatorial- current system has been described by Sverdrup and associates (1942, pp. 708-712). In brief, the major surface currents of this region are the North and South Equatorial Currents flowing toward the west, with the eastward-flowing Equatorial Coun- tercurrent sandwiched in between. Although the width of the Countercurrent (CC) may vary with longitude and season, its southern and northern boundaries are ordinarily near latitudes 5° X. and 10° X. in the Central Pacific. The South Eqiuitorial Current (SEC) is therefore on both sides of the Equator, while the North Equatorial Current (NEC) is confined entirely to the Xorthern Hemisphere. The prevailing east to southeast tradewinds, together with the Coriolis force resulting from the earth's rotation, induce a divergence of the surface waters at the Equator that is accompanied by up- welling. Under certain conditions, described by Cromwell (1953) a convergence may be formed, between the Equator and the southern boundary of the CC, which, we hypothesize, may tend to concentrate plankton and, consec|uently. the tinia forage organisnis. Over the range of latitude sampled (17° X".- 14° S.), there are therefore certain natural sub- divisions of the environment that may be estab- lished on the basis of the features mentioned above. These may be defined as follows: (1) The XEC from the northern limit of our sampling (17° X.) to the northern boundary of the CC; (2) the CC, with its boundaries determined at the time of each crossing from vertical temperature sections;' (3) a zone of convergence in the SEC extending — accoril- ing to our definition — from the southern boundary of the CC to latitude 11^° X.; (4) a zone of diver- gence or upwelling in the SEC along the Equator from latitude 1J^° X. to latitude \)^ ■ ^^ i ^ 5 SEC 4 DIV CONV AREA 2 CC FinuRE 11. — Variations with the current system in (A) yellowfin and bigeye catch on longline gear, (B) average volunie of food per stomach, and (C) average volume of food per pound of body weight. Boundaries for each di- vision of the current .system are defined in the text. Part A i.s derived from cruises 7, 11, and 18 of the Iltiiih M. Smith, cruises 11, 12, 13, 14, and 15 of the John If. Mann- ing, cruise 1 of the Charle.<; II. Gilbert, and cruise 1 of the Cavalieri. Xumberof ob.servations, as stations fished (part A) or .stomachs examined (parts B and ('), is shown in parentheses. volume of stomach contents.' The divergence zone at tlie Equator produced good catches of yellowfin, but these fish contained the lowest food volumes. On the basis of both the average volume of food per stomach and the average volume of food per pound of l)ody weiglit — disregarding the three stomachs collected from the NEC — we judge that the yellowfin captured in areas 2, 3, and 5 were equally well fed. In the bigeye, there is a suggestion of parallel variation in catch rate and volume of stomach contents. This species was tlie best fed in areas 1 and 2, which were also the areas of best catch. The bigeye from near the Equator (area 4), where catches were poorest, con- tained the lowest food volumes. Table 7 illustrates variations in certain food components as related to the system of currents. The consumption of Crustacea by yellowfin is rougldy in accordance with the varying abundance of zooplankton as determined from oiu' plankton surveys (King and Demond 1953, King 1954). Tiieir utilization l)y l)igeye is quite different, iiowever, and may be related to differences in the kinds of organisms involved. In tlie food of yellowfin, for example, the crustacean fraction was principally amphi|)ods, with isopods and crab larvae of some importance; the bigeye had fed chiefly on shrimp and euphausids. The complex variations in the consumption of squid, Bramidae (chiefly Collijbu-y drachme), Gempylidae (chiefly Gempylus serpens), and total fish are difficult to understand, since we lack information on the lati- tudinal variations in abundance of these forage organisms. We should like ne.xt to exan)ine in greater detail the differences between the CC (urea 2) and the convergent zone (area 3) with respect to volume and composition of food utilized as related to deptli of capture of the tinias. As previously stated, the CC is a region of relatively good catch for l)igeye and of poor catch for yellowfin. Bigeye from this region contained about 50 percent more food in their stomachs than did the yellowfin, but they averaged somewhat larger in body size. • It was previously reported (Reiiiljes :iiicl Kiiiix I9.i:t) thai on one eniisc (eruise n, Hugh \f. Smith) there was some iiuliealion that for yellowfin the average volume of stomaeh contents varied directly with the ("ateli rate. 76 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 7. — Variations in average volume per stornach of the major food categories as related to the current system (Boundaries of each area or division of current system are defined in text; volume is measured in cc] Food organisms Area 1, NEC Area 2, CC Area 3. Conv. Area 4, Div. Area 5, SEC Crustaceans: Yellowfin - - - 0.3 2.1 38.0 73.1 2.7 1.3 64.9 75.4 8.5 41.0 8.1 6.9 20.2 3.5 0.6 3.3 30.7 32.3 12.5 8.1 62.2 82.6 16.4 22.5 4.2 18.1 18.3 2.9 1.8 33. 1 29.2 4. 1 0.7 44.0 45.6 6.6 0.6 5.0 11.7 16.6 0.8 0.1 34.3 38.9 3.8 Squids: 16.6 11.9 Other molluslts: 6. 1 7.1 8.0 155. 7 0.7 5.0 4.0 37.2 1.4 Fish (total): 91.6 80.6 Bramidae: 9. 1 5.4 Gempylidae: 12.8 16.5 Thunnidae: Yellowfin - - . - -- - 1.5 28.3 Sudidae: 0.2 5.7 1.2 0.1 1. 1 1.3 0.6 1.6 2.5 10.9 Molidae: 44.6 5.2 6. 1 Other foods: 0.2 5.7 0.2 106,6 126.4 0.7 0.1 84.8 77.3 0.9 0.3 All foods: 42 3 201.9 106.6 152.0 115.3 97.9 Number of stomachs examined; 3 24 147 149 138 153 0.3 1.3 86 58 136 146 108 143 1.0 1.1 164 47 141 145 119 139 0.9 0.9 98 14 144 153 129 164 0.7 0.5 88 23 140 138 Average weight (lbs.): 118 Bigeye - - 120 Yellowfin - 1.0 Bigeye - -- . . . ._ 0.8 On the basis of average volume of food per pound of body weight there was little difTerence between the two species. The region of convergence has yielded the best yellowfin catches but has produced consistently poor bigeye catches. In this region the bigeye had about 20 percent more food in their stomachs than did the yellowfin, but the bigeye were also larger in average body size. Again the two species were almost identical with respect to average volume of food per pound of body weight. In the CC, tlie thermocline occurs at shallow to moderate depths, while in the convergent zone it lies much deeper. Accompanying changes in the deptli and velocity of the surface currents may greatly affect the fishing depth of the longliiie. In the region of sliallow thermocline it is possible that, as a result of the streaming of the line caused by the marked sliear between the moving surface waters and the relatively quiet waters below the thermocline, all hooks may be fishing at about the same level (Murphy and Shomura 1953b), and no marked difference might be expected in the food between the various hook levels. In a region of deep tliermocline the longline can hang vertically and lie entirely within the homogenous surface layer. A marked difference in hook depth and possible differences in the stomach contents of the catch may then result. Data have been assembled in table 8 and figure 12 to illustrate the variations in average volume per stomach for the major food categories with deptli of capture of the tanas in these two ocean areas. In the CC there is greater change in the food of yellowfin with depth than in the convergent zone; this is evidenced by a consistent increase with deptli, in the CC, in the utilization of Bramidae, Gempylidae, and total fish. In the bigeye the only important and consistent variation shown in the CC is a marked increase with deptli in the amount of Crustacea eaten and a decrease in the importance of Gempylidae, as contrasted with FOOD OK BIGEYE AND YELLOWFIN TUNA 77 ^YELLOWFIN BIGEYE 0 160 I u < Z O 120 ui 80 P 40 UI Z y- z 0 1^120 CD 3 100 O CRUSTACEA i _a SQUID I 80 - 60 TOTAL FISH i HOOKS CONV CC CONV , IL _ _^ 186 2 85 CC SHALLOW INTERMEDIATE CONV CC 384 DEEP i 4 I «| J I i J i CC ALL DEPTHS COMBINED FioiRE 12. — Variations in average volume per stomach of the major food categories with depth of capture of 250 yellowfin and 105 bigeye tuna taken by longline in the C'ountercurrent and the convergent zone. the increase with deptli in Gempyhdae as noted for yellowfin. In the convergent zone there is a similar increase with depth in the utilization of Crustacea and Bramidae. The major foods are of about equal importance in both areas. There is no indication that the tunas have one set of foods in the CC and another in the converfient zone. Tlie main difference between the two species is the much greater consumption of Crustacea by the bigeye in both the CC and tlie convergent zone. If we may con- sider the longline catch rate as an index to abun- dance, it would appear that the bigeye responds in a different manner than the yellowfin to the more favorable foraging conditions whicli, we hypothesize, e.xist in tiie convergent zone. OTHER VARIATIONS IN VOLUME OF STOMACH CONTENTS ^\^len the stomach-content volumes are clas- sified according to an arbitrary scale, the results (table 9) indicate for both species a rather low percentage of empty or near-empty stomachs; the average stomach contained a relatively small amount of food. This may mean that feeding is almost continuous, as contrasted with an irregular or spasmodic feeding habit, and that these fish have a high rate of digestion. For instance, it is hard to believe that a food volume of less than 100 CC, which was found in more than 50 percent of the stomachs (table 9), constitutes a daily or even semidaily ration for these large active fish. Unfortunately, our food studies provide no in- formation on rate of food consumption or digestion. In longline fishing, the gear is ordinarily set at daybreak and is hauled in during the afternoon. The time of landing is known, but not the time that the fish took the hook. On some cruises, 50 percent or more of the tuna are dead when landed. One might assume that these fish hail been on the line for a longer period of time than the fish that were landed alive. On the basis of this hypothesis we examined the records from certain cruises for which we had the greatest number of observations supplying information on condition when landed. These data, as summarized in table 10, seem to indicate that the fish that were dead when landed contained larger volumes of food, on the average, than those that were landed alive. Although we cannot satisfactorily explain this difi"erence, we believe that it may be related to the tendency for more dead fish to occur on the deep hooks than on the liooks fishing at shallow and intermediate depths; and in the yellowfin, at least, we have found an increase in volume of stomach contents with depth of capture (table 3). A combination of these factoi-s might produce the results shown in table 10. 78 IISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Variations in average volume of food per stomach in pelation to depth of capture, comparing yellowfin and bigeye tuna taken by longline in the Countercurrent and convergent zone Table 8. [Organisms making up less than 1 percent of the total food volume for each depth category were omitted from table; volume is measured in cc] Convergent zone Countercurrent Yellowfin Bigeye Yellowfin Bigeye Food organisms Hooks 1 and 6 (shal- low) Hooks 2 and 5 (inter- medi- ate) Hooks 3 and 4 (deep) Hooks 1 and 6 (shal- low) Hooks 2 and 5 (inter- medi- ate) Hooks 3 and 4 (deep) Hooks 1 and 6 (shal- low) Hooks 2 and 6 (inter- medi- ate) Hooks 3and^ (deep) Hooks 1 and 6 (shal- low) Hooks 2 and 5 (inter- medi- ate) Hook.'i 3 and 4 (deep) Crustacea (total) ._ 0.6 0.6 0.8 0.5 2.1 4.3 4.2 28.4 1.7 2.9 3.7 0.1 0.5 0.3 0.1 0.6 3.1 2.6 Squids (total) 25.5 27.6 39.4 9.2 47.7 52.1 37.4 46.5 110.0 138.5 47.8 Loliginidae: 1.8 4.6 12.3 '"'i.'i 2.8 18.0 43.0 39.6 4.9 11.0 14.6 3.1 20.1 5.2 2.5 1.3 Enoploteuthidae: 1.2 3.3 2.1 1.8 1.9 3.4 7.1 5.5 Omniastrephidae: 5.5 7.3 4.6 2.0 2.1 5.1 5.0 8.8 5.0 7.3 2.9 1.4 3.6 3.4 2.9 8.1 103.1 17.6 4.5 Cranchiidae: 12.0 14.3 2.0 8.4 2.3 15.4 10.6 2.1 0.9 12.3 7.5 1.9 2.5 1.4 3.3 1.5 2.6 1.5 1 1 Argonautidae: 1.6 10.3 54.7 1.0 3.5 53. 8 1.2 74.2 0.9 12.9 1.6 68.0 1.1 Fish (total) 32.5 114.6 70.0 1.2 2.7 0.6 1.8 0.5 111.9 73.3 107.9 60.7 Sternoptychidae: 9.7 Unidentified Sternoptychidae 0.6 9.0 3.6 1.1 Sudidae: Unidentified Sudidae 26.3 4.2 2.1 Exocoetidae: 1.7 0.1 3.4 1.4 3.6 „fU*iJf i^Mm) 2.0 1.2 Carangidae: 1.8 2.4 2.4 15.3 1.1 1.0 Bramidae: 10.1 1.4 2.6 8.5 4.6 0.4 3.5 8.2" 1.3 9.9 14.9 7.7 4.0 12.4 4.1 0.6 61.7 4.5 1.4 17.9 4.6 3.5 2 4 1.3 Gempylidae: 2.0 1.2 1.3 1.7 5.2 1.3 ' " 0.6 39.0 4.2 0.4 2.3 0.1 1.2 4.3 1.0 15.5 3.2 9.5 9.5 9.5 0.4 2.3 Unidentified Gempylidae . .... 2.9 Nomeidae: 0.8 8.1 4.4 12.8 15.8 4.2 1.1 28.2 41.1 6.4 9.2 0.3 1.0 Echeneidae: 1.4 0.9 0.3 3.8 0.8 0.2 1.2 0.2 Balistidae: 4 3 0.8 Ostraciidae: 1.9 2.3 3.3 0.7 1.3 Molidae; Fanzania sp Other foods 4.0 0.5 95.6 6.8 1.7 0.6 161.9 0.4 90.8 0.8 130.4 '42.2 0.7 166.0 0.1 All foods 115.3 66.5 109.5 25 135 106 1.0 183.4 248.6 103.9 Number of stomachs examined 49 141 120 0.8 49 141 120 O.g 61 141 120 1.1 3 146 139 0.3 14 153 164 1.0 28 141 130 0.9 9 131 97 0.7 25 139 115 1.4 7 153 162 1.1 13 146 144 1.7 22 140 Average weight (lbs.) 127 Average volume of food per pound of body weight (cc.). 0.8 FOOD OF BIGEYE AND YELLOWFIN TUNA 79 Table 9. — Distribution of the volume of stomarh contents of /,S9 yelloivjin and 166 bigeye caught by longline fishing in the central Pacific Less than 140 cm long 140 cm. or larger Volume (cc.) Number Percent of total number Accumu- lated per- centage Number Percent of toUl number Accumu- lated per- centage Empty; (0-0.9): 6 2 17 * 21 9 46 10 49 11 31 13 15 12 4 2 0 0 3.2 3.2 9.0 6.3 11.2 14.3 24.0 15.8 26.1 17.5 16.5 20.6 8.0 19 0 2.1 3.2 0.0 0.0 3.2 3.2 12.2 9.S 23.4 23.8 47.4 39.6 73.5 57.1 90.0 77.7 98.0 96.7 100.0 100.0 4 6 17 10 22 9 50 13 66 1.6 5.8 6.7 9.7 8.7 9.4 19.9 12.6 9fi f Bigeve _ 5 8 1.0-9.9: Yellowfln 8 3 Bigeye 15 5 10.0-24.9: Yellowfin 17 0 Bigeve _ 24 9 25.0-19.9: YelloHfln... 36 1 Bigeve _ 37 5 50.0-99.9: Yellowfin A3 1 18 17 .1 100.0-199.9: Yellowfin... 56 24 30 17 3 5 3 1 22.2 23.3 11.9 16.5 1.2 4.9 1.2 1.0 85 3 78 3 200.0^99.9: Bigeve 94 8 500.0-999.9 Bigeve 99 7 Yellowfln 100 0 Bigeye 10(1 0 Total: Yellowfin 188 63 251 103 T.\BLE 10. — Summary of data relating average volume of stomach contents to condition of fish, whether dead or alive, at time of landing Yellowfin Bigeye Cruise Average volume of stomach contents Number of stomachs examined Average fork length Average volume of stomach contents Number of stomachs examined Average fork length Hugh M. Smitl> cruise 11: Landed dead cc. 93.7 50.6 96.6 71.4 64 23 37 32 cm. 135 141 143 144 cc. 174.9 83.6 11 24 cm. 152 Landed alive John H. Manning cruise 14: Landed dead . Landed alive Jotin R. Manning cruise 15: Landed dead 100.9 280.5 9 10 119 Landed alive 13" SUMMARY AND CONCLUSIONS 1. This study is based on the quantitative analysis of the stomacii contents of 166 bigeye tuna {Parathunnus sibi) and of 439 yellowfiii tuna {Neothunnus macropterus) caught at tlie same time or nearly the same time as the bigeye. 2. These tuna were captured in the central Pacific during the period October 1950-Juiic 1953 by means of longline-gear fishing at deptlis of 150 to 500 feet. 3. The food of the yellowfin consisted of fish (62 percent by volume), squid (29 percent), other moilusks (7 percent), and crustaceans (1 percent); the food of bigeye consisted of fish (62 percent), squid (33 percent), other moilusks (3 percent), and crustaceans (2 percent). 4. Both species of tuna apj)ear to utilize a great variety of animal food, ranging from small plank- ton organisms to large squid and fish. Food items of major importance to both spt^cies were pomfret {Collybus drachme), snake mackerel (Gempylus ser- pens), and squid of the families Ommastrephidae and Loliginidae. 5. This great diversity of diet suggests that many forms of fish, squid, and shrimp — if available through culture or capture might be eflTective as live i)ait or longline bait in tuna fishing. 80 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 6. Stomatopod crustaceans, common in the food of yellowfin, were completely lacking from the bigeye stomachs. The young of other tunas, mostly skipjack, formed a much more important part of the yellowfin diet than of the bigeye diet. 7. In both species, the larger tuna had more food in their stomachs than did the smaller fish, but the larger fish contained less food per pound of body weight than did the smaller fish. There were few completely empty stomachs. 8. In both tunas, the smaller individuals con- sumed a greater proportion by volume of crusta- ceans and fish and a lesser proportion of mollusks than the larger size group. The same fish families were prominent in the diet of botli size groups. 9. There was an increase in volume of stomach contents with depth of capture for the yellowfin; in the bigeye, the largest volumes were found in specimens from intermediate depths. There was no marked variation in composition of stomach contents over the range of depth sampled (esti- mated at 150 to 500 feet), which may be evidence that both the forage organisms and the tuna range rather freely tlnougliout this water layer. 10. In both yellowfin and bigeye, fish were con- sumed in greatest amount during the period April- July, and in least amount during August and Sep- tember. There was little correspondence between the two species in the seasonal variation in the other major food items. 11. In respect to longitudinal variations in the food, the two species were similar in the lower volume of total fish in the stomach contents of those tunas captured in the central region (140°- 150° W. longitude) of the sampled area. The utilization of specific foods did not vary with longitude in any regular pattern for the two species. 12. When classified according to natural subdi- visions of the equatorial current system, the volume of stomach contents in the bigeye varied directly witli the longline catch rate, while in tlie yellowfin there was little change in volume of stomach con- tents with even a marked change in catch rate. 13. Tuna ttiat were dead when landed con- tained, on the average, more food in their stom- achs than those landed alive. 14. Despite the difl'erences that we have pointed out, tlie foods of the yellowfin and bigeye are re- markably similar. We conchuh', tlierefore, that when occupying the same general area the two species have essentially the same feeding habits. If there is any marked food selection, it must be exercised b\' seeking different areas for feeding. LITERATURE CITED BERfi, L. S. 1947. Classification of fishes both recent and fossil. J. VV. Edwards, Ann Arbor, Mich. 517 pp. Brock, V. E. 1949. A preliminary report on Parathunnus sibi in Hawaiian waters and a key to the tunas and tuna- like fishes of Hawaii. Pacific Science, vol. 3, No. 3, pp. 271-277. Cromwell, Townsenu 1951. Mid-Pacific oceanography, January through March 1950. U. S. Fish and Wildlife Service, Spec. Sci. Rept. — Fisheries No. 54. 77 pp. 1953. Circulation in a meridional plane in the central equatorial Pacific. Journal of Marine Research, vol. 12, No. 2, pp. 196-213. Kanagawa Prefectire Fisheries Experiment Station 1951. Report of South Seas tuna fishery investigations. 162 pp. (Partial tran.slation from the Japanese by W. G. Van Campen in the files of Pacific Oceanic Fishery Investigations.) King, J. E. 1954. Variations in zooplankton abundance in the cen- tral equatorial Pacific, 1950-52. Fifth Meeting, Indo- Pacific Fisheries Council, Symposium on marine and fresh-water plankton in the Indo- Pacific, pp. 10-17. KiNfj, J. E., AND Joan Demond 1953. Zooplankton abundance in the central Pacific. U. S. Fish and Wildlife Service, Fishery Bull. 54, vol. 82, pp. 111-144. MlRPHY, G. I., AND R. S. ShOMURA 1953a. Longline fishing for deep-swimming tunas in the central Pacific, 1950-51. U. S. Fish and Wildlife Service, Spec. Sci. Rept. — Fisheries No. 98. 47 pp. 1953b. Longline fishing for deep-swimming tunas in the central Pacific, January-June. 1952. U. S. Fish and VV'ildlife Service, Spec. Sci. Rept. — Fisheries No. 108, 32 pp. 1955. Longline fishing for deep-swimming tunas in the central Pacific, August-November 1952. U. S. Fish and Wildlife Service, Spec. Sci. Rept. — Fisheries No. 137. 42 pp. Nakamira, Hiroshi 1949. The tunas and their fisheries. Tokyo: Takeuchi Shobo. Tran.slated from the Japanese by W. G. Van Campen, U. S. Fish and Wildlife Service, Spec. Sci. Rept.— Fisheries No. 82, 1952. 115 pp. NisKA, E. L. 1953. Construction details of tuna longline gear used by Pacific Oceanic Fishery Investigations. U. S. Fish and Wildlife Service, Commercial Fisheries Review, vol. 15, No. 6, pp. 1-6. FOOD OF BIGEYE AND YELLOWFIN TUNA 81 Orsr, Tamio I')54. Analysis of the Hawaiian longline. fishery, 1948-52. U. S. Fish and Wildlife Service, Commer- cial Fisheries Review, vol. 10, Xo. 9, pp. 1-17. RKIST.IKS, .J. \V., and J. E. KiN Number Percent Cubic centi- meters Percent ' Vert e brat a— Continued 3 13 1 5 2 27 4 1 1 6 1 4 1 13 3 1 0.2 1.4 0.2 0.9 0.2 3.0 0.7 0.2 3.4 10.7 7.0 321.3 31.5 295.3 6.6 1.9 1 7 1 0.6 3.6 0.5 12.0 Hemirhamphidao (halfbeaks) Exocoetidae (flying fishes): 0.7 Par^erocoetus sp_ 0.7 6 1 5 1 3.0 0.6 64.8 1.0 0.3 Bregmacerotidae: Lophotidae (oarfishes) 1 2 1 1 1 1 1 1 0.6 0.6 0.6 0.6 2.8 40.0 190.0 3.9 Trachypteridae (ribbon fishes): TrachypteTus sp 0.2 Regalecidae (oarfishes): 0.9 10 1 5 3 1 2 0.7 0.2 0.5 45.0 94.0 21.6 0.1 0.2 Diretmidae Caulolepidae: A. COTTiUtUS 9 1 2 1 1.2 0.6 2i.6 6.4 27 2 3 1 0.7 0.2 94.0 0.2 0.2 Holoeentridae (squirrel fishes): 1 1 0.6 0.1 Zeidae (John Dories): 2 2 0.5 5.3 Caproidae; Antigonia sp .1. capros , . ._ 11 150 1 1 0.6 0.6 33.0 55.4 0.1 Athcrinidae (sityt'rsides) ; • a 0.2 Polynemidae (fhreadflns) / 1 2 1 0.2 0.2 7.2 29.6 Priacanthidae (ciitalufas): Prica7}thiis cnieiitatus 1 1 0.6 6.8 Apngonidae (cardinal fishes): y^ Pnrascombrops peUucida -^ r 1 7 8 1 2 2 0.2 0.5 0.5 0,6 6,9 4,6 Scorn bropidae: 4 1 1 1 1 1 0.6 0.6 0.6 9.2 1.9 1.2 Sco mbrops sp Unidentified Scombropidae Carangidae (jacks): 1 1 1 1 0.2 0.2 88,0 118,0 0.2 0.3 1 1 0.6 8.6 1 1 1.012 88 2 244 2 3 2 1 8 1 1 190 33 1 1 66 2 1 2 1 1 0.2 0.2 43.3 7.5 0.2 0.2 15.0 0.5 0.2 0.5 0.2 0.2 115,7 1.8 3. 465. 9 786.3 3.0 4,2 675,2 118.1 1.4 3.3 1.5 21.0 0.3 Unidentified Carangidae Bramidae (pomfrets): CoUyhus drachme 7.8 1.8 76 34 26 22 15.7 13.3 2. 550. 8 835.5 11.4 3.7 P ocellatus 1 18 1 10 0.6 6.0 14.8 476.9 1.5 0.3 2.1 Coryphaenidae (dolphins): Unidentified Coryphaenidae 3 2 1.2 2.6 Leiognathidae MuUidae (goat fishes): 1 1 0.2 1.2 1 i 0.6 3.7 Chat'todontidae (butterfly fishes) 5 3 26 12 n 31 1 202 1 10 7 3 2 10 3 1 10 1 80 1 I 1 0.7 0.5 2.3 0.7 0.2 9.2 1,8.1 0,2 0.2 0.2 29.3 4.4 39.5 24.8 12.8 97.9 21.9 2.119.5 0.4 21.5 36 r I 1 7 1 1 3 0.6 0.6 1.8 7.6 2.0 29.9 Chanipsodontidae' Ckampsodon sp Chiasmodontidac 0.1 Acanthurida'' (sureeon fisjics): . \ 0.2 Geiiipylidae (snake mackerels): 4.7 86 2 1 43 i 25.9 0.6 0.6 2. 023. 7 2.4 4 A 9.1 1 2 18 48 1 1 18 0.6 0.6 0.6 10.8 5.5 2.8 80.0 .■(76 :< Rexea solandrii _ " i' 149 i 62 0.'? 14.1 i.8' 7X6.2 OA UnidimtificMl (Jcmpylidae i.8 1.7 See footnote at end of table. FOOD OF BIGEYE AND YELLOWFIN TUNA 85 Table II.— Check-list of food organisms found in the stomachs of 439 yellowfin and 166 bigeye tuna captured on longline in the Central Pacific, 19S0-.53— Continued {Family names of fishes are as given in Berg 1947) Yellowfin Bigeye Food organisms Number of organ- isms Stomach.'! in which occurred Aggregate total volume Number of organ- isms Stomachs in which occurred Aggregate total volume Number Percent Cubic centi- meters Percent ' Number Percent Cubic centi- meters Percent • Vertebrala— Continued Sconibridju- (mackerels); Scomber sp ♦-— -•- 9 4 41 14 2 41 2 3 16 4 1 12 0.5 0.7 3.6 0 9 0.2 2.7 21.4 68.6 102.3 30.8 10.0 79.0 0.3 0.2 Nonn'ida<' (rudder fishes): •. 34 8 6 3 3.6 1.8 76.9 22.8 0 3 Cuhiceps sp ---■ 0 1 C. tfinmpsoJti _.. __ 0.2 4 1 15 3 2 1 10 1 1.2 0.6 6.0 0.6 10.4 5.6 75.0 17.5 .\Ioitodactytv3 sp. _ 58 2S 5.7 122.6 0.3 0 3 Thunnldae (tuna fishes): .Sardi^p --- 48 2 1 68 1 6 16 4 4 12 26 8 23 fi 1 8 11 33 78 3 18 11 ) 9 3 ) ^ ' 3 13 19 17 2 1 12 1 4 8 2 3 11 20 5 12 3 1 7 7 14 29 3 13 10 6 2 2 7 1 5 1 10 12 3.9 0.5 0.2 2.7 0.2 0.9 1.8 0.5 0.7 2.5 4.6 1.1 2.7 0.7 0.5 1.6 1.6 3.2 6.6 0.7 3.0 2.3 1.4 0.5 0.5 1.6 0.2 1.1 0.2 2.3 2.7 3,497.1 962.0 32.2 1,961.0 70.0 12.2 144.8 26.0 14.3 145.8 116.3 72.6 412.5 134.0 6.8 71.9 27.8 78.3 322.7 15.3 243.4 27.1 77.0 57.5 215.7 lis. 1 10.0 66.1 54.0 162.3 4,318.8 7.8 2.2 Seothunnu^ macropttruit 1 1 0.6 662.0 3 0 Germn aialunijn 4.4 0.2 2 2 1.2 224.0 1 0 rnidenllflod Thunnldae 1 1 0.6 1.0 E'^heneidae (renioras): 0.3 Rfmara sp 0.3 0.3 0.2 0.9 0.3 Unidentified Echeneidae. 7 3 i.8 S.6 Balistidae (trigger fishes): Balistes sp. li. iiycteris li. riuQens . Xanthkhfbijs sp_.. I'nidentified Balistidae ._ 0.2 Monacanthidae (file fishes).-. .. . Ostraciidae (trunk fishes); Ostrncioti sp 0.2 0.7 15 8 1 2 1 2 2 1 2 1 1.2 1.2 0.6 1.2 0.6 23.0 29.7 2.5 69.5 0.6 0 1 Lactoria s\) 0.5 0 3 Unidentified O.straciidae 0.2 0.1 0.5 0.3 Lagocephalus sp L...^... ? Telrodon sp . / I'nidentified Tetrodontidae. 1 1 0.6 16.6 Diodontidae (porcupine fishes); C.ailinis >. / 0.1 0.1 0.4 9.7 Unidentified Diodontidae / 1 2 1 1 2 0.6 0.6 1.6 0.1 113 0 Molidae (sun fishes); Ranzajtja SI) . . . .... n (I Unidentified Molidae 151 0 n " Antennariidae (frogfishes): 1 1,160 1 213 0.2 48.5 0.8 2.968.9 44, 679. 5 439 Other and unidentified fish 6.6 511 81 48.8 3. 884. 0 22, 297. 3 166 17.4 Total food Number of stomachs examined ' Given only when 0.1 percent or greater. U. S. GOVERNMENT PRINTING OFFICE 1956 O— 388734 UNITED STATES DEPAETMENT OF THE INTERIOR FISH AND WILDLIFE SERVICE CORRECTIONS for Fishery Biilletin 108, COMPARATIVE STUDY OF FOOD OF BIGEYE AND YELLOWFIN TUNA IN THE CENTRAL PACIFIC Page 62, second coliimn, line 10: for "latitudes" insert "longitudes." Pages 69 and 73: Figures 9 and 10 are transposed (the captions are correctly numbered). Page "JB, table 8: Under Atherinidae, for "Atherinus insularum" insert "Pranesus insularum." Page Qk, table 11: Under Exocoetidae, for " Paraexocoetus" insert "Parexocoetus ." Under Atherinldae, for "A. insularum" insert "Pranesus insularum." Under Acanthuridae, f or "''^epatus sp." insert "Acanthurus sp." Under Gempylidae, for "Neoephinnula" insert " Neoepinnula . " Page 85, table 11: Under Thimnidae, for "Sardi" insert "Sarda." Under Tetrodontidae for "Sphoeroides lagocephalus" insert "Lagocephalus lagocephalus . " Under Diodontidae, for "Cheilomycteris" insert "Cheilomycterus." 11067 P f LIFE HISTORY OF LAKE HERRING OF GREEN BAY, LAKE MICHIGAN By Stanford H. Smith, Fishery Research Biologist The lake herring, or shallowwater cisco, Leu- cichthys artedi (LeSueur) , occurs in all of the Great Lakes and in many inland lakes of the St. Law- rence, Hudson River, and upper Mississippi River drainages (Hubbs and Lagler, 1949), and has rather general distribution throughout Canada and Alaska in lakes and some rivers, and in Hudson and James Bays (Dymond 1933, 1943, and 1947). Close relatives of the lake herring have a circumpolar distribution in the glaciated areas of the Northern Hemisphere. The lake herring is a member of the family Coregonidae, a complex and not well understood group of fish. Much confusion resulted from early attempts to describe tliis group in the Great Lakes (see Koelz 1929; pp. 311-314). The dis- agreement stemmed botli from the fact that early workers studied only small numbers of specimens from one or a few localities and from the high degree of individual and geographic variability in size, shape, and taxonomic counts that charac- terizes this group. Koelz made a comprehensive ta.xonomic study of coregonids inhabiting the Great Lakes and Lake Nipigon based on about 15,000 specimens from many parts of each lake. He recognized the high degree of variabihty in the group and was able to organize the confused taxonomy. What had been described as several species by comparisons of a few specimens often were found to be representatives of a single species that varied greatly in form over its range. Koelz recognized the different species inhabiting the several lakes and thus established a system of nomenclature which has since been adequate for the species of the Great Lakes. He recognized all coregonids of the Great Lakes as belonging to the family Coregonidae and the genera Coregonus (Artedi) Linnaeus, Leucichthys Dybowski, and Prosopium Mihier that had been described from studies of coregonids over their entire range. A few authors have deviated recently from the system of classification used by Koelz and have placed Leucichthys and Prosopium in the genus Coregonus. I prefer to retain Leucichthys as a genus because it represents a well-defined group in North America. The Leucichthys group in Europe is ascribed to the subgenus Aryyrosomus; however, European workers have written me that these fish are distinct from other coregonids of that continent. The consolidation of all groups under the single genus Coregonus disregards the recognizable divergence of the phyletic lines represented by the three genera. It is true that the high degree of morphological plasticity characteristic of the coregonids sometimes causes morphometric and even gross appearances to approximate or, in isolated instances, to overlap each other. This superficial parallelism may occasionally hide the distinctness of the groups, but it cannot overrule the primary genetic divergence that is so clearly shown by the distributional pattern of each group. For each genus there is a central range where its members are highly variable {Coregonus in Europe, Prosopium in northwestern North America, Leuci- chthys in northeastern North America), and where they are usually divided into several spe«?s7" Range extensions of each group are characterized by lesser morphological variability and at the extremes only one or two relatively stable species remain. Ambient morphological divergences in isolated populations of one group may in some instances parallel developments common among members of another group and thereby tend to obscure the distinctness of the groups. Such occurrences cannot, however, be interpreted as incomplete separation of the groups. 1 believe the separate genera describe these phyletic lines in the clearest and most useful manner and shoidd be retained in keeping with this basic purpose of modern taxonomy. 87 88 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Because of its varied form in different localities, the lake herring is known by more than one com- mon name. Names used in this work where other authors are quoted are sometimes cisco or tullibee rather than lake herring. These names are most often applied to the deep-bodied forms that occur in inland lakes and Lakes Erie and Ontario. Most lake herring from the upper Great Lakes, however, are of the characteristically shallow-bodied form that is most commonly termed "lake herring." The lake herring is of major importance in the commercial fishery of Green Bay. Fluctuations in its abundance bring a degree of economic uncer- tainty to the people who depend upon this fish for part of their livelihood. Although the lake herring has been important in the commercial catch of Green Bay, little has been known about it. Knowledge of this species provides greater under- standing of its reactions to changing environ- mental conditions, and also is required to develop management principles that would allow maximum utilization of the species without depleting the population. A study of the life history of the Green Bay lake herring was initiated in 1948 when field collec- tions of scales were made by Dr. Ralph Hile, of the United States Fish and Wildlife Service, as part of a cooperative project with the Wisconsin Conservation Department for the study of Green Bay fish populations. After 1950, field work was carried on by the author with the help of Leonard S. Joeris and Donald Mraz of the Sturgeon Bay field station of the Service's Great Lakes Fishery Investigations. During 1952, the research vessel FWS Cisco, operated by the Great Lakes Fishery Investigations, was available for approximately 1 week each in May, July, and October for the study of the distribution of lake herring in Green Bay. Some material on the lake herring of Green Bay was collected during parts of two other cruises of the Cisco in May and June. The author is most grateful to Drs. Ralph Hile, John Van Oosten, and James W. Moffett, U. S. Fish and Wildlife Service, and to Dr. Karl F. Lagler, University of Michigan, for valuable guidance during the conduct of the study. GENERAL FEATURES OF GREEN BAY Green Bay is a nearly detached arm of Lake Michigan with its long axis roughly parallel to the northeast shore of the lake. Morphometric features of Green Bay and Lake Michigan are compared in table 1 . The two bodies of water are similar in that they are long and narrow, but they differ greatly in depth and area. The greatest length of Green Bay is about 118 miles on a northeast-southwest axis between the upper end of Big Bay de Noc and the city of Green Bay, Wis. (fig. 1). The greatest width, about 23 miles, is on a northwest-southeast axis in the region of the northern island passages. The area of Green Bay included within a line drawn between the town of Fairport and the tip of the Door Peninsula near Gills Rock is approximately 1,590 square miles. The greatest depth, about 160 feet, is just northwest of Washington Island. The bay is relatively shallow — mean depth, 51 feet. One- third of its area is less than 30 feet deep and only 1 1 percent is more than 100 feet deep. Table I. — Morphometric features of Lake Michigan and Green Bay [Data from the U, S, Lake Survey Chart Nos. 7 and 70, 1953 edition] Measurement Greatest length (miles) . Greatest width (miles). Shoreline length (miles) Area (squarf miles) Volume (cubic miles) __ Greatest depth (feet)... Mean depth (feet) LakeMich- igan 307 ng 1,661 22.400 1,165 932 274 Green Bay 118 23 379 1,590 15 160 51 Four major channels in the northern island area with depths of 45 to 130 feet connect Green Bay with Lake Michigan. The manmadc Sturgeon Bay Canal which is 160 feet wide and 20 feet deep joins the two bodies of water in the southern area at Sturgeon Bay. A study being carried on by the Great Lakes Fishery Investigations of the United States Fish and Wildlife Service has pro- vided some data about the exchange of water between Green Bay and Lake Michigan. Al- though not as comprehensive as might be desired, the data do give a general idea of the water -ex- change system between the lake and bay, and of water movements within the bay. An outstanding feature of the water movements in Green Bay is tlie high degree of irregularity in direction and velocity. The direction and rate of water movements are believed to be governed . mainly by wind and barometric pressure. Flow of water into the bay from rivers is believed to be of minor importance in the major water move- ments except during spring runoff. Movement LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 89 Figure 1. — Map of Green Bay showing locations of experimental gill-net -stations. Triangles, shallow-water stations (A, 30 feet; B, 40 feet); squares, 60-foot stations; circles, 90-foot stations. 90 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE of Lake Michigan water into Green Bay is charac- terized more by surges than by a regular movement. Surges into the bay result primarily from seiche action set in motion by wind and pressure changes over Lake Michigan. The resultant currents in tlie bay cause a tremendous amount of mixing. In the northern passages the sequence of inflow, mixing, and outflow result in a great amount of water excliange between the bay and the lake. Evidence of a high degree of exchange in the northern area is found in the relatively clear. Lake Michigan-type water that lacks the deep green color produced by dense phytoplankton growth characteristic of the remainder of the bay. Def- inite lines of demarcation cannot be made on this basis, however, because of mixing of water masses. Clear lake water is sometimes observed in the Sturgeon Bay area, but here a sharp line of demar- cation is usually present between the two types of water. This condition indicates that little mixing occui-s before the lake water is returned with an outgoing surge through the canal. In addition to water movements propagated by currents and water-level changes in Lake Michi- gan, the water in the bay itself is subject to indig- enous seiches and currents caused by local condi- tions. The systems operating simultaneously in lake and bay, as they must most of the time, result in extremely complex and irregular water move- ments. The water level in Green Bay is subject to almost continuous chaiige. A change of a foot an hour is not uncommon and occasionally a drop of several feet in the southern end of the bay strands fishing boats in shallow harbors. Although a complex resonance pattern is characteristic of water-level charts of Green Bay, peaks occur at intervals of about 12 hours. The peaks show no relation to the movements of the moon. Typical spacing of the peaks within the 24-hour period can be completely disrupted by severe storms after which a new system is established with peaks occurring at different hours of the day but again at 12-hour intervals. Some of the effects of water movements on the water temperatures in Green Bay will be shown later in a discussion of the distribution of lake lierring. ECONOMIC IMPORTANCE OF THE FISHERY Green Bay supports one of the most productive commercial fisheries of the Great Lakes and the lake herring is a major contributor to the catch. Hile, Lunger, and Buettner (1953) showed that on the average 28.8 percent of the total pounds of all species taken in the State of Michigan waters of Green Bay consisted of herring. In 1952, the last year for which complete statistics and values are available, the lake herring catch of Green Bay (both Wisconsin and Michigan) amounted to 9,121,600 pounds and had a value to the fishermen of $456,080. This catch represented 94.1 percent of the production of this species in all of Lake Michigan and 38.7 percent of the lake herring production of all United States waters of the Great Lakes. The commercial production of the lake herring in Green Bay is characterized by wide annual and seasonal fluctuations. The catch in Michigan waters of Green Bay ranged from 1,515,000 to 11,850,000 pounds (average 5,841,000 pounds) from 1891 to 1908 (Hile, Lunger, and Buettner, 1953) and averaged 82.4 percent of the total pounds of all species taken. In a later period (1929-49) there was a marked drop in the produc- tion to between 160,000 and 2,668,000 pounds (average 1,070,000 pounds) which contributed an average of 29.9 percent to the catch of all species. The production of lake herring in Michigan and Wisconsin waters during the years for which reliable records are available for both States (table 2) show wide variation seasonally and annually. Fluctuations of the catch are influ- enced primarily by weather, availability and abundance of other species with higher market value, and the abundance of lake herring itself. Thus, the causes of fluctuations are difficult to ascertain, but the great difference between the 1891-1908 and 1929-49 data on the Michigan waters of Green Bay (82-percent drop in average production) shown by Hile, Lunger, and Buettner indicates that the population must be subject to wide variations. The present study, however, has been conducted in years (1948-52) when total production has been high and relatively stable LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 91 (6,320,000 to 9,122,000 pounds) compared to the 10-year period after 1936. Table 2. — Commercial catch of lake herring in Wisconsin and Michigan waters of Green Bay, by quarters, 1936-53 [In thousands of pounds] Year January- March April- June July-Sep- tember October- December Total 1936 1,121 1,475 1,422 381 653 382 299 330 249 581 854 1,686 2,260 1,354 1,788 1,322 1,818 587 897 1,127 779 513 399 347 249 428 292 791 1,381 1,223 1,576 1,622 1,065 593 675 842 254 246 142 158 146 129 119 303 333 1,003 687 338 303 376 747 563 384 1, 597 1,254 665 449 291 442 182 223 131 1,289 2,294 1,832 3,288 3,041 3,663 5,049 6,066 4,081 3,869 1937 4, 102 1938 3,008 1939 1,.')01 1940 1,489 1941 1,300 1942 849 1943 1944 .. .. 1,284 1,005 1945 3,664 1946 5 216 1947 5,285 1948 7,462 1949 6,320 1950 1951 6,892 7,711 1952 9,122 19.'J3 . 5,894 Average, 1936-53 Percentage... 1,031 24.4 822 19.5 376 8.9 1,991 47.2 4,220 Note. — These data are from summaries of commeicial catch records made by the V. S. Fish and Wildlife Service for Michigan waters and by the Wisconsin Conservation Department. Data for Wisconsin prior to 1942 included lake herring taken in that area of Lake Michigan adjacent to Green Bay, but catches in this area arc characteristically small and are not believed to influence trends. Because of highly seasonal production and rapid deterioration in handling and storage, the lake herring brings a low average price (5 cents per pound to the fishermen of Green Bay in 1952) and much of the catch is used for animal food. Given better markets and improved handling, the species may become a more important source of human food. The lake herring has some small value as a sport fish. Its habit of feeding principally on small planktonic organisms and its disinclination to strike at lures has caused it to be overlooked by anglers using conventional methods. During recent years, however, fishermen have found that when lake herring are feeding on mayflies tliej' will also strike at artificial flies. A sports fishery during the period of mayfly emergence is growing rapidly in popularity in the northern areas of Lakes Huron and Michigan. Some large lake herring are also taken with minnows as bait. A certain amount of angling for lake herring is carried on through the ice botli on the Great Lakes and on inland lakes. COLLECTION OF DATA Scale samples and data on weight, length, sex, and state of development of se.x organs w'erc ob- tained on 4,390 specimens. Collections made between May 26, 1948, and January 22, 1952, w^ere taken from commeicial pound nets and gill nets as indicated in table 3. Scale samples of May, July, and October, 1952, were from fish captured in experimental gill nets. Table 4 lists all fish taken in experimental gill nets for whi(rh length and weight measurements and sex determi- nations were made; in some of the May collec- tions, however, weight and sex data are missing. Table 3. — Collections of Green Bay lake herring from which scale samples were taken, 1948-5S Date Locality Gear used ' Num- ber of nsh IH8 Mav26. Point Comfort .-. Schumachers Point... Schumachers Point... Point Comfort 262 Oct. 12... 238-inch gill net 152 19(9 Feb. 16 345 May 13 . do 200 13 Pensaukee do 241 Oct. 5 do do 283 19S0 Feb, 22 Schumachers Point do 341 27 .. do-— 166 June 21 Fish Creek do 62 22. do. do 25 July 13 ... do 43 Sept. 14 OUls Rock do 201 Nov. 29.... Fo.x do 107 30 2?8-inch gill net 108 Dec. 4 Sister Bay.. 112 1951 Feb. 20 2?8-inch gill net 168 20.-.- do 29 20 Ingallston. . do 223 22 .- Schumachers Point... Point Comfort do 189 May 8. do 143 June 152 0111s Rock .. do 11 19 do 80 21... Gills Rock Fox Gills Rock do do do 26 Aug. 20 59 29 15 Nov. 11 Pensaukee 2^2-inch giU net 80 Dec. 12 Gills Rock 79 I9SI Jan. 21 Escanaba Pound net 90 22 Pensaukee.. Station D . do 92 Mav 8 2-lnch gill net 44 11 113 Julv 21 Station L Station C do do 30 24 19 Oct. 22 23 . Station B Station I^ do do 46 19 24 Station I do 187 Total 4,390 ' See text, p. 95, for comment* on mesh sizes of pound nets. » This is a selected sample. All other samples are either random or repre- sent the entire catch of one net. 388748 O — 57- 92 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 4. — Lake herring taken in experimental gill nets in 1952 Date' Sta- tion! Num- ber of flsh Date Sta- tion! Num- ber of flsh May 2. D I D E F O E F G A B C J K L J K 80 28 58 26 115 118 85 31 31 13 32 5 81 19 46 78 51 July 21 L H I C D A B C D H I J K L 85 ^ 6 22 22 10 8 24 11 24 19 11 27 6 11 Oct. 22 15 22 22 22 46 22 143 22 23 19 25 24 140 25 24.. 25 26 25 Total ... 25 Tnnp n 179 130 12 12 139 July 21 2,039 21 ' Weight and sex data lacking for some May collections. 2 See figure 1 for location. Total length (tip of the snout to the end of the tail, lobes compressed) was recorded to the nearest 0.1 inch. Weights measured on a spring scale, with 18-ounce capacity, were recorded to the nearest 0.1 ounce. All lengths are given in inches and weights in ounces unless otherwise stated. Samples from the commercial fishery were cap- tured in standard fishing gear designed primarily for lake herring. Netting of pound nets used in the lake herring (and smelt, Osmerus mordax) fishery customarily has meshes (in the pot) of 1)4 to 2 inches, extension measure as manufactured. Nets of these mesh sizes are capable of capturing lake herring smaller than any taken from them during this study. Consequently, mesh size need not be considered as a selective factor in the treat- ment of samples from pound nets. Most small- mesh gill nets used in the Green Bay herring fishery have a mesh size of 2% inches (allowable range 2)^ to 2% inches, depending on season, loca- tion, and conditions) extension measure. One col- lection in southern Green Bay was taken from a 2)Mnch-mesh gUl net on November 11, 1951. Experimental gill nets used to collect lake herring in the summer and fall of 1952 are described in Vertical Distribution in Green Bay (p. 128). All experimental gill nets were fished from the Service's research vessel Cisco. Analyses and discussions in this report include all data that are believed pertinent to the solution of each particular problem. The exclusion of data of doubtful value in some instances causes dis- crepancies in the number of specimens listed in diff'erent tables. Whenever the excluded data are extensive or may influence results under alter- nate considerations, the reason for their omission is given. All collections of data used in this report are either taken from the entire catch of a net or are random samples unless otherwise stated. EXAMINATION OF SCALES Scales for age and growth analysis were taken when possible from the left side of the body in the area just above the lateral line and below the in- sertion of the dorsal fin. Van Oosten (1929, p. 274) stated that this area was selected "* * * after a careful examination had shown that its scales were less variable in shape and size, when compared one with another, than those of other parts of the body." Since the scales of lake herring are loosely attached and are frequently lost in nature, a liberal sample was taken to ensure the inclusion of non- regenerated scales. The scales from each fish were placed in an envelope on which were recorded the species, locality, date, length, weight, sex, condi- tion of sex organs, gear, and name of collector. The "key" scales required to establish the body- scale relation were removed from approximately the center of the area from which routine samples were taken. The location was the same as that used by Van Oosten — the fourth row above the lateral line and immediately below the base of the first ray of the dorsal fin. Some scales were mounted on glass slides in a glycerin-gelatin medium. Plastic impressions were made of the others. Each slide carried three or four scales of normal shape and without evi- dence of regeneration. The label on the slide bore the data shown on the envelope from which the scales were taken. Plastic impressions of scales were made by placing six or eight dry, uncleaned scales sculptured side down on a 1- by 3-inch strip of cellulose acetate bearing a serial number corresponding to that on the scale envelope. A second plastic strip was placed over the scales and the two strips were passed through a roller press set at the crushing pressure of cellulose acetate. (See Smith 1954.) The sec- ond strip of plastic holds the scales in position and ensures an even impression which produces a light, clear image. The numbered plastic strips bearing scale impressions were returned to the envelope and thus were not separated from the original data. LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 93 Before the plastic-impression method was adopted, careful microscopic comparisons were made of the scales and their impressions to be certain that replication was complete and without distortion. Butler and Smith (1953) who studied the reliability of scale impressions in age and growth studies found that growth calculations made from scale impressions did not differ signif- icantly from those made from the scales them- selves. About 500 of the scales used in this study were mounted in a glycerin-gelatin medium ; plastic impressions were made of the remaining 3,900. AH key scales used to establish the body- scale relation were mounted in glycerin-gelatin. Scale measurements for growth computations were made from the magnified (X41) scale image projected on the screen of a microprojection de- vice (described by Moffett 1952) and recorded to the nearest millimeter. The scale to be meas- ured was oriented so that a line on the viewing "screen bisected the image at its greatest antero- posterior diameter. Measurements of the total diameter and of diameters of growth fields cir- cumscribed by annuli were made along this line. The total diameter was measured from the extreme anterior to the extreme posterior margins of the scale. Diameters of growth fields were measured from the inside edge of the first complete circulus outside the annulus. Scale measurements of each fish were entered on IBM (International Business Machine) cards along with coded information concerning each fish. All subsequent computations and tabula- tions were made by means of the 602A IBM calculator and the 404 IBM tabulator at the Statistical Research Laboratory of the University of Michigan. Ages were determined by counting the annuli or year-marks on the scales. Van Oosten (1929) clearly established the validity of this method for the age determination of the lake herring of Saginaw Bay. More recent authors reporting on this species (Carlander 1945; Cooper 1937; Fry 1937; Hile 1931, 1936; Pritchard 1931 ; Stone 1938; and others) have accepted the use of scale mark- uigs for age analysis of lake herring. Nothing in the data on the Green Bay lake herring gives cause to question the validity of scales for age determination. Nevertheless, cer- tain difficulties of interpretation were encountered. Accessory checks, or false annuli, occurred on scales of nearly all fish after the second year of life. The general appearance of these checks and their location with respect to the annuli on either side left little doubt as to their identity; however, the possibility of some errors of age determination cannot be discounted. The regular appearance of accessory checks is not confined to the Green Bay stock. These false annuli on cisco scales have been reported by HUe (1936) in the cisco of Muskellunge Lake and by Fry (1937) in Lake Nipissing. Bauch (1949) described a fast-slow-fast growth pattern in a population of "kleinen Marane," Coregonus albula L. (the European coregonid most similar to the lake herring), in Mochelsee. He attributed the midseason check in the scales to oxygen depletion and an accumulation of hydrogen sulfide in the hypolimnion which forced these fish, nor- mally inhabitants of the deeper waters in summer, to live in upper strata where less favorable tem- perature conditions exist. Data on the Green Bay herring are inadequate to show the cause of accessory checks or even the time of their forma- tion. Seemingly the formation of checks varies from fish to fish and possibly' according to season and locality. The characteristics of the annulus on the scales of Green Bay lake herring are similar to those described for scales in other populations. The circular ridges on the scale start forming on the anterior margin of the scale and grow posteriorly along the lateral fields. When growth stops com- pletely and resumes again, growth of the un- finished circuli is not completed; instead a new circulus is started which encompasses the ends of those left incomplete at the cessation of growth. Fish having scales without an annulus are designated as belonging to age group 0, those with one annulus to age group I, * * *. For convenience, each fish is held to pass into the next higher age group on January 1. Since annulus formation does not actually take place until spring or early summer, the convention requires that a "virtual" annulus be credited at the edge of the scale from January 1 until the new annulus is visible. Year classes are identified by year of hatching (spring) rather than year of egg deposi- tion (fall). Thus, it is always possible to de termine the year class of a fish by subtracting its age from the year of capture; for example, a fish 94 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 5. — Age composition of lake herring taken in commercial pound nets, by quarters, 1949-52 [Dominant age groups indicated by asterisk] Quarter and year Number offish Percentage in age group— I II III IV V VI VII age 1 January-March: 1949 345 505 437 178 12.2 10.9 7.1 3.9 •74.2 •74.8 •80. 1 •87.1 13.6 13.5 12.6 8.4 4 01 1950 . 0.8 4.05 1951 0.2 4.05 1952 0.6 4.06 All years, 1949-52 1,465 0.1 9.2 •77.8 12.6 0.3 4.04 April-June: 1948 262 439 87 246 1.5 •53.6 16.2 10.3 16.7 42.2 •73.3 •72.4 •67.5 2.7 9.1 14.9 ID. 1 3.46 1949 1.4 2.4 0.4 3.96 1950 4.09 1961 4.9 0.4 3.86 All years, 1948-51 1,034 1.5 25.3 •64.0 8.2 0.9 0.1 3.82 July-September: 1950 244 73 2.9 8.2 2.9 30.1 ♦50.8 27.4 40.5 •32.9 2.9 1.4 3.38 1951 - - - - 2.89 All years, 1960-51 317 4.1 9.2 •45.4 38.8 2.5 3.06 October-December; 1949 278 219 0.7 0.5 6.8 5.9 •76.3 •68.9 15.1 24.7 0.7 0.4 3.09 1950 3.18 All years, 1949-50 497 0.6 6.5 •73.0 19.3 0.4 0.2 3 13 ' Average number of annull. belonging to age group III captured in 1949 be- longs to the 1946 year class. AGE COMPOSITION The principal characteristics of the age com- position of Green Bay lake herring taken in the commercial fishery (tables 5 and 6, and fig. 2) are the shift from older to younger fish during the calendar year and a strong tendency for the same age group to be dominant year after year during the same season. In pound-net collections made during the first quarter (January-March), age group IV was Table 6. — Age composition of lake herring taken in com- mercial gill nets, in 1948. 1950, and 1951, and experi- mental gill nets in 1952, by quarters [Dominant age groups Indicated by asterisic] Quarter and year Num- ber of flsh Percentage in age group— Aver- age I II III IV V VI age 1 166 154 48 3.0 3.2 35.4 •80.1 •85.1 •64.6 16.9 11.1 6.6' 4.14 April-June: 1952 4.09 July-September: 1952. . 3.65 October- December: 1948 152 108 80 250 'i.9" 3.9 4.6 2.5 0.4 •52.6 •72.2 •85.0 •60.8 38.2 19.4 10.0 37.2 6.3 1.9 2.5 1.6 3 45 1950 3. 15 1951 3 13 1952 3 40 All years, 1948-52.... 590 0.3 2.4 •64.1 30.5 2.7 3.33 100 80 60 40 20 I 60 I 20 uj 80 O < P 60 z uj O 40 X UJ °- 20 0 60 60 40 20 APRIL - JUNE JULY - SEPTEMBER IV V AGE GROUP ' Average number of annuli. Figure 2. — Age composition of lake herring taken in com- mercial pound nets (solid line), and commercial and experimental gill nets (broken line) in various quarters, 1948-52. LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 95 without exception strongly represented (74.2 to 87.1 percent) and made up 77.8 percent of all fish taken over the period 1949-52. In April-June collections the IV-group was still strongest in 3 years (67.5 to 73.3 percent) and made up 42.2 percent of the sample in the remaining year. The percentage of IV-group fish dropped from 77.8 in the first quarter to 64.0 in the second, whereas the Ill-group increased from 9.2 to 25.3 percent. Age group III was dominant in the summer quar- ter (July-September) in one of the two samples (50.8 percent) as well as in combined data of 1950-51 (45.4 percent). The transition to domi- nance by age group III was complete in the fourth quarter (October-December) , where it maintained this position in both years (68.9 and 76.3 percent) and in combined data for 1949-50 (73.0 percent). The dominance of the Ill-group (73.0 percent), which advances to age group IV on January 1, is only slightly less than that of the IV-group of the first quarter of the following year (77.8 percent). The mid-year shift of dominance from age group IV to age group III is also shown clearly by the average ages (table 5). The much less extensive data on gill-net sam- ples ' (table 6 and fig. 2) suggest that the trend of age composition is much the same as for pound nets. Age group IV was dominant, but the average age was decreasing in the first three quarters and age group III was dominant in all samples of the fourth quarter. Despite similar trends in the seasonal shift of age composition, gill nets in general took older fish than did pound nets. The small differences where large numbers of fish were concerned, how- ever, indicated that during the years of this study both gears were cropping a similar segment of the population. The age composition of the commercial catch demonstrated for Green Bay requires that a differ- ent year class be a major contributor to the fishery each year. The fishery, in turn, must then be very sensitive to fluctuations in success of year classes. Because of the resulting instability in the economy of small fishing communities it would be advantageous to devise some method of predicting good and poor year classes before they enter the fishery so that problems of high or low production ' Collections from experimental and commercial gear are shown together In Elll-net data. Figures presented In a later discussion on length at capture show that lake herring taken In the two types of gears at the same time of year have similar length distributions. could be anticipated. Unfortunately, this study has been conducted during a period of high and relatively stable production (see Economic Impor- tance of the Fishery, p. 90) and no fluctuations or means of their detection were discernible. The catch and abundance (expressed as catch per unit- of-effort), however, are normally subject to wide fluctuations (Hile, Lunger, and Buettner, 1953). The age composition of a representative sample of an entire population should normally show a preponderance of fish in the youngest age group, with progressively decreasing numbers as age increases. This pattern of diminishing numbers with age must exist in lake herring populations (even though it has never been demonstrated), for a population that regularly has fewer young fish than old must soon disappear. Since young lake herring have to be abundant, their scanty repre- sentation in samples of the population must be attributed either to the inability of collecting gear to capture them or to their absence from the area sampled. It is believed that the scarcity of young herring in the 1948-52 samples was largely the result of their scarcity on the fishing grounds. A principal gear of capture, the pound net, was fully capable of taking lake herring as young as 1 or 2 years old had they been present in abundance. Pound nets from which lake herring were taken for this study were also designed to capture smelt. Because of their small size and slender form, smelt require smaller mesh sizes than do the lake herring and yellow perch {Perca flavescens), which constitute important portions of the commercial catch. Mesh sizes ranging from 1 }i to 2 inches, extension measure as manufactured, made even smaller by treatment with preservative, have been used in Green Bay since smelt became an important com- mercial species about 1940 (Hile, Lunger, and Buettner, 1953). Although this mesh was far smaller than was previously considered satisfac- tory to catch commercial-sized herring, its intro- duction did not result in any continuous appear- ance of smaller herring in the catch even though it regularly captured yearling smelt and perch. In southern Green Bay, large numbers of trout- perch (Percopsis omiscomaycus) 3 to 4 inches long are regularly taken. The ability of pound nets to catch young herring was clearly demonstrated in the winter of 1944-45 when, according to Hile, Lunger, and Buettner, 96 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE large numbers of "pin" herring were taken in pound nets. Scales from 78 specimens taken at Escanaba on May 27, 1945, revealed that all were fish of the 1943 year class and were nearing the end of their second year of life. Inasmuch as this has been the only phenomenal occurrence of small herring in smelt-type pound nets since they came into common use, it can be assumed that the appearance in numbers of young herring in 1945 could have resulted from a successful hatch in 1943 and that the abnormally plentiful young herring extended beyond their normal range into the shallow-water areas in which pound nets are located. Although lake herring average about 5 inches long at the end of their first year of life, which is within the size range of other small fish taken, none were ever present in our pound-net samples. The lake herring is a relatively short-lived species. Hile (1936) reported the maximum age of XII in Trout Lake, Wisconsin. Although no other author has reported a fish this old, fish in age group XI have been reported by Fry (1937) in Lake Nipissing and by Hile in Clear Lake. Lake herring in age group X have been reported by Carlander (1945), Eddy and Carlander (1942), Stone (1938), and Van Oosten (1929). The low- est maximum age reached in any population was reported by Hile for Muskellunge Lake where the oldest fish belonged to age group IV. The oldest age groups in these populations are represented in the samples by only one or two individuals; in most lake herring stocks heavy mortality starts between the third and seventh years of life. The oldest lake herring taken in Green Bay were two Vll-group fish caught in pound nets in June 1951.^ Only 17 representatives of age group VI were recorded during the course of this study. The observed age compositions of several North American lake herring populations show that age groups II to V are best represented in the samples and that of these age groups III is usually the most common. Some of the differences among samples from various populations were undoubtedly the result of selectivity of collecting gear. It appears, nevertheless, that fish are much shorter lived in some populations than in others. Hile (1936) collected fish from several lakes with the same gill ' One of the VH-group fish was in a selected sample collected on June 15, 1951, and does not appear In discussions dealing with age. The other VII group flsh wa.s In the June 19, 1951, collection. nets, and his data should be well adapted to a comparison of age composition in different bodies of water. Hile's data show that age groups II and III were best represented in the Muskellunge and Clear Lake populations, but that the oldest fish taken in Muskellunge Lake belonged to age group rV, whereas in Clear Lake ciscoes lived as long as 1 1 years and age group VII made up more than 11 percent of the samples. The difference between these two lakes in observed age composi- tion is as great as that recorded elsewhere in the literature. It is possible that differences reported by other authors can be real and that the longev- ity does vary with local conditions. Van Oosten (1929) showed that age group III (age group IV under his system of age designation) predominated in his samples from Saginaw Bay, all of which were taken from pound nets during the period October to December. This same age group dominated samples taken from Green Bay pound nets during the same time of the year (table 5). SIZE AT CAPTURE The lengths of lake herring captured in pound nets (table 7) and gill nets (table 8), varied both as to average and range among collections of the same year and of different years. Mean lengths for samples, however, show no distinct seasonal pattern, which is in marked disagreement with the well-established, seasonal changes in age composi- tion (see p. 94). The data on age would suggest that the consistently older fish taken during the first half of the year should be longer than the predominantly younger fish taken in the second half. The discrepancy is explained by the length frequencies of age groups (table 9) which show a wide overlap of length distribution where length groups are frequently represented by fish of three ages. Differences between mean lengths of age groups III and V were only 0.4 to 0.7 inch in different years. Thus, lake herring of these age groups are similar in length regardless of age and no great changes in length should be expected to follow changes in age composition. That there is a greater growth than is indicated by the average lengths of age groups is brought out in a later discussion of computed growth. The apparently poor growth suggested by the similar average lengths of different-aged fish in the com- mercial catch must be due either to a strong LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 97 Table 7. — Length distribution of lake herring taken in pound nets, by month and year, 1948-6S Total length May 1948 1949 1950 1951 Jan- uary (Inches Feb. May Oct. Feb. Jane July Sept. Nov. Dec. Feb. May June Aog. 1952 1 1 2 1 2 1 3 6 21 S 6 2 3 7 10 6 2 1 6 .S to 6 fi 1 I 1 3 3 6 10 11 4 2 7 5 to 7 9 1 4 5 7 12 29 82 100 27 10 2 2 1 1 8 0 to 8 4 1 2 4 10 60 181 146 26 7 1 2 i' 8 47 49 28 6 1 1 2 2 2 3 9 20 27 21 12 3 8 5 to 8 9 1 5 41 98 67 28 18 2 1 1 2 8 36 172 190 81 11 3 2 1 1 1 9 14 35 16 9 2 3 8 16 69 82 24 6 1 ■ 1 35 160 95 44 8 2 4 12 103 186 111 19 5 1 9 5 to 9 9 3 21 45 32 3 1 2 3 12 64 29 3 1 4 100 to 10.4 8 10.5 to 10.9 -- 5« 11.0 to 11.4 73 11 5 to 11 9 34 12 0 to 12 4 4 12 5 to 12 9 2 13 0 to 13 4 1 i 1 14 5 to 14 g 1 1 1 1 1 Number of flsh 262 10.5 34S 10.5 441 10.7 283 10.8 S07 10.6 87 10.8 43 9.5 201 10.9 107 10.8 112 10.8 441 10.8 143 11.2 106 10.7 74 9.3 182 11.1 Table 8. — Length distribution of lake herring taken in gill nets^ by month and year^ 194S-5S 1 Collections from 2^i-tnch-mesh commercial gill nets. ' Collections from 2>2-inch-mesh commercial gill nets. ' Collections from 2-inch-mesh experimental gill nets. Total length (Inches) October 1948 > November 1950' 1951 1952 • Feb.i Nov.« May July Oct. 7.5to7.9 1 1 1 8 0 to 8 4 8.5 to 8 9 1 9.0to9.4 1 1 9.5 to 9.9.. _ 2 12 44 87 15 6 2 4 26 63 46 13 3 1 5 16 17 5 6 10.0 to 10 4 4 18 33 55 23 14 3 4 21 33 26 14 5 1 1 9 10.5 to 10.9 5 20 28 16 9 1 1 64 11.0 to 11.4.- _ 104 11.5 to 11.9 58 12.0 to 12.4 5 12 5 to 12 9 13.0 to 13.4 1 13.5 to 13.9 14 0 to 14 4 1 Number offish 152 11.7 108 11.3 168 11.1 80 11.7 157 10.9 49 10.6 252 11.2 Table 9. — Length distribution of lake herring, by age group, taken from pound nets in January and February, 1949-SS Total length 1949 1950 1951 1952 (inches) III IV V III IV V VI II III IV V HI IV V VI 8.0 to 8 4 1 8.5 to 8.9 2 3 9 19 17 6 1 1 1 3 10 45 152 115 18 6 9.0 to 9.4 1 24 123 71 32 3 2 2" 14 16 5 27 138 143 59 5 1 i 4 6 48 65 28 3 1 9.5 to 9 9 9 23 8 2 10.0 to 10.4 .. 15 29 16 5 1 2 11 13 4 1 4 14 25 7 1 1 2 1 2 2 3 1 4 5 4 1 10.5 to 10.9 ll.fl to 11.4 11.5 to 11.9 12.0 to 12.4 12.5 to 12.9 1 13.0 to 13.4 14.0 to 14 4 14.5 to 14.9 Number of fish Average length 42 10.3 256 10.4 47 10.7 66 10.3 378 10.5 68 10.9 13.0 1 8.1 31 10.5 350 10.8 55 11.2 7 10.8 155 11.1 15 11.2 1 12.6 98 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE modification of the population by the fishery (that is, a selective destruction of the larger fish) or to a differential distribution of the fish according to size so that only a certain segment of the popula- tion is represented in the fishery. Since the second condition obviously would contribute to the first, it may be assumed that the commercial fishery exerts a strong modifying effect on the population. Natural mortality, of course, may also play an important but unmeasurable role in this process. A progressive increase in length of lake herring of each age group in successive years from 1949 to 1952 (table 9) indicates that more rapid growth took place in the later years. This trend is also brought out in a later section on annual fluctua- tions in growth rate. Small as variations were in the average lengths of lake herring collected at different times of the year (tables 7 and 8), November-May collections taken at about the same time but often at con- siderable distances apart, showed still smaller differences of no more than 0.2 inch. This similarity was not always present, however, for in collections of other months (June-October) large differences sometimes occurred. Examples of these small differences in average lengths of herring taken in different areas are given in the following table: Date Location Average length (Inches) May 13... Suamlco 10 8 13 Pensaukee 10 7 Feb. 11... lUO Schumachers Point 10 6 27 10.5 Nov. 29 Fox 10 8 Dec. 4 Feb. 20.-. 1961 Ingallston 10 7 20- Pensaukee Schumachers Point Escanaba 22 - 10 9 Jan. 21... 196B 22 The weight.of Green Bay lake herring at capture presents much the same picture as does length. Weights of fish of a given age are distributed over a wide range and each weight group is frequently represented by fish of three ages (table 10). Differences between age groups III and V varied only 0.6 ounce to 1.2 ounces in different years as would be expected when differences in length were small. GROWTH BODY-SCALE RELATION AND CALCULATION OF GROWTH Van Oosten (1929) established the validity of computations of the growth of lake herring from the diameters of the entire scale and of growth fields within the several annuli. Since the publi- cation of his work, most investigators reporting on growth of this species have accepted Van Oosten's conclusions. The relation between body length and the anterior scale radius of lake herring was determined for the tullibee of Lake of the Woods by Carlander (1945). Carlander used the anterior radius be- cause he found annuli difficult to locate in the posterior field. He demonstrated that the rela- tion between scale radius and standard length was described satisfactorily by a third-degree equation. From a comparison of results of calculations from diameters and anterior radii Van Oosten (1929, p. 327) found that "* * * the diameter of a scale grows in length more nearly proportional with the body than does the anterior radius [and] * * * that the diameter dimension is less variable than the anterior radius * * *." Since no difficulty was experienced in locating annuli in the posterior field of scales of Green Bay lake herring, diameter measurements were used in this study to take advantage of the simple, direct-proportional rela- tionship determined by Van Oosten. It was held desirable, nevertheless, to study the body-scale relation of the Green Bay lake herring to make certain that the procedure was valid in this stock. If direct-proportion computations are to be valid, the body-scale ratio must be the same for all lengths of fish from the time of completion of the first annulus. Van Oosten (1929) found that after formation of this annulus the ratio of total scale diameter to body length was so nearly constant in the herring of Saginaw Bay that an assumption of constancy could be made. In the Green Bay lake herring the body-scale ratio exhibited no trend with increase in fish length (table 11). A f-test to determine whether such variations as did occur represented a significant trend confirmed the validity of the assumption that the ratio does not change with length (0.8 Males - - 11.0 11. 1 1952: 1 Males - 10. s 10.6 I Collected from 2-inch-mesh experimental gill nets. ' Collected from 2H-inch and 2>4-inch-mesh commercial gill nets. Effect of gear selection on estimation of growth From a review of tlie literature and from his own data Hile (1936, pp. 298, 306) held that— * * * in general a sparse representation in a sample of a young age group whose average length is near the lower limit of effectiveness of the nets used, is a source of suspicion as to the reliability of the sample of that particular group. If this same sparsely represented group gives calculated growths that are in serious disagreement with those of the older age groups it should be eliminated from the data used for the study of growth in the population as a whole. 104 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Hile concluded that — * * * if these selected groups are eliminated the re- maining growth ■ data can be considered accurate and trustworthy within very narrow limits. If Hile's assumptions are correct we should find close agreement between growth of best-repre- sented age groups of herring talten in gill nets and herring of the same ages taken in the less-selective pound nets. This expectation is fulfilled by the data of table 17. Growth of the best represented age groups (III, IV, and V) was almost identical in pound-net and gill-net samples. In shorter age groups, I and II, the greater calculated lengths of herring from gill-net samples indicate that the larger, faster-growing individuals are selected by gill nets. This tendency for herring caught in gill nets to be larger than those taken in pound nets is still present though somewhat reduced in age group III. Because the effects of gill-net selection extends to ages as high as the Ill-group (which is frequently dominant), most detailed analyses of growth in later sections have been based on pound- net samples alone. Table 17. — Comparison of growth of lake herring taken in pound nets and in gill nets, by age groups [Calculated total length in inches] Net' Num- ber of fish Length at end of year of life— 1 2 3 4 5 6 Age group I: 16 2 78 16 906 404 2,018 475 208 69 11 1 6.1 6.2 5.2 5.8 6.4 6.6 5.3 5.2 5.0 4.9 5.0 5.3 Gill Age group H: 8.1 9.0 81 8.2 7.8 7.8 7.4 7.3 7.8 7.1 Gill Age group III: 9.9 10.0 9.6 9.5 8 9 9.0 9.7 8 6 Gill Age group IV: 10.6 10.7 10.1 10.2 11.0 10.2 Gill Age group V: Pound - 11. 0 11.2 12.1 11.3 Gill Age group VI: 13.1 Gill 12.0 ' Collections from pound nets in 1948-52, and from 2 to 2)4-inch-mesh experimental and commercial gill nets in 1948 and 1950-52. Seasonal differences The apparently slow growth indicated by small differences between lengths of lake herring of different age groups at capture, brought out in a previous discussion of the length frequencies of age groups, again suggests the possibility of se- lective destruction of fish of more rapid growth by the commercial fishery. If such a selective destruction is taking place and is strong, it should result in growth differences detectable in samples taken in the same year but several months apart. That selective destruction was sufficiently great to influence estimates of growth is indicated by the data of table 18. In every comparison, except the third year of life, in the Ill-group taken in 1949, fish caught earlier in the year had higher calculated lengths than did those taken later. In 14 of 18 comparisons the advantage of the early- season over the late-season fish amounted to 0.4 inch or more. Because of the seasonal differences in growth patterns in fish of the same age group it is necessary to stratify samples according to sea- sons when making discriminating comparisons. Table 18. — Comparison of growth of lake herring, by age group, taken in pound nets at different seasons, 1949-51 [Calculated total length in inches) Location and date of capture Num- ber of flsb Length at cap- ture Length at end of year of life— 1 2 3 4 SooTHEKN Green Bay Age group III: Feb. 16, 1949. 42 212 256 42 23 73 133 31 172 23 10.3 10.8 10.4 U.2 10.2 10.7 10.5 11.0 10.7 11.1 6.0 6.2 5.5 4.8 5.5 5.3 5.0 4.9 5.2 4.8 8.6 7.9 7.9 7.4 8.4 7.8 7.7 7.3 7.7 7.3 10.3 10.3 9.4 9.0 10.2 9.7 9.3 8 9 9.4 9.0 Oct. 5, 1949 Age group IV: Feb. 16,1949 Oct. 6, 1949 - 10.4 10.3 NoKTHERN Green Bay Age group III: Feb. 27, 1950 - Nov 29 1950 Age group IV: Feb. 27, 1950 --- 10.6 Nov 29, 1950 10.1 Feb. 20, 1951 10.7 Aug. 20, 1951 10.3 Geographic differences That environmental conditions must differ in the various parts of Green Bay is obvious (see General Features of Green Bay, p. 88) . If environ- mental conditions influence growth and if the pop- ulation is not regularly mixed by active migration or passive transport with currents, differences in the growth of lake herring captured in various sections of Green Bay should be detectable. Differences between growth in northern and southern waters of the bay are indicated by com- parisons of lake herring taken in pound nets at the same time of year at locations separated by considerable distances (table 19). In 10 compari- sons of size at capture for fish of the same age, northern fish were shorter in six, and longer in two; lengths of the remaining two groups were LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 105 Table 19. — Comparison of growth of lake herring, by age groups, taken in pound nets at the same time of year at different locations [Calculated total length In inches] Date and locality Area Number offish Length at capture Length at end of year of lite— Length increment 1 2 3 4 5 1 2 3 4 S Feb. 22-27, 1950: Age group III: North 23 33 245 133 9 59 73 78 31 23 6 22 172 154 41 12 79 76 8 7 10.2 10.3 10.5 10.5 11.5 10.8 10.7 10.8 11.0 10.9 10.0 10.7 10.7 10.9 11.1 11.5 11.1 11.1 11.0 11.5 5.5 5.9 5.0 5.4 4.8 5.0 5.3 5.5 4.9 6.2 5.3 5.8 5.2 5.4 4.8 5.0 5.0 5.5 4.5 5.1 8.4 8.6 7.5 7.9 7.4 7.2 7.8 7.9 7.3 7.5 7.9 8.7 7.7 7.9 7.2 7.5 7.7 8.0 6.9 7.7 10.2 10.3 9.3 9.5 9.1 8.7 9.7 9.7 8.9 9.0 10.0 10.7 9.6 9.7 8.8 9.0 9.7 9.8 8.6 9.3 5.5 5.9 5.0 S.4 4.8 6.0 6.3 5.5 4.9 5.2 5.3 6.8 5.2 5.4 4.8 5.0 6.0 6.6 4.5 5.1 2.9 2.7 2.5 2.5 2.6 2.2 2.5 2.4 2.4 2.3 2.6 2.9 2.5 2.5 2.4 2.6 2.7 2.5 2.4 2.6 1.8 1.7 1.8 1.6 1.7 1.5 1.9 1.8 1.6 1.5 2.1 2.0 1.9 1.8 1.6 1.5 2.0 1.8 1.7 1.6 South Age group IV: North 10.5 10.5 10.5 9.9 11.5 10.8 1.2 1.0 1.4 1.2 South Age group V. North 1.0 Schumachers Point South 0.9 Nov. 29-Dec. 4, 1950: Age group III: North Sister Bav South Age group IV; North 10.1 10.1 1.2 1.1 South Feb. 20-22, 1951: Age group III: North South Age group IV: North 10.7 10.9 10.1 10.4 11.1 11.1 9.8 10.6 11.1 11.5 11.0 11.5 1.1 1.2 1.3 1.4 1.4 1.3 1.2 1.3 Schumachers Point South Age group V: Ingallston North 1.0 South 1.1 Jan. 21-22, 1952: Age group III: North.. - Pensaukee South Age group IV: Escanaba North 1.2 Pensaukee South 0.9 equal in the two areas. Differences between growth of lake herring from northern and southern localities are much more apparent in the calcu- lated lengths. Without exception northern fish grew less in their first year than did southern fish. Although growth increments of the northern fish were predominantly larger than those of southern fish in the second year and were without exception greater in the third year, the initial handicap of slower growth in the first year was overcome by the end of the third j^ear of life in only 2 of 10 pairs of samples. By the end of the fourth year, however, the initial differences in size in the two areas had largely disappeared. The significance of this comparison may be questionable in the light of information brought out in a later discussion (Growth Compensation, p. 109), that fish with poor first-year growth also tend to be slightly shorter at capture than fish having better growth in the first year. It is possible then that differences between calculated lengths of lake herring in northern and southern samples may be a reflection of differences in the length at capture. That such an explanation is not adequate is indicated, however, in the data of table 20 which gives comparisons of the growth histories of fish of the same age in the same }^-inch length interval. Northern Green Bay fish of the same length and age as the southern Green Bay fish at capture tended to be shorter than the southern at the ends of their first, second, and third years of life; but after the first growing season northern fish usually grew more than southern fish. This similarity of growth differences in selected length intervals and entire age groups is evidence that northern and southern fish do have different patterns of growth. The hypothesis of a north-south gradient is suggested by the fact that differences in first year's growth are greater in samples taken farther apart. Annual fluctuations in growth rate Since calculated growth histories of lake herring in Green Bay differ according to season and geo- graphical location, studies of annual fluctuations in growth must be based on samples taken in the same location at the same time each j'ear. The series of samples that best met these require- ments were taken in the southern part of Green Bay in January or February in the years 1949 to 1952. The materials for the study of annual fluctuations in the growth based on these coUec- 106 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 20. — Comparison of growth of lake herring, of same age and length at capture, taken in pound nets at the same time of year at different locations [Calculated total length In inches] Length group and locality Number offish Length at capture Length at end of year of life- Length Increment Feb. 22-27, 1950; Age group IV; 10.&-10.4 in.: Escanaba (north) . _ _ Schumachers Point (souths Age group IV; 10.5-10.9 in.: Escanaba (north) Schumachers Point (south). Age group IV; 11.0-11.4 In.: Escanaba (north) Schumachers Point (south). Nov. 29-Dec. 4, 1950: Age group III; 10.0-10.4 In.: Fox (north) Sister Bay (south) Age group III; 10.5-10.9 in.: Fox (north) Sister Bay (south) Age group III; 11.0-11.4 In.: Fox (north)-- --. Sister Bay (south) Feb. 20-22, 1951; Age group IV; 10.0-10.4 In.: Ingallston (north) Schumachers Point (south). Age group IV; 10.5-10.9 in.: Ingallston (north) Schumachers Point (south). Age group IV; 11.0-11.4 in.: Ingallston (north) Schumachers Point (south). Jan. 21-22, 1952: Age group IV; 10.5-10.9 in.: Escanaba (north) Pensaukee (south) Age group IV; 11.0-11.4 in.: Escanaba (north) Pensaulcee (soutli) Age group IV; 11.5-11.9 In.; Escanaba (north) Pensaukee (soutli) 37 106 10.2 10.3 10.7 10.7 11.3 11.2 10.3 10.3 10.7 10.7 11.1 11.1 10.3 10.2 10.7 10.7 11.1 11.1 10.7 10.8 11.2 11.1 11.7 11.6 5.0 5.2 5.0 5.5 5.1 5.6 6.1 4.9 5.4 5.5 5.2 5.6 5.1 5.2 5.1 5.4 5.4 5.4 4.8 5.3 5.1 6.6 5.2 6.8 7.3 7.6 7.6 8.0 7.9 8.3 7.5 7.3 7.9 7.9 7.9 8.1 7.4 7.5 7.7 7.8 8.0 8.0 7.6 7.7 7.8 8.2 8.3 8.4 9.0 9.2 9.4 9.9 10.0 9.3 9.3 9.7 9.0 9.0 9.4 9.5 9.8 9.8 9.4 9.6 9.8 10.2 10.3 10.2 10.3 10.7 10.7 11.3 11.2 10.3 10.2 10.7 10.7 11.1 11.1 10.7 10.8 11.2 11.1 11.7 11.6 5.0 5.2 6.0 6.5 6.1 6.6 5.1 4.9 5.4 6.6 6.2 5.6 5.1 6.2 5.1 5.4 5.4 5.4 4.8 6.3 5.1 5.6 5.2 5.8 2.3 2.4 2.6 2.5 2.8 2.7 2.4 2.4 2.5 2.4 2.7 2.6 2.3 2.3 2.6 2.4 2.6 2.6 2.7 2.4 2.7 2.6 2.1 2.6 1.7 1.6 1.2 1.1 1.3 1.1 1.4 1.2 1.9 1.9 1.3 1.2 1.3 1.2 1.3 1.3 1.3 1.2 1.4 1.2 1.5 1.3 tions (table 21) are so arranged that in each section of the table the vertical columns show the calculated growth in different years of life but in the same calendar year, the horizontal rows give a comparison of the growth in different calendar years for the same year of life, and each diagonal row gives the growth history of a single year class. For the quantitative determination of annual fluctuations of growth the data were subjected to the analysis described by Hile (1941), a procedure involving the determination of the percentage change in growth from each year to the next. The chain of estimates thus obtained was then ad- justed to a mean of 0.0 for the period of years cov- ered by the data (table 22). The fluctuations show a trend toward an improvement of growth during the period covered and show a possible tendency to be cyclic. From a value slightly below average in 1944 ( — 2.1 percent), growth declined to a minimum of —6.5 in 1946 (fig. 5). The year 1947 was the first in a 4-year period of improvement that culminated in growth 9.1 per- cent above average in 1950. Figure 5. — Fluctuation of growth in length of lake herring from the 1944-51 mean. Temperature is commonly considered an im- portant factor in the determination of fluctuations in growth. Hile (1936, pp. 276-280) discussed the possible influence of air temperature on the growth ofcisco populations in northeastern Wis- consin lakes and cited works of several authors who found a positive correlation between summer temperatures and the amount of growth of sev- eral European species of coregonids. Concerning the Wisconsin cisco populations Hile concluded — The failure of variations in the amount of growth in different calendar years to show any close general depend- LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 107 ence on either annual variations in temperature or annual variations in population density suggests that possibly these variations in growth depend closely on both factors, and that the failure of these factors fo operate in the same direction in the same year tends to obscure the effect of each of them. Van Oosten (1929) found no correlation between annual fluctuations in first-year growth and annual fluctuations in the air. temperatures during the growing season for Saginaw Bay lake herring. More recently, Svardson (1951) has shown that the growth of whitefish in Sweden was greater in hot summers than in cool. The data for the Green Bay lake herring (table 22) give no evidence of a definite relation between fluctuations of growth and deviations of mean air temperatures or population density over the 8-year period 1944 to 1951. Table 21. — Annual calculated growth increments of lake herring from pound nets in southern Green Bay in January or February, 1949-52 .\gc group and year Annual growth increment (inches) in— of life 1944 1945 1946 1947 1948 1949 1950 1951 Age group IH: 1.7 2.7 5.8 1.0 1.6 2.5 5.5 0.8 1.2 1.5 2.6 1.7 2.9 6.0 1.0 1.8 2.5 2.0 2.9 2.0 2.6 5.9 6.0 Age group IV: 4th vcar 1.2 1.8 1.3 1.5 2.5 5.4 2.4 5.4 5.5 Age group V: 51 h year 0.9 1.4 1.6 1.1 1.3 0.9 i.i 1.5 2.5 5.1 1.3 2.2 5.0 2.4 5.0 5.1 Number of flsh in age group: III 42 245 12 33 154 7 22 76 6 IV 256 59 V 47 Table 22. — Deviation of growth, air temperature during the growing season (May-October), and abundance of lake herring in southern Green Bay, from the average for the 8-year period, 1944-51 Year Percentage growth devi- ation Mean tem- perature de- viation " F. ' Abundance = 1M4 -2.1 -4.1 -6.5 -3.8 -2.1 3.1 9.1 6.7 2.0 -1.8 -0.1 1.4 0.8 1.8 -2.1 -2,0 -63 IW.V _ ^ H»4ti . . ._ -36 24 1947 24 1948 28 1949 ...: 14 \\IH\ 5 Wb\ . . 4 ' Mean monthly deviations of air temperatures for the period May-October recorded by t^ S. Weather Bureau at Green Bay, Wisconsin. 2 Percentage deviation fioni average catch per unil-iif-efTort in pound and gill net.s computed from Wisconsin commercial catch records for southern Green Hay (Wisconsin commercial fishing district M-1). Discrepancies in calculated growth The systematic discrepancies among calculated growth histories of different age groups already noted for the Green Bay lake herring are a frequent, almost regular, characteristic of data on growth of fish. These differences occur among different age groups of the same year class as well as among age groups of different year classes. The pattern of the discrepancies varies from species to species and stock to stock. Most common is that which goes under the name of Lee's phenomenon of "apparent decrease of growth," in which the esti- mates of lengtii at the end of various years of life decrease with increase in the age of the fish on which the estimate is based. In this "typical" situation, tlie calculated lengths in the earlier years of life show the greatest disagreements. More recent authors have tended to depart from this definition and to apply the term "I^ee's phenom- enon" to all systematic discrepancies among calculated lengths. The literature on causes of Lee's phenomenon, in both the restricted and the broader sense, is extensive and to a considerable degree contro- versial. A review of the subject at this time could serve little purpose.' It may be useful, never- theless, to list the principal factors that liave been offered in explanation of systematic discrepancies in calculated lengths. These several factors, the significance of which will become clearer from later discussions, are as follows: 1. Use of wrong formula for growth calcula- tions. 2. Selective action of fishing gear. 3. Biological segregation on basis of size or maturity. 4. Higher mortality rate (natural or in the fishery) of the fish with the more rapid growth. In the consideration of discrepancies among calculated lengths of Green Bay lake herring, the first of these items is not significant since the validity of the method of calculation was estab- lished by a study of the body-scale relationship. The effects of gear selectivity (item 2) can be rendered insignificant by confining studies of growth discrepancies to samples taken by pound nets, which, as has been pointed out, were capable of capturing fish smaller than the smallest herring « See Van Oosten (1929) and lllle (1936) for detailed discussions of the prob- lem. 388748 O — 57- 108 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 23. — Calculated total length of lake herring at the end of each year of life, by age group and year class, 1943-50 [Pound-net samples only. Length in inches] Age group Num- ber of flsh Length at end of year of lite— of capture 1 2 3 4 5 6 7 1950 year class: I I II III I II III IV II III IV V II III IV V VI III IV V VI IV V VI VII V VI 6 8 35 2 20 119 155 19 340 661 15 4 326 594 83 1 141 619 88 1 110 89 6 1 7 3 4.3 5.7 4.4 5.9 6.2 5.8 5.4 6.3 5.8 6.4 6.2 4.8 5.9 6.6 5.2 4.9 4.8 5.9 6.4 4.9 6.1 5.1 5.1 6.0 5.0 6.2 4.8 1949 year class: 1950 1951 7.3 8.7 1952 10.8 1948 year class; 1950 8.6 8.0 7.9 8.9 8.0 7.7 7.2 9.7 8.1 7.7 7.4 7.8 8.2 7.9 7.3 9.7 7.7 7.4 7.9 8.3 8.0 7.5 1951 9.9 9.7 11.1 1947 year class: 1949 1950 9.8 9.6 8.9 1951 10.8 10.2 1952 11.2 1946 year class: 1948 1949 9.9 9.3 9.0 9.6 10.2 9.5 8.8 11.9 9.6 8.8 9.7 10.0 9.6 8.9 1050 10.5 10.3 10.8 1951 11.3 11.9 1952 12.6 1945 year class: 1948 1949 10.6 10.0 13.3 10.6 9.9 11.1 U.3 10.6 10.1 1960 10.9 14.8 1951... 16.3 1944 year class: 194S 1949 10.9 12.3 12.6 11.5 11.0 1950 1951 1943 year class: 1948 13.3 13.6 14.6 1949 12.0 appearing in the samples. Such discrepancies as do appear, therefore, are to be attributed princi- pally to factors 3 and 4. The inconsistencies among the calculated growth histories of the different age groups of the several year classes ' of the Green Bay lake herring (table 23) differ from Lee's phenomenon as originally described (Lee 1920). It is true that the estimates of length for a particular year of life did tend to decrease with increase in age of fish on which estimates were based. On the other hand, the size of the differences did not decrease with increase in the number of years of life as is charac- teristic of Lee's phenomenon. In all but one comparison between age groups represented by 15 or more fish the estimate of first-year length decreased with increase of age (the one exception is in age groups IV and V of the 1944 year class). The trends were similar for the second-, third-, and fourth-year calculated lengths, but exceptions were more numerous. ' The most discriminating comparisons are those among different age groups of the same year class, since these are not biased by annual fluctua- tions in growth. It is believed that the discrepancies among the calculated lengths of the age groups of the Green Bay lake herring represent the combined effects of segregation according to size within the popula- tion and of selective destruction of the faster- growing individuals in the fishery made possible by that segregation: . Because of the connection between these two factors it is difficult to judge their relative importance. - In fact, an attempt to separate the two is not desirable, since they are essentially parts of a single process. In the younger age groups, only the largest fish (a small percentage of the total) enter the pound- net fishery. (Note the small representation of age groups I and II in collections — table 17). Se- lection in the gill-net fishery is similar (table 17), but the effects of selective destruction probably occur later in gill nets than in pound nets. This biological selection (plus gear selection in gill nets) leads to the overestimation of the rate of growth in those age groups. At the same time, destruc- tion of the larger, fast-growing fish modifies the growth characteristics exhibited by the remaining stock. As members of a year class grow older, bias to the immediate sample resulting from the selective capture of the larger fish declines, but the cumulative effects of destruction of the faster-growing individuals become increasingly important. Ten populations of Leucichthys artedi, for which various authors have given figures of calculated growth of different age groups, have all exhibited Lee's phenomenon to some degree. Disagree- ments were large in only one of four cisco popula- tions in northeast Wisconsin (Hile 1936). In the Irondequoit Bay cisco population the growth rate decreased with increased age among the younger age groups, but differences were random at the higher ages (Stone 1938). Fry (1937) found only small discrepancies among the estimates of the first-year growth of the Lake Nipissing cisco, but disagreements were large in later years. The variation in the nature of the discrepancies in calculated growth of fish of different age in the several populations leads to the conclusion that the causes of Lee's phenomenon are not the same in all populations. Principal explanations of the phenomenon in lake herring advanced by various authors are — 1 . Selective action of gill nets used in collecting samples. LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 109 2. Segregation as to maturity during the spawning run. 3. Segregation as to size, independent of ma- turity. 4. Higher mortality rate among fast-growing fish than among slow-growing. Hile (1936) found that discrepancies in the calculated lengths of ciscoes in three of four Wisconsin lakes were the result of faulty sampling traceable to selective action of gill nets. Carlander (1945) attributed Lee's phenomenon in ciscoes of Lake of the Woods, Minnesota, to the selectivity of large-mesh gill nets, as well as to differential mortality of fast- and slow-growing fish as pro- posed by Hile (1936). Eddy and Carlander (1942) also found the phenomenon in ciscoes of Gull Lake, Minnesota. Van Oosten (1929) and Cooper (1937), whose samples came from spawning-run lake herring in Saginaw Bay and Blind Lake, respectively, offered similar explanations of Lee's phenomenon. Their views are expressed adequately in the following quotation (p. 570) from Cooper's paper: * * * the lake herring first reaches maturity during its third, fourth, or fifth year of life, depending upon individual rate of growth; the more rapidly growing individuals of any one year class attain maturity first. It follows that the youngest year groups were represented in the catch (from the spawning grounds) only by their biggest indi- viduals and, as older age groups were considered, more and more of those fish that had been smaller individuals in their earlier years appeared in the older groups. There- fore the younger age groups contained a larger proportion of fast-growing fish than did the older groups and, con- sequently, the computed lengths for the early years of life would be greater in the younger age groups than in the older. The persistence of the phenomenon in the older age groups (in groups in which all individuals are mature) may be explained on the basis of differential mortality, that is, on the assumption that the more rapidly growing fish die off earlier in life than the more slowly growing fish. In Green Bay, as has been pointed out, segregation by size (^and hence by rate of growth within a year class) appears to take place at all seasons. Evidence was presented by Hile (1936) that a high natural mortality rate was correlated with rapid growth in the cisco population of Silver Lake. Cooper has suggested differential mortality as a possible factor in Lee's phenomenon. Hile also advanced the hypothesis that, if there was scgriigation of fast- and slow-growing fish with depth, the gill nets which were always fished on the bottom could not take equal samples of both. Fry (1937) demonstrated that faster-growing young fish were found in deeper waters of Lake Nipissing during the summer and were joined in successive years by more and more of the slower-growing members of the same year class. Behavior of this type explains why Lee's phe- nomenon might be found in samples taken in a certain location at a particular period of the year. Although a difference in seasonal distribution of fast- and slow-growing lake herring may exist in Green Ba}' and may be contributing to Lee's phenomenon there, it cannot be the main causative agent, because the phenomenon exists in samples collected at different depths and at different locations in the same and different seasons. Growth compensation Growth compensation — the tendency for the smaller fish at a particular age to have the more rapid subsequent growth — seems to be common among fish (V&n Oosten 1929 ; Eddy and Carlander, 1942). The existence of growth compensation was mentioned in 4 of 14 publications on growth of lake herring (Carlander 1945, in Lake of the Woods tullibee; Eddy and Carlander, 1942, in the tullibee of 17 Minnesota lakes; Hile 1936, in the cisco of four Wisconsin lakes; and Van Oosten 1929, in the Saginaw Bay lake herring). Growth compensation seems to be a general occurrence in North American coregonids. It has been shown in the following stocks: Lake Michigan kiyi (Leucichthys kiyi) by Deason and Hile (1947); Reighard's chub (Z. reighaidi), longjaw cisco (Z. alpenae) and bloater (L. hoyi) of Lake Michigan by Jobes (1943, 1949a, and 1949b); Lake Huron whitefish by Van Oosten (1939) ; and Lake Superior longjaw {L. zenithicus) by Van Oosten (1937). McHugh (1941) did not find growth compensation in several populations of Rocky Mountain white- fish (Prosopium vnlliamsoni) . Of the authors who mentioned growth com- pensation in studies of lake herring only Hile (1936) and Van Oosten (1929) discussed its char- acteristics in any detail. Carlander (1945, p. 129) stated that— As was demonstrated for the ciscoes by Van (losten (1929) and Hile (1936), growth compensation occurs in the Lake of the Woods tullibee but the compensation is not great enough to overcome any advantage in length which large individuals may hold at the end of the first year. 110 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 24. — Calculated growth of lake herring grouped by size in different years of life [Based on IV-group fish of Feb. 16, 1949, sample collected at Schumachers Point. Terminal groups contain 15 percent and middle groups 35 percent of total number of fish. Mean lengths for the year of life of grouping and corresponding growth increments are italicized. Maximum difference between lengths in parentheses] Total length (inches) Number offish Length at end of year of life— Length Increment 1 2 3 4 I 1 2 3 4 1st year of life: 3.7to4.9 38 90 90 38 38 90 90 38 38 90 90 38 38 90 90 38 i.e S.l 5.7 e.s (1.6) 5.0 5.3 5.6 6.0 (1.0) 5.0 6.3 5.6 ,5.8 (0.8) S.2 5.4 5.6 5.7 (0.5) 7.4 7.7 8.1 8.5 (1.1) 7.1 7.7 8.1 s.e (1.5) 7.2 7.7 8.1 8.4 (1.2) 7.4 7.8 8.0 8.3 (0.9) 9.1 9.3 9.5 9.8 (0.7) 8.9 9.3 9.6 9.9 (1.0) «.8 9.g 9.6 W.l (1.3) 8.9 9.2 9.5 10.0 (1.1) 10.3 10.3 10.5 10.8 (0.5) 10.1 10.4 10.5 10.9 (0.8) 10.0 10.2 10.5 11.1 (1.1) 9.9 10.1 10.6 11. S (1.3) i.6 6.1 5.7 2.8 2.5 2.4 2.3 1.7 1.6 1.4 1.3 1.2 4 9 to 5 5 . 1.0 1.0 6.0 to 6.7... - --- 1.0 2d year of life: 6.0 to 7.4 -- 5.0 5.3 5.6 6.0 I. I s.i 1.6 i.6 1.8 1.6 1.4 1.3 1.2 7 4 to 7 9 .-- __- - 1.1 79to84. 0.9 8 4 to 9 4 1.0 3d year of life: 5.0 5.3 6.6 5.8 2.2 2.4 2.5 2.6 1.6 l.B l.B 1.7 1.2 9 0 to 9 4 . . . 1.0 94to9.9 - 0.9 9.9 to 10.9 -- - 1.0 4th year of life: 9 3 to 10 0 5.2 5.4 5.6 5.7 2.2 2.4 2.4 2.6 1.5 1.4 1.5 1.7 1.0 10 0 to 10.4 1.0 10 4 to 10.9 1.1 10 9 to 12 1 - - l.t 1 Length at capture. The characteristics of growth compensation brought out by these authors for this species were similar, in that the shorter fish at the end of the first year of life tended to grow more in the following year than did the longer first-year fish. The studies demonstrated further that the initial advantage of the longer first-year fish was not completely overcome. This type of compensatory growth was also found in the Green Bay herring (table 24), Previous investigators have examined the phe- nomenon of growth compensation by dividing fish into different length groups according to the first year's growth and comparing subsequent growth of these groups. It is not to be anticipated, how- ever, that these first-j^car groupings will retain their identities in subsequent years; that is, indi- vidual growth will vary sufficiently so that a new grouping on a similar basis in later years will show some exchange of fish between the original groups. In lake herring both previous and subsequent growth of fish of the same length in a particular year of life varied widely (table 25). For example, the 47 lake herring that were 7.0 to 7.4 inches long in the second year of life had ranged from 3.5 to 5.9 inches in their first year and from 8,5 to 10.9 inches in their third year. Because of the tendency for fish of a given length in a particular year of life to derive from fish of a considerable length range in earlier years and, in turn, to contribute to a wide range of length in subsequent years, it is to be anticipated that the growth of fish of different length groups will vary according to the 3'ear of life in which the grouping is made. This expectation is met by the data of table 24 in which length groupings of fish of a single age group are made on a similar basis (see caption of table) for each year of life. The maximum difference (difference between mean lengths of the terminal group) without exception was greatest for the year of grouping, and de- creased consistently in previous and subsequent years of life. The decrease from the year of grouping toward earlier years reflects the diverse origin of the fish with respect to their positions in the length distributions in those earlier years. Tlie decrease in the maximum difference in years of life following the year of grouping represents a tendency toward convergence of size. Further information on these growth relation- ships is to be had from the annual growth incre- ments of length shown at the right of table 24. Here it is seen that the increments in each year of life preceding the year of grouping tended to fall in the same order as in the grouping year itself, l)ut that in subsequent years the increments tended to fall in the reverse order. As a general biological phenomenon, growth compensation may reflect principles holding for LAKE HERRING OF GREEN BAY. LAKE MICHIGAN 111 Table 25. — Subsequent and/or previous frequency distribution of the calculated length of lake herring that had the same cal- culated length at the end of a particular year of life (Based on all age group III Ash of tbe 1950 pound-net collections] Year of grouping and calculated total length (inches) Length frequency at end of year of life— 1 2 3 1 2 3 1 2 3 I 2 3 1st year of life: 54 5 0 to *) 4 90 •> 5 to 5 9 101 65 6 *) to 6 9 5 22 12 9 6 1 17 40 30 2 1 3 23 51 18 2 2 1 7 0 to 7 4 1 5 28 26 4 2 & 0 to K 4 » 6 15 19 10 3 4 17 44 20 5 2 10 35 41 10 2 1 90to94 2 21 23 10 S to 10 9 17 2 2d year of life: 1 4 22 17 3 1 12 40 23 5 1 9 30 51 28 1 4 .S to 4 9 6 2 18 26 4 7 0 to 7 4 47 7 5 to 7 9 81 120 9 16 16 5 1 2 21 40 16 2 56 6 56 46 11 1 10 10 0 to 10 4 29 16 - 1 3(1 year of life: 1 1 13 16 10 2 4 0 to 4 4 2 19 44 35 21 1 12 21 41 23 5 3 5 10 17 2 1 2 13 21 6 70to74 16 40 56 10 8 16 46 29 3 1 2 11 16 5 2 7 5 to 7 9 43 122 102 37 fish and also for other animals as well. Hile (1941, p. 305) stated that— A wealth of experimental evidence supports the view that among animals in general the inherent capacity for growth is lost chiefly through its exercise, and, conversely, the failure to grow does not entail necessarily the loss of the natural ability to grow. In support of this statement he cited the work of several authors on such widely separated groups as mammals, fish, salamanders, and insects. Hodgson (1929), on the other hand, demonstrated that compensatory growth could be a perfectly natural result of comparisons of fish of different age (fish that have the same number of anniili that were of different ages because they hatclied at different times during the season). Hodgson explained his view that growth compensation is "apparent" by comparing identical hypothetical growth curves that started at different points along the time axis.* Later Hile (1941) applied Hodgson's principle to the sigmoid growth curve of the rock bass to explain the variety of relations!iips among the annual increments of different yearling size groups. A similar use of the growth curve of the Green Bay lake herring is presented in figure 6. Here the two growth curves are identical but fish .1 hatched and started to grow at time 0^, whereas fish B hatched and started to grow at time Oa. At time ' Hodgson (1929) felt that a bimodal length-frequency distribution of first- year sea herring resulted from a long irregular hatching period and estimated that hatching extended over about 3 months. Hile (19.36) also attributed a bimodal first-year lenpth-frcquoncy distribution of cl.scoes In certain year classes in two Wisconsin laltes to irregular weather conditions during the hatching period that resulted in irregular and prolonged batching. It has been impo.ssible to learn anything about the hatching of lake herring in Green Bay. but it is believed that hatching may extend over a period of several weeks since spawning occurs over a period of 4 to 6 weeks. 112 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 26. — Growth exhibited by lake herring that were the same length at the end of different years of life [Total length in inches] I I- O z 111 ^ ^^ / e ^' / c / 11 b 1 a d T T I ME T+l Fic.uBE 6. — The effects of differences in age (time of hatching) on the amount of growth during a later grow- ing season. T, which may mark the end of any growth period, fish A has attained length ac and fish B has at- tained tlie lesser length ab. During the interval from T to T+l fish A, which had expended more of its ability to grow at time T, added the incre- ment of length /fif which is less than the increment ef added by younger fish B. This explanation of growth compensation as an apparent phenomenon is based on tlie premise that all fish have the same growth curve. Under this concept, the growth of fish during a particular time interval depends principally on the size it had attained at the beginning of the interval. The data presented in table 26 shows that this assumption is essentially correct, for fish of the same lengtli at different ages tend to grow the same amount in tlie following year. In a review of Hodgson's (1929) treatment of growth compensation, Ford (193.3) offered the Length group and Num- ber of flsh Year of life at which computed lengths are grouped Length at end of year of life— Length increment year of capture 2 3 4 3 4 8.0 to 8.4 inches: 1949 207 128 133 / 81 \ 69 46 104 I 57 \ 56 { 1 { 201 ( ' \ 221 { 146 { 2^ i I { I 2 3 2 3 2 3 2 3 I 2 3 2 3 2 3 8.2 '8"2" "8."2' 8.6 ""8." 7" "8.' 7" "8." 6' 9.2 "9."i" "9.3' '"9.2" 9.7 8.2 9.7 8.2 9.8 8.2 10.0 8.8 10.0 8.8 10.1 8.8 10.2 8.7 10.4 9.2 10.3 9.2 10.6 9.2 10.5 9.3 "9.'8' "9.6" ""9.'7" io.'i" "io.o" io.'i' 'io.'i' 'io.'s" 'i6.'4' io.'e' io.r 1.5 ■'i.'s' "i.'e' 1.4 "i.l' "i'V "i.'e' 1.2 ■"i.'2' 'i.'s' "i.'s' 1950- 1951 . 8..'; to 8.9 inches: 1949 . 1950 1951- 1952 9,0 to 9.4 inches: 1949- 1950 1.1 1951- 1.2 1952.. 1.4 1.4 Unweighted aver- 1.39 1.35 criticism that Hodgson employed curves of iden- tical shape, although it is well known that growth may vary from individual to individual. Ford demonstrated that Hodgson's explanation could be supported from comparisons of growth along dissimilar curves starting at different points along the time axis. From the hypothetical curves of figure 7 it can be shown that dissimilar curves starting at the same point on the time axis will also exhibit growth compensation. This compensation depends on the fact proved earlier that size, not age, at the start of a period of growth determines the amount of growth that will be made during the period. The form of three of the five curves of figure 7 is iden- tical with that of the growth curves of figure 6, namely, OaCQ, OC, and Oabf. The curves Ocg and 06/ represent individual fish ^4 and B whose growth up to time T departed from the typical. Fish A grew more (ac) and fish B grew less (ab) at time T than the typical fish which would follow curve OC. Since, however, length is more important than age as a determiner of growth within a period (table 26), fish A may be expected subsequently to grow along the curve eg or, in other words, to follow the same course as a normal fish hatched at time LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 113 I I- o z u Figure 7. — The effect of size at the start of a growing season on the amoui of growth during that season. 0.4. Similarly, the growth after time T of fish B should be that of a r .rmal fish that hatched at Ob. The compensatory effects of differences in the growth of these two typical fish (hatched at the same time) during the period T—T-\-\ is identical with the compensation between the two typical fish hatched at 0^ and Ob in figure 6. General growth in length Distorting influences of the negative correlation between individual length of life anti rate of growth make it impossible to establisli a general curve that might repres nt the growth history of a typical, or "avei, ^e," fish. Only growth of particular age groups can be shown. Curves in figure 8, based on data of all pound-net collections (table 17), are believed to be the most reliable means of representing the geneial growth of Green Bay lake herring taken in the commercial fishery. Age groups I, II, and VI are omitted from the data . IV V - III /^ - /^ 7 / / ./ 1 1 1 i 3 ACE Figure 8. — Calculated growth in length of age groups III. IV. and V of Green Bay lake herring as determined for all fish of these age groups taken in pound nets. 1948-62. because of the bias introduced by their almost com- plete absence from samples during some seasons, and because they are represented by small num- bers of fish. LENGTH-WEIGHT RELATION GENERAL RELATIONSHIP The variation of the volume of an object of con- stant shape with the cube of any linear dimension is a well known principle of mathematics. It can also be said that the weight of an object must vary with the cube of any linear dimension if the shape and specific gravity are both constant. If, however, the shape or specific gravity changes the relationship does not hold, but other relatively simple relationships ordinarily can be used to in- terpret the cha^-^es. The usefulness ■ f the "cube law" in the study of the weight of animals was recognized by Herbert Spencer in 1871 according to a discussion of its application in this field by Thompson (1942). Hile (1936) reviewed the use of this principle in studies of the relation between the length and weight of fish. The condition coefficient "C" determined by the formula ('=WIU (C = the coefficient, U'=weight, and Z,=length) is widely employed by fishery workers as an index of changes in the form of fish 114 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE that result from such phenomena as maturation and release of sex products or variations in the amount of fat or flesh. If the cube relationship is maintained throughout life then C is an un- biased expression of condition and it is possible to compare the coefficients of fish of different length. Thompson (1942) pointed out, however, that, "* * * inasmuch as the animal is continually apt to change its body proportions during life, k [his sj^mbol for condition coefficient] also is continually subject to change." In this situation C becomes a function of length and the C values of fish of different length are not directly comparable as measures of departure from the "normal" for the stock. Hile (1936, p. 23S) stated that— Although the cube law does appear to apply to the length- weight relationship in some species * * *, these instances appear to be the exceptions, for the * * * inadequacy of the cube law in describing the length-weight relationship in fishes have been repeated by numerous investigators and on many forms of fishes. The situation is further complicated by the fact that not only does the length-weight relation de- viate from the cube law, but it is not the same for different populations of the same species and it varies from year to year within the same popula- tions (Hile 1936). The relation between length and weight in most populations of fish is represented satisfactorily by the formula W=cL", where Tr=weight, Z=length, and c and n are constants. However, since the relation between length and weight in a popula- tion varies with respect to sex, season, method of capture, and year of capture, as will be shown later, no single equation can describe the situation at all times and any general relationship that might be established is of necessity artificial. Nevertheless, a general length-weight equation based on all available data, regardless of sex, maturity, collecting gear, or season of capture, can be useful as an estimate of the average situation. An estimate of the length-weight relation of Green Bay lake herring based on all data is log H'-:— 2.4386-^3.0729 log L, where W equals weight in ounces, and L equals total length in inches. Data upon which this estimate was based are shown in table 27. The weights computed from the mean length of fish in each length group are the basis of the curve in figure 9; the empirical data are shown by dots. Comparisons of calculated and actual weights e 8 10 12 TOTAL LENGTH (INCHES) Figure 9. — Length-weight relation of the lake herring of Green Bay. The dots show the empirical data; the curve is the graph of the equation given in the text. prove that this formula does not describe the empirical data precisely. Calculated weights are generally less than actual weights for fish under 9 inches, greater for fish between 9 and 12 inches, and less for fish longer than 12 inches. A close fit was hardly to be expected in view of the known heterogeneity of the material. Few fish under 9 inches and over 12 inches were taken, and these were not equally represented in all seasons (table 7). This variation in representation to- gether with seasonal differences in weights of fish of the same length are responsible for the irregu- larities. SEASONAL CHANGES IN WEIGHT The study of seasonal changes in weight of Green Bay lake herring (table 28) is restricted to fish captured in the same calendar year (1949), in the same area (extreme southern Green Bay), and in the same gear (pound nets). In the 12 h^igth intervals represented by 3 or more fish on all three collection dates the October fish were heaviest in 11 and the May fish were lightest in 11. Febru- ary specimens were, of course, cliaracteristically intermediate (10 of 12 comparisons). Over the length range at which all dates were represented, the October specimens averaged 4.8 percent LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 115 Table 27. — Relation between the total length and weight of Green Bay lake herring |AU collections combined] Table 28. — Seasonal changes in weight of lake herring taken in pound nets in southern Green Bay during 1949 [Weight In ounces] Total length (Inches) Number offish Weight (ounces) Empirical Calculated 5.8 to 5.9 1 4 2 6 8 S 5 11 5 4 12 7 17 16 12 11 15 21 28 45 137 211 328 431 495 497 503 376 206 146 103 59 39 24 17 15 6 6 3 1 3 3 2 1 1 1 0.90 1.07 1.05 1.16 1.26 1.36 1.48 1.62 1.86 1.86 2.15 2.47 2.48 2.76 2.77 2.85 3.19 3.33 3.32 3.73 3.98 4.16 4.45 4.64 4.97 5.19 5. 51 5.88 6.23 6.64 7.03 7.51 8.10 8.58 9.20 9.94 9.98 11.31 11.53 9.10 11.63 13.00 15.25 14.20 14.50 14.90 0.80 6.0 to 6.1 0.91 6.2 to 6.3 1.02 6.4 to 6.5 1 10 6.6 to 6.7 1.24 6.8 to 6.9 1 32 7.0 to 7.1 - 1 47 7.2 to 7.3 1.59 7.4 to 7.5 1 76 7.6 to 7.7 1.86 7.8 to 7.9 2 05 8.0 to 8.1 2 19 8.2 to 8.3- 2.38 8.4 to 8.5 2 57 8.6 to 8.7 - 2.76 8.8 to 8.9 ... 2 93 9.0 to 9.1 3 18 9.2 to 9.3- - 3.41 9.4 to 9.5 3 62 9.6 to 9.7 3 87 4. 11 10.0 to 10.1 4 37 10.2 to 10.3 - - 4.64 10.4 to 10.5 . . . 4 93 10.6 to 10.7 5 23 108 to 10.9..- -.- 5.52 5 84 11.2 to 11.3 - 6. 17 11.4 to 11.5 - - 6.51 11.6 to 11.7 6 87 11.8 to 11.9 - 7.22 12.0 to 12.1 7 69 12.2 to 12.3 - 8.04 12.4 to 12.5 - -.. 8.42 12.6 to 12.7 8 84 12.8 to 12.9 - 9.31 9 71 13.2 to 13.3 13.4 to 13.5 - . . . 10 73 13.6 to 13.7 11 10 13.8 to 13.9.-.- 11.58 14.0 to 14.1 12 19 14.6 to 14.7 - - 13 77 14.8 to 14.9 . . - 14 68 15.6 to 15.7 - 17 23 16.6 to 16.7.-.- - 20 83 above and the May fish 4.6 percent below the unweighted mean for the tliree dates. The February specimens were sliglitly (0.2 percent) above tiie mean. Seasonal changes in weights of fish are often associated with, and are used to follow, the de- velopment and release of sex products. Thomp- son (1942) showed a weight cycle for plaice fol- lowing the spawning cycle, but he pointed out that immature fish also experience a seasonal weight fluctuation similar to that of mature fish. These seasonal changes, he believed, indicate a cycle of relative well-being originating in the variation of conditions that influence the addition or removal of body fat or tissue. SEX DIFFERENCES IN WEIGHT Because of the demonstrated seasonal changes in weight, studies of sex differences in weight are best made on samples taken within a short period February 16 May 13 October 5 Total length (inches) Num- ber offish Aver- age weight Num- ber offish Aver- welght Num- ber offish Aver- age weight 7.8 to 7.9 1 1 3 2 1 4 7 3 4 8 12 27 34 30 41 51 14 12 9 7 3 2 1 1 1 1 1.80 8.0 to 8.1 2L10 8.2 to 8.3 1.97 8.4 to 8.5 Z35 8.8 to 8.9 . - 2.90 9.0 to 9.1 i 1 3 3 8 26 46 69 73 75 62 36 23 5 4 5 2.80 3.60 3.03 3.30 3.48 3.85 4. 15 4.27 4.53 4.82 5.05 5.45 5.67 6.34 6.20 7.02 2.20 9.2 to 9.3 1 2 7 26 50 73 66 34 32 22 15 9 3 3 2 2.90 3.45 3.57 3.86 4.01 4.30 4.45 4.71 5.00 5.42 5.68 5.85 6.23 6.76 6.65 3.22 9.4 to 9.5 3.45 9.6 to 9.7- 3.70 9.8 to 9.9 3.52 10.0 to 10.1. . . . 4. 11 10.2 to 10.3... 4.33 10.4 to 10.5 4 82 10.6 to 10.7 5. 18 108 to 10.9 5.27 11.0 to 11.1 11.2 to 11.3 5.70 5.98 11.4 to 11.5 6.30 11.6 to 11.7 . . 6.62 11.8 to 11.9... 7.12 12.0 to 12.1 7.61 12.2 to 12.3 8.06 12.4 to 12.5. 8.55 13.0 to 13.1 9.10 13.2 to 13.3 .... 9.30 13.4 to 13.5 12.20 14.0 to 14.1, 13.40 15.6 to 15.7 1 14.50 of time. Actually the comparisons offered by the data in table 29 are based on collections of single days. On none of these dates were sex differences large. The weights of male and female lake herring of corresponding length were nearly the same in February (males 0.5 percent lighter than females). Females were the lighter in May (1.8 percent) but were heavier in October (3.4 percent). Sex differences probably are greater at the time of spawning in the latter part of Xovember; un- fortunately, adequate samples were not available for study of this point. Carlander (1945) found no significant difference in condition coefficients of male and female tullibee from Ijake of the Woods. Direct comparison of weights of male and female tullibee from Gull Lake (Eddy and Carlander, 1942) showed the females to be slightly heavier for their length than the males, but the difference was small. These authors did not consider possible seasonal varia- tions in their presentation. Two of four popula- tions of ciscoes in northeastern Wisconsin lakes (collections were made only during the summer) showed no differences in weigiit between sexes; in the other two stocks the males were the heavier in one and the females were the lieavier in the other (Hile 1936). Van Oosten (1929) found little difl'erence between average condition coefficients 116 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 29. — Weights of lake herring by sex and location taken in pound nets in southern Green Bay in 1949 [Weight in ounces] February 16, 1949 (Schumachers Point) May 13, 1949 (Pensaukee and Suamlco) October 6, 1949 (Pensaukee) Total length (inches) Males Females Males Females Males Females Number offish Average weight Number offish Average weight Number offish Average weight Number offish Average weight Number offish Average weight Number offish Average weight 7 8 to 7 9 1 1 3 2 2 1 1.80 g.OtoSl 2.10 8.2 to 8.3 1.96 8 4 to 8 6 3.35 8.6to8.7 -- 2.30 8.8 to 8.9 2.90 90to9.1 1 2.80 1 2 3 1 3 2 6 9 11 19 19 17 4 5 3 2 2.20 3. 55 3.40 4.30 3 47 4.10 4.05 4.72 5.10 5.15 5.70 5.83 6.07 6.40 6.63 7.70 9 2 to 9 3 1 2 4 21 39 48 37 22 21 12 12 7 3 3 2 2.90 3.45 3.47 3.86 4.01 4.30 4.44 4.65 4.97 6.38 5.75 5.94 6.23 6.76 6.65 1 1 1 2 11 21 33 29 38 23 18 5 1 2 3 3.60 3.20 3.60 3.50 3.82 4.11 4.34 4.55 4.88 6.16 6.41 5.90 7.10 6.80 7.20 2 4 2 1 t 18 23 11 22 34 10 7 6 5 3 2 1 1 2.90 94 to95 2 2 6 15 25 36 44 37 39 18 18 •1 2 2.95 3.20 3.48 3.86 4.17 4.21 4.52 4.76 4.99 6.48 6.61 6.16 6.60 5.70 3.60 9.6 to 9.7 - 3 5 11 25 29 12 11 10 3 2 3.70 3.86 4.00 4.32 4.48 4.82 5.07 6.48 4.43 5.55 3 40 9.8 to9.9. 3.70 10.0 to 10.1 4.11 10.2 to 10.3 4.61 10 4 to 10 5 4.88 10.6 to 10.7 5.23 10.8 to 10.9. 5.46 11 0 to 11.1 5.71 ll,2toll.3 -- --- 6.05 11 4 to 11.5 6.40 11.6 to 11.7 6.78 11 8 to 11 9 7.41 12 0 to 12.1 7.58 12 2 to 12 3 8.07 12 4 to 12 5 8.55 13 0 to 13 1 9. 10 13 2 to 13 3 9.30 1 12.20 14 0 to 14 1 I 13.40 1 14.50 of male and female lake herring of the spawning run (October-November) in Saginaw Bay (all lengths combined). Seasonal variations in differ- ences of weight between the sexes in related species have been reported by Jobes for Levcichthys reighardi (1943), L. alpenae (1949a), and L. hoyi (1949b) , and by Deason and Hile for L. kiyi (1947). Comparisons by Bauch (1949) of the mean condi- tion coefficients of Coregonus albnla of Mochelsee showed that females were slightly heavier than males during all seasons. In spawning-run samples of the same species from Keitelesee (Jarvi 1920) ripe females were heaviest for their length and spent females were lightest (only slightly lighter than males). ANNUAL DIFFERENCES IN WEIGHT Annual fluctuations in the length-weight relation of Green Bay lake herring captured at the same time of year (January or February) in 1949 to 1952 generally were small (table 30). Weights of fish of the same length showed an upward trend from 1949 to 1952. The amount of change from year to year is indicated roughly by the following Table 30. — Weights of lake herring taken in pound nets during February 1949-51 and January 1952 [Weight in ounces] 1949 1960 1951 1952 Total length (inches) Num- ber of fish Aver- age weight Num- ber of fish Aver- age weight Num- ber of fish Aver- age weight Num- ber of fish Aver- weight 8.0 to 8.1 1 1 1 1 2 1 9 13 29 40 81 61 78 53 14 10 6 4 2 2 2.40 2.60 2.60 2.80 3.60 3.30 3.70 3.88 4.21 4.60 4.71 5.02 5.16 5.41 6.71 6.17 6.37 6.62 7.42 7.80 8.55 8 6 to 8 7 8 8 to 8.9 2 2 3 5 7 27 48 74 101 77 62 43 22 19 3 5 2.30 2.45 3.13 3.02 3.56 3.71 4.01 4.35 4.67 4.79 6.00 6.30 5.55 5.87 6.80 6.46 9.0 to 9.1 9 2 to 9.3 1 2 7 26 50 73 66 34 32 22 15 9 3 3 2 2.90 3.45 3.57 3.86 4.01 4.30 4.46 4.71 6.00 6.42 5.68 5.85 6.23 6.76 6.65 9.4 to 9.5 2 2.60 9 6 to 9 7 9.8 to 9.9 3 3 2 7 22 30 38 29 19 11 10 1 2 1 2 4.03 10.0 to 10.1 10.2 to 10.3 10.4 to 10.5 10.6 to 10.7 10.8 to 10.9 11.0 to 11.1 11.2 to 11.3 11.4tQll.5 11.6 to 11.7 11.8 to 11.9 12 0 to 12 1 4.03 4.70 4.80 6.04 6.17 i.K 5.74 6.12 6.20 6.60 6.80 12.2 to 12.3 7.50 7.90 6.45 12.4 to 12.5 7.80 12 6 to 12.7 7.80 12 8 to 12 9 8.70 13 0 to 13 1 2 10.10 13 2 to 13 3 11.00 14 0 to 14 1 1 11.40 14.20 Mean deviation from average percent. - -2.3 -1.8 1.3 1.7 LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 117 mean percentages of deviations from the average weight for all years: 1949, —2.3 percent; 1950, — 1.8 percent; 1951, 1.3 percent; and 1952, 1.7 percent. This period of increasing weight was also one of generally improving growth rate (table 22). Hile (1936) found that the length-weight relation and condition coefficient varied from year to year in three of four populations of ciscoes in north- eastern Wisconsin lakes. Annual differences in the length-weight relation were reported by Deason and Hile (1947) for Leucichthys Iciyi and by Jobes for L. reighardi (1943), L. alpenae (1949a), and L. hnyi (1949b). INFLUENCE OF METHOD OF CAPTURE ON WEIGHT Discussions of seasonal and annual fluctuations, and sex differences in the length-weight relation have been based entirely on fish taken from pound nets. Gill-net samples were omitted from these comparisons because of the bias to length-weight data introduced by gill-net selectivity. Farran (1936) treated this problem in detail and estab- lished limits of selectivity (in terms of length and girth) of different sizes of mesh of gill nets in cap- turing marine herring. Deason and Hile (1947) demonstrated that within a sample of kiyi from Lake Michigan that was homogeneous as to age, sex, and locality and date of capture, the coeffi- cient of condition decreased with increase in length of fish taken by gill nets of the same mesh size but increased in fish of the same length with increase of mesh size. Although materials for the study of effects of gear selection on length-weight data in the Green Bay lake herring are scanty, those that are avail- able (table 31) demonstrate conclusively that gill nets tend to take heavier fish than do pound nets operating in the same area and season, but because of the small numbers of fish on which the individual averages are based, a number of exceptions oc- curred. The records for females taken during the spawning season show almost no difference be- tween samples from the two gears. The extent of the bias in the remaining comparisons is suffi- ciently great, however, to make exclusion of the gill-net samples desirable in detailed studies of the length-weight relation. Table 31. — Weights of lake herring taken in gill nets and in pound nets at different times of the year, 1950 and 1961 [Weight in ounces) November February Total length Oill net 1 Pound net ' Om net » Pound net ' (inches) Num- ber of fish .Aver- age weight Num- ber of ash Aver- age weight Num- ber of fish Aver- age weight Num- ber of fish Aver- age weight Males: 9.7 to 9.9 2 3 8 12 13 4 2 4.00 4.13 4.83 5.05 5.51 5.95 6.00 1 3 4 8 26 19 4 3 1 3.90 4.56 4.65 5.25 5.63 5.98 6.47 6.56 7.70 1 3 5 17 21 16 4.00 10.0 to 10.2 10.3 to 10.5 10.6 to 10.8 10.9 to 11. 1 11.2 to 11.4 11.5 to 11.7 1 2 11 13 13 10 4 1 1 1 4.70 5.10 5.32 5.69 5.86 6.60 6.70 7.40 8.60 9.00 3.86 4.74 5.10 5.30 5.61 11.8 to 12.0 12.1 to 12.3 12.4 to 12.6 1 7.90 12.7 to 12.9 Total or aver- 57 1 1 5.99 2.30 2.30 4.10 44 5.16 69 5.66 64 5.24 Females: 7.6 to 7.8 . 8.2 to 8.4,. 9.1 to 9.3 9.7 to 9.9 -. 1 7 8 16 22 5 1 4.10 4.38 4.90 5.30 5.67 6.40 6.30 1 2 6 20 32 22 8 2 3 2 3.60 4.60 4.95 5.12 5.50 5.82 6.43 7.25 8.36 8.05 10.0 to 10.2 10.3 to 10.5_ 10.6 to 10.8 10.9 to 11. K ... 11.2 to 11.4 11.5 to 11.7 11.8 to 12.0 1 1 4 5 7 9 5 9 1 3 1 1 5.20 5.30 5.65 5.68 5.92 6.76 7.20 7.67 8.00 8.80 9.40 10.50 3 18 31 42 12 12 4 1 1 4.50 4.72 4.86 5.23 5.44 6.16 6.97 12.1 to 12.3. 12.4 to 12.6 1 8.30 8.00 9.20 12.7 to 12.9 2 9.35 13.0 to 13.2 1 9.40 13.6 to 13.8 Total or aver- SO 6. 65 63 5.54 98 5.67 125 5.30 ' Collected from a 2?8-lnch-mesh gill net at Oconto on November 30, 1950. 2 Collected from a pound net at Fox on November 29, 1950. > Collected from a 25s-lnch-mesh gill net at Pensaukee on February 20, 1951. * Collected from a pound net at Schumachers Point on February 22, 1951. GENERAL GROWTH IN WEIGHT The differences in weight according to sex, season, and year of capture detract from the use- fulness of the general growth curves of the Green Bay lake herring. The best means of depicting growth is to compute weights for the calculated lengths of the best-represented age groups for all pound-net data (table 17). Weights for age groups III, IV, and V calculated from the general length-weight relation (p. 114) are given in table 32. Growth curves for these age groups are given in figure 10. 118 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE u z «04 I O ^3 I- O u V / IV / III / // ■ /// / - /// - /// ■J ■ I 0 I 2 3 4 5 6 AGE Figure 10. — Calculated growth in weight of age groups III, IV, and V of Green Bay lake herring as determined from calculated lengths for all fish taken from pound nets, 1948-52, and the length-weight relation for all fish of these age groups. Table 32. — Calculated growth in weight for the Green Bay lake herring of age groups III, IV, and V [Calculated from the general length-weight relation, p. 114, and lengths of fish taken in pound nets In 1948-52, table 17, p. 104] Age group Weight (ounces) at age— 1 2 3 4 5 III 0.65 .61 .51 2.25 2.01 1.71 4.18 3.68 3.01 IV 5.15 4.44 V-. 5 77 REPRODUCTION AND EARLY GROWTH SEX COMPOSITION As is common among fish, data on sex composi- tion in lake herring populations are highly variable. Some factors that may contribute to variability of sex composition in samples from a population are: /. Segregation of the sexes through various periods of the year including segregation resulting from sex differences in age and size at maturity. 2. Differences in mortality (natural or fishing) between the sexes. S. Gear selectivity in relation to sex differ- ences in activity and morphology. To evaluate any of tliese factors would be diffi- cult, particularly since they are interrelated and some or all of them may affect the sex composition of a sample. Reports of various authors on different popula- tions of lake herring (table 3.3) show sex composi- tion, expressed as percentage of females, ranging from 29 percent in Blind Lake (Cooper 1937) to 73 percent in Trout Lake (Hile 1936). The Blind Lake collection was made during the spawning period but the paucity of females is not character- istic of spawning fish as may be seen by the sex composition of other samples collected during the spawning period — in Swains Lake (67 percent) and Saginaw Bay (51 percent). Six out of 1 1 lake herring populations for which data have been published on the change of sex composition in relation to age (table 33) show a rise in the proportion of females with increase in age (Clear I^ake, Gidl Ijake, Lake of the Woods, Muskelliinge Lake, Swains Lake, and Trout Lake), 2 populations show a downward trend (Blind Lake and Saginaw Bay), and 3 exhibit no clear trend (Irondequoit Bay, Lake Nipissing, and Silver Lake). The 6 populations exhibiting an increase in the percentage of females with age were col- lected with gill nets and 1 (Swains Lake) was sampled exclusively during the spawning period. Of the 2 populations with a downward trend, 1 was sampled with pound nets (Saginaw Bay) and the other with gill nets (Blind Lake), and both sets of data were based on spawning-run collec- tions. One of the 3 populations showing no trend was sampled with pound nets (Irondequoit Bay) and the other 2 were sampled with gill nets, and all represent samples from more than 1 month and year. It is obvious from these comparisons that the relation of sex composition to age as reported for different stocks is not clearly influenced by collecting gear or sexual activity at time of collection. Some information on possible sources of bias in determining the sex composition of a population is brought out in the Green Bay data on fluctua- tions in the sex ratio according to age, gear of collection, and depth, season, and year of capture. In pound-net samples, which made up the bulk of the Green Bay collections, the percentage of females was consistently higher in February than during other months of the year, and since no trend was shown in sex composition during the other months, the data for all but the February samples are combined in table 34. This seasonal LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 119 Table 33. — Changes in sex composition icilh age for different lake herring populations [Number of fish in parentheses] Slate and body of water Percentage females in age group — All ages Gear used Investigator 0 I II III IV V VI VII VIII IX x+ Michigan: 100 (1) 67 (3) 35 (23) 55 (11) 29 (66) 54 (818) 24 (38) 50 (1.434) 20 (5) 52 (112) 59 (63) 38 (13) 56 (350) 47 (39) 0 (2) 53 (160) 78 (368) 30 (10) 50 (539) 64 (22) 65 (342) 63 (38) 37 (43) 55 (293) 45 (20) 20 (10) 45 (124) 71 (24) 71 (168) 67 (12) 57 (68) 49 (255) 50 (26) 0 (1) 42 (19) 68 (28) 80 (10) 100 (4) 59 (76) 36 (67) 66 (50) 28 (162) 51 (2, 950) 67 (84) 63 (658) 54 (421) 57 (494) 53 (1. 524) 51 (440) 58 (861) 55 (496) 73 (1,101) gill pound gUl -..do- ...do pound gill ...do ...do - -..do ...do Cooper (1937). 20 (5) 100 (1) Van Oosten (1929), 100 (3) 100 (1) Brown and Moflett (1942). Minnesota: Gull Lake 0 (2) 49 (101) 39 (64) 53 (156) 47 (102) 57 (472) 62 (26) 62 (97) 17 (24) 54 (120) 66 (216) 55 (334) 48 (95) 60 (361) 56 (86) 67 (520) Eddy and Carlander (1942). 48 (80) 0 (2) 64 (11) 76 (21) 68 (31) 100 (1) 92 (13) 75 (4) 0 (1) 50 (2) 60 (5) 100 (4) Carlander (1945). Stone (1938). Ontario" Lake Xipissing Fry (1937). Wisconsin: Clear Lake 42 (69) 50 (26) 52 (66) 50 (2) Hlle (1936). Do. 56 (133) 93 (80) 67 (24) 92 (12) 100 (1) 100 (4) Do. 80 (5) 100 (2) 33 (3) Do. Table 34. — Sea; composition of lake herring taken in pound nets, 1948-5B [Number of fish in parentheses; males at left, females at right] Time of collection Percentage females in age group— I II III IV V VI VII All fish 1 1948: May .. ... ....... 75 (1:3) 60 (57:84) 74 (11:31) 61 (112:172) 64 (20:36) 46 (153:128) 71 (9:22) 47 (31:27) 71 (2:5) 55 (50:60) 67 (85:171) 57 (157:206) 61 (148:230) 47 (114:102) 57 (148:200) 43 (107:82) 48 (81:74) 71 (2:5) 68 (15:32) 60 (17:25) 54 (31:37) 45 (11:9) 56 (24:31) 19 (21:5) 73 (4:11) 58 1940: (110:152) 68 100 (0:2) 63 (7:12) 33 (2:1) 50 (2:2) 0 (2:0) 100 (0:1) 100 (0:1) 0 (1:0) (111:234) 59 1950: (298:426) 60 43 (4:3) 50 (10:10) 100 (0:1) 44 (15:12) (201:306) 46 1951: (294:252) 59 100 (0:2) 100 (0:1) (182:257) 42 1952" January (178:130) 51 (90:92) 1949-52' January-February 100 (0:1) 53 (33:37) 69 (42:94) 54 (352:411) 59 (462:675) 51 (428:450) 60 (74:111) 46 (51:44) 50 (3:3) 33 (4:2) 60 1948-51: May-December 63 (4:7) 100 (0:1) (584:889) 52 (880:960) ' Includes flsh of unknown age. diflFerencc in th(^ percentage of females appears in the data for individual age groups as well as in the data for all ages combined. The percentage of females in samples of lake herring taken from pound nets also showed a clear tendency to decrease during the period 1949 to 1952 (table 84). This trend is present in the best-presented age groups (III and IV') as well as in tlie data for all ages combined in both the Januarv-P\'l)ruarv samples and the samples from the remaining montlis. The change in sex composition witli increase in age of lake herring taken in pound nets was irreg- ular, but a downward trend in the percentage of females is evident in most series (table 34) and is conspicuous where data of all years for compara- able periods have been combined (bottom of table). This trend would suggest that young females might be taken in the pound-net fishery at a higher rate than young males. A sex differ- ence in mortality of this kind should result in a progressive reduction in the proportion of females within a year class. That this expectation is fulfilled consistentlv in the January-February 120 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE collections is demonstrated by the following tabulation of percentages of females in samples of four different year classes at various ages (where represented by 25 fish or more). Table 35. — iSex composition of lake herring taken in gill nets, 1948-52 [Number of flsh In parentheses; males at left, females at right] Year class Percentage females and age group In— 1949 1950 1961 1952 67 (IV) 74 (III) 54 (V) 61 (IV) 64 (HI) 1946 66 (V) 57 (IV) 71 (III) 1947 48 (IV) The corresponding tabulation for collections made in months other than January and February demonstrates a similar trend but does contain one exception, the V-group of the 1944 year class. Year class Percentage females and age group In— 1948 1949 1950 1952 1944 55 (IV) 60 (III) 68 (V) 57 (IV) 61 (III) 1946 47 (IV) 46 (III) 19 (V) 1947 43 (IV) In gill-net collections the percentage of females varied widely from sample to sample (table 35). Although the available data are insufficient for a study of annual and seasonal trends, they offer no evidence of disagreement with the trends established in pound-net data. The change in sex composition with age of gill-net caught fish, however, is the reverse of that of fish taken by pound nets. The gill-net samples show a clear tendency toward a higher percentage of females with increasing age. This progressive destruction by gill nets of females in the older age groups should tend to counteract the effect of pound nets in cropping younger females at a faster rate than males. The combined effects of the selective destruction of the two fishing gears in determining differences in sex composition with season and age cannot be evaluated with data at hand. It is clear, however, that females are cropped more heavily than males during the winter (January-February) fishery by both pound nets and gill nets. The effects of this destruction of females are counteracted in part by the greater destruction of males in the remaining months of the year. Records of the sex composition of samples of lake herring taken at various levels between the surface and bottom (see description of oblique Time of Percentage females in age group— All ages coUectlon ' I II in IV V VI 1948' October 33 (4:2) 40 (3:2) 73 (22:58) 43 (44:33) 100 (0:5) 30 (47:20) 20 (4:1) 53 (8:9) 42 (88:64) 74 (15:43) 62 (8:13) 56 (58:75) 50 (4:4) 44 (74:57) 68 (10:21) 42 (64:39) 63 (3:5) 0 (2:0) 63 (10:17) 71 1960: November- .- 1951: 100 (0:2) (44:108) 47 (57:50) 5 „9 33 (2:1) (69:%) 33 1962: 53 (8:9) 100 (0:1) (64:26) 46 July (87:70) 61 100 (0:1) 75 (1:3) (19:30) 43 (143:109) AU dates 100 (0:2) 40 (9:6) 47 (213:190) 53 (223:252) 69 (24:34) 100 (0:1) 51 (473:491) 1 Collection from commercial gUl nets— 1948, 1960, 1951; collection from experimental gill nets— 1962. gUl-net sets in Vertical Distribution in Green Bay, p. 128), yielded no evidence of segregation of the sexes according to depth in June or July, but they indicated a strong tendency in October toward a higher percentage of females in the deeper strata than in the shallower (table 36). This trend was much stronger in samples from nets fished in 60 feet of water than in 90 feet (fig. 11). 30 ^S 60 DEPTH (FEET) Figure 11. — Sex composition of lake herring taken at vari- ous depths in October 1952 in 2-inch-mesh experimental gill nets set obliquely from surface to the bottom. Open circles indicate the data for 60-foot stations and solid circles the data for 90-foot stations; the regression lines were fitted by least squares. LAKE HERRING OF GREEN BAY. LAKE MICHIGAN 121 Table 36.- — Sex composition of lake herring taken al various depths in experimental S-inch-mesh gill nets in 195S (Number of fish In parentheses] I Month, depth, and station ■ Date taken Percentage females at— All depths 0-15 feet 15-30 feet 30-45 feet 45-60 feet 60-75 feet 75-90 feet June: 60 feet: K 12 12 11 SO (4) 0 (1) SO (4) too (1) 67 (6) 71 (17) SO (10) 63 (8) 60 (25) 63 (32) 58 (19) L 63 (0) (46) 90 feet- J 48 (23) 47 (15) 54 (6) (81) All depths - . - .._ 50 (4) 50 (6) 0 (1) 64 (33) 38 (8) SO (4) 35 (31) 73 (30) 100 (1) 40 (5) 44 (9) 62 (65) 80 (10) 33 (6) 38 (16) 63 (30) 48 (23) 47 (IS) 58 July: 60 feet: C 24 22 21 21 27 22 21 (146) 58 (0) (19) H 40 CO) 0 (1) (0) 33 (3) 44 (25) (10) K 35 (SI) L 61 (0) (85) 90 feet: D. .- 0 (1) 0 (2) 43 (21) SO (4) SO (2) 67 (27) SO I "(6) (0) "(oj 40 (15) 56 (18) (6) 38 J... (0) 100 (1) (0) 100 (2) (24) 56 (78) All depths-- so (2) 47 (15) 43 (44) 38 (37) 33 (40) 27 (30) 35 (49) 45 (11) 33 (40) 23 (30) 45 (31) 51 (88) 54 (95) 38 (24) 64 (33) 52 30feet:A 22 22 22 24 25 25 23 24 25 (273) 47 (0) 100 (2) 42 (36) 18 (40) 42 (26) 33 (30) 50 (8) 33 (40) 33 (30) (IS) 40 feet: B>..- _ 46 (46) 60 feet: C 44 (41) 57 (35) 60 (30) SO (30) 62 (29) 64 (25) 59 (44) 60 (30) 45 (143) H 40 (140) K,... - 48 (130) L 43 (139) 90 feet: D 47 I . .. (0) 43 (40) 37 (30) (0) 53 (57) SO (28) (0) 38 (8) 47 (30) (6) 100 (2) 48 (31) (19) 42 J (187) 40 (179) All depths 35 (296) 34 (212) 48 (206) 57 (213) 45 (38) 52 (33) 43 (998) • See figure 1 for location. ' Depths intervals 0-20 and 20-10. Despite the clear-cut change in sex composition with increase of depth, the vaHdity of an assump- tion that sexes are segregated according to depth in October is questionable. Since the gill nets used to obtain these collections were stationary the activity of the fish was a primary determinant of the number of fish taken by them. Accordingly, it is possible that changes in sex composition with depth do not reflect a corresponding difference in the actual relative abundance of males and females but that they are merely "apparent changes" traceable to sex differences in activity. In other words, the males may have been much more active than the females near the surface, whereas the activity of the sexes may have been equal or nearly equal at the greater depths. No evidence on the question of sex differences in activity is available from the present study or from published reports on the lake herring. Evidence has been published, however, that the males of the related kiyi of Lake Michigan become much more active during the spawning period (Hile and Deason, 1947). If a similar behavior is assumed for the lake herring, and if it is assumed further that the heightened activity of males starts in advance of the spawning period and that the fish near the surface are the ones closest to the spawning state (the lake herring is a pelagic spawner), then sex tlifferences in activity rather tlian true segregation can explain the relation of sex composition of lake 122 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE herring to depth in October samples which were taken about 3 weeks before spawning starts. AGE AND SIZE AT MATURITY A lake herring was considered immature if it was not in spawning condition when captured during the spawning season, or if the state of the gonads indicated that it would not spawn during the next spawning period following its capture. As most small lake herring captured in Green Bay were taken within a few months before the spawn- ing period, at a time when all mature fish had well-developed gonads, little difficulty was ex- perienced in distinguishing the immature individuals. Published statements as to the age at which the lake herring matures frequently have been indefinite because of considerable individual varia- tion among fish and because of questionable dependability of samples as a result of gear selection or segregation on the basis of maturity. Hile (1936) suspected that his estimates of per- centage of maturity in the younger age groups were too high if the faster-growing fish of each age group matured first, since his gill nets did not take the smaller members of those age groups. Van Oosten (1929), who sampled only fish of the spawning run, felt that since immature fish did not participate in spawning activities they were not properly represented in the samples. A summary of published data on the maturity of lake herring (table 37) shows that the age at which most fish mature in different populations varies from I to IV. Although lake herring maturing in the first jear of life (age group 0) have never been reported, maturity in the second year (age group I) is common. The reason for later maturity in some populations is not clearly understood . The Green Bay collections contained relatively few immature lake herring, all of which were in age groups 0, I, and II (table 38). The two 0-group fish taken (one male and one female) were immature. In age group I, 32 percent of the males and 1 1 percent of the females were mature. By the next year (age group II) most fish of both sexes had reached maturity (97 percent of the males; 88 percent of the females). This tendency for males to mature sooner than females was also found in tlic lake lierring of Saginaw Bay (Van Oosten 1929) and Irondequoit Bay (Stone 1938). The three 2-year-old ciscoes taken in Lake Ontario by Pritchard (1930) were all mature females. The average lengths of mature and immature fish indicate that the larger members of an age group are more likely to be mature (table 38). Table 37. — Age at which lake herring of different popula- tions reach sexual maturity (Arranged according to age at maturity] Age group in which— Body of water Few flsh mature Some fish mature Most or all flsh mature Investigator Clear Lake, Wis I n II II II II III III III IV IV IV Hile (1936). Green Bay, Lake Michigan I I I I Present work. Carlander (1945) Saginaw Bay, Lake Huron. Trout, Silver, and Muskel- lunge Lakes, Wis. Lake Erie .. - Van Oosten (1929). Hile (1936). Clemens (1922). Irondequoit Bay, Lake Ontario. Blind Lake, Mich '11 . II Stone (1938). Clemens (1922). Cahn (1927). Lake Ontario II III III Pritchard (1931) Dymond (1933). Manitoba Lakes Bajkov (1930). Table 38.- -Relations among age, lerigth, and sexual maturity in the lake herring of Green Bay [Total length in inches] Mature Immature Percent- Age group ' Number of flsh Total length Number offish Total length age mature Males: 0 1 17 1 1 16 4 6.0 6.9 8.0 5.8 7.8 8.2 0 I 8 33 8.2 9.9 32 II 97 Females: 0 0 I . . 2 30 8.6 9.3 11 II _ 88 > All flsh older than age group II were mature. HATCHING AND EARLY GROWTH Almost nothing is known about the incubation, hatching, and early development of lake herring in nature. Cahn (1927) collected unhatched eggs in Lake Oconomowoc, Wisconsin, in March, but had no positive evidence as to the time of hatching. Pritchard (1930) observed that hatching takes place during April and early May in the Bay of Quinte, Lake Ontario, and he made daily collec- tions of the growing fry from May 9 to June 1. These fry were found among reeds in shallow- water areas of protected bays, but apparently they moved toward the open water as they grew. On June 1 when the last individuals were collected, they were 20 millimeters long. After that date tliey could not be located again. Greeley and Greene (1931) collected young-of-the-year lake LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 123 herring in the St. Lawrence River near Ogdensburg, N. Y., on June 6, 17, and 18, and at Waddington, N. Y., on June 28, in 1930. Since adult lake herring are unknown in the river, it is presumed that these fish came 78 miles downstream from Lake Ontario. The lengths of these young herring on different dates of collection were as follows: Date Number or ash Length Average Range June 6 June 17, 18 15 142 76 21 mm 28 mm 36 mm 18 to 28 mm. 24 to 32 mm June 28 The mean length of 21 millimeters of the June 6 collection corresponds closely with the length of 20 mm. recorded by Pritchard (1930) on June 1. Cahn (1927) took three young of the year with an average length of 62.5 mm. in a gill net fished on the bottom at 52 feet in Lake Oconomowoc on June 20, 1922. Fry (1937) caught 0-group ciscoes in the region of the thermocline in Lake Nipissing during late August 1933 to 1935. Hile (1937) found 17 young lake herring (average length, 65 mm.) washed upon the shore of Trout Lake during late summer. Reighard (1915) re- ported similar recoveries of young in Douglas Lake and Ward (1896) found small lake herring washed up on the shore of Lake Michigan follow- ing storms. Records of a few recoveries of small lake herring from shallow-water areas of the Great Lakes are on file in the Fish Division of the Mu- seum of Zoology, University of Michigan. Knowledge of the distribution and habits of young-of-the-year lake herring is scanty in Green Bay also. In spite of a constant lookout for them during all field work and attempts to locate them with midwater and bottom trawls during the sum- mer of 1952 only two young-of-the-year lake her- ring have been taken from Green Bay. Botli were captured in a 1 -inch-mesh gill net from 17 fathoms of water off Gills Rock on December 12, 1951. One was a male 6.0 inches long and the other a female 5.8 inches long. Otter trawls of the same construction as those used in a search for small lake herring in Green Bay caught large numbers of yearling lake herring in Lake Superior in 1953 (often more than 1,000 in a 10-minute tow). The best catches were made at 5 to 15 fathoms in the Apostle Islands area near Bayfield, Wis., and along the southeastern shore of the Keweenaw Peninsula. These collections have not yet been studied, but examination of a few specimens showed that they were just starting their second year of life. Young-of-the-year lake herring 10 to 12 mm. long were taken in surface plankton tows on May 29 and 30 near Bayfield, Wis. These fry match the descriptions by Prichard (1930) and Fish (1932) of lake herring of the same length from Lakes Erie and Ontario. Further evidence that fry of the genus Levcichthys may be pelagic was obtained wlien fry of either L. hoyi or L. reighardi (my tentative identification) were collected by the author with a dip net in open water on Lake Michigan near North Manitou Island on July 30, 1952. FECUNDITY The number of eggs produced by female lake herring varies widely both within and between populations. Jordan and Evermann (1902) and Bean (1902) carried similar accounts of what must be the same fish — a 2K-pound female tullibee from the "western territories of Canada" — that held 23,700 eggs. Cahn (1927), using the volumetric method, estimated the number of eggs of a 465- gram (about 1 pound) female from Lake Ocono- mowoc at 15,238. Bajkov (1930) stated that the tullibee of the Canadian prairie provinces carry 15,000 to 20,000 eggs. Brown and Moffett (1942), using a partial- weight method, estimated the number of eggs in ovaries of 9 ciscoes from Swains Lake. The results of their study were as follows: Average Range Number of eggs 30,328 23,272 to 37,272. 16.7 In 1S.2 to 16.2 In. Weight offish 1.721b 1.48 to 1.861b. No correlation was found between number of eggs and size or age of these fish. Scott (1951) also used the partial-weight method to estimate the egg count of 12 Il-group and 6 Ill-group ciscoes from Lake Erie. His findings are summarized as follows : Age group II: -N'umber of eggs Total length of flsh ' Weight of fish Age group III: Numner of eggs Total length of fish ' Weight of flsh Average 29.225.. 13.4 In. 1.181b.. 23,017.. 15.3 in.. 1.65 lb- Range 16,000 to 42,500. 11.7 to 14.4 in. 0.65 to 1.50 1b. 14.200 to 38,600. 13.8 to 1(1.6 In. 0.06 to 2.21 lb. ' This paper gave only standard lengths. Estimates of total length In this stock have been based on the assumption that the ratio of total lengtb to standard length was 1.19— a value near the middle of the range o( conversion factors listed by Carlander (1950). 124 FISHERY BXJLLETm OF THE FISH AND WILDLIFE SERVICE Scott found that the number of eggs tended to increase with length and weight of the female. He pointed out, however, that the apparent decrease in the number of eggs with increasing age of the fish, as shown by his data, may be in error, since all age group III fish were ripe when collected and unknown numbers of eggs were lost in handling them. Stone (1938) estimated (by the volumetric method) the number of eggs of 104 Irondequoit Bay lake herring to average 24,095; the mean length of these fish was 13.4 inches. The average number of eggs per fish in the different age groups ranged from 13,723 in the 2-year-olds (average length, 11.9 in.) to 48,999 in the 8-year-olds (aver- age length, 16.7 in.). The number of eggs was estimated by the dry- weight method for 72 Green Bay lake herring. This method was developed by Paul H. Eschmeyer and George F. Lunger, of the Service's Great Lakes Fishery Investigations, in studies of the fecundity of lake trout; they have not pub- lished an account of the procedure. The general procedure with lake herring ovaries (which differs somewhat from that followed by Eschmeyer and Lunger for lake trout) is as follows: The formalin-preserved ovaries are broken up thoroughly and the larger pieces of connective tissue are removed; the remaining materials are dried at 60° C. until there is no further weight loss; a sample of 100 eggs is removed and weighed (weighing is facilitated if the dried material is allowed first to reach moisture equilibrium with the atmos- phere) ; the total number of eggs is computed from ovary weight and sample weight. The dependability of the method was tested by making 38 estimates from 100-egg samples of 19 ovaries for which actual counts were made. The advantage of the dry-weight method in reducing error is clearly shown in the followmg comparisons: Method Percentage error Investigator Mean Range 0.3 6.7 5.1 0. 1 to 2. 2 0.4 to 21.0 3.0 to 14.0 Wet weight-- Volumetric Brown and Moflett (1942). Stone (1938). The number of eggs per fish in the Green Ba3'^ lake herring (table 39; fig. 12) varied widely but nevertheless exhibited a tendency to increase with length of the fish. In the entire sample of 72 fish with an average length of 11.2 inches, the average number of eggs was 6,375. Fish of different age but of the same length showed no diff'erence in egg 10 I I TOTAL LENGTH Figure 12. — Relation between length of Green Bay lake herring and number of eggs. The dots represent indi- vidual fish and circles are averages for 0.3-inch length groups; the line was fitted by least squares to the means of the 0.3-inch groups. number (details of analysis are not presented here). The relative number of eggs (expressed as number of eggs per ounce of body weight), contrary to the actual number, showed a downward trend with increase in length (fig. 13). For the entire sample (61 fish — 11 fish not weighed) the average number of eggs per ounce of fish was 1,012 (table 39). This value is below those for the ciscoes of Swains TOTAL LENGTH (INCHES) Figure 13. — Relation between length of Green Bay lake herring and the number of eggs per ounce of body weight. The dots represent individual fish and circles are aver- ages for 0.3-inch length groups; the hne was fitted by least squares to the means of the 0.3-inch groups. LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 125 Lake (1,103 eggs per ounce of fish). Lake Erie (1,546 eggs per ounce of age group II fish), and Irondequoit Bay (1,369 eggs per ounce of fish) as computed from data given by Brown and Moffett (1942), Scott (1951), and Stone (1938), respec- tively. Table 39. — Relation between the length of the individual lake herring and the number, weight, and size of the eggs it produces (Number of fish In parentheses] Total length (Inches) Number of eggs per fish Num- ber of eggs per ounce of flshi Average egg diameter (millimeters) Aver- age Range Octo- ber Novem- ber 8 5 to 8 7 3,748 (1) 5.985 (4) 5,182 (2) 6,662 (15) 6,079 (16) 5,790 (14) 6.140 (U) 7,663 (4) 8,109 (1) 8,368 (2) 8,061 (1) 5.304 (1) 1,102 (1) 1,202 (4) 1,027 (1) 1,156 (16) 976 (16) 918 (11) 851 (9) 986 (3) 977 (1) 1.61 (1) 1.55 (2) 1.68 (1) 1.59 (12) 1.63 (11) 1.65 (10) 1.64 (3) lOOtolO.2 4,419 to 7,641 5,025 to 6,339 3,968 to 11,212 3,471 to 9,102 3,783 to 9,924 4,602 to 8,120 5,085 to 10,250 1.86 10.3 to 10.5 (2) 1.99 10.6 to 10.8 - (1) 1.75 10.9 to 11.1 - (3) 1.95 11 2 to 11.4 (4) 1.91 11.5 to 11.7 (3) 1.94 11.8 to 12.0 - (1) 1.75 12 1 to 12 3 (1) 1 87 12 7 to 12.9 6,294 to 10,442 (1) 1.87 13 0 to 13 2 (2) 1.98 13 6 to 13 8 (1) 1 98 (1) AU lengths 6,375 (72) 3,471 to 11,212 1,012 (61) 1.62 (52) 1.88 (20) I Records of weight were lacking for 11 flsh. The average egg diameters showed no tendency to change with increase of length but were larger in fish of the spa^vning run in November (1.88 mm.) than in the prespawning October specimens (1.62 mm.). Other analyses revealed no correla- tion between egg diameter and total number of eggs in individual fish. SPAWNING Time and factors of spawning According to available records, lake herring in the latitude of the Great Lakes spawn sometime between mid-November and mid-December, and spawning activity at one location usually covers a period of 1 to 2 weeks. That the spawning date may differ with latitude is indicated by Dymond (1933), who found evidence that the lake herring of Hudson and James Bays spawn as early as September 10. Water temperature unquestion- ably is an important factor influencing the time of 388748 O 57 6 spawning. Cahn (1927) stated that ciscoes did not begin to spawn in Lakes Mendota and Oco- nomowoc, Wisconsin, until water temperature had dropped below 4° C, and that the temperature was either 3.1° or 3.0° C. at the time spawning ended (5 years of observations). To verify this apparent relation between temperature and spawn- ing, Cahn (p. 100) held 25 ciscoes in tanks with the following results: * * * The water was kept at a temperature of 4.5° C. during a period of four months [weeks?], covering the breeding season. In spite of the fact that fifteen of the confined fish were females, all heavy with eggs, not a single egg was laid during this time. In a second tank, exactly similar to the first, and with the same water sup- ply, but cooled by means of ice to a temperature of 3.5° C, females from the first tank spawned within ten minutes after transfer. \ second experiment consisted in transferring two females into the second tank while the water was 4.5° C. After two hours in this tank, a large piece of ice was added and a careful record of the temperature kept. The first female spawned with the temperature at 3.6° C, the second at 3.4° C. Monti (1929) found that whitefish did not spawn in Italian lakes where winter temperatures remained above 7° or 8° C. Evidence supporting the hypothesis of a critical breeding temperature was given by Pritchard (1930). During the spawn- ing period of the lake herring, which starts in raid- November, the temperatures at a hatchery intake near Belleville in the Bay of Quinte, Lake Ontario, were — Date Temper- ature (°C.) Date Temper- ature (°C.) Nov. 15 6.1 7.8 6.1 4.4 Nov. 23 - 4.4 16. 24. 3.3 17 25 3.3 18 Stone (1938) recorded a temperature of 3.8° C. shortly before ciscoes started to spawn in Ironde- quoit Bay, Lake Ontario. He observed also that spawning started earlier in the southern end of the bay where water temperatures dropped sooner than in the northern end. Washburn ' reported water temperatures near 3.3° C. during cisco spawning in Birch Lake, Michigan. Brown and MofFett (1942) found spawning at its peak in Swains Lake, Mich., on December 14, 1937, when • Washbrnn, Oeorge N. 1944. Experimental gill netting In HIrch Lake, Cass County, Michigan, Michigan Department of Conservation, Institute tor Fisheries Research, Rept. No. 948, 33 pp. (Typewritten.] 126 FISHERY BtJLLETIN OF THE FISH AND WILDLIFE SERVICE the surface-water temperature was "* * * prac- tically at the freezing point." They expressed the belief that spawning may have continued for several days after ice covered the lake. Although the lake herring of Green Bay spawns from mid-November to mid-December, considera- able variation in the progress of spawning activi- ties does take place. It is believed, however, that some spawning is going on every year some place in the bay during this entire period. An example of the dynamic situation during the progress of spawning in Green Bay is offered by records of catches of a single pound net at Sister Bay in December 1950. On December 2 the catch consisted mostly of ripe fish ready to spawn; on December 3 about 50 percent of the fish were spent; on December 4 all of the 112 lake herrmg examined from the catch of this net were spent; furthermore, their gonads were in an advanced state of recovery — a condition typical of that found in February. This observation suggests that (1) Lake herring move in schools during the spawning period in Green Bay; (2) fish of one school do not necessarily complete spawning in the place at which they have started; and (.3) schools in one area do not all spawn at the same time. Differences in progress of spawning between various groups of fish most probably are the re- sult of differences in the temperature regime in the several parts of this hydrographically complex bay (see General Features of Green Bay, p. 88). Available temperature records are inadequate for study of local differences during the spawning season. Records that have been made, however, show that the temperature drops through the 4° to 3° C. range during the last half of November and the first half of December when spawning takes place. Spawning grounds Most reports, particularly those concerning in- land lakes, indicate that eggs of lake herring are laid in shoal areas 3 to 10 feet deep (Bean 1902; Cahn 1927; Pritchard 1930; Stone 1938). Al- though no evidence was given that spawning did not occur in deeper water, the regularly observed movement of fish into shoal areas and back to deep water clearly indicates that the shallower region must be the preferred spawning area. Wagner (1911, p. 76) reported that in Green Lake, Wisconsin, which is 237 feet deep, "Local fishermen generally believe that spawning takes place at a depth of about seventy feet, on marly bottom, but this is somewhat doubtful." Koelz (1929) related that lake herring spawn in water 60 feet deep at the western end of Lake Erie and in water 30 to 150 feet deep at the eastern end. In Lake Ontario, Koelz said a deep-water form spawns in 90 to 180 feet whereas shallow-water lake herring spawn in 60 feet of water. In Green Bay, spawning fish are most concentrated in water 10 to 60 feet deep but catches of both ripe and spent fish are observed from nets fished at depths down to at least 140 feet. Apparently, spawning takes place over practically all depths and in all sections of the bay. Spawning lake herring in general show no pref- erence for a particular bottom type. Spawning has been reported over boulders, gravel, sand, marl, clay, mud, and aquatic vegetation. In Green Bay, spawning takes place over areas of boulders, sand, and mud, with no clear indication of preference. The failure to select particular bottom types probably stems from the fact that lake herring are pelagic spawners. Evidence that eggs are extruded a considerable distance above the bottom is given by Pritchard (1930), who found eggs evenly scattered over the bottom with no evidence of local concentrations that would be expected if eggs were deposited near the bottom. Spawning behavior Few observations have been made of the spawn- ing act of lake herring. Cahn (1927) described spawning activity of ciscoes in Lake Oconomowoc as being slow and deliberate with no chasing or darting about. In contrast with Cahn's observa- tion, Bean (1902) described the night-time spawn- ing activity of the tullibee in New York as being accompanied by "* * * constant loud splashing and fluttering." This type of activity has also been reported by Washburn (see footnote 9, p. 125) in Birch Lake, Michigan, where "The fish were seen darting about singly and in pairs, oc- casionally coming to the surface and splashing the water. The appearance of these fish on shoals would take place just before dark at or about sunset and continue until 10:00 p. m." Fishermen of Green Bay tell of similar jumping and splashing activity of lake herring during the spawning pe- riod. No observations have been reported on LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 127 the part this activity plays in the spawning process. Brown and MofTett (1942, p. 149) ob- served ciscoes breaking water in vSwains Lake in the early evening during the peak of the spawning period and remarked, "There was no concentra- tion of these fish. Approximately as many were seen over deep water as were observed over the shallows." The greatest depth of Swains Lake is 64 feet. Progress of spawning by age, sex, and size Although the lake herring population of Green Bay, taken as a whole, may show irregular prog- ress of spawning, there seems to be some pattern in the progress of spawning of individual segments of the population. The data of the only two samples that contained a good representation of both ripe and spent fish (table 40) show that, without excep- tion, the longer fish tended to spawn before the shorter fish and that the percentage of spent males was greater than that of females. In both sexes the upward trend in the percentage of spent fish with increase of age suggests that older fish spawn earlier than do younger fish. Table 40. — Comparison of ripe and spent lake herring in spawning-run collections from Green Bay, by sex and age group, November 2^30, 1950 [Length in inches] Ripe Spent Age group Number offish Total length (Inches) Number offish Total length (inches) Percent- age spent Males: I 1 2 29 8 9.7 10.8 10.8 11.1 0 II 2 45 12 2 11.1 11.1 11.2 11.8 50 Ill IV V Females: II 2 38 14 10.7 10.8 U.O 0 Ill , IV : 38 18 11.4 11.9 50 56 Predation on eggs Possibly the greatest mortality in the life cycle of lake herring takes place immediately after the eggs are laid. A common predator on these eggs is the lake herring itself. Stone (1938) found cisco eggs in 23 of 34 cisco stomachs collected during the spawning season. Pritchard (1931) found cisco eggs in 6 of 46 cisco stomachs. In Green Bay, 16 stomachs of 19 feeding lake herring taken on November 28, 1950, contained from 1 to 33 herring eggs. Although the lake herring com- monly eats its own eggs, other species seem to to make greater inroads. Stone found from a few to 200 cisco eggs in stomachs of 20 of 36 brown bullhead (Ameiurus nebiUosus) and believed this fish to be an important predator in Irondequoit Bay. Pritchard noted cisco eggs in the brown bullhead in Lake Ontario, but the yellow perch was a heavier consumer of cisco eggs (average of 275 eggs per stomach) during the peak of spawn- ing. He also found cisco eggs in whitefish stom- achs, but the numbers were small as whitefish were not present during the main spawning period. Rawson (1930) also reported that whitefish feed on cisco eggs. Jordan and Evermann (1902) found that the mud-puppy (Necturus maculosus) con- sumes cisco eggs in Lake Erie. These reports of predation on cisco eggs have been mostly inci- dental and have not been based on a special study of this problem. Since lake herring eggs, after being laid, lie unprotected on the bottom, varia- tion in the amount of predation at this stage may influence the relative strength of a year class. DISTRIBUTION AND MOVEMENTS The distribution of lake herring during the summer months has been a subject of much com- ment in the literature (Cahn 1927; Fry 1937: Hile 1936; Hile and Juday, 1941; Koelz 1929; Nelson and Hasler, 1942; Pearse 1921; Reighard 1915; Scott 1931 ; Stone 1938; Van Oosten 1930; Wagner 1911). Although the observations of various authors are not exactly comparable because char- acteristics of the bodies of water studied were different, the distribution is similar in all lakes of the same type. Upon the warming of surface waters in the spring the lake herring, a steno- thermic, cold-water animal, avoids this change by vacating shallow water. As warming continues and a thermocline develops, undesirable or intol- erable temperatures of the epilimnion may cause the lake herring to be restricted to the thermocline and hypolimnion. In the southern portion of their range lake her- ring are rarely found in lakes that do not develop thermoclines or where the hypolimnion becomes unusually warm. In Indian Village Lake, Indiana, near the extreme southern limit of the range, they have adapted themselves to conditions that might be considered intolerable elsewhere (Scott 1931). In lakes in which either the oxygen becomes depleted or undesirable gases are formed in the 128 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE hypolimnion, lake herring are forced to inhabit the area of the thermocline. In years of such extreme stagnation lake herring must choose be- tween the epUimnion with adequate oxygen and unsuitable temperatures, or the hypolimnion with inadequate oxygen and suitable temperatures. Cahn (1927) reported that in situations of this type heavy mortalities occur in southern Wiscon- sin lakes. During a period of extreme stagnation in Snow Lake, Indiana, Scott observed ciscoes coming to the surface, apparently in a state of asphyxiation. These fish recovered quickly, how- ever, and returned to deeper water. According to Koelz (1929) the lake herring of Lakes Erie and Huron follow the normal pattern by descend- ing into deep water during the midsummer months. Koelz reported that lake herring were taken in Lake Superior 1 mOe off Grand Marais, Minn., in floating gill nets all year except late July and early August. Surface temperatures in this region are always relatively low. Among the authors who have reported on the vertical distribution of the lake herring, only a few have given limnological records or experi- mental data from which judgment can be formed as to limiting values of the controlling factors. Most detailed consideration of the problem was given by Hile (1936) and Fry (1937). From Hile's data on the vertical distribution of ciscoes and on temperature and oxygen conditions during the general period of his fishing operations (he had no limnological records on the actual dates of lifting gill nets) in Muskellunge and Silver Lakes, it may be seen that ciscoes were taken only rarely at temperatures above 17° to 18° C, which marked the upper limit of their distribution, or at oxygen concentrations below 3 or 4 parts per million at the deeper limit of their distribution. Fry com- mented that he seldom took ciscoes in Lake Nipissing in water 20° C. or warmer. He men- tioned oxygen depletion as a possible limiting factor for the lower limit of distribution, but con- sidered carbon dioxide concentration to be of greater importance in making the hypolimnion uninhabitable. Hile did not mention carbon diox- ide as a factor in the distribution of the cisco, but in a later publication on the bathymetric distribu- tion of fish in several lakes of northeastern Wiscon- sin (Hile and Juday, 1941) skepticism was ex- pressed as to the influence of both carbon dioxide and pH concentrations on the distribution of various species in those waters. Cahn's (1927) aquarium experiments indicated that ciscoes avoided temperatures above 17° C. Of the several possible limiting factors men- tioned by earlier investigators only temperature can be held important in Green Bay. Oxygen concentrations in the deeper waters of the bay during the summer of 1952 were always above 7 p. p. m. and the pH fell within the range 7.8 to 8.2. Although determinations were not made of carbon dioxide concentrations in deep water during the summer, the values for oxygen and pH constitute prima-facie evidence that carbon di- oxide was not present in excessive amounts. VERTICAL DISTRIBUTION IN GREEN BAY The occurrence of lake herring in commercial nets in Green Bay gives some information about the vertical distribution of the herring. In months of cool weather (September or October to May or June) lake herring are commonly taken in pound nets fished in shallow areas and in gill nets fished at all depths. During other months, how- ever, nets set in shallow water make only oc- casional catches, usually following a storm, and gill nets fished on the bottom in deeper water take few lake herring. In 1952, a study was undertaken to determine the distribution of the lake herring before, during, and after the summer period when, according to the fishermen, the lake herring "disappear." Oblique gUl-net sets,'" similar to those used by Fry (1937), were employed to determine the depth at which lake herring were located. In these sets 140 linear feet of gill netting were fished in every 15-foot stratum. At station B, where the water was 40 feet deep, 140 feet of gill netting were fished in each 20-foot stratum. One end of the gang of nets was tied to an anchor and the other end to a 15-gallon-drum float. The depths at which segments of the nets fished were con- trolled by gallon-jug floats attached to the nets with lines whose lengths were multiples of 15 feet. To hold the nets tight, an anchor rope about equal in length to the gang was tied to the 15-gallon drum, puUed against the first anchor, and set with a long buoy line (see figure 14 for a diagram of an oblique set in 60 feet of water). " These experimental nylon gill-nets were 280 feet long and 6 feet deep, and had mesh sizes of 1, H, and 2 inches, eitension measure. The 2-inch-mesh nets were used most eitensively. i LAKE HERRING OF GREEN BAY, LAKE MICHIGAN Table 41. — Vertical distribution of lake herring taken in oblique sets of gill nets in different periods [Mesb slies, extension measure, In Inches] 129 Month, station, and depth date set and lifted Mesh size Number of fish Percentage of catch at— 0-15 feet 15-30 feet 30-45 feet 45-60 feet 60-76 feet 75-90 feet EARLY mat: A (30ft.) 3-4 3- 4 3- 4 1-2 8-9 8- 9 8- 9 10-11 10-11 10-11 5- 6 5- 6 6- 7 7-8 7- 8 24-25 24-25 24-25 21-22 21-22 21-22 10-1 1 11-12 11-12 23-24 23-24 27-27 21-22 21-22 19-21 19-21 19-21 21-22 21-22 21-22 22-23 23-24 23-24 24-25 24-25 24-25 IVi IM IM 1 J^ 2 2 2 IH Hi 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 0 1 0 80 5 7 45 26 115 118 0 28 0 0 0 13 32 5 85 31 31 82 18 46 0 19 6 10 24 152 90 197 15 46 379 19 189 439 1,220 645 446 0 100 0 1 0 42 9 65 30 70 0 7 0 0 0 85 94 60 60 87 87 0 17 0 0 0 0 0 0 1 1 0 100 96 45 58 52 29 14 34 11 0 0 0 2 0 0 4 23 7 12 0 7 0 0 0 15 6 40 34 10 13 1 6 9 0 5 0 0 0 1 3 13 0 4 37 42 33 36 21 36 52 B (40ft. )i C (60 (t.) 0 11 20 0 45 12 11 10 0 14 0 0 0 0 29 20 29 33 0 28 8 0 11 0 0 0 D(90ft) 39 40 29 7 18 Do 20 Do 0 Do --.. 2 E (60 ft ) F(90ft) 19 5 O(60ft.) H (60 ft ) I (90ft.) 21 0 40 J (90 ft ) 0 K (60 ft ) L(60ft.) - LATE MAY. A (30ft.) - B (40 ft.)' - C(60ft) 0 4 0 0 14 33 37 0 2 0 0 39 44 54 E (60 ft.) - 0 3 Q(60ft.) J (90ft.) 28 18 L(60ft) A (30 ft.) 42 17 40 21 7 78 S3 53 0 60 63 12 18 34 D (90 ft.) 17 66 H(60ft.) - I (90 ft.) 8 14 8 J (90 ft.) 65 K (60ft.)._ OCTOBER: A (30 ft.).- - B (40ft.)i C (60 ft.) 11 0 9 19 28 23 21 7 0 6 13 26 7 16 D(90ft.) 0 0 H (60ft.)._. I (90 ft.) 2 9 1 J (90 ft.) - 2 K (60 ft.) L(60ft.) Depth Intervals 0-20 and 20-40, one 280-foot net used from surface to bottom. i^si^^sgMgg;ss^i:^sg::^^?gj;^^^^t^^i^^^^j^!j^^ Figure 14. — Method of setting a gill net in an oblique position. Horizontal scale much reduced. It was obviously impossible to avoid some sagging in the gang. The amount of sag was lessened by the action of currents which are almost always fairly strong in Green Bay. Oblique sets of gill nets were fished at 12 stations in representative areas throughout Green Bay (fig. 1) from early May to late October 1952. One station (A) was established in 30 feet of water and another (B) in 40 feet in the shallow southern portion of the bay. Six stations (C, E, G, H, K, L) were located in 60 feet of water, and 4 stations (D, F, I, J) in 90 feet. No lake herring were taken at the shallow- water station A (30 feet) in southern Green Bay and only one was caught at station B (40 feet) in early May (table 41); only a few were taken in late May (A, 13 fish; B, 32 fish) at which time they were found mostly in the upper 15 to 20 feet of water. No lake herring were obtained at the 30-foot station in July. A few were again taken at both stations in October (A, 15 fish; B, 46 fish) when lake herring were concentrated near the surface. Since the distribution of fish varied randomly among individual 60- and 90-foot stations in southern, central, and northern Green Bay during any one season, data for stations of equal depth have been combined to show seasonal differences in distribution. The graphical representations of distribution of lake herring at 60-foot stations 130 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE OCTOBER 21 -2S Figure 15.— Vertical distribution of lake herring taken in oblique sets of gill nets in 60 feet of water. The full width of the panel for each time period is 100 percent. OCTOBER 22-25 Figure 16. — Vertical distribution of lake herring taken in oblique sets of gill nets in 90 feet of water. The full width of the panel for each time period is 100 percent. (fig. 15) and at 90-foot stations (fig. 16) are given as unweighted mean percentages for all stations where lake herring were caught. Despite the concentration of lake herring in the upper 15-foot stratum at 60-foot stations (E and G) in early May (67.9 percent, fig. 15), evidence at 90-foot stations where nets were fished on the same and other dates in early May (table 41, fig. 16) suggests a possibly random distribution. In late May the lake herring were concentrated in the top 15-foot stratum at both 60- and 90-foot stations (69.0 and 87.1 percent); most of the remaining fish were at the 15- to 30-foot level (29.0 percent at 60-foot stations and 9.7 percent at 90-foot stations). Lake herring exhibited a strong tendency to move into water deeper than 30 feet in June (84.5 percent at 60-foot stations and 98.7 percent at the 90-foot station); the tendency toward concentration below 30 feet was still greater in July (94.4 and 99.3 percent at 60- and 90-foot stations). Lake herring were present in fair numbers except in deepest water (75 to 90 feet) in October but showed a decided tendency to be concentrated in the upper 30 feet (74.6 and 66.8 percent at 60- and 90-foot stations). The general seasonal trend in vertical distribu- tions from May to October may be summarized as follows. The first change was from a variable pattern in early May to a pelagic distribution in late May. In June and July the lake herring had descended to depths greater than 30 feet and by October they had resumed the pelagic habitat with the greatest concentration between the surface and 30 feet. The distribution of lake herring during the spawning period in November and December has been brought out in the discussion of spawning activity. Distribution during winter and early spring is subject of much speculation. The few observations that have been made lead to the conclusion that schools of lake herring may be found at any depth. The vertical distribution of lake herring showed no relation to temperature, except in the avoid- ance of water with temperature near or above 20° C. In early May when the water was rela- tively cool (3.2° to 7.6° C. at stations where lake herring were caught) and varied little with depth (table 42), the distribution of herring was largely random (table 41). Late May temperatures are available for only the shallow-water stations (A and B); the surface temperatures at these loca- tions were 14.1° and 14.5° C. and bottom tem- peratures were 13.0° and 11.6° V. Although the lake herring had moved toward greater depths in June, it cannot be assumed that increasing water temperature near the surface was the cause. The temperatures from the surface to 30 feet were between 12.3° and 15.1° C. (total range at all stations) — not greatly different from those at stations A and B in late May. The lake herring continued to be concentrated below 30 feet in July. During this month water tempera- tures at less than 30 feet (fig. 17 and table 42) usually were within the range of 18.3° to 21.5° C. (the range from 15.6° at 20 feet to 18.6° at the surface on July 24 represents a transitory situation following a severe storm). Since these July tem- peratures from the surface to 30 feet were mostly near or above the values considered critical for the lake herring (see p. 128), it is probable that tem- perature conditions held the lake herring in greater depths in July. This view is supported by the absence of lake herring at shallow-water station A LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 131 Table 42.— Water temperature at or near experimental gill-net stations [Temperature records made with a bathythermograpb. Depth of cast (feet) is Indicated by deepest temperature recorded] Period and station ' Day of month Temperature (° C.) at depth of— Ofeet 10 feet 20 feet 30 feet 40 feet SO feet 60 feet 70 feet 80 feet 90 feet 100 feet E*RLY may: A - - 3 3 1 8 10 11 11 10 11 5 6 7 7 24 24 10 11 11 11 23 24 24 23 24 22 27 21 21 22 19 21 19 21 19 21 21 21 21 22 22 23 23 23 24 24 24 24 12.1 13.1 3.2 9.2 6.6 6.0 6.9 7.6 7.5 5.5 7.S 8.6 6.0 14.1 14.5 12.9 13.1 13.6 15.1 20.5 19.8 19.6 19.9 18.6 20.2 20.6 20.3 21.3 21.5 20.5 21.4 20.3 20.2 20.7 20.5 8.4 8 9 9.2 9.4 9.9 9.8 10 2 10 1 10 0 9.5 9.7 8.5 9.1 8.4 3.2 8.0 6.6 6.0 5.9 7.0 7.1 S.4 7.5 8.2 6.0 13.8 13.6 12.9 13.1 13.5 15.0 20.5 19.8 19.3 19.9 16.9 20.1 20.5 20.2 21.1 21.5 19.9 21.2 19.7 19.9 19 4 20.4 8.4 8.9 9.2 9.4 9.9 9.8 10.2 10.1 10.0 9.5 9.6 8.3 8.6 7.4 3.2 6.9 6.5 6 0 5.6 6.6 6.9 5.1 7.4 8.0 5.3 13.1 13.5 12 8 12.5 13.0 14.4 20.5 19.7 18.3 19 6 15.6 19.9 20.2 19.9 21.0 21.2 19 2 19.7 19.1 19.5 18.7 19.9 8.4 8.9 9.2 9.4 9.9 9.8 10 2 10.1 10.0 9.4 9.4 8.3 B - 6.9 3.2 6.6 6.5 6.0 5.5 5.9 6.8 4.5 7.3 7.6 5.2 13.0 13.3 12.8 12.3 12.6 13.7 20.3 19.7 11.4 18.5 14.7 11.5 20.0 11.7 20.4 19.5 18.5 18.5 19.0 18.6 15.6 14.5 8.4 8.9 9.2 9.4 9.9 9.7 10.2 10.0 10 0 9.3 9.4 8.2 D 3.2 6.2 6.5 6.0 5.2 4.9 6.6 4.2 7.2 6.6 5.2 3 2 5.8 3.2 .<> 4 3.2 5.3 5.9 6.0 4.3 3.2 5.2 3.2 5.1 D SO E 6. 5 6 3 E 6.0 4.8 4.8 6.1 4.2 7.1 6.1 5.2 6.0 4.3 4.7 5.3 4.2 7.0 S.3 5.2 F 4.3 4.3 0 0 I 4.2 6.9 4.2 5.1 4.2 SO 4.2 J 5.0 K L LATE MAV: A - B 11.6 12.6 11.6 11.5 11.2 JUNE: J 12 2 10.8 9.3 7.3 11.1 9.1 7.7 6.2 8.4 7.5 6.0 7.6 6.0 7.3 K - K L ji'LT: A A B 10.2 12.0 14.2 9.8 17.7 9.2 18.0 14.6 13.9 14.5 17.0 14.5 14.0 12.1 C .... 9.6 13.0 8.8 10.0 8.9 12.6 10.1 10 6 12.9 11.0 10 4 11.5 10 0 9.0 9.6 7.9 8.1 8.4 8.5 8.4 9.9 11 5 10.2 9.6 9.6 9.1 7.8 7.5 8.4 7.9 7.2 8.9 10 1 10.1 9.4 9.0 8.7 7.6 7.4 C D 7.S 7.3 7 2 H 7.2 6.8 8.3 10.0 6.9 6.8 8.0 9.0 6.8 I 8.7 J 8.0 J. 8.7 K K L L 9.1 OCTOBER: A B... 8.9 9.2 9.4 9.8 9.7 10.1 10 0 10 0 9.3 9.4 8.1 C 9.2 9.4 9.7 9.7 10.1 10.0 10.0 9.3 9.4 8.1 9.2 9.4 9.7 9.6 10.0 10.0 10.0 9.3 9.4 8.1 c D ._ 9.6 9.5 10.0 10.0 9.9 9.3 9.4 9.5 9.3 9.5 B.S D H- I... 10.0 9.8 9.3 10.0 9.8 9.3 10.0 1- . J K 9.3 L . I See figure 1 for location. where the temperatures on July 23 and 24 were 19.7° to 20.5° C. In October when the lake herring were again most plentiful from the surface to 30 feet, the water temperature was generally cool (8.1° to 10.2° C.) and the temperature gradi- ents from top to bottom were insignificant (great- est difference 0.4°). The one previous study of the vertical distribu- tion of the lake herring in relation to size of fish was made by Fry (1937) in I^ake Nipissing. Fry found that the movement from shallow water to the hypolimnion was not a mass descent but was "* * * an orderly succession of certain groups of individuals which migrate in order of size and sex." Consideration of the distribution in rela- tion to sex was made earlier (p. 120) in the dis- cussion of sex composition. In the summary o the length of lake herring taken at various depths (table 43) the sexes are combined, as no sex differ- ences in length were found for fish taken in the same depth of water. Despite certain exceptions (usually at depths in which the catches were small) the average size of lake herring tended to decrease with increase in depth of water in all collecting periods. Davidoff " found a similar tendency for the larger ciscoes of Myors Lake, Indiana, to be near the surface. The seasonal changes in the distribution of the lake herring must be a major cause of the highly " Davidofl, Edwin B. 1953. Growth, response to netting, and bathy- metric distribution of the Cisco. Levcichthysarledii (Le Sueur) In Myers Lake. Indiana. M. S. thesis. Department of Fisheries. University of Michigan. (Typewritten.) 132 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE D H I J K J I L 10 IS 20 TEMPERATURE (°c) 10 IS 20 10 15 20 15 20 25 Figure 17. — Relation between temperature and vertical distribution of lake herring July 19-27, 1952. See figure 1 for locations of stations. The full width of the panel for each station represents 100 percent. Table 43. — Length of lake herring taken in oblique sets of gill nets, by depth and season, 1952 12-inch-mesh gill nets, all stations combined. Length in inches] May 1-11 May 21-25 June 10-12 July 19-27 October 21-26 Depth (feet) Number offish Total length Number offish Total length Number offish Total length Number offish Total length Number offish Total length n to 15 137 30 48 66 26 7 11.0 11.0 10 9 10.9 10.8 10.7 119 40 3 2 11.1 10.9 10.9 10.5 4 6 34 65 23 15 11.1 10.5 10.8 10 9 10.8 10.7 2 31 87 95 24 33 11.1 11.0 10.7 10.6 10.7 10.7 279 210 206 212 38 33 11.3 15 to 30— 11.2 30 to 45-- 11.0 46 to 60 10.8 60 to 75 11.1 76to90- 1 10.7 10.7 seasonal character of the fishery. Nearly half (47.2 percent) of the commercial catch is made during the fourth quarter of the year (fall) and a fourth (24.4 percent) during the first quarter (winter). (See table 2.) Production is much lower in the spring (19.5 percent) and summer quarters (8.9 percent). Principal gear for taking lake herring are pound nets which fish from the surface to the bottom and are seldom set at depths greater than 35 to 40 feet, and gill nets which are set on the bottom at all depths but are effective only 6 to 1 1 feet above the bottom.'^ Thus, pound nets can take herring only when the fish are in the shallower inshore waters and gill nets can capture them only when the fish are near the bottom. From figures 15 and 16 it may be seen that in June and July " In State of Michigan waters gill nets may not be more than 11 feet deep (distance from float line to lead line) ; In Wisconsin the greatest depth allowed (stated In numbers of meshes) is about 6 feet. most herring were in water too deep to be taken by pound nets. The same condition most probably held in August, September, and even early October. During the same period the lake herring should have been available to gill nets set in deeper water. Examination of the nets as they were lifted from oblique sets revealed, however, that most of the herring taken in the bottom 15- foot stratum were caught in the upper half of the net section. Because the fish are some 6 to 8 feet above the bottom, commercial fishing with gill nets during the summer period is not productive. The distribution pattern in which some, and at times most, lake herring are above the bottom and out- side the 30- to 40-foot contour provides them a considerable degree of protection from commercial exploitation. In view of the relative inefficiency of present fishing gear, the probability of depletion by the present fishery is small. LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 133 MOVEMENTS AND ACTIVITY Available evidence shows that the lake herring, as well as other coregonids in the Great Lakes, does not undergo extensive migrations. Principal source of information is the study of Smith and Van Oosten (1940), who reported the following percentage recaptures from Lake Michigan fish tagged from 1929 to 1931 near Port Washington, Wis.: 5.4 percent from 593 lake herring; 22.1 percent from 457 whitefish; 5.7 percent from 106 chubs {Leucichthys spp., other than lake herring); and 20.0 percent from 35 pilots, or round white- fish {Prosopium cylindraceum quadril late rale). Lake herring were not recovered at distances greater than 50 miles from point of tagging, while lake trout and rainbow trout tagged in the same study were recaptured at distances as great as 125 to 225 miles away. The percentage distribution of recoveries was as follows: Miles traveled from point of release Lake herring Whitefish Chubs Pilots 1 to 10 69 28 3 67 29 I 3 100 100 11 to 25 26 to 50 51 to 75 Jarvi's (1920) study of the "kleine Marane," a species similar to the lake herring, in Keitelesee, Finland," disclosed the presence of distinct stocks, with respect to growth and age composition, in different basins. In view of these differences he concluded that the movements of the "kleine Marane" must be limited and that the few ob- served migrations probably resulted from unusual temporary conditions. Local movements of the lake herring that have been observed probably are the result of thermal conditions or represent spawning and feeding ac- tivities. The vertical movement accompanying thermal stratification is not as great as the hori- zontal distances that must be traveled in Green Bay when the fish abandon the .warming shallow- water areas to seek colder water. This distance amounts to about 10 miles in northern Green Bay and 25 miles in southern Green Bay. Similar dis- tances are covered in the return to shallow-water areas prior to and accompanying spawning. In Lake Erie the summer and spawning movements, according to the distribution described by Van " The greatest length of Keitelesee Is 72 kilometers, or about 45 miles. Oosten (1930) must involve distances of 100 miles or more. Cahn (1927) found that theciscoes of Oconomo- woc Lake were closer to the surface at night than during the day, and he interpreted this diurnal mi- gration as a feeding movement. Jarvi (1920) ob- served the same diurnal movement in the "kleine Marane" of Keitelesee. He believed that the local horizontal movements of schools of "kleine Ma- rane," as well as the diurnal movements, were asso- ciated with feeding. Similar movements of lake herring schools in Green Bay are shown by the highly erratic catches of nets fished in the same locality day after day. A good example is found in the catches at Sister Bay on December 2, 3, and 4, 1950, where a pound net apparently took members of three different schools on three suc- cessive days (see p. 126). There is some evidence that the strong currents, which are common in Green Bay, are responsible for movement of lake herring. These movements are reported by commercial fishermen who occa- sionally, following summer storms, take lake her- ring in shallow-water areas where they are not normally found during the summer period. These occurrences indicate that lake herring can be trans- ported by currents. SUMMARY 1 . The lake herring occurs in many of the deeper, colder lakes of the northeastern section of the United States, over most of Canada and Alaska, and also in Hudson and James Bays. It is rarely found in rivers. 2. Green Bay is one of the most productive com- mercial fishing areas in the Great Lakes and the lake herring is a major contributor to the total catch in the bay. In 1952 Green Bay produced 38.7 percent of the total take of lake herring from all United States waters of the Great Lakes. The commercial catch fluctuates widely, but this study was conducted during years (1948-52) when production was high and relatively stable. 3. Green Bay is 118 miles long and 23 miles wide. Water exchange with Lake Michigan is rela- tively free in the northern end of the bay, but prac- tically nil in the southern section. Water move- ments in the bay are complex and often are of considerable magnitude. They result in an un- stable, almost continually changing environment within the bay. 134 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 4. Scales for studies of age, growth, and year- class strength were collected from 4,390 lake herring taken from pound and gill nets. Investi- gation of most phases of the life history was based on catches of pound nets, which are less selective with respect to size of fish than are gill nets. Length records were obtained for all and weight and sex data for most of the 2,039 lake herring taken from experimental gill nets. 5. Age determinations were made by examining the magnified image of scales projected on a screen. Fish with no annulus (year-mark) were assigned to age group 0, those with 1 annulus to age group I * * *. All fish were considered to pass into the next higher age group on January 1. 6. The maximum age of lake herring reported in any population is XII ; the oldest fish from Green Bay belonged to age group VII. The best-repre- sented age groups in the various populations for which there are published records are age groups II to V; age groups III and IV were the most plen- tiful in Green Bay. The commercial catch in Green Bay was dominated by age group IV in the period January to June and by age group III in July to December. 7. The age composition of lake herring from pound nets was not representative of the popula- tion, as young fish were seldom taken even though the mesh sizes (1)2 to 2 inches, extension measure) were small enough to hold them. Yearling lake herring, as a rule, do not inhabit the relatively shallow, inshore areas where pound nets are fished. 8. The length of lake herring from the commer- cial pound nets and gill nets varied little from season to season. Even during the summer period of rapid growth the effects of individual increases in length were largely compensated by the selec- tive destruction of the larger lake herring in the fishery and by the shift to a lower average age. 9. The relation between the total body length in inches (L) and the magnified (X41) scale diameter in millimeters (S) of Green Bay lake herring is described by the formula L = 0.01615 + 0.05486 S Since the intercept is so small, its value was assumed to be 0, and lengths at the end of various years of life were calculated from scale measure- ments by direct proportion. 10. Annuli are formed on scales of the Green Bay lake herring in May and June. The progress of annulus formation is irregular, possibly because of different local environmental influences. The younger age groups and the smaller fish within an age group tend to form annuli earliest. 11. Growth within the season was described by a sigmoid curve. Growth started about the first of May and terminated near the end of October, with the fastest growth in July. 12. Males and females grew at the same rate. 13. Selective destruction of fast-growing in- dividuals was so great that seasonal differences in style of growth were detectable, that is, lake herring taken early in the year had grown faster in earlier years than had fish of the same age group captured later in the same year. 14. Calculated length at the end of the first year of life increased from north to south. These first-year differences, almost surely of environ- mental origin, were rapidly reduced by com- pensatory growth in later years of life. 15. Annual fluctuations in growth in length indicated that conditions affecting growth of lake herring in Green Bay changed little from year to year. The growth rate was below average and decreasing from 1944 to 1946, improved from 1946 through 1950, and then declined somewhat in 1951. Growth was well above average during the period 1949-51. 16. The different age groups exhibited sys- tematic discrepancies in calculated growth re- sembling those commonly termed Lee's phenom- enon of "apparent decrease of growth rate." Selective destruction of the larger, faster-growing fish by the commercial fishery was held to be the most important of the various factors that may have contributed to the discrepancies. 17. Growth compensation takes place in Green Bay lake herring. It was shown that growth compensation will appear in the calculated growth of fish that follow identical growth curves but that are hatched at different times in the season. It was also demonstrated that length rather than age is the primary determinant of subsequent growth of the individual, and hence that growth compensation can occur among fish whose growth curves are different. 18. The general length-weight relation of the Green Bay lake herring is described by the equation log W^= -2.4386 + 3.0729 log L, LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 135 where W is weight in ounces and L is total length in inches. 19. Weight varied according to sex and to method, season, and year of capture. 20. Females were relatively more abundant in samples taken from pound nets in February than in May to December. They were also more plentiful in the younger age groups than in the older. The selective destruction of females in younger age groups may be a major factor in the progressive decline with increased age in the percentage of females in a year class. 21. The percentage of females in the Green Bay lake herring population declined continuously from 1949 to 1952. 22. The percentage of females in collections taken in oblique sets of gill nets increased with depth of water in October. This change in sex composition with depth may reflect an actual difference in the distribution of the sexes, but a difference in the activity of the sexes may have been a major factor. 23. Some Green Bay lake herring matured during their second year of life, and all had reached maturity by the end of the third year. 24. Lake herring spawn in Green Bay between mid-November and mid-December, but spawning of an individual school of fish may be completed in a fraction of this period. Fish of the same school do not necessarily complete spawning in one location. Within a school, the older fish and the larger fish of an age group tended to spawn first and the males spawned earlier than the females. 25. Lake herring are pelagic spawners; the eggs are broadcast and settle unprotected to the bottom. Inshore areas are preferred, but there is evidence that lake herring may spawn in Green Bay over water as deep as 140 feet. 26. The literature indicates that lake herring hatch in early spring (April-May) and that newly hatched fry are pelagic. Young-of-the-year lake herring have rarely been collected. They prob- ably lead a bathypelagic existence where they are relatively immune from capture by the usual methods of collection. 27. Although the number of eggs produced by female lake herring (range, 3,471 to 11,212) varied widely for fish of the same total length as well as of different lengths, the number of eggs tended to increase with length of the fish. The relative number of eggs (i. e.. number per ounce of bodv weight) tended to decrease with increase of length. 28. The lake herring of Green Bay were ran- domly distributed from top to bottom in early May, but they were concentrated in the upper 15 to 30 feet in late May. They descended to deeper water in June and were restricted to strata more than 30 feet below the surface in July when tem- peratures in shallower water were unfavorable (17° C. and above). In October, lake herring were again found at all levels but were most abundant in the upper 30 feet. 29. Lake herring are not migratory but they sometimes move considerable distances to avoid unfavorable temperatures. Local movements are probably associated with feeding or represent passive transport by currents. 136 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE LITERATURE CITED Bajkov, Alexander. 1930. Fishing industry and fisheries investigations in the prairie provinces. Trans. American Fish- eries Soc, vol. 60, pp. 215-237. Bauch, Gerd. 1949. Untersuchiingen Uber das VVachstum der kleinen Marane (Coreqonus albula L.) in den Gewassern Mitteleuropas. Abhandlungen Fisch- erei, Deutschen Forschungsanstalt fiir Fischerei, Lieferung 2, S. 239-326. BEAINf, Tarleton H. 1902. Food and game fishes of New York. New York Forest, Fish, and Game Comm., Rept. No. 7, pp. 251-460. Brown, C. J. D., and J. W. Moffett. 1942. Observations on the number of eggs and feed- ing habits of the cisco (Leucichthys arledi) in Swains Lake, Jack.son County, Michigan. Co- peia 1942, No. 3, pp. 149-152. October 8. Butler, Robert L., and Lloyd L. Smith, Jr. 1953. A method for cellulose acetate impressions of fish scales with a measurement of its reliability. United States Department of the Interior, Fish and Wildlife Service, Progressive Fish-Culturist, vol. 15, No. 4, pp. 175-178. October. Cahn, Alvin Robert. 1927. An ecological study of southern Wisconsin fishes. Illinois Biological Monographs, vol. XI, No. 1, 151 pp. January. [Contributions from Zoological Lab. of the University of Illinois, No. 297.] Carlander, Kenneth D. 1945. Growth, length-weight relationship and popu- lation fluctuations of the tullibee, Leucichthys arledi tullibee (Richardson), with reference to the commercial fisheries, Lake of the Woods, Minne- sota. Trans. American Fisheries Soc, vol. 73 (1943), pp. 125-136. 1950. Handbook of freshwater fishery biology, v -f281 pp. Wm. C. Brown Co., Dubuque, Iowa. Clemens, Wilbert A. 1922. A study of the ciscoes of Lake Erie. Univer- sity of Toronto Studies, Biol. Ser., No. 20, Pub- lications Ontario Fisheries Research Lab., No. 2, pp. 27-37. Toronto. Cooper, Gerald P. 1937. Age, growth, and morphometry of the cisco, Leucichthys arledi (Le Sueur), in Blind Lake, Washtenaw County, Michigan. Papers Michi- gan Acad. Sciences, Arts, and Letters, vol. 22 (1936), Zoology, pp. 563-571. Ann Arbor, Mich. Dannevig, Alf, and Gunnar Dannevig. 1937. The season in which "winter" zone.s in the scales of trout from .southern Norway are formed. Jour, du Con.seil, Con.seil Permanent Inter- national pour L'E.xploration de la Mer, vol. 12, No. 2, pp. 192-198. Copenhague. Deason, Hilary J., and Ralph Hilb. 1947. Age and growth of the kiyi, Leucichthys kiyi Koelz, in Lake Michigan. Trans. American Fisheries Soc, vol. 74 (1944), pp. 88-142. Dymond, J. R. 1933. Biological and oceanographic conditions in Hudson Bay. 8. The coregonine fishes of Hud.son and James Bays. Biological Board of Canada, Contributions Canadian Biology and Fisheries, vol. 8, No. 1 (Ser. A, general. No. 28), pp. 1-12. Toronto. 1943. The coregonine fishes of northwestern Canada. Contributions Royal Ontario Museum Zoology, No. 24, pp. 171-232. 1947. A list of the freshwater fishes of Canada east of the Rocky Mountains. Royal Ontario Mu- seum Zoology, Misc. Publications No. 1, 36 pp. Toronto. Eddy, Samuel, and Kenneth D. Carlander. 1942. Growth rate studies of Minnesota fish. Min- nesota Department of Conservation, Division of Fish and Game, Bureau Fishery Research, In- vestigational Rept. No. 28, 64 pp. Farran, G. p. 1936. On the mesh of herring drift-nets in relation to the condition factor of the fish. Jour, du Con- seil, Conseil Permanent International pour L'Ex- ploration de la Mer, vol. 11, No. 1, pp. 43-52. Copenhague. Fish, Marie Poland. 1932. Contributions to the early life histories of sixty-two species of fishes from Lake Erie and its tributary waters. United States Bureau of Fisheries, vol. 47 (1935), Bull. No. 10 (1932), pp. 293-398. Ford, E. 1933. An account of the herring investigations con- ducted at Plymouth during the years from 1924 to 1933. Jour. Marine Biological Association of the United Kingdom, vol. 19, N. S., pp. 305-384. I'rey, David G. 1942. Studies on Wisconsin carp. 1. Influence of age, size, and sex on time of annulus formation by 1936 year class. Copeia 1942, No. 4, pp. 214-223. December 28. Fry, F. E. J. 1937. The summer migration of the cisco, Leucich- thys arledi (Le Sueur), in Lake Nipissing, On- tario. University of Toronto Studies, Biol. Ser., No. 44, Publications Ontario Fisheries Research Lab., No. 55, 91 pp. Toronto. Greeley, John R., and C. Willard Greene. 1931. A biological survey of the St. Lawrence water- shed. II. Fishes of the area. Supplement to 20th Annual Rept. for 1930, New York Conserva- tion Department, pp. 44-94. Hansen, Donald F. 1937. The date of annual ring formation in ths scales of the white crappie. Trans. American Fisheries Soc, vol. 66 (1936), pp. 227-236. LAKE HERRING OF GREEN BAY, LAKE MICHIGAN 137 HiLE, Ralph. 193]. The rate of growth of fishes of Indiana. In- vestigation of Indiana Lakes. II. Indiana De- partment of Conservation, Division of Fish and Game, Publ. No. 107, pp. 7-55. 1936. Age and growth of the eisco, Leucichthyx artedi (Le Sueur), in the lakes of the northeastern high- lands, Wisconsin. United States Bureau of FisherieF, vol. 48, Bull. No. 19, pp. 211-317. 1937. Morphometry of the cisco, Leucichthys artedi (Le Sueur), in the lakes of the northeastern high- lands, Wisconsin. Internationale Revue der gesamten Hydrobiologie und Hydrographie, Band 36, S. 57-130. 1941. Age and growth of the rock bass, Ambloplites rupeslris (Rafinesque), in Nebish Lake, Wiscon- sin. Trans. Wisconsin Acad. Sciences, Arts, and Letters, vol. 33, pp. 189-337. Madison. HiLE, Ralph, and Hilary J. Deason. 1947. Distribution, abundance, and spawning season and grounds of the kiyi, Leucichthys kiyi Koelz, in Lake Michigan. Trans. American Fisheries Soc, vol. 74 (1944), pp. 143-165. HiLE, Ralph, and Chancy Juday. 1941. Bathymetric distribution of fish in lakes of the northeastern highlands, Wisconsin. Trans. Wis- consin Acad. Sciences, Arts, and Letters, vol. 33, pp. 147-187. HiLE, Ralph, George F. Li'noer, and Howard J. Bt'ETTNER. 1953. Fluctuations in the fisheries of State of Michi- gan waters of Green Bay. United States Depart- ment of the Interior, Fish and Wildlife Service, vol. 54, Fishery Bull. 75, pp. 1-34. Hodgson, William C. 1924. Investigations into the age, length and matur- ity of the herring of the southern North Sea. Part I. Some observations on the scales and growth of the English herring. Ministry of Agriculture and Fisheries, Fishery Investigations, ser. II, vol. 7, No. 8, 36 pp. London. 1929. Investigations into the age, length and matur- ity of the herring of the southern North Sea. Part III. The composition of the catches from 1923 to 1928. Ministry of Agriculture and Fisheries, Fishery Investigations, ser. II, vol. 11, No. 7, 75 pp. London. Hi'BFs, Carl L., and Karl F. Lagler. 1949. Fishes of the Great Lakes region. Cranbrook Institute of Science, Bull. No. 26. xi+186 pp. Cranbrook Press, Bloomfield Hills, Mich. Jarvi, T. H. 1920. Die kleine Marane Coregonus albula L. im Keitelesee. Annales Academiae Scientiarum Fennicae, ser. A, vol. 14, No. 1, 302 pp. [Suomalai- sen Tiedeakatemian Kustantama.] Helsinki. Jobes, Frank W. 1943. The age, growth, and bathymetric distribution of Reighard's chub, Leucichthys reighardi Koelz, in Lake Michigan. Trans. American Fisheries Soc, vol. 72 (1942), pp. 108-135. Jobes, Frank W. — Continued 1949a. The age, growth, and distribution of the long- jaw Cisco, Leucichthys alpenae Koelz, in I.,ake Michigan. Trans. American Fisheries Soc, vol. 76 (1946), pp. 215-247. 1949b. The age, growth, and bathymetric distribu- tion of the bloater, Leucichthys hoyi (Gill), in Lake Michigan. Papers Michigan Acad. Sciences, Arts, and Letters, vol. 33 (1947), Part 2: Zoology, pp. 135-172. Jordan, David S., and Barton W. Evermann. 1902. American food and game fishes. 572 pp. Doubleday, Page, and Co., New York. Koelz, Walter. 1929. Coregonid fishes of the Great Lakes. Bull. United States Bureau of Fisheries, vol. 43 (1927), Part 2, Document No. 1048, pp. 297-643. 1931. The coregonid fishes of northeastern .America. Papers Michigan Acad. Sciences, Arts, and I^etters, vol. 13 (1930), pp. 303-432. Lee, Rosa M. 1920. A review of the methods of age and growth determination in fishes by means of scales. Ministry of Agriculture and Fisheries, Fishery Investigations, ser. II, Sea Fisheries, vol. 4, No. 2, 32 pp. London. McHuGH, J. L. 1941. Growth of the Rocky Mountain whitefish. Jour. Fisheries Research Board of Canada, vol. 5, No. 4, pp. 337-343. Moffett, James W. 1952. The study and interpretation of fish scales. The Science Counselor, vol. 15, No. 2, pp. 40-42. Monti, Rina. 1929. Biologia dei coregoni nei laghi italiani. Nota prima: La natura dei laghi nostri e I'introduzione dei coregoni. Istituto di Anatomia Comparata della R. Universita di Milano, Rivista di biologia, vol. 11, fasc 3-4 e 5-6, pp. 1-19. Istituto Editoriale Scientifico, Milano. Nelson, Merlin N., and Arthur D. Hasler. 1942. The growth, food, distribution and relative abundance of the fishes of Lake Geneva, Wiscon- sin, in 1941. Trans. Wisconsin Acad. Sciences, Arts, and Letters, vol. 34, pp. 137-148. Pearse, a. S. 1921. The distribution and food of the fishes of three Wisconsin lakes in summer. University of Wisconsin Studies in Science, No. 3, 61 pp. Madison. Pritchard, Andrew L. 1930. Spawning habits and fry of the cisco (Leucich- thys artedi) in Lake Ontario. Biological Board of Canada, Contributions Canadian Biology and Fisheries, N. S., vol. 6 (1930), No. 9, pp. 227-240. Toronto. 1931. Taxonomic and life history studies of the ciscoes of Lake Ontario. University of Toronto Studies, Publications Ontario Fisheries Re- search Lab., No. 41, 78 pp. 138 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Rawson, D. S. 1930. The bottom fauna of Lake Simcoe and its role in the ecology of the lake. University of Toronto Studies, Publications Ontario Fisheries Research Lab., No. 40, 183 pp. Reighard, Jacob. 1915. An ecological reconnaissance of the fishes of Douglas Lake, Cheboygan County, Michigan, in midsummer. Bull. United States Bureau of Fisheries, vol. 33, Document No. 814, pp. 215- 249. Scott, W. B. 1951. Fluctuations in abundance of the Lake Erie Cisco {Leucichlhys artedi) population. Contri- butions Royal Ontario Museum of Zoology, No. 32, 41 pp. Scott, Will. 1931. The lakes of northeastern Indiana. Investi- gations of Indiana Lakes. III. Indiana De- partment of Conservation, Division of Fish and Game, Publ. No. 107, pp. 57-145. Smith, Oliver B., and John Van Oostbn. 1940. Tagging experiments with lake trout, white- fish, and other species of fish from Lake Michigan. Trans. American Fisheries Soc, vol. 69 (1939), pp. 63-84. Smith, Stanford H. 1954. Method of producing plastic impressions of fish scales without using heat. United States Department of the Interior, Fish and Wildlife Service, Progressive Fish-Culturist, vol. 16, No. 2, pp. 75-78. April. Stone, Udell B. 1938. Growth, habits, and fecundity of the ciscoes of Irondequoit Bay, New York. Trans. Ameri- can Fisheries Soc, vol. 67 (1937), pp. 234-245. SVARDSON, GUNNAR. 1951. The coregonid problem. III. Whitefish from the Baltic, successfully introduced into fresh waters in the north of Sweden. Fishery Board of Sweden, Institute of Freshwater Research, Rept. No. 32, Ann. Rept. 1950, pp. 79-125. Drottningholm. Thompson, D'Arcy W. 1942. On growth and form. 1,116 pp. The Mac- millan Co., New York. Van Oosten, John. 1923. The whitefishes {Coregonus clupeaformis) . A study of the scales of whitefishes of known ages. Zoologica, Scientific Contributions of the New York Zoological Soc, vol. 2, No. 17, pp. 380-412. 1929. Life history of the lake herring {Leucichlhys artedi LeSueur) of Lake Huron as revealed by its scales, with a critique of the scale method. Bull. U. S. Bureau of Fisheries, vol. 44 (1928), pp. 265-428. 1930. The disappearance of the Lake Erie cisco — a preliminary report. Trans. American Fisheries Soc, vol. 60, pp. 204-214. 1937. The age, growth, and sex ratio of the Lake Superior longjaw, Leucichlhys zenilhicus (Jordan and Evermann). Papers Michigan Acad. Sci- ences, Arts, and Letters, vol. 22 (1936), Zoology, pp. 691-711. 1939. The age, growth, sexual maturity, and sex ratio of the common whitefish, Coregonus clupea- formis (Mitchill), of Lake Huron. Papers Michigan Acad. Sciences, Arts, and Letters, vol. 24 (1938), Part II, pp. 195-221. Wagner, George. 1911. The Cisco of Green Lake, Wisconsin. Bull. Wisconsin Natural History Soc, vol. 9, Nos. 1-2, pp. 73-77. Milwaukee, Wis. Ward, Henry B. 1896. A biological examination of Lake Michigan in the Traverse Bay region. Bull. Michigan Fish Coram., No. 6, pp. 1-71. Lansing. WOHLSCHLAG, DoNALD E. 1953. Some characteristics of the fish populations in an Arctic Alaskan lake. In Current Biological Research in the Alaskan Arctic. Stanford LTni- versity Publications, University series. Biological Sciences, vol. 11, No. 1, pp. 19-29. U S. GOVERNMENT PRINTING OFFICE 1957 O — 388748 OBSERVATIONS ON THE DEVELOPMENT OF THE ATLANTIC SAILFISH ISTIOPHORUS AMERIC AN US (CITVIER), WITH NOTES ON AN UNIDEN- TIFIED SPECIES OF ISTIOPHORID Bv Jack W. Gehringer, Fishery Research Biologist The South Atlantic Fishery Investigations, con- ducted by the U. S. Fish and Wildhfe Service in cooperation with the U. S. Navy Hydrographic Office, the Office of Naval Research, the Georgia State Game and Fish Commission, and the Florida State Board of Conservation (through the Marine Laboratory, University of Miami), has engaged since July 1952 in a biological, chemical, and phy- sical oceanographic study of the waters between Cape Hatteras and the Florida Straits from the coast to considerably bej'ond the axis of the Gulf Stream. Field operations are conducted with the research vessel Theodore A\ Gill. Biological specimens are collected with standard half-meter silk nets, high- speed metal nets (Arnold and Gehrhiger, 1952), a continuous plankton sampler, 18-inch diameter dip nets equipped with 10-foot bamboo handles and lined with )^-inch nylon mesh, and trolling and hand lines. During hydrographic observations, at which time the vessel is drifting, dip-net operations are carried out, aided at night by flood and spotlights. Dip-netting sometimes produces relatively rare fish larvae and juveniles. Such was the case on July 29, 1953, between 1700 and 1900 hours dur- ing the occupation of regular station 30 (approxi- mately 90 miles east of Brunswick, Ga.) on Theodore N. Gill cruise 3, when several small istiophorids were captured. Dip-netting and sur- face tows on that station and on subsequent sta- tions produced a total of 26 specimens ranging in standard length from 3.4 to 38.8 mm. Since little has been pubUshed on the early life history of the sailfish and other istiophorids, in- formation that could be obtained from the speci- mens is of considerable value. There was a dearth of material in the 3.8-9.4 mm. range in our collections, however. The United States Na- tional Museum kindly loaned their small istio- phorid specimens, most of which were in the 3.8- 9.4 mm. range, including some from the Gulf of Mexico. I decided to include in the study all available material both from the waters off the South Atlantic Coast of the United States and from the Gulf of Mexico. Additional specimens were loaned by the Gulf Fishery Investigations (Arnold 1955) and various other organizations. Subsequent Theodore N. Gill cruises produced sev- eral more specimens; one was removed from the stomach of a small swordfish, Xiphias gladius (Arata 1954), and one was taken from the stomach of another small istiophorid. Three mounted specimens of Atlantic sailfish, Istiophorus ameri- canus (Cuvier), 374 to 625 mm. in standard length, were photographed and measured. In total, 168 specimens were examined. During my examination of the material I found two groups of fishes to be involved. Those below approximately 10 mm. in standard length did not separate into two groups on any character or group of characters examined, or location or time of cap- ture. The specimens exceeding approximately 10 mm. in standard length separated on some mor- phometric measurements into two distinct groups which converged at approximately 10 mm. The converging of the two groups at 10 mm. precludes the positive identification, by species, of speci- mens below 10 mm., so far as my observations are concerned. Beyond 10 mm. one group traces through development to the Atlantic sailfish, Istiophorus americanus (Cuvier). The other group, with a maximum size of 45.0 mm. stand- ard length, has not been identified. Lack of specimens exceeding 45 mm. makes positive iden- tification impossible. For these specimens I pre- sent selected measurements, counts, and figures, and discuss them with reference to Atlantic sail- fish specimens of similar sizes. The unidentified species is represented by 15 specimens exceeding 10 mm. Several specimens below 10 mm., which were taken at the same time as these, possibly belong to the same group. Though the taxonomy of the istiophorids is in 139 140 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE question, it is generally accepted that there is but one species of Atlantic sailfish: Istiophorus ameri- canus (Cuvier). In accordance with Bailey's re- view (1951) of the authorship of Cuvier and Valenciennes (1831), the single authority is used here as opposed to the common use of both names. Other members of the staff and members of the crew of the Theodore N. Gill assisted in the collec- tion of specimens and various other aspects of the study. Special thanks are extended to Isaac Ginsburg for loan of specimens and for critical reading of the manuscript; to Leonard P. Schultz of the U. S. National Museum for information and loan of specimens; to Royal D. Suttkus of Tulane University, Loren Woods, of the Chicago Natural History Museum, Giles W. Mead, Robert M. Yount of Myrtle Beach, S. C, and Tony Seaman of Morehead City, N. C, for loan of specimens; to Stewart Springer for data ; to Albert W. Collier, Jr., and Edgar L. Arnold, Jr., for loan of specimens and critical reading of the manuscript; and to Frank T. Knapp of the Georgia Game and Fish Commission and George F. Arata, Jr., of the Florida Board of Conservation, for critical reading of the manuscript. METHODS AND DATA METHODS OF MEASUREMENT Measurements from the specimens were made with a binocular, stereoscopic microscope and a micrometer eyepiece, except for the three mounted specimens, whose measurements were made with vernier calipers. Measurements of specimens less Figure 1. — Areas of capture of specimens (excluding Beebe's and 3 mounted specimens) indicated by circled dots, the 100-fathom curve by dotted lines, and the approximate axis of the Gulf Stream by arrows. ATLANTIC SAILFISH 141 than 100 mm. in standard length were recorded to the nearest 0.1 mm., and tliose of specimens more than 100 mm. in standard length to the nearest millimeter. Measurements of Beebe's (1941), Voss's (1953), and Baughman's (1941a) specimens were taken from their papers. DEFINITIONS OF TERMS Standard length. — Tip of snout to tip of urost.vle or most posterior e.xtension of hypural segment. Total length. — Tip of snout to tip of caudal fin, or finfold. Head length. — Tip of snout to posterior extension of fleshy margin of opercle. Width of head. — Measurement of widest portion of brain case, at point where dorso-lateral keel of pterotic spine joins orbital crest. Depth of head. — Vertical measurement of head at posterior angle of jaw. Snout length. — Tip of snout to anterior margin of eye. Lower jaw length. — Tip of mandible to posterior angle of the jaw. Snout extension. — Tip of snout to mandible tip fwith mouth clo.sed). Eye diameter. — Horizontal measurement of eye. Pterotic spine length. — Tip of spine to point of attachment of dorso-medial keel of spine to head. Main preopercular spine length. — Tip of main preoper- cular spine to vertical at posterior edge of preopercle. Pectoral fin length. — Tip of pectoral fin to insertion. Pelvic fin length. — Tip of pelvic fin to insertion. Dorsal fin-ray lengths. — Tips to insertions. Teeth. — Number of teeth on one side each of the upper and lower jaws, Pterotic spine serrations. — Numbers of serrations on keels of left and right pterotic spines. Ratio of secondary preopercular spines. — Number of spines, in ratio, on left side of head. Number of dorsal, anal, and pectoral fin rays. — Counts are total numbers with no distinction made between spines and soft rays. MEASUREMENTS AND MERISTIC COUNTS Table 1 gives selected measurements and meristic counts from the 19 specimens described or figured; table 2 gives selected measurements from 13 Florida specimens described by Voss (1953); table 3 gives selected measurements from two specimens described by Beebe (1941), and 3 mounted specimens from the Gulf of Mexico, described by Baughman (1941a). Voss's, Beebe's, and Baughman's data are plotted with mine on the respective graphs. Table 1. — Selected measurements and meristic counts from 19 specimens described and figured Measurements and counts for specimen N 3. — 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Measurements in mm: Standard length 3.6 3.8 2.0 1.1 1.4 0.5 1.0 0.0 0.7 0.4 1.2 4.7 4.9 2.2 1.3 1.2 0.7 1.2 0.1 0.7 0.7 1.3 0.8 0.1 none none none none none none 10 5.6 6.0 2.7 1.4 1.5 1.0 1.7 0.1 0.8 1.0 1.8 0.8 0.1 none none none none none none 13 6.4 7.1 3.5 1.7 1.8 1.3 2.4 0.1 1.1 0.8 2.1 1.2 0.2 8.1 9.6 3.9 2.0 2.3 1.2 2.3 0.2 1.3 0.6 2.4 1.3 0.6 9.5 11.0 4.7 2.2 2.5 1.5 2.7 0.2 1.5 0.6 1.5 1.5 1.4 1.1 11.3 12.5 5.6 2.3 2.5 2.4 3.3 0.4 1.3 1.0 2.0 1.2 1.2 0.9 1.3 1.6 14.6 16.7 7.0 2.6 3.1 3.2 4.7 0.7 1.7 1.0 1.8 1.8 2.4 1.9 2.4 2.3 2.1 18.2 20.6 9.5 2.7 3.4 5.0 5.3 2.1 1.8 0.8 2.3 2.0 3.6 3.2 3.6 3.9 3.7 20.9 23.8 10.7 2.9 3.8 5.7 5.9 2.0 2.1 1.1 3.0 2.2 4.5 2.9 4.5 4.8 4.7 4.5 3.6 51 43 45-44 18-19 18-17 2:2 49 23 17 27.4 30.7 12.9 3.1 3.7 7.4 6.9 3.3 2.3 0.9 2.4 2.7 5.8 5.4 7.1 7.2 6.3 42 52 48-43 15-14 17-16 " 'si 25 18 38 8 42.9 19.9 3.4 4.5 12.9 8.4 7.6 2.7 1.0 2.7 3.8 8.7 5.9 10.6 11.5 11.1 6.6 9.3 55 SO 22-27 14-15 12-17 1:0 51 24 20 56.2 61.0 29.3 3.6 5.1 21.6 10.8 14.6 3.1 0.6 2.7 5.4 13.5 8.4 15.6 18.2 17.4 16.8 15.6 86 60 64.1 68.9 32.6 3.9 5.7 23.9 11.1 16.6 3.0 1.0 2.1 5.1 14.5 7.7 17.3 18.6 18.6 18.0 16.2 60 85 .. 43 101 54' 8 16 31 4 1 ...... 119 374 419 168 "'"35" 127 55 79 8 none none 21 '11.3 13.1 5.2 2.3 2.6 1.7 3.1 0.1 1.6 1.0 2.3 1.5 1.8 1.2 ■21.0 24.6 9.0 3.4 4.2 3.0 5.1 0.3 2.6 0.9 1.6 3.0 6.1 6.3 6.3 6.2 6.0 5.7 4.8 36 43 33-10 11-16 12-15 '45.0 Total length 50.5 15.2 Head width Head depth 4.8 7.2 5.2 8.5 0.8 Eye diameter Pterutlc spine length Main preopercular spine length 3.6 0.6 1.8 6.0 Pelvic fln length Dorsal fln-ray lengths: 5th none none none none none none none 13.5 12.0 15.6 26 16.4 15th 1.2 16.2 16.1 15.0 Counts: Number of teeth one side of jaw: 14 18 28 -- 10 -- 26 25 27 26 28... 43 48 37-36 16-21 1&-16 1:1 SO 23 18 45 46 37-38 16-15 18-18 2:2 52 22 16 34 34 34-34 16-18 15-18 2:1 J 42 16 14 102 SO Pterotic spine serrations: Dorso-lateral keel (left & right) 22-28 17-14 30-34 ... 18 .. 30 -. 12 none none none none 35-42 Dorso-medial keel (left & right) 14 10 Ventral keel (left & 14... 2:1 50 22 18 -. 11 12-13 Ratio of secondary pre- opercular spines (upper: 1:0 fold fold fold 1:0 fold fold fold 1:1 fold fold fold 2:1 fold' 16 2:1 42 10 15 2:1 "40 22 16 Number of dorsal fln rays. - Number of anal fin rays . . . Number of pectoral fln rays 53 24 18 65 24 20 51 »20 18 44 23 20 49 24 18 > Indicates unidentified species. ' Indicates questionable values. 142 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 2. — Selected measurements from 13 Florida specimens described by Voss (1963) (In millimeters] Specimen No. — Standard length Total length Head length Snout length Eye diameter Pterotic spine length Preoper- cular spine length Lower jaw length Pelvic fin length 1 3.9 4.8 5.5 6.3 8.0 15.0 19.5 20.0 20.5 29 5 37.5 70.0 208.0 4.2 5.4 6.5 7.3 ft 6 16.7 22.0 21.5 22.0 32.5 41.0 76.0 234.0 1.5 2.2 2.7 3.8 4.2 7.5 ft 7 9.8 10.9 15.9 18.6 39 0 94.0 0.5 0. B 0.8 1.2 1.3 3.8 5.0 5.0 5.7 10.0 12.4 28.7 76.0 0.5 0.7 0.9 1.1 1.3 1.6 2.0 0.4 0.8 0.9 0.8 0.8 0.8 0.7 0.8 1.0 1.0 1.1 0.4 0.7 0.3 1.4 1.9 1.1 2.6 2.5 2.3 2.3 2.6 2.8 2.2 0.7 1.2 1.8 1.9 2.3 5.1 5.6 6.2 5.8 7.5 8.9 12.5 37.0 3 0.2 4 5 0.8 6 1.6 7 4.0 g 2.2 9 2.0 2.4 2.8 3.7 3.5 10 --- 6.4 11 12 -- 15.0 13 46.0 Table 3.- — Selected measurements from 2 specimens described by Beebe (1941) and 3 mounted specimens described by Baughman {1941a) [In millimeters] Item Standard length. - Head length Snout length. Eye diameter Pterotic spine length Main preopercular spine length. Lower jaw length Snout extension Pelvic fin length Dorsal fln height Number of dorsal rays Number of pectoral rays Measurements from Beebe No. 1 No. 2 437 169 140 12 130 Measurements from Baughman No. 1 No. 2 No. 544 233 175 16 849.5 314 228.5 22 116 '264" 851 324 232 23 1 Measurements are conversions of Beebe's percent-of-standard-length values. DESCRIPTION OF SPECIMENS The importance of line drawings, in a develop- mental series of a fish, to portray metamorphic changes is well understood ; therefore I have chosen for illustration (to scale) only those sizes at which important changes are apparent. A complete description is given of the smallest specimen, and for other specimens a brief summary of important changes which have occurred from the preceding size is presented. A series of specimens ranging in standard length from 3.6 to 374 mm. is figured and discussed. Those exceeding 10 mm. in standard length separated into two readily distinguishable groups, one tracing through development to the Atlantic sailfish, Istiophorus americanus (Cuvier), and the other remaining unidentified. Those below 10 mm. did not separate into two distinct groups; hence they are treated as one. Drawings of three specimens of the unidentified species are pre- sented (figs. 23, 24, and 25), with discussion lim- ited to variations from sailfish. Various authors, Gunther (1873-74), Lutken (1880), Goode (1883), Beebe (1941), LaMonte and Marcy (1941), and Voss (1953), have pub- lished figures and descriptions of small sailfish singly or in series. Those described by Voss from waters off southern Florida constitute the most complete series. I am unable to explain the differences between my findings and those of the authors cited. For comparison with the series of illustrations of sailfish material studied, I include as figure 21 a photograph of a 437-mm. sailfish, from off Texas, described by Beebe (1941), and as figure 22 a line drawing of an adult sailfish to portray general outlines. LARVA, 3.6 MILLIMETERS (Fig. 2) Although the specimen is damaged, it is the smallest complete larva in the material studied, and is smaller than any specimen previously described. Head. — The jaws are equal, and the snout is not produced. The orbital crest originates anterior to the nostril, curves over the eye (with a large spine over the eye), and continues posteriorly as a serrated ridge continuous with the dorsolateral keel of the pterotic spine. The pterotic spine has 3 serrated keels (dorsolateral, dorsomedial, and ventral in position), is directed posteriorly and parallel to sagittal plane of the body, and prom- inent serrations on the dorsolateral keel diminish to minute notches on the ridge connecting with the orbital crest. The main preopercular spine has 3 serrated keels (dorsomedial, lateral, and ventromedial in position), arises from the posterior ventral edge of preopercle, is directed posteriorly at approximately a 45-degree angle to sagittal ATLANTIC SAILFISH 143 <^'''^ *l ^Umnfc Figure 2. — Larva, 3.6 millimeters long (U. S. Nat. Mus. No. 111814). plane of the body (when opercle is closed), and the ventromedial keel terminates anteriorly at the posterior angle of the lower jaw. The upper secondary preopercular spine is serrated and arises from posterior edge of the preopercle on a serrated keel continuous with the dorsomedial keel of the main preopercular spine. (See figure 3 for arrange- ment of head spines in the dorsal view of a 3.8- mm. specimen.) The keel on the face of the oper- cle is serrated and possesses a dominant, acute, and medially situated protuberance. Keels on lower jaw are serrated, and two in number; one arising from the lateral surface of the posterior portion of the jaw and possessing a dominant protuberance, and the other comprising the lower edge of the lower jaw. Teeth are few in number, large, with the anterior ones tusklike. The eye is large, with a diameter one-third the length of the head. The nostril is a single opening in the swollen anterior-basal portion of the orbital crest. Body. — The body is short and deep, the visceral sac distended with food, and mj'omeres and the rodlike urostyle are prominent. Fins. — The dorsal, caudal, and anal are a con- tinuous fold with indistinct supporting structures developing in the caudal portion. The pectoral is round, with indistinct supporting structures. No pelvic fins are present. Pigmentation. — Pigment has faded from long preservation, and that remaining is limited to a Figure 3. — Larva, 3.8 mm. {Theodore N. Gill collections), dorsal view of head showing spination and pigmentation. PS, pterotic spine; USPS, upper secondary preopercular spine; PFS, spine on face of preopercle; MPS, main preopercular spine (directed at 90° from sagittal plane of body as opercles are open, angle is 45° when opercles ure closed); OC, orbital crest (showing heavy spine over eye). 389133 O— 57- 144 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE few melanophores (faded to brown color) on the dorsal surface of the brain case. The eyes are black with white pupils (preservation caused a reversal of color from that of the live condition of white eyes and black pupils). LARVA, 4.7 MILLIMETERS (Fig. 4) The visceral sac was nearly empty and not so distended as in other specimens of similar size. The snout is slightly produced and extends beyond the tip of the mandible. The nostril opening has elongated. Pelvic fins are now present, as buds, and indistinct supporting structures have appeared in the dorsal and anal portions of the finfold (which is now notched anterior to the caudal portion), and some rays are discernible in the ventral por- tion of the caudal. LARVA, 5.6 MILLIMETERS (Fig. 5) The snout and mandible have elongated slightly. Fanglike teeth are present on the tip of the snout. The nostril is partially divided by developing flaps. A lower secondary preopercular spine h?.s appeared on the ventromedial keel of the main preopercular spine. Separation of the finfold into dorsal, caudal, and anal portions is distinct but not complete. Additional caudal rays have appeared, the urostyle has turned upward, and supporting structures in dorsal, anal, and pectoral fins are further developed, but not yet discernible as rays. LARVA, 5.4 MILLIMETERS (Fig. 6) The snout is more elongated and extends farther beyond the tip of the mandible. Several palatine teeth have appeared, and although snout fangs are missing (appear broken off) they are present on other specimens of similar size. Each nostril is now divided (by a flap of skin) into two openings. Two upper secondary preopercular spines are now present. (At this size the height of spinous con- dition of the head is reached, although the pterotic spines are blunt and appear deformed on this specimen.) Rays are discernible in dorsal and anal fins, the pectoral fin has 16 rays, additional rays are present in the caudal fin, separation of dorsal, anal, and caudal fins is complete, and the pelvic fins have elongated. The pattern of pig- mentation is more distinct (probably owing to better color preservation). Figure 4. — Larva, 4.7 millimeters long (USNM 163332). Figure 5.— Larva, 5.6 millimeters long (USNM 163333). ATLANTIC SAILFISH 145 FioiRE 6.— Larva, 6.4 millinipters long (USX.M 111814). LARVA, 8.1 MILLIMETERS (Fig. 7) The snout extends farther beyond the tip of the mandible. Many teeth of varied sizes are present in both jaws, and fanglike teeth occur on the tip of the mandible. There are 42 rays in the dorsal fin, 10 in tiie anal, 15 in the pectoral, and 2 in the pelvic (posterior rays in dorsal fin are less clearly defined than anterior ones), and the caudal fin is notched. The pattern of pigmentation extends over a greater area. LARVA, 9.5 MILLIMETERS (Fig. 8) The anterior opening of each nostril has a collar (formed by a flap of skm). A second lower secondary preopercular spine is present on the left main preopercular spine, but a similar spine was not found on the right main spine, or on either spine of other specimens of similar size. The dorsal fin has increased in height, and when depressed fits into a groove formed from a raised dermal flap on each side of fin (groove ends at FiGiRE 7.— Larva, 8.1 millimeters long (USNM 111814). Figure 8.— Larva, 9.5 millimeters long (USXM 1 11814). 146 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE approximately the 25th ray) ; the anal fin has 22 rays; the pectoral fin is quite angular; and the pelvic fins have elongated. Pigment is present on lower middle portion of the dorsal fin. SAILFISH LARVA, 11.3 MILLIMETERS (Fig. 9) The snout has elongated, but there has been a corresponding elongation of the mandible, and the snout extension remains unchanged. The diameter of the eye is approximately one-quarter the head length (one-third on the previous speci- mens). The keel on the face of the preopercle is damaged so that no serrations are evident. The dorsal fin has 50 rays, the anal fin 22 rays, and the pectoral fin 18 rays (the full complement of rays has been reached in these 3 fins). The pelvic fins are slightly shorter on this specimen. A groove for the pelvic fins has appeared as an indentation of the belly. SAILFISH LARVA, 14.6 MILLIMETERS (Fig. 10) There is marked elongation of the snout and extension beyond tip of mandible. The teeth are more numerous, with those on the lower jaw more closely set than those on the upper. There is a marked increase in height of dorsal fin, and the the anal fin has 23 rays. A 3-toothed (or 3- pronged) scale, arising from the pectoral girdle, is situated on each side of the body just below the tip of the pterotic spine. There is a general increase in density of pigment. SAILFISH LARVA, 18.2 MILLIMETERS (Fig. 11) The snout is markedly produced (approximately one-half the head length) and extends two-fifths its length beyond the tip of the mandible. Teeth are more numerous. The eye diameter is approxi- mately one-fifth the head length. There is a marked heightening in anterior portion of the dorsal fin and a change in shape, and pelvic fins have lengthened and a third ray has appeared (constituting the full complement). A second 3-toothed scale is found on each side of the body at the pectoral girdle. The caudal peduncle is proportionately more slender. Pigmentation ex- tends over the entire anterior portion of the dorsal fin. Figure 9.— Sailflsh larva, 11.3 millimeters long (USNM 163333). ^^Jr^S^ Figure 10.— Sailfish larva, 14.6 millimeters long. From Theodore N. Gill collections. ATLANTIC SAILFISH 147 SAILFISH LARVA. 20.9 MILLIMETERS (Fig. 12) All rays in tlie dorsal and anal fins extend (for the first time) to fin margins, and the dorsal fin groove extends the entire length of the dorsal fin (previously approximately two-thirds tlie length of the fin). SAILFISH LARVA, 27.4 MILLIMETERS (Fig. 13) The snout is three-fifths tlie head length. Fangs liave disappeared from the snout tip. The eye diameter is approximately one-sixth the head length (previously one-fifth). A small pore is discernible on each side of the snout just anterior to the nostril. The dorsal fin has increased in height, and the anal fin is indented in central portion. SAILFISH LARVA, 38.8 MILLIMETERS (Fig. 14) The snout extends for one-half its length beyond the tip of the mandible. Teetii are fewer in number in the part of the upper jaw extending beyond tip of mandible. Two pores are present on each side of the snout, one anterior to and one below each nostril. The dorsal fin is higher and more uneven in outline, and an anal fin groove has appeared (formed by a dermal flap on each side of fin). The lateral line is discernible for the first time. The pigmentation has developed into distinct patterns on the body and the dorsal fin. SAILFISH LARVA, 56.2 MILLIMETERS (Fig. 15) The snout is markedly produced (three-quarters the head length) and extends for approximately Figure 11. — Sailfish larva, 18.2 millimeters long. From Theodore N. Gill collections. Figure 12. — Sailfish larva, 20.9 millimeters long. From Theodore N. Gill collections. 148 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Figure 13. — Sailfish larva, 27.4 millimeters long. From Theodore N. Gill collections. Figure 14. — Sailfish larva, 38.8 millimeters long. From Theodore N. Oill collections. Figure 15.— Sailfish larva, 56.2 millimeters long (USNM 163413). ATLANTIC SAILFISH 149 t\vo-thii-(ls of its length boyoiul the tip of tlic maiulible. Tlie margin of the preopercle is serrated, but no secondary preopercular spines are discernible. The eye diameter is one-tenth the head length. The body is slimmer. The dorsal fin is higher, the pectoral fin longer and more angular, the aiu^l fin has a pronounced notch in its middle portion, and the pelvic fin has increased in length. Dermal spines are present over opercle, preopercle, and body except for the area covered by the pectoral fin (when depressed), but only the tips of the spines protrude through the skin (spines are discernible on a 43-mm. speci- men). Dermal spines are fully described in discussion of 64.1-mm. specimen. Pattern of pigmentation on the body is more pronounced. SAILFISH LARVA, 64.1 MILLIMETERS (Figs. 16 and 17) Teeth on the snout beyond mandible tip are weak and few in number, and palatine teeth are FiGURK 16. — Sailfish larva, 64.1 millimeters long (Alaska collections); view of head and teeth, orbital crest, small pores on snout, and serrations on lower jaw. present in two patches on each side on upper jaw (one below tlie nostril and one near the mandible tip). The arrangement of teeth in the lower jaw and posterior portion of upper jaw is portrayed in figure 16. Several pores are present on the snout near the nostrils (fig. 16). The minute dermal spines present on the opercle, preopercle, and uniformly over the body arise from ill-defined plates. The spines are narrow-based, acutely tipped cones which protrude through the skin (fig. 17). The interspinous distance varies from one to two times the spine height. Figure 17. — Sailfish larva, 64.1 millimeters long (^/asfca collections) ; oblique view of dermal spines, with tips of spines protruding through the sliin. SAILFISH LARVA, 101 MILLIMETERS (Figs. 18 and 19) Although the specimen is in poor condition and incomplete, many important characters remain. The snout is four-fifths the head length and extends for three-quarters its length beyond the tip of the mandible. Teeth are few in number in the portion of the upper jaw extending beyond the mandible tip. The serrated keels have dis- appeared from the lower jaw. (Although not shown in figure 18. serrations on the orbital crest Figure 18.— Sailfish larva, 101 millimeters long (USNM 107200). 150 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE and pterotic and main preopercular spines remain distinct.) The dermal spines are long, slender, with a gradual taper, and arise from the centers of irregularly shaped (though generally rounded) plates which appear slightly superior to connective tissues. The plates (or bases) have 3 or 4 con- centric raised ridges connected at random by numerous raised radials. The concentric ridges make a continuous spiral on some plates, but on others are entirely separate ridges (fig. 19). FiCiURE 19. — Sailfish larva, 101 millimeters long (USNM 107200): Above, dorsal view of dermal scale; below, oblique view. Diameter of scale approximately 0.6 mil- limeter. SAILFISH JUVENILE, 374 MILLIMETERS (Fig. 20) Observations are from a mounted specimen, and only those that appear accurate are presented (the first dorsal fin and the pelvic fins are artificial). There is no evidence of spines or serrated keels on the head. The only evidence of teeth are minute spines present mainly on the ventral surface of the snout in the portion extending beyond the tip of the mandible. Tlie caudal fin lobes are long and falcate, and the anal fin is separated. GROWTH AND DEVELOPMENT CHANGES IN RATES OF GROWTH OF VARIOUS BODY PARTS Several of the numerous measurements and meristic counts taken from the 168 specimens examined were selected to portray changes in rates of growth of various body parts. Original meas- urements were used in establishing the curves appearing in figures 26 to 35. As the graphs are largely self-explanatory, only the salient points are summarized. The curves are drawn to include all specimens less than 10 mm. in standard length, but only the Atlantic sailfish beyond 10 mm. It was considered that insufficient data were available to fit curves for the unidentified species. The fol- lowing comments apply only to the Atlantic sailfish. The rate of increase in head length approximates that of the standard length, with indications of a slightly higher rate in specimens smaller than 10 mm. (fig. 26). The rate of increase in head width approximates that of the standard length in specimens smaller than 10 mm. and falls well below it in specimens between 10 and 100 mm. (The high value for the 101-mm. specimen has been disregarded in drawing the curve, fig. 27.) The rate of increase in head depth approximates that of the standard length in specimens smaller than 10 mm., decreases and falls below it in specimens between 10 and 40 mm., and increases to approximate it in specimens exceeding 40 mm. (fig. 28). The rate of increase in snout length exceeds, considerably, that of the standard length in specimens smaller than 10 mm., decreases some- what (but still exceeds the standard length rate) in specimens between 10 and 100 mm., and falls slightly below it in specimens exceeding 100 mm. (fig. 29). The rate of increase in lower jaw length exceeds that of the standard length in specimens smaller than 10 mm., decreases and falls below it in specimens between 10 and 100 mm., and increases to approximate it in specimens exceeding 100 mm. (fig. 30). ATLANTIC SAILFISH 151 FiGiRE 20. — Young sailfish, 374 millimeters long. (Mounted specimen.) Captured in surf at Myrtle Beaoli. S. C. by hand, Aug. 3, 1952, by Robert M. and John Yount. The rate of increase in snout extension is con- stant and much higher than that of the standard length in specimens smaller than 50 mm., decreases gradually in specimens larger than 50 mm., so that it falls below that of the standard length in speci- mens exceeding 100 mm. (fig. 31). The rate of increase in eye diameter exceeds that of the standard length Ln specimens smaller than 10 mm., decreases to considerably less than it in specimens between 10 and 100 mm., and increases (but remains slightly below it) in specimens exceeding 100 mm. (fig. 32). The rate of increase in pterotic spine length is initially much higher than that of the standard length, but decreases sharply, and spine growth has ceased in specimens approximately 7 mm long (fig. 33). The rate of increase in the main preopercular spine length is initially much higher than that of the standard length, but decreases sharply, and spine growth has ceased in specimens approxi- mately 10 mm. long (fig. 33). The rate of increase in pelvic fin length exceeds, considerably, that of the standard length in specimens smaller than 10 mm., but decreased gradually in specimens between 10 and 20 mm., after which it approximates the standard length rate (fig. 34). The rate of increase in length of the longest dorsal ray (13th or 15th) exceeds that of the .■isni 3S <)— 57- 152 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE standard length in specimens between 10 and 20 mm., but decreases in specimens between 20 and 40 mm . , after which it approximates the standard length rate (fig. 35). DIVERGENCE IN MEASUREMENTS OF BODY PARTS FOR UNIDENTIFIED SPECIES The length of the snout shows the greatest divergence, which is evident at a smaller size than are other divergent characters. Since the snout length is reflected in the standard length, several graphs were constructed using "standard length minus snout length" as a base for comparison. The "weight" of snout length was thus removed from the base, and comparison of fish with similar body lengths was possible. The divergence in snout length (fig. 36), length of lower jaw (fig. 37), and snout extension (fig. 38), is more clearly defined than in plots against standard length (figs. 29, 30, and 31). When the eye diameter is plotted against snout length, the result is sub- stantially the same: divergence first evident at a standard length of approximately 10 mm. (fig. 39). Measurements of head depth, head width, aiul eye diameter, when plotted against "standard length minus snout length," revealed no marked divergence, at least below 20 mm. Values for the unidentified species plot either just above or among the higher values for the sailfish (also noticed in the plots of these values against stand- ard length, figs. 28, 27, and 32). Fk:t"Re21. — Sailfish, juvenile 437 iiiilliineters long, "20 inches total length, 437 mm. standard length; taken 3 miles off Aransas Pass, Texas, Aug. 31, 1941, by Aubrey Xelson " (Photograph courtesy of William Beebe.) J<^ FiQiRE 22. — General shape and proportions of adult sailfish. ATLANTIC SAILFISH 153 KiGiRE 23.— I'liiticiilifit'd spi'cics of larva, 1 1 .3 millimeters long. Theodore N. Gill t-oUeei ioi. FiniRE 24. — Unidentified .species of larva, 21.0 millimeters long, Theodore N. Gill collections FlGlTRE 25. — Unidentified species of larva, 45.0 millimeters long, Theodore X. Gill collect ions. 154 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 1 1 1 1 1 1 1 1 1 1 1 1 1 M 1 1 1 1 1 1 I I 1 1 _ I 600 - - 400 - g 200 - / y. 100 - y ^ -_ 60 - / - 40 - J/ / /■ / - 20 ■■■e 10 -_ /■• • - 6 - . yi ^•^ - 4 - - 2 1 o 1 1 1 1 i t 1 1 1 1 1 J 1 1 1 1 1 1 1 1 Mil 4 6 10 20 40 60 100 200 400 600 STANDARD LENGTH IN MM. Figure 26. — Relation of head length to standard length. Dots represent study specimens; x's, Beebe's specimens; circles, Voss's specimens; triangles, Baughman's specimens; and large black dots, the unindentified species. ~1 I I I I II I I "1 1 1 I I I M I "T 1 1 I I I I I I S a < e I I I I I I I I I J 1 I J I I I 4 6 10 20 40 60 100 200 400 600 STANDARD LENGTH IN MM. FiGUHE 27. — Relation of head width to standard length. Large black dots represent the unidentified species. ATLANTIC SAILFISH 155 60 40 20 n 1 — I — I I M 1 1 1 1 1 — I Mill -| 1 1 I I II I :H'- J I I I I I I I [ I I I r I r I I I I 1 I 10 20 40 60 100 STANDARD LENGTH IN MM. Figure 28. — Relation of head depth to standard length. Large black dots represent the unidentified species. DEVELOPMENT OF FIN RAYS Tables 4, 5, and 6 show the numbers of rays in the dorsal, anal, and pectoral fins of specimens of different sizes. Counts listed include both the Atlantic sailfish and the unidentified species. The following discussion of the numbers of rays in the fins applies only to the Atlantic sailfish for speci- mens exceeding 10 mm. Specimens below 10 mm. include both the Atlantic sailfish and the un- identified species. The number of rays in the dorsal fin of speci- mens which exceeded 10 mm. in length ranged from 47 to 57 (table 4). In those exceeding 26 mm. in length tiie full complement of rays is present, and the number ranged from 49 to 57, witii 75 percent having 49 to 53. The smallest specimen with a complement of 49 rays or more is 1 1 ..3 mm. in length, and tiie largest with fewer than this number is 16.2 mm. Table 4. —Dulrib ttion of specimens by length anc by number of rays in he dorsal Jin Length Number of specimens with ray count of— 42 43 44 45 46 47 48 49 50 SI 52 53 54 55 56 57 10.0-13.9 mm ... 14.0-17.9 mm.... 1 1 0 0 0 0 13 0 0 u 0 12 0 1 1 0 0 12 0 0 0 0 1 1 0 0 0 0 1 0 1 0 0 0 0 0 0 0 1 1 0 1(12) 3 1 0 0 3 0 1 0 0 5 0 2 1 0 ■I 0 0 0 0 3 0 0 0 0 1 0 0 0 0 1 0 0 0 0 «1 0 0 18.0-21.9 mm 0 22.0-25.9 mm.... 26.0-101 mm 0 1 Total ' 1 '3 ■3 12 1 1 3 0 3(1 2) 7 6 3(11) 3 1 1 1 1 1 Unidentified specimens. ' Specimen from Beebe (1941). 156 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE s g 10 _ 1 1 1 1 M 1 1 1 1 1 1 1 1 M 1 1 1 1 1 1 M 1 ; FIN RAY _ - y - A-'' y^ - / / 0 / : - / - - / 3 - / / 0 • • • : - - ■■>■ _ - 1 1 r 1 1 J 1 1 1 1 II 1 1 1 t 1 1 1 1.. 10 20 40 60 100 STANDARD LENGTH IN MM. FinuRE 29. — Relation of snout length to standard length. Dots represent study specimens; I's, Beebe's specimen circles, Voss's specimens; triangles, Baughman's specimens; and large black dots, the unidentified species. ■ The number of rays in the anal fin of specimens between 6.0 and 9.9 mm. ranged from 20 to 22 (table 5). In sailfish specimens between 10 and 25.9 mm. the number ranged from 20 to 23. In specimens between 26 and 101 mm. the number ranged from 24 to 28 (excluding the 101-nim. specimen with a damaged anal fin whicli has 20 recognizable rays), with 92 percent having 24 to 26. I consider that the break at the 26-mm. size results from an inadequate sample. The smallest specimen with a complement of 22 or more rays is 9.5 mm., and the largest with fewer than this number (excluding the 101-mm. specimen men- tioned above) is 23.3 mm. 60 40 10 s Z 6 I •- 4 < 2 o ATLANTIC SAILFISH T 1 1 1 I I I I I 1 1 1 1 I MM 157 1 1 — I — I I I I J I 1 I 1 1 1 J I I I 1 1 1 I I I I 1 1 STANDARD LENGTH IN MM. Fi(!URE 30. — Relation of lower jaw length to standard length. Dut.s represent study specimens; x, Beebe's specimen; circles, Voss's specimens; triangles, Baughman's specimens; and large black dots, the unidentified species. Table 5. — Dislribution of specimens by length and by number of rays in the anal Jin Length Number of specimens with ray count of— 16 17 18 19 20 21 22 23 24 26 26 27 28 6.0-9.9 mm 0 '4 0 0 0 0 0 0 0 0 0 0 0 1 2 0 0 0 0 0 ' 1 0 0 0 0 2 1('2) 0 1 0 1 0 2 1 0 1 0 1 1 0 1 1 0 U'l) 3 1 1 0 0 0 0 0 0 0 90 1) 0 0 0 0 0 20 1) 0 0 0 0 0 1 0 0 0 0 0 0 0 10.0-13.9 mm 0 14.0-17.9 mm 0 18.0-21.9 mm.-- - 22.0-25.9 nmi 0 0 26.0-101 mm.- 1 Total- '4 0 '2 1 1 5('2) 4 401) 40 2) 901) 20 1) 1 0 1 ' Unidentified specimens. Tlu' mimlier of rays in the pectoral fin of speei- nieiis i)et\veen 6.0 and 9.9 mm. ranged from l(j to 18 (table 6). Sailfisli exceetling 10 mm. in length had complements ranging from 16 rays to 20. In specimens exceeding 26 mm., the number ranged from 18 to 20, with 82 percent having 18 rays. The smallest specimen with a complement of 18 lays or more is 7.9 mm. in length, and the largest with fewer than this number is 23.3 mm. 158 600 400 200 - 60 40 Z o OT z Ui I- X bJ 3 O .2 - FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 1 1 1 — I Mill T 1 1 I M II I 1 I I I MM I .!. _l L ...jl^jM.!, I. L !. I lI J I II 1 I I J I I I I I I 10 20 40 60 100 STANDARD LENGTH IN MM. Figure 31. Table 6. -Relation of snout extension to standard length. Dots represent .study specimens; x's, Beebe's specimens; and large black dots, the unidentified species. -Distribution of specimens by length and by num- ber of rays in the pectoral Jin Length Number of specimens with ray count of— 14 15 16 17 18 19 20 21 22 6.0-9.9 mm 10.0-13.9 mm 14.0-17.9 mm 0 '1 0 0 0 0 '1 0 0 0 0 0 0 8 30 2) 2 0 1 0 0 1 1 1 1 0 0 2 20 4) 2 1 0 1401) 0 '1 1 0 0 0 101) 0 ' 1 0 20 1) 0 0 0 0 0 0 0 0 n 18.0-21.9 mm. 0 22.0-25.9 mm 26.O-101 mm 0 1 1 Total 0 140 2) 20 1) 210 S) 201) 30 3) 0 11 ■ Unidentified specimens. 'Specimen from Beebe (1941). SECONDARY PREOPERCULAR SPINES The numbers of upper and lower secondary preopercular spines, in ratios (for specimens with recognizable spines) are shown in table 7. A pat- tern (herein called ratio, upper: lower) prevails in the number of upper and lower spines for size groups, although considerable overlap of ratios is found. Of primary interest is the size at which the 2:1 ratio (apparently the iieiglit of normal spine development) occurs. Although several specimens smaller than 5.5 mm. have a 2:1 ratio, this ratio does not become dominant until a speci- ATLANTIC SAILFISH 159 men size of iippi'oximately 6.5 mm. Ri)ecimeMs exceeding 10 mm. have considerable variation, and on specimens exceeding 40 mm. the spines are difficult to discern. Until such time as identilication of specimens less than 10 mm. to species is accomplished, the pattern must be presented as representing one group. Table 7. — Distribution of specimens by length and by ratios of secondary preopercular spines standard length Number of specimens with spine ratio (upper:lower) of— 1:0 1:1 2:1 other ' 1 3 3.8 mm 2 1 3 2 1 5 1 1 4.0 mm .- 4.1 mm 1 4.3 mm ,. 2 1 4.5 mm - - . . - . . 4.6 mm . 3 2 2 3 1 .. 1 4.9 mm . . 3 5.1 mm - _ . . . 1 (2:0) 1 1 2 1 2 2 1 2 4 2 2 1 1 2 2 1 2 3 1 1 5.6 mm 1 5.7 mm. 1 2 .. 2 5.9 mm 1 (2:0) 1 (2:0) 1 6.1 mm 6.2 mm . 6.4 mm 2 6.5 mm 6.7 mm. . . 1 1 (2:2) 6.8mm. 6.9 mm 7.0 mm 3 1 2 1 1 3 7.3 mm 1 (2:0) 1 8.0 mm 8.1 mm... 1 (2:2) 8.6 mm . 1 1 1 1 »2 22 1 1 (3:1) 9.2 mm 9.4 mm 1 (2:3) 9.5 mm.. 10.3 mm. 10.6mm 10.8 mm 2 1 (2:2) 1 (2:2) 22 '1 1 11.4mm.. 12.3mm.. 2 1 (2:3) 2 1 (2:2) 12.8 mm.. 1 >2 13.0 mm 13.9 mm « 1 (3:2) 14.6mm-. 1 16.2 mm 1 1 1 16.9 mm 1 (2:2) 17.8 mm 18.2 mm 1 (2:2) 1 (2:2) 1 (3:1) 1 (3:3) 20.9mm 23.3 mm.. 29.6 mm 32.4 mm. 1 1 43.7 mm 1 (2:2) 1 Ratio shown In parentheses. 'Unlden tlfled sp€ clmens. PTEROTIC SPINE SERRATIONS The number of serrations on the dorsomedial and ventral keels of the pterotic spine ranges from 10 to 20, with but few exceptions, throughout the size range of specimens examined. The number of serrations on the dorsolateral keel in- creases from a range of 18-25 on 4-mm. specimens, to 24-42 on 10-mm. specimens, to 28-44 on 25-mm, specimens, and holds relatively stable between 44 and 49 on larger specimens.' We should expect less variation in number of serrations on keels of specimens exceeding 10 mm. in length, since growth of pterotic spines ceases when specimens are approximately 7 mm. in length (fig. 33). TEETH The number of teeth in relation to specimen size is shown in figure 40. For the sizes shown, there is an increase in number of teeth with an increase in specimen size; and the range is narrower in the sm.aller specimens than in the larger ones (many teeth of all sizes in larger specimens, but only a few well-developed ones in the smaller ones). Teeth were present in all specimens examined, but the 101-mm. specimen was badly damaged, and an accurate count could not be made. The remaining teeth of this specimen were not so large, relatively, as those of smaller specimens. Counts of 102 and 110 for one-half the upper jaw of the 45.0-mm. specimens of the unidentified species exceed those for Atlantic sailfish specimens to 64 mm . DEVELOPMENT OF PIGMENTATION Observations on the development of pigment include all specimens below 10 mm. in length, and only the Atlantic sailfish specimens above 10 mm. Notes on variations from this pattern of development in the luiidentified species follow the observations on the Atlantic sailfish. There are a few large melanophores on the dorsal surface of the brain case of the 3. 4-mm. specimen. There is a gradual increase in the pigmentation (consisting of small chromatophores) extending to the dorsal surfaces of the snout and the body on specimens approximately 4 mm. long, down the sides of the head and body posteriorly to the anus at approximately 5 mm., and to the ' Specimens of unidcntiru'd specie.'' 45.0 mm. long had counts of 37 and 50. otherwise, all counts for the species fell within those for Atlantic sailHsh. 160 FISHERY BULLETIN OF THE FISH AND "WILDLIFE SERVICE 60 40 1 6 (E 4 -| 1 1 1 I I I I I "T — I I I I I n I I I I 1 1 I I I I r I I I I I I I I J L_l_ _U 10 20 40 60 STANDARD LENGTH IN MM. Figure 32. — Relation of eye diameter to standard length. Dots represent study specimens; x's, Beebe's specimens; circles, Voss's specimens; triangles, Baughman's specimens; and large black dots, the unidentified species. caudal fin, witli increasing density, at appro.xi- mately 10 mm. The preopercle and opprcle are less densely pigmented than the dorsal part of the snout or body at any particular size. At the 6-mm. size, scattered melanophores appear in the pattern of chromatophores over the dorsal surface of the body. In specimens smaller than 10 mm., the dorsal surface of the brain case is pigmented by scattered melanophores. In specimens exceeding 10 mm., the pigmenta- tion changes little except for a gradual increase in density. Generally it is as follows: Upper jaw and sides of head, blue-black; mandible, non- pigmented; eye, silver with black pupil; upper body, dark blue to black; and lower sides of body anterior to the anus and caudal, blue. The belly is a silvery white, fins are usually ti-anslucent (except for the dorsal), and spines are nonpig- mented. Pigmentation on the dorsal fin develops from a scattering of chromatophores on the lower central portion at approximately 10 mm. to generally dense areas (with scattered less dense areas) at approximately 20 mm. Tips of dorsal fin rays are nonpigmented, and pigment on the dorsal fin extends posteriorly to approximately the 35th ray. Bars (or blotches) of chromato- phores appear on the body at approximately 35 mm. and persist through the size range of speci- mens examined. Color notes on fresh specimens (15-20 mm. in standard length) are as follows: Dorsal surface of head and body, steel-blue; sides of head and upper opercles, blue-black; eye, silver-wliite with a blue tinge and black pupil; ventral sides of body from anus posterior, blue; anal and pectoral fins, hyaline; caudal fin, translucent white; dorsal fin anterior portion generally blue-black with yellow and white streaks on rays, and posterior portion, hyaline; and ventral fins are tinged with yellow. The principal variations from this pattern of development in the unidentified species are as follows: (1) In general, the pigment on the dorsal ATLANTIC SAILFISH 161 o 111 .6 - < 2 S X 1- o z liJ < ID z O I IE T 1 1 — I Mill 1 1 1 — I I M I I Mill .-..-i*: /:• H — I I I I MM H — I I I I III ^... •• o I I I I I I I J I I I II II J I I I I 11 10 20 40 60 100 STANDARD LENGTH IN MM. FicURE 33. — Relation of head-spine lengths to standard length. Dots represent study specimens; x, Beebe's specimen; circles, Voss's specimens; and large black dots, the unidentified species. fin is present farther posteriorly than on tlie sail- fisli, and (2) there are no dark bars or blotches on the sides of the body, as are present on sailfish exceeding approximately 35 mm. in standard length. Color notes on live 35-45 mm. specimens of the unidentified species do not vary noticeably from those given above for the Atlantic sailfish. SUMMARY OF GENERAL GROWTH AND DEVELOP- MENT OF THE ATLANTIC SAILFISH = 1. From 3.4 to 7 mm., head spination develops • Points 1 and 2 are suniniarifS of development of all specimens below 11) mm., since no separation to species was made below this size. rapidly, and the pterotic spine growth ceases at approximately 7 mm. 2. From 7 to 10 mm., head spination reaches its maximum and no further development occurs, the snout begins to elongate, and rays begin to appear in the fins. 3. From 10 to 20 mm., the snout elongates farther and the snout extension increases, and the dorsal, anal, pelvic, and pectoral fins develop their full complement of rays. 4. From 20 to 50 mm., the snout elongates and snout extension increases, the dorsal, anal, pec- 162 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 60 40 o z lU > _l bJ -l 1 — I — I I I I I I 1 1 — I — I I I I 1 1 1 1 — I — I I I I I / ••/ I I 1 1 J I I I I I I 10 20 40 60 100 STANDARD LENGTH IN MM. 200 Figure 34. — Relation of pelvic fin length to standard length. Dots represent study specimens; x, Beebe's specimen: circles, Voss's specimens; triangles, Baughman's .specimens; and large black dots, the unidentified species. toral, and pelvic fins develop in size and shape, and dermal spines appear. 5. From 50 to 100 mm., the snout elongates farther, but the snout extension stabilizes, the dorsal and anal fins further develop in size, shape, and progress toward their eventual division, and dermal spines develop further. On the basis of the foregoing observations on growth and development, I have divided the specimens less than 100 mm. in length into three categories. The size range below 7 mm. lias been desig- nated "early larval," that period during which the head spines are developing (by 7 mm. all except the pterotic spine have ceased growing), and finfolds have little differentiation of rays. The size range from 7 to 20 mm. has been des- ignated "midlarval," that period during which all spine development ceases (at approximately 10 mm.), fins receive their full complement of rays and undergo changes in size and shape, and the snout begins to elongate. 600 - 400 200 20 >- < tn UJ O 2 o -l — I — I I I I 1 1 ATLANTIC SAILFISH 1 1 — I I I I I I 163 "T "T 1 — TTT / / I :a / J 1 1 1 1 1 1 1 1 1 1 10 20 40 60 100 STANDARD LENGTH IN MM. Figure 35. — Relation of length of longest dorsal fin-ray to standard length. Dots represent study specimens; x's, Beebe's specimens; and large black dots, the unidentified species. Tlie size range between 20 and 100 mm. has been designated "late larval," that period din-ing which head spines begin to disappear (although pterotic and main preopercular spines persist at 101 mm.), fins undergo other changes in size and shape (toward eventual division in dorsal and anal fins), dermal spines develop, and jaw teeth begin to disappear. While juvenile characters are developing within this range, it is my opinion that the important larval characters which per- sist preclude the use of the term "juvenile" for specimens below 100 mm. SUMMARY COMPARISON OF ATLANTIC SAILFISH WITH UNIDENTIFIED SPECIES A summary of the ])rincipal differences between specimens of the Atlantic sailfish and the imiden- tified species at comparalile sizes is outlined in table 8. 164 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 8. — Comparison of certain characters from selected specimens of the Atlantic sailfish and the unidentified species Character Size Snout length Number of fin rays: Dorsal Anal Pectoral Size Snout length Snout extension Mandibular keels Number of fin rays: Dorsal Anal. -- Pectoral Longest dorsal fin ray . Dorsal fin, anterior lobe. Dorsal fin, pigment. ., Pelvic fin rays. Size Snout length Snout extension . . . Mandibular keels. Atlantic sailfish 11.3 mm ;s length of head . 20.9 mm-.-- -. ii length of head . . . ii length of snout . . Noticeably serrate . Number of fin rays: Dorsal- Anal Pectoral Longest dorsal fin ray. Dorsal fin, anterior lobe. Dorsal fin, pigment. -. Pelvic fin rays.. Body pigment-. Dermal spines-- 49 23 17 Number 13 to 15 Terminates at about the 25th ray. Extends posteriorly to 25th ray. Third is twice length of first. 38.8-56.2 mm ?5 to ?i length of head.. ii to ?4 length of snout. Present (minutely ser- rate) . Unidentified species 51-53---- ---- - 24 18-20 Number 13 to 15 Terminates at about the 40th ray. Extends posteriorly to 33-37th ray. Third is twice length of first. Distinct barred or blotched pattern. Discernible at 43 mm. as spines which pro- trude slightly through skin, uniformly dis- tributed over body. 11.3 mm. \i length of head. 42. 16. 14. 21.0 mm. H length of head. M 0 length of snout. Minutely serrate. 44. 23. 20. Number 5. Terminates at about the 37th ray. Extends posteriorly to 37th ray. First and third equal in length. 45.0 mm. W length of head. M length of snout. Absent. 49. 24. 18. Number 13. Terminates at about the 40th ray. Extends posteriorly to the 4Qth ray. Third and first equal in length. No bars or blotches. Distinct spines arising from individual base plates, uniformly dis- tributed except for ir- regularly sized patches on body above lateral line (skin has worm track appearance). (Dermal spines resem- ble those in Beebe, 1941, text figure 2 for Istiophonts greyi Jordan & Hill— 84 mm.) FOOD HABITS Prior to Beebe's report (1941) that copepods are the primary food of small sailfish, and Voss's ob- servations (1953) on the food of postlarval and juvenile sailfish, little is found on the subject in the literature. The stomachs of 32 istiophorid specimens from the Theodore N. Gill collections were examined, and stomach contents are listed in table 9. With reference to this table it should be noted that copepods constituted the food of speci- mens less than 6 mm. long. At this size fish larvae also were eaten, and no specimen exceeding 13 mm. had copepods in its stomach. Voss (1953) also found evidence of change in the diet of young sailfish from copepods to fish larvae at a size of approximately 6 mm. Of particular interest are the small istiophorids removed from the stomachs of three of the speci- mens (one 10.2 mm. long from a 21.0-mm. speci- men, one 6.6 mm. long from a 13.0-mm. specimen, and part of one with a head 2.4 mm. long from a 16.2-mm. specimen). One specimen 6.0 mm. long removed from the stomach of a 21.9-mm. sword- fish {Xiphias gladius), Arata fl954), had copepods in its stomach. An unidentified species of fish oc- curred frequently in the stomachs of several speci- mens taken 90 miles east of Brunswick, Ga., July 29, 1953, and flying fish predominated in the stom- achs of others taken 150 miles east of Charleston, S. C, August 10, 1953. During the collection of the latter, small flying fish were also dipnetted. Small istiophorids were "relatively abundant" in the water when the above collections were made. Some fish larvae were larger than half the length of the fish that had eaten them. These data add support to the theory advanced by previous 1 — T 1 1 "^ m 1 60 40 - 20 - <»' 10 - *o ' 6 - 0- •CO 4 - 0-* • 2 - =«■■ • » • ••* • * • • _1_ I -J I I I I I I I J u I I I 2 4 6 10 20 40 60 STANDARD LENGTH MINUS SNOUT LENGTH, IN MM. FiGCRE 36. — Relation of snout length to standard length minus snout length. Large black dots represent the unidentified species. ATLANTIC SAILFISH 165 20 - I 1 I 1 1 1 1 I 1 ■ 1 1 1 1 1 I I 1 s z 10 - - 7 _ o • - I y- ~ O - • o — 7 UJ e _ _l i* * O , -35°N., 144°-173°E. in Nov. 1936- Feb. 1937; operational data, catch rates, water temperatures to 200 m; 1936 tana, landing statistics for the prefecture. 1941a. Albacore fishery. Chiba-ken suisan shiken- jo Katsuura bimj6 jigyo hokoku (1938) :1-21. [J.P] Results of participation in cooperative summer albacore longlining explorations, three cruises to 28°-43°N., 175°E.-175'' W., May- Aug. 1938; description of gear, operational data, finances, catch and prices, with water temperatures and spe- ific gravities to 200 m. 1941b. The skipjack fishery. Chiba-ken suisan shi- kenjo Katsuura bunjo jigyo hokoku (1938): 22-25. [J, P] Japan: albacore and skipjack fishing con- ditions correlated with water temperature. 1941c. Tuna fishery. Chiba-ken suisan shikenj6 Katsuura bimjo jigyo hokoku (1938) : 29-37. [J.P] Results of three longlining cruises to 29°- 37°N., 155°E.-178°W. in Nov. 1938-Feb. 1939; construction of gear, operational data, water temperatures to 200 m; catch and catch rates, principally albacore and bigeye. 1941d. Albacore fishery. Chiba-ken suisan shiken- jo Katsuura bunjo jigyo hokoku (1939) :1-13. [J.P] Results of participation in cooperative summer albacore longlining explorations, three cruises to 32°-45°N., 176°E.-176°W. in May-Oct. 1939; construction of gear, catch and prices; fishing logs with opera- tional data, water temperatures and speci- fic gravities to 200 m; albacore catch rates. 1941e. Skipjack fishery. Chiba-ken suisan shiken- jo Katsuura bunjo jigyo hokoku (1939) :14-17. [J.P] Landings of pole-and-line skipjack and al- bacore in the prefecture by months for 1938 and 1939; discussion of fishing con- ditions in each month relative to water temperatures. BIBLIOGRAPHY ON THE TUNAS 185 Chiba Prefectural fisheries experimental Sta- tion, KATSUURA Branch. — Continued 194 If. Tuna fishery. Chiba-ken suisan shikenjo Katsuura bunjo jigyo hokoku (1939) : 21-29. [J. P] Results of three longlining cruises, one for albacore to 30°-36° N., ITl'-lSO" E. in Nov.-Dec. and two for yellowfin, to 2°-6'' N., 130°-134° E. in April-May and to 5°-10° N., 131°-134° E. in 10° N., 131°-134° E. in Mar.-Apr. 1940; opera- tional data, water temperatures to 200 m., catch and catch rates for albacore, yel- lowfin, and bigeye. Chilton, Cyrus h. 1949. "Little tuna" of the Atlantic and Gulf coasts. Fish. Leafl., Wash. 353:1-5. [P] Euthynnus alletteratus : description. Chu, Yuanting T. 1931. Index piscium sinensium. Biol. Bull., Shang- hai 1:107-108. Auocis rochei, Neothunnus macropterus : synonymy, distribution. Cleaver, Fred c, and bell m. shimada. 1950. Japanese skipjack {Katsuwomis pelamis) fishing methods. Comm. Fish. Rev. 12(11) :1- 27. [P] History, biology and ecology, fishery for bait, fishing gear, fishing techniques, han- dling of catch, fishing grounds and sea- sons. Clemens, W. a., and G. V. Wilby. 1946. Fishes of the Pacific Coast of Canada. Bull. Fish. Res. Bd. Can. 48:164-167. Katsuwonus pelamis, Thunnus alalunga: description, distribution, food, records of capture in Canadian Pacific waters, fig- ured. CONNER, G. 1930. The five tunas and Mexico. Fish. Bull., Sac- ramento 20:75-89. Statistical records concerning the tuna catch and industry. CONRAD, M. G. 1937. The brain of the swordfish {Xiphias gtodius). Amer. Mus. Novit. 900:1-4. Comparison between brains of Xiphias, Scomber, Thunnus, and Euthynnus. COnseil International Pour L' exploration de la MER AND commission INTERNATIONALE POUR L' exploration SCIENTIFIQUE DE LA MER MfiDITERRANfiE. 1933. Conference d'experts pour I'examen des m6thodes scientifiques et techniques k ap- pliquer k I'^tude des poissons de la famille des Thonid^s. Rapp. Cons. Explor. Mer 84:92-103. [P] Brief reports of discussions of standardl- CONSEiL International. — Continued zation of techniques for tuna biology; T. thynnus, G. alalunga: nomenclature, sys- tematics, tagging, morphometric measure- ments, meristic counts, maturity, age and growth, statistics. COPLEY, H. 1947. Fish records and observations. Nat. E. Afr., Nairobi 2:9-10. Neothunnus macropterus: East African coast. CORWiN, Genevieve A. 1930. A bibliography of the tunas. Fish Bull., Sac- ramento 22:103 p. tPl Arranged alphabetically by author with brief annotations; covers the literature on tuna through 1929. Has a subject index, and a Ust of abbreviations used for period- icals cited. COWAN, Ian M. 1938. Some fish records from the coast of British Columbia. Copeia 1938(2) :97. Germo alalunga: recorded. CRANE, JOCELYN. 1936. Notes on the biology and ecology of giant tima, Thunnus thynnus Linnaeus, observed at Portland, Maine. Zoologica, N. Y., 21:207-212. Description, food, parasites, sex. De Beaufort, L. F., and W. M. Chapman. 1951. The fishes of the Indo- Australian tirchipelago. Volume 9:215-227. Leiden, E. J. Brill. [P] Euthynnus pelamis, Euthynnus allettera- tus affinis, Thunnus sibi, Thunnus macro- pterus, Thunnus tonggol, Auxis thazard. Figured: Euthynnus alletteratus, Thunnus macropterus. De Buen, Fernando. 1931. El supuesto paso por el Estrecho de Gibral- tar del atun en su emigraciOn gen^tica. Rapp. Comm. int. Mer M6dit. 6:405-409. Tuna (T. thynnus) : reproduction, migra- tion; Mediterranean Sea, Atlantic Ocean. 1930. Ictlologia espafiola: Scombriformes y Thun- niformes. Bol. Oceanogr. Pesc, Madr. 15(2): 33-53. [P] Key, distribution, Spanish and Basque common names: Germ,o alalunga, Thun- nus thynnus, Neothunnus albacora, Para- thunnus obesus, Auxis thazard, Euthyn- nus alletteratus, Katsuwonus pelamis. Sys- tematics, figures. 1932. Formas ontogtoicas de peces (Nota pri- mera). Notas Inst. esp. Oceanogr. Ser. II, no. 57:38 p. [P] Morphometries and descriptions: Auxis thazard, Sarda sarda, Thunnus thynnus. 1936. Fauna ictioldgica. Cat&logo de los peces ib6r- icos: de la planicie continental, aguas dulces. 186 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE De Buen, Fernando. — Continued peldgicos y de los abismos pr6ximos. Segunda parte. Notas Inst. esp. Oceanogr., Ser. II (89) ; : 91-149. [P] Synonymy: Germo alalunga, Thunnus thynnus, Neothunnus albacora, Auxis tha- zard, Eiithynnus alletteratus, Katsuwomis pelamis. 1937. Aires de ponte du thon (Thunnus thynnus L.). Int. Congr. Zool. 12:2123-2136. [P] Thunnus thynnus: spawning areas. De Buen, Feknando, and F. Frade. 1932. Clef dichotomique pour une classification ra- pide des poissons scombriformes. Rapp. Comm. int. Mer M6dit. 7(N.S.) :71-74. Classification, keys. De Jong, J. K. 1940. A preliminary investigation of the spawning habits of some fishes of the Java sea. Treubia 17(4):325-326. Euthynnus alUtteratus : frequencies of egg diameter measurements; resorption of eggs noted. De La tourrasse, Guy. 1951. La p§che aux thons sur la Cote Basque Fran- caise et son Evolution r^cente. Rev. Trav. Off. Peches marit. 17 (66fasc. 1) :42 p. [P] Thunnus thynnus and Oermo alalunga: distribution; fishing methods and boat design, trolling and livebait fishing. Delsman, H. C. 1931. Fish eggs and larvae from the Java Sea. Treubia 13(3/4) :407-409. Eggs and larvae believed to be those of Scomber (Delsman, Treubia 8(3/4:395-399), reidentified as Thyn- nus thunnina. 1933. Tunny in the North Sea. Nature 132(3338) : 640-641. Short note on the occurrence of Thun- nus thynnus in the North Sea since 1911. Delsman, H. C, and J. G. F. Hardenburg. 1934. De Indische zeevischen en zeevisscherij. Bibl. Ned. ind. nat. Ver. 6:330-343. Euthynnus alletteratus, E. pelamys, Neo- thunnus macropterus, N. rcwus: descrip- tion, distribution, key, Malayan names; spawning of E. alletteratus and descrip- tion of eggs and larvae; spawning of N. rarus and description of eggs; food of E. pelamys; E. alletteratus, and N. macrop- terus figured. DIEUZEIDE, R. 1930. Sur quelques scombriniens des cotes Alg6ri- ennes. Bull. Sta. Aquic. Peche Castiglione, 1929 (2e fasc.) :i33, 150-151, 159. Thunnus thynnus : classification. DIEUZEIDE, R. — Continued 1931. La p§che du thon a la ligne dans la bale de Castiglione. Bull. Sta. Aquic. PSche Castigli- one, 1930 (2e fasc.) : 107-127. Fishing gear and methods; Thunmis thyn- nus: Mediterranean. Domantay, Jose S. 1940a. The catching of live bait for tuna fishing in Mindanao. Philipp. J. Sci. 73(3) :337-342. [P] Description and figures of gear used; men- tions 16 species, mostly sardines, ancho- vies, and small scombroid and carangoid fishes, used as bait. 1940b. Tuna fishing in southern Mindanao. Philipp. J. Sci. 73(4) : 423-435. [P] Auxis thazard, Euthynnus yaito, Katsu- wonus pelamis, Neothunnus itosihi, N. ma- cropterus, Parathunntts sibi: distribution, figured; livebait fishing methods, gear, and boats. Dontcheff, Y., and R. Legendre. 1948. Thon blanc ou germon. Composition chimi- que et valeur alimentaire du germon. Rev. Trav. Off. Pgch. marit 11(44 fasc.4) : 447-462. Thunnus germo: chemical analysis. Dung, Dorothy I. Y., and William F. Rovce. 1953. Morphometric measurements of Pacific scom- brids. Spec. sci. Rept: Fish. U. S. Fish Wildl. : 95. [P] Morphometric data on: Neothunnus ma- cropterus, Parathunnus sibi, Germo ala- lunga, Katsuivonus pelamis, Thunnus thyn- nus, Thunnus orientalis, Thunnus maccoyii, Kishinoella tonggol, Euthynnus affinis, Gymnosarda nuda. ECKLES, Howard H. 1949a. Fishery exploration in the Hawaiian Islands (August to October 1948, by the vessel Oregon of the Pacific Exploration Company). Comm. Fish. Rev. 11(6) :l-9. [P] Euthynnus yaito, Katsuwonus pelamis, Neothunnus macropterus:. recorded; K. pelamis and N. macropterus: figured. 1949b. Observations on juvenile oceanic skipjack (Katsuwonus pelamis) from Hawaiian waters and sierra mackerel from the eastern Pacific. Fish. Bull., U. S. 51(48) : 245-250. [P] Katsuwonus pelamis: anatomy, descrip- tions, figfures and records of capture of juveniles; spawning; juveniles noted in stomachs of adults. Ego, Kenji, and TAMIO Otsu. 1952. Japanese tuna-mothership expeditions in the western equatorial Pacific Ocean, June 1950 to June 1951. Comm. Fish. Rev. 14(6):1-19. [P] ■ i-'"' Catch statistics, prices. ■ BIBLIOGRAPHY ON THE TUNAS 187 Ehrenbaum, E. 1934. Thunfische in den nordeuropaischen Gewas- sern. Fischmarkt 2(5) :116-119. Cuxhaven. Thunntis thynnus: migration, catch sta- tistics. ESPENSHADE, ADA V. 1948. Japanese fisheries production, 1908-1946. Fish. Leafl., Wash. 279:40 p. [P] Production of important species of fish In coastal waters (metric tons) : bonito (katsuo), tuna {maguro). Also chart showing pre-war areas fished for tima; table showing production of important species from offshore fisheries: bonito, tuna. Farina, Luigi. 1931a. L'attuale crisi dell' industria delle tonnare, cause e remidi. Boll. Pesca Piscic Idrobiol. 7(5):752-9. Mediterranean; statistics on trap catches. 1931b. Remarques sur les madragues des c6tes frangaises de I'Afrique du Nord. Bull. Soc. oc^anogr. Fr. 11(62) : 1115-6. Fishing gear and methods: traps; Medi- terranean; Atlantic Ocean. Federation of Japan Tuna and Bonito Fisheries Cooperative associations. 1951a. The present condition of the tuna fisheries. Katsuo to maguro 16:2-10. (In: Spec. sci. Rep: Fish U. S. Fish Wildl. 79) [P] Statistical tables on catch through 1949, size and composition of the fleet, number of fishermen, mothership operations, im- ports and exports. 1951b. The 1950 catch. Katsuo to maguro 19:2-8. [J.P] Statistics on tima catch, size and compo- sition of the livebait and longline fleets, average annual catch per vessel ton; comparisons with earlier years; mother- ship operations. 1952. The present condition of the tuna fisheries. Katsuo to maguro 26:2-12. [J, P] Tables of statistics on tuna catch, fleet, prices, exports, featuring 1947-50 average values. 1953a. The 1952 tuna catch. Katsuo to maguro 37:2-8. [J, P] Tables of statistics on catch of longline and livebait fisheries, catch per vessel ton, mothership operations, exports and imports. 1953b. The present condition of the tuna fisher- ies. Katsuo to maguro 39:2-13. [J, P] Strength and composition of fleet, num- ber of fishermen, catch statistics, catch Federation of Japan tuna. — Continued per vessel ton, price trends, operating costs, mothership operations, imports and exports; statistics through 1952. FERREIRA, ERNESTO. 1932. La pesca dell' albacora nelle Azzore. Note 1st. Biol. mar. Rovigno, No. 1. Thunnus gcrmo: fishing methods and gear, Atlantic Ocean. FICK, H. 1937. Der Fang von Thunf ischen und Heringshaien. Fischmarkt 5(1). Thunnus thynnus: North Sea. Fish, Marie p. 1948. Sonic fishes of the Pacific. Tech. Rep. Woods Hole oceanogr. Inst. 2:87-91. [P] Auxis thazard, Euthynnus, Germo ala- lunga, Katsuwonus pelaniis, Neothunnus macropterus, Thunnus thynnus: distribu- tion, English common names, synonymy of K. pelamis, G. alalunga, T. thynnus; air bladders of G. alalunga, N. macropterus, and T. thynnus described; Japanese com- :-■■ ','■« ■ mon names of Euthynnus and T. thynnus; vertical distribution of Parathunnus me- bachi noted. Fisheries Society of Japan (Dai Nippon Suisankai). 1931. Illustrations of Japanese aquatic plants and animals, v.l. Tokyo. [P] Descriptions, figures of : Katsuwonus va- gans, Auxis thazard, Euthynnus yaito, Thunnus orientalis, Parathunnus sibi, Neo- thunnus macropterus, Germo germo, Sar- da orientalis. Fitch, John E. 1950. Notes on some Pacific fishes. CaUf. Fish Game 36(2) :65. [P] Stomach contents of Neothunnus macropterus. 1953. Extensions to known geographical distribu- tions of some marine fishes on the Pacific coast. Calif. Fish Game 39(4) : 539-52. [P] Euthynnus yaito: first record from the American Pacific coast. Flett, a. 1944. A report on livebait fishing for tuna in Aus- tralia. J. Commonw. sci. industr. Res. Organ. Aust. 17(1): 59-64 Tuna - livebait fishing. FOOD and agriculture organization of the United Nations. 1949a. General aspects of the world's tuna fisher- ies. Fish. Bull. F. A. O. 2(4) : 81-108. [P] World landings, distribution. 1949b. Recommended scientific and common names of important food fishes. A. Scombri- formes. Fish. Div., FAO, UN. 98 p. Auxis thazard, Euthynnus alletteraUta, 188 FISHERY BTJLLETIN OF THE FISH AND WILDLIFE SERVICE Food and Agriculture. — Continued Germo alalunga, Katsuwonus pelamis, Ne- othunnus macropterus, Thunnus thynnus: distribution; synonymy, world-wide com- mon names and recommended nomencla- ture. FORMOSA Government-General Fisheries experiment Station. 1930. Northern oceanographlc conditions and skip- jack fishing. Taiwan sotokufu suisan shikenjo suisan shiken hokoku, kaiyS chosa (1928) :67- 70. [J, P] Fishing conditions in relation to water temperature, specific gfravity, and cur- rents. 1931. Northern oceanographic conditions and skip- jack fishing. Taiwan sotokufu suisan shikenjo suisan shiken hokoku, kalyo chosa (1929) :28- 30. [J, P] Fishing conditions in relation to water temperature, specific gravity, and cur- rents. 1932. Northern oceanographic conditions {ind skip- jack fishing. Taiwan sStokufu suisan shlkenjd suisan shiken hokoku, kaiyo chosa (1930) :10- 11. [J,P] Fishing conditions in relation to water temperature, specific gravity, and cur- rents. 1933a. Oceanographic conditions and skipjack fishing in northern Formosa. Taiwan sotokufu suisan shikenjo jigyo hokoku, kaiyo chosa (1931) : 13-15. [J, P] Fishing conditions in relation to currents, surface water temperature, and specific gravity. 1933b. Experimental fishing and investigation in southern waters by the Shonan Maru. Taiwan suisan shikenjo jigyo hokoku, gyorobu ( 1931 ) : 1-50. [J,P] Yellowfin tuna: Indo-Pacific region; lengfth-weight data, fishing conditions in relation to oceanography and weather, catch per unit of effort, distribution, stomach contents. 1934. Oceanographic conditions and skipjack fish- ing in northern Formosa. Taiwan sotokufu suisan shikenjo jigryo hokoku, kaiyo chosa (1932): 10-12. [J, P] Fishing conditions in relation to currents, surface water temperature, and specific gravity. FortuniC, V. 1930. Crtice o ribarstvu uopde, a nada sve u po- druCju bivse republike dubrovaCke. 82 p. Fishing methods and gear, Adriatic Sea. FOWLER, Henry W. 1931. The fishes of Oceania. Supplement 1. Mem. Bishop Mus. 11(5) :325. [P] Euthynnus alletteratus, E. pelamis, Germo alalunga, G. macropterus, G. sihi, Thun- nus thunnus: listed, synonymy of G. ma- cropterus. 1933. Description of a new long-finned tima (S»ma- thunnus guildi) from Tahiti. Proc. Acad. nat. Sci. Philad. 85:163-164. Descriptions of a new genus Semathunnus and new species, Semathunnus guildi; Sem,athunnus distinguished from Neothun- nus. 1934. The fishes of Oceania. Supplement 2. Mem. Bishop Mus. 11(6) :400. [P] Euthynnus pelamis, Semathunnus guildi, S. itosibi, Thunnus orientalis: listed, syn- onymy. 1936. A synopsis of the fishes of China. VI. The mackerel and related fishes. Hongkong Nat. 7:61-80, 186-202. Thunnus thynnus, Neothunnus macrop- terus, Auxis thazard, keys, description, synonymy. 1938. The fishes of the George Vanderbilt South Pacific Expedition, 1937. Monogr. Acad. nat. Sci. Philad. 2:31-33, 253, 277. Auxis thazard, Euthynnus Kneatus, E. pel- amis: description, synonymy: Auxis tfca- zard, E. alletteratus, E. lineatus, E. pelam- is, Germo germo, Neothunnus macropter- us, Parathunnus sibi, Thunnus thynnus: recorded from Pacific. 1944. Results of the Fifth George Vanderbilt Ex- pedition (1941). Monogr. Acad. nat. Sci. Philad. 6:349, 373-4, 378, 498. Auxis thazard, Euthynnus lineatus, Kat- suwonus pelamis, Thunnus thynnus: re- cords of capture; synonymy. Pacific re- cords of A. thazard, E. alletteratus, E. lineatus, Germ,o alalunga, K. pelamis, Neo- thunnus argentivittatus, Thunnus thyn- nus; description of T. thynnus; figure of E. lineatus. 1949. The fishes of Oceania. Supplement 3. Mem. Bishop Mus. 12(2):73-74. [P] Auxis thazard, Euthynnus wallisi, Katsu- wonus vagans, Neothunnus macropterus, Parathunnus sibi: listed, synonymy. Frade, F. 1930a. Anomalies chez le thon rouge. Bull. Soc. portug. Sci. nat. 11(1) :l-5. [P] Describes abnormal structures or forma- tions in swim-bladder and head of Thun- nus thynnus. BIBLIOGRAPHY ON THE TUNAS 189 Frade, F. — Continued 1930b. L'anomalie faciale du thon rouge et son importance pour I'^tude des migrations. Bull. Soc. portug. Sci. nat. 11(2):7-10. [P] Discusses a certain anomaly, consisting of grooves on the side of the head, among red tunnies on the Portugese coast. 1931a. Donn^es biomdtriques pour I'^tude du thon rouge de I'Algarve. Bull. Soc. portug. Sci. nat. 11(7) :89-130. [P] Comparison of Atlantic and Tunisian T. thynnus. 1931b. Domi^es biom^triques sur trois espfeces de thons de I'Atlantique oriental. Rapp. Cons. Explor. Mer 10:117-126. [P] Thunniis thynnus, Parathunnus ohesus Lowe, Neothunnus albacora Lowe: mor- phometries. 1931c. Neothu7inus albacora (Lowe 1839). Faime ichthyol. Allan. N. 8. Description, synonymy, figure. 1931d. Sur le nombre des rayons des nageoires et de pinnules branchiales chez le thon rouge At- lantique. Bull. Soc. portug. Sci. nat. 11(10): 139-144. [P] Thunnus thynnus: anatomy. 1932. Sur les caractSres ost^ologiques k utiliser pour la determination des Thonid^s de I'Atlan- tique oriental et de la M^diterran^e. Rapp. Comm. int. Mer M6dit. 7:79-90. Osteology and specific identification of Thunniinae. 1935. Recherches biom^triques sur la maturity sexuelle du thon rouge. Trav. Stat. Biol, marit. Lisbonne 41. Thunnus thynnus: sexual maturity. 1937a. Recherches biom^triques sur la maturity sexuelle du thon rouge. Int. Congr. Zool. 12: 2137-2142. [P] Thunnus thynnus: sexual maturity. 1937b. Recherches sur la maturity sexuelle du thon rouge de I'Atlantique et de la M6diterran6e. Bull. Soc. portug. Sci. nat. 12:243-250. [P] Thunntis thynnus: sexual maturity. 1953. Sur r6tat de maturity sexuelle d'un germon pris en Mer Tyrrhenienne. J. Cons. int. Ex- plor. Mer 19(1) :72-76. [P] Thunnus germo: histological study of ma- turity of Mediterranean albacore, compar- ison with Atlantic population. Frade, f., and F. de Buen. 1932. Clef de classification principalement d'aprfes la morphologic interne. Poissona Scombri- Frade, F., and F. DE BUEN. — Continued formes (excepts la famille Scombridae). Rapp. Comm. int. Mer M6dit. 7(N.S.) :69-70. Classification; keys; anatomy, external and internal. Frade, Fernando, and S. Mana^as. 1933. Sur l'6tat de maturity des gonades chez le thon rouge gSn^tique. C. R. Ass. Anat., Apr. 10-12, 1933:1-15. [P] Thunnus thynnus: figures of ovaries and testes. Fraser-Brunner, a. 1949. On the fishes of the genus Euthynnus. Ann. Mag. nat. Hist. 2(20) : 622-628. [P] Euthynnus af finis af finis, E. af finis linea- tus, E. af finis yaito: classification, distri- bution, key, figures, synonymy. 1950. The fishes of the family Scombridae. Ann. Mag. nat. Hist. 3(26) :131-163. [P] Allothutmus fallai, Auxis thazard, Euthyn- nus af finis, E. pelamis, Thunnus alalunga, T. albacora, T. obesus, T. thynnus, T. tonggol, T. zacalles: classification, de- scription, distribution, key, figures, syno- nymy. FUJII, T. 1932. A study of the tunny fishery of Hokkaido. Bull. Sch. Fish. Hokkaido 2(l):32-47 [Je]. Thunnus thynnus: vertical migrations, variation of yield of the fishery in rela- tion to oceanographic changes. FUKUDA, M., and S. Iizuka. 1940a. Experimental tuna fishing. Kumamoto- ken suisan shikenjo jigyo hokoku 1938:15-20. [J.P] Bigeye tuna, black tuna: Ryukyu Islands; catch in relation to water temperature. 1940b. Skipjack tagging experiment. Kumamoto- ken suisan shikenjo jigyo hokoku 1938:21. [J,P] Japan: release records of tagged skipjack. Galtsoff, Paul S. 1952. Staining of growth rings in the vertebrae of tuna (Thunnus thynnus). Copeia 1952(2) : 103-105. [P] Ganssle, David, and H. B. Clemens. 1953. California-tagged albacore recovered off Japan. Calif. Fish Game 39(4) :443. Thunnus germo: tagging, migration; Pacific Ocean — northeast, northwest. Genovese, Sebastiano. 1952. Osservazioni idrologiche eseguite nella ton- nara del "Tono" (Milazzo) durante la cam- pagna di pesca 1952. Boll. Pesca Piscic. Idro- biol. 7 (U.S.), Fasc. 2:196-206. [P] Air temperature, water temperature, baro- 190 FISHERY BTJLLE3TIN OF THE FISH AND WILDLIFE SERVICE Genovese, Sebastiano. — Continued metric pressure, salinity, density, oxygen, transparency, currents during the fishing season (May-June) of a liorth Sicilian tuna trap for Thunnus thyitnus. 1953. Osservazioni idrologiche eseguite nelld toii- nara Capo San Marco (Sciacca) durante la campagna di pesca 1953. Boll. Pesca Piscic. Idrobiol. 8(n.s.), Fasc. 2:241-251. [P] Meteorological, hydrological data and catch records at a south Sicilian tuna trap; Thunnus thynnus: fishing season, average weight, notes on sexual maturity. GiNSBURG, Isaac. 1953. The taxonomic status and nomenclature of some Atlantic and Pacific populations of yellowfin and bluefin tunas. Copeia 1953(1) : 1-10. [P] Thunnus thynnus, T. secundodorsalis, T. saliens, T. albacares, T. subulatus, T. cata- linae, T. macropterus : synonymy. GODSIL, H. C. 1936. Tuna tagging. J. Cons. int. Explor. Mer ll(l):94-47. [P]. Description and figures of a strap-disk opercular tag used on yellowfin and skip- jack off California, together with the tools technique for applying it. 1937. The five tunas. Fish. Bull.; Sacramento 49:24-33. [P]. Catch statistics for yellovsrfin, bluefin, skipjack, albacore, and bonito in and ad- jacent to California waters. 1938a. The high seas tuna fishery of California. Fish. Bull., Sacramento 51:1-40. [P]. Yellowfin and skipjack fishing methods, capture of livebait, handling of catch. 1938b. Tuna tagging. Calif. Fish Game 24:245- 250. Skipjack, yellowfin tuna: tagging meth- ods and release records. 1938c. Tuna tags. J. Cons. Int. Explor. Mer 13(2): 217-220. [P] i Reports tagging of approximately 4,000 tuna on American Pacific coasts with, no recoveries; results of tests exposing ena- meled silver tags to sea water. 1945. The Pacific tunas. Calif. Fish Game 31 (4) ; 185-194. [P] Keys and figures for Katsuwonus pela- mis, Sarda lineolata,' Thunnus thynnus, Neothunnus macropterus, Thunnus alalun- ga, and Pan-athxmnua mebachi. GODSiL^ H. C. — Continued ; 1948. A preliminary population study of the yel- I lowfLn tuna and the albacore. Fish Bull., Sacramento 70:90 p. [P] Neothunnus macropterus, Thunnus gernio: morphometric data; population relation- ship of Japanese, Hawaiian, and Califor- nia fish analyzed; methods of taking mor- phometric measurements described. 1949a. A progress report on the tuna investiga- tions. Calif. Fish Game 35(1) :5-9. [P] Albacore, yellowfin tuna: summary of population studies based on morphometric analysis. 1949b. The tunas. In: The commercial fish catch off California for the year 1947 with an his- torical review 1916-1947. Fish Bull., Sacra- mento 74:11-27. [P] Catch statistics and distribution for Neo- ■ .ifrt'J . ■"■ thunnus macropterus, Katsuivonua pela- mis, Thunnus germo, Thunnus thynnus; fishing methods briefly described. GODSIL, H. C, and R. D. Byers. 1944. A systematic study of the Pacific tunas. Fish Bull., Sacramento 60:131 p. [P] Katsuwonus pelamis, Neothunnus macrop- terus, Parathunnus mebachi, Thunnus ger- mo, T. thynnus: proportional measure- ments, methods of measurement, internal anatomy, key, figures, description, classl- ' ' ' ■ ' f ication, counts of meristic characters, anatomical differences between species listed; population relationships discussed for all except P. mebachi. ''o .iv>r! GODSiL, H. C, and E. C. GREENHOOD. 1948. Some observations on the tunas of the Hawaiian region. Calif. Div. Fish Game. Bur. Mar. Fish. Mimeogr. Rep. Albacore, black skipjack, skipjack, yellow- fin tuna : distribution. 1951. A comparison of the populations of yellow- fin tuna, Neothunnus macropterus, from the eastern and central Pacific. Fish Bull.j Sacra- mento 82:1-33. [P] Comparative study of morphometric meas- urements of specimens from Hawaii, Pal- myra, Fiji, and the Pacific Coast of North America. A study of the homogeneity of the central Pacific stocks is also included. 1952. Observations on the occurrence of tunas in the eastern and central Pacific. Calif. Fish Game 38 ( 2 ) : 239-249. [P] Distribution of Thunnus germo, Neothun- nus macropterus, Katsuwonus pelamis. GODSIL, H. C, and E. K. Holmbekg. i 1950. A comparison of the bluefin ' tunas, genus BIBLIOGRAPHY ON THE TUNAS 191 GODSIL, H. C, and E. K. Holmberc}. — Continued Thunntis,. from New England, Australia, and California. Fish. Bull., Sacramento 77:1-55. [P] Atlantic bluefin (T. thynnus), California bluefin (T. thynnus), and Australian blue- fin (T. maccoyii) compared. Graham, David h. 1938. Fishes of Otago Harbour and adjacent seas with additions to previous records. Trans, roy. Sec. N.Z. 38(3) :414. Auxis thazard: listed. Green HOOD, E. C. 1952. Results of the examination of four small yellowfin tuna, Neothunnus macropterus. Calif. Fish Game 38(2) :157-163. [P] Morphometric and anatomical compari- son of three specimens from Hawaiian waters with one from Costa Rica, ranging 216 to 302 mm. in length. HadZi, J. 1934 ( ? ) . §to znamo danas o 2ivotu tunja. Rlbar- skiKalendar: 29-35. Tuna: Adriatic Sea. Hart, J. L., and H. J. Hollister. 1947. Notes on the albacore fishery. Progr. Rep. biol. Stas. Nanaimo and Prince Rupert 71:3-4. [P] Albacore catch correlated with water temperature and area; stomach contents. Hart, J. L., et al. 1948. Accumulated data on albacore (Thunnus alalunga). Circ. biol. Stas. Nanaimo and Prince Rupert 12 :5 p. [P] Thunnus alalunga: food, sizes landed in B. C, log records. Hasegawa, Kiichi. 1937. Progress report of experimental tuna fish- ing in waters adjacent to Woleai. Nany6 suisan joho 1:3-7. (Pacific Oceanic Fishery Investigations Translation No. 7. In: Spec, sci. Rep.: Fish. U. S. Fish Wildl. 46). [P] Yellowfin, bigeye, skipjack: longline catches in western Carolines; oceanogra- phic data from the fishing stations. hasegawa, KIMPEI. 1938. On a report of investigations of summer albacore. Kaiyo gyogyo 3(4) :14-31. [J] Germo alalunga : Pacific Ocean-northwest. Hatai, Shinkishi, et al. 1941. A symposium on the investigation of skip- jack and tuna spawning grounds. Kagaku nanyo 4(1) :64-75. (Pacific Oceanic Fish- ery Investigations Translation No. 16. In: Spec. sci. Rep: Fish. U. S. Fish Wildl. 18). [P] Skipjack: Japan, Indonesian waters, South Seas; eggs, juveniles, food, migration, Hatai, Shinkishi et al. — Continued sexual maturity, method of determining male and female skipjack. Black tuna: Japan, Philippine region; probable spawn- ing areas and season, sexual maturity, eggs. Yellowfin tuna: sexual maturity and probable spawning season in the Indo-Pacific area. Bigeye tuna: juveniles. Heldt, H. 1930. Le thon rouge et sa pgche, nouveaux aspects de la question. Bull. Sta. oc^anogr. Salamm- b6 18:69 p. [P] Thunnus thynnus: synonymy, anatomy, distribution, food, migrations, tropisms, spawning, growth, catch statistics, bibliog- raphy. 1931. Thunnus thynnus. In: Faune Ichthyologi- que de I'Atlantique Nord, No. 7. [P] Distribution, figure, description, brief synonymy. 1931. Le thon rouge et sa peche, ^l^ments d'un nouveau rapport. Bull. Sta. oc6anogr. Salamm- b6 21: 165 p. [P] Thunnus thynnus: figiared, synonymy, morphology, compared with T. secundo- dorsalis, biometry, meristic counts, distri- bution, migrations, spawning, growth, fish- ing methods, utilization, catch statistics, bibliography. 1932a. Rep^rage des bancs de thons par avion. Notes Sta. oc^anogr. Salammbo 26:12 p. [P] Aerial scouting for tuna schools in trap "• fishery off Moroccan coast; discussion of application of aircraft to tuna fishing and to the study of tuna migrations. 1932b. Le thon rouge et sa pfiche, rapport pour 1931. Bull. Sta. oc^anogr. Salammbd 29: 168 p. [P] Thunnus thynnus: figured, synonymy, "facial anomaly," meristic counts, distri- bution, migrations; tags, hooks, and har- poons figured; spawning, growth; fishing methods and gear, especially purse seine and trap; catch statistics, bibliography. 1934. Le thon rouge et sa p6che. Rapp. Comm. int. Mer M6dit. 8:187-255. Thunnus thynnus: food, migration, spawn- ing, statistics, development, bibliography. 1937. Le thon rouge et sa pfiche, rapport pour les ann^es 1933, 1934, et 1935 (9e Rapport). Rapp. Comm. int. Mer M^dit. 10:235-315. Thunnus thynnus: bionomics and fishery. 1938. Le thon rouge et sa pfiche. Rapp. Comm. int. Mer M6dit. 11 : 311-358. Thunnus thynnus: bionomics. 192 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Heldt, H. — Continued 1943. fitudes sur le thon, la daurade, et les muges. Histoires d'^cailles et d'hamcQons. Broch Sta. oc^anogrr. Salammbd 1. Thunnus thynnus: growth; migration; Mediterranean Sea. 1950. Le germon (Germo alalunga Gmelin). fitude biolopque d'aprfes l'6xamen des 6cailles. Ann. biol., Copenhague 7:63-64. [P] Scale reading; age and growth. Herald, Earl S. 1949. Pipefishes and seahorses as food for tuna. Calif. Fish Game 35 (4): 329. [P] Euthynnus yaito, yellowfin tima: stomach contents. 1951. Pseudofins on the caudal peduncle of juve- nile scombroids. Calif. Fish Game 37(3): 335-337. [P] Auoeis thazard, Katsuwonus pelamis. Herre, Albert W. C. T. 1932. A check list of fishes recorded from Tahiti. J. Pan-Pacif. Res. Instn. 7(1) :3. [P] Euthynnus alletteratus, E. pelamis, Neo- thunnus macropterus: listed. 1933. A check list of fishes from Dumaguete, Oriental Negros, P.I., and its immediate vicin- ity. J. Pan-Pacif. Res. Instn. 8(4) :7. [P] Euthynnus yaito, Katsuwonus pelamis: lUted. 1935. A check list of the fishes of the Pelew Islands. Mid-Pacif. Mag. 47(2) : 164. [P] Katsuwonus pelamis, Neothunnus macrop- terus : listed. 1936. Fishes of the Crane Pacific Expedition. Zool. Ser. Field Mus. nat. Hist. 21:105-107. Katsuwonus pelamis, Neothunnus macrop- terus, Thunnus thynnus: distribution, syn- onymy, observations on N. m^wropterua fin lengths. 1940. Distribution of the mackerel-like fishes in the western Pacific north of the Equator. Proc. Pacif. Sci. Congr. 6th, vol. 3:211-215. [P] Auxis thazard, Euthynnus alletteratus, E. yaito, Germ.0 alalunga, Katsuwonus pela- mis, Neothunnus macropterus, N. rarus, Parathunnus sibi, Thunnus thynnus: dis- tribution. 1953. Check list of Philippine fishes. Res. Rep. U. S. Fish Wildl. 20:977. Synonymy, range. HERRE, Albert W. C. T., and A. F. Umali. 1948. English and local common names of Philip- pine fishes. Circ. U. S. Fish Wildl. Serv. 14: 128 p. [P] Herre, Albert W. C. T., and A. F. Umali. — Continued Auxis thazard, Euthynnus yaito, Oerm,o alalunga, Katsuwonus peUumis, Neothun- nus macropterus: listed. Hiatt, R. W., and V. E. BROCK. 1948. On the herding of prey and schooling of the black skipjack, Euthynnus yaito Klshi- nouye. Pac. Sci. 2(4) :297-298. [P] Euthynnus yaito: observations of herding of scads, Decapterus sanctae-helenae, in the Marshall Is. Higashi, Hideo. 1940a. Utilization of fishery byproducts from the South Seas (3). Nanyo suisan 6(7) :13-20. [J,P] Bigeye tuna, black tuna, skipjack, yellow- fin tima: ratio of viscera weight to body weight. 1940b. Utilization of fishery byproducts from the South Seas (7). Nanyo suisan 6(12) : 10-13. EJ.P] Skipjack: ratio of viscera weight to body weight; proportional measurements of various body parts. 1941a. Utilization of fishery byproducts from the South Seas (10). Nanyo suisan 7(3):32-39. [J.P] Katsuwonus vagans, Neothunnus macrop- terus: proportional measurements of vari- ous body parts; age analysis. 1941b. Utilization of fishery byproducts from the South Seas (14). Nanyo suisan 7(8) :36-43. [J,P] Bigeye tima, yellowfin tima : length-weight data; proportional measurements of vari- ous body parts; liver figured. 1942. Record of experiments on fishes of the South Seas. Nanyo suisan 8(11) : 13-27 [J,P] Katsuwonus vagans, Neothunnus macrop- terus, Parathunnus sibi: length-weight data; proportional measurements of vari- ous body parts. HiGASHi, Hideo, and M. Hirai. 1948. The nicotinic acid content of fish. Contrib. cent. Fish Sta. Japan (1946-1948) 18:129-132. Skipjack, yellowfin tuna: nicotinic acid content of various body parts. Hildebrand, Samuel F. 1946. A descriptive catalog of the shore fishes of Peru. Bull. U. S. nat. Mus. 189:361-372. [P] Euthynnus alletteratus, Katsuwonus pela- mis, Thunnus macropterus: classification; description, synonymy; distribution, food, key. Thunnus germo, T. thynnus: key, occurrences recorded. BIBLIOGRAPHY ON THE TUNAS 193 HIRATSUKA, HITOSHI, and KAKUJI IMAIZUMI. 1934. Experimental fishing and investigation in southern waters by the Shonan Maru. Taiwan sotokufu suisan shikenjo jigyo hokoku (gyo- robu) 1932:97-164. [J,P] Yellowfin tuna: Indo-Pacific region; length-weight data, fishing conditions in relation to oceanography and weather; catch per unit of effort; distribution. HIRATSUKA, HITOSHI, and KIYOJI ItO. 1934. Report on experimental tuna fishing in the Celebes Sea. Taiwan sotokufu suisan shiken- jo jigyo hokoku 1934:1-28. [J,P] Yellowfin tuna: length-weight data; fish- ing conditions in relation to oceanography and weather; catch per unit of effort; dis- tribution. HIRATSUKA, HITOSHI, and SEIICHI MORITA. 1935. Correlation between length and weight of yellowfin tuna. Taiwan suisan zasshi 241: 8-10. (Pacific Oceanic Fishery Investiga- tions Translation No. 26. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 22) . [P] Neothunnus macropterus : Celebes Sea; morphometries. 1936. Correlation between length and weight of yellowfin tuna from the Celebes Sea. Suisan kenkyu shi 31 ( 1 ) : 67-68. [J] Neothunnus macropterus : length-weight relationship; Celebes Sea. HIKTZ, M. 1933. O tuni i tunolovu. Priroda 23(10) : 318-320. Adriatic Sea, fishing gear and methods. HORIGUCHI, YOSHISHIGE, D. KAKIMOTO, and KENiCHi Kashiwada. 1950. The distribution of inosite in the skipjack (Katsuwonus pelamis). Kagoshima sui sen ken ho 1:41-46. [J] Chemical analysis. HORIGUCHI, YOSHISHKJE, KENICHI KASHIWADA, and Daiichi Kakimoto. 1953. Biochemical studies on skipjack, Katsuwo- nus vagans. II. Contents of inorganic sub- stances in pyloric caeca. Bull. Jap. Soc. sci. Fish. 18(7) :279-282. [Je,P] Qualitative and quantitative analysis of the inorganic content of the pyloric caeca; quantitative results compared with those from analysis of muscle tissue. lEHISA, SATORU. 1939. Catch of tunny in the seas south of Kyushu. Bull. Jap. Soc. sci. Fish. 8(3) :143-144. [Je,P] Thunnus orientalis: catches correlated v«th water temperature. IKEBE, KenzO. 1938. Progress report on skipjack baitfish hold- ing experiments. Nanyo suisan joho 2(4): 2-4. [J.P] Skipjack live bait fishing: mortality of baitfish ( Spratelloides delicatulusf) held in pound net at Palau; water temperature, salinity, specific gravity. 1939a-1940. Four papers on the morphometry and age of tropical tunas. Nanyo suisan joho 3(10) :4 p; 4(1) :3 p; 4(2) :4 p; 4(5) :5 p. Pa- cific Oceanic Fishery Investigations Transla- tion No. 34. In: Spec. sci. Rep: Fish U. S. Fish Wildl. 22). [P] Length and weight data on Neothunnus macropterus, Makaira mitsukurii, Para- thU7inus mebachi, and Thunnus germo from Palau, the Marshalls, and Saipan; ages given based on Aikawa's age-size tables. 1939b. On the age of yellowfin taken in Palau waters. Nanyo suissin joho 3(10) :4-8. [J.P] Length-weight data, body condition, sex- ual maturity; age analysis based on size groups according to Aikawa's tables. 1940a. Age and measurements of timas in Palau waters. Nanyo suisan joho 4(1) :2-4. [J,P] Bigeye tuna, yellowfin tuna, striped mar- lin: length-weight data, condition factor; age analysis of yellowfin tuna based on size groups in accordance with Aikawa's tables. 1940b. Measurements of yellowfin tuna taken south of the Marshall Islands. Nanyo suisan j6h6 4 (2) :2-5. [J,P] Length-weight data on longline-caught fish, a total of 75 from 6 stations; age analysis based on size according to Aika- wa's tables. 1940c. Measurements of albacore and yellowfin tuna taken in Saipan waters. Nany6 suisan joho 4(5) :63-67. [J,P] Lengths and weights (gutted) of 8 alba- core and 58 yellowfin taken on longUnes north of Saipan; age analysis according to Aikawa's size age tables. 1940d. Investigation of tunas in Palau waters. Nany6 suisan joho 4(6): 2-4. [ J,P] Catches at 14 longlining stations near 7° N., 134°E.; surface temperatures and cur- rents noted; catch rates for yellowfin and spearfishes combined. 1941a. A survey of tuna fishing grounds in the Marshall and Caroline islands. Nanyo suisan joho 5(l):6-9. (Pacific Oceanic Fish- 194 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE IKEBE, KenzO. — Continued ery Investigations Translation No. 15 In: Spec. sci. Rep.: Fisli. U. S. Fish Wild. 47) [P] Catches from exploratory longlining in the Equatorial Countercurrent; currents at 10 fishing stations recorded; catch rates for tunas and marlins combined. 1941b. Measurements of yellowftn tuna from the Equatorial Countercurrent area. Nanyo suisan joho 5(3) :5-13. [J,P] Lengths, weights, sex, and estimated age (from Aikawa's tables) of 188 longline- caught yellowfin from 8 stations in south- em Carolines waters. 1941c. A contribution to the study of tuna spawn- ing grounds. Nanyo suisan joho 5(4): 9-12. [J,P] South Seas: probable tuna spawning grounds; lengths, weights, and estimated ages of 20 juvenile yellowfin, with dates and positions of capture. 1942. Report of the investigation of tuna fishing in the Timor, Arafura, and Banda seas. Nanyo suisan 8(1) :29-41. (Pacific Oceanic Fishery Investigations Translation No. 48. In: Spec. Sci. Rep: Fish. U. S. Fish Wildl. 45) [P] Bigeye tuna, yellowfin tuna: longline fish- ing conditions in relation to oceano- graphy; catches and catch rates at 10 stations. IKEBE, KenzO, and Takeshi matsumoto. 1937. Progress report on experimental skipjack fishing near Yap. Nanyo suisan joho 1(4) : 3-9. (Pacific Oceanic Fishery Investigations Trans- lation No. 6. In: Spec. sci. Rep.: Fish. U. S. Fildl. 46) [P] Results of 10 days' livebait fishing around Yap: fishing logs; lengths, weights, sex, and condition factor for 83 skipjack; water temperatures and salinities to 200 m. at 8 stations. 1938. Report of a skipjack bait investigation in Saipan waters. Nanyo suisan joho 6:2-12. (Pacific Oceanic Fishery Investigations Trans- lation No. 30. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 44). [P] Common names and descriptions of the species used for livebait at Saipan; results of intensive experimental fishing for live- bait. IKEDA, NOBUYA. 1932. The bait problem and the development of our skipjack and mackerel fisheries. Miyagi no suisan 1:9-29. [J] Katsuwontis pelamis. Pacific Ocean-north- west; livebait fishing. IKEDA, NOBtTYA, and SEIJI AND6. 1933. A consideration of skipjack fishing conditions off northeastern Japan in 1930. Gyoro kenkyti- kai kaiho 5. [J] KatsHwonus pelamis. Pacific Ocean-north- west. IKEHAEA, Isaac I. 1953. Live-bait fishing for tuna in the central Pa- cific. Spec. sci. Rep.: Fish. U. S. Fish Wildl. 107:20 p. [P] Availability and characteristics of live- bait species of the Hawaiian, Line, and Phoenix islands; results of exploratory livebait fishing around these groups; size frequency distribution of yellowfin tima caught by livebait fishing in Line and Phoenix islands; evaluation of baiting grounds in the area. IMAI, Sadahiko. 1950. Studies on flying fishes. 1. On the young of flying fishes eaten by tuna. Kagoshima sui sen ken ho 1:137-148. [J] Tuna: food. IMAIZUMI, Kakuji. 1937. An account of the investigation of tuna fishing grounds in the East Philippine Sea. Taiwan suisan zasshi 271:6-23. [J,P] Popular account of an exploratory tuna longlining expedition: total catch and catch rates by species for 17 stations; brief remarks on maturity of yellowfin, distribution of catch rates, and oceano- graphic conditions. IMAMURA, YUTAKA. 1949. The skipjack fishery. Suisan koza 6:17-94. (Pacific Oceanic Fishery Investigations Trtms- lation No. 32. In: Spec, sci Rep.: Fish. U. S. Fish Wildl. 49). [P] Auxis hira, A. maru, Euthynntis yaito: Japan; description, distribution, habits. Katsuwonus pelamis: Japan; anatomy, description, migration, spawning areas and seasons, food, populations, habits, natural enemies, fishing conditions in relation to oceanography. 1953. The tuna fishery. Suisan koza 6:104 p. Tokyo. [J,P] General account of tuna livebait fishing, purse seining, gillnetting, trolling, and longlining; tables of operating and econ- omic data on Japanese longliners. INANAMI, YOSHIYUKI. 1940a. Relationship of viscera weight to body weight in yellowfin tuna. Nanyo suisan j6- h6 4(2):2-7. [J,P] Length, weight, body depth, body breadth. BIBLIOGRAPHY ON THE TUNAS 195 INANAMI, YOSHIYUKI.— Continued and weight of gills and viscera for 13 large longline-caught yellowfin; percen- tage of gill-and-viscera weight in body weight calculated. 1940b. Oceanography and fishing conditions in the sea area centered on Palau. Nanyo suisan jo- ho4(3):5-7. [J,P] Bigeye tuna, yellowfin tuna: longline fish- ing conditions in relation to currents and water color. 1940c. Tuna fishing conditions and currents along the eastern coast of the Palau Islands. Nanyo suisan joho 4(2) :7-10. [J,P] Bigeye tuna, yellowfin tuna: longline fish- ing conditions in relation to local currents at 47 stations within 30 miles of the coast. 1941. Report of oceanographic changes and fishing conditions in Palau waters. Nanyo suisan jo- ho 5(2) :2-6. (Pacific Oceanic Fishery Investi- gations Translation No. 3. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 42) [P] Describes the effects of a southward shift of the Equatorial Counter-current on oceanographic conditions and on the skip- jack fishing at Palau. 1942a. Oceanographic conditions and yellowfin tuna fishing grounds in South Sea Islands wa- ters. Nanyo suisan joho 6(l):2-5. [J,P] Location of longline fishing grounds in re- lation to currents, transparency, water color, and water temperature in the equa- torial current system between 130°E. and 170°E. longitude. 1942b. Skipjack fishing conditions at Saipan, Truk, and Ponape. Nanyo suisan joho 6(1) :5-7. [J,P] Seasonal fluctuations in commercial catch and the size of fish taken. 1942c. Small skipjack caught at Truk. Nanyo sui- san joho 6(1) :7. [J,P] Records and measurements of two juve- niles. 1942d. Grounds fished by tima boats operating in the Inner South Seas. Nanyo suisan joho 6(1) : 7-9. [J,P] Albacore, bigeye, skipjack, yellowfin tu- na: fishing conditions in relation to water temperature; seasonal shifts in equatorial longlining grounds at 150°E. to leO'E. longitude. INOUE, MOTOO. 1953. Albacore fishing conditions and oceanogra- phic conditions in the 1952-53 longlining sea- son. Tokai daigaku sangyo kagaku kenkyusho INOUE, MOTOO. — Continued suisan kenkyubu gyogyo shiryo No. 3:17 p. [J.P] Thunnus gerino: fishing conditions in re- lation to oceanography, catch per unit of effort; Pacific Ocean — northwest. ISAWA, TAKAO. 1935. On the tuna of the Japan Sea coast of Hok- kaido. Hokkaido sui shi junpo 1935:727-731. [J] Tuna: distribution; Sea of Japan. IWATE PREFECTUKE FISHERIES EXPERIMENT STATION. 1953a. South Seas tuna fishing experiment report. 1:44 p. [J.P] Results of an exploratory longlining trip aroxmd 10°N., 170°W. Fishing logs; cur- rents, salinities and water temperatures to 300 m., plankton collections; distribu- tion of catch rates for N. macropterus, P. sibi, and black marlin; length frequen- cy curves, sex ratios and apparent ma- turity, notes on stomach contents, catch by branch line number and estimated depth; data on shipboard refrigeration and the prices received from each species landed. 1953b. South Seas tuna fishing experiment report. 2:31 p. [J,P] Results of an experimental longlining trip around 4°N., 175°W. Fishing logs; cur- rents, water temperatures and salinities to 300 m., plankton collections; length fre- quency distributions for N. macropterus, P. sibij and black marlin; catch rates; catch by branch line numbers and esti- mated depts; prices received for each spe- cies landed; sex ratios ajid apparent ma- turity. JAPANESE Bureau of Fisheries. 1933. Report of the southern fisheries investigation for 1931. Bur. Fish. Min. Agr. and For. (1931) 1933:96 p. [J,P] Results of tuna longlining and purse sein- ing by boats of the training ship Hakuyo Maru In the Celebes Sea and Indian Ocean; shipboard tuna canning experiment; com- plete logistical and operational data; catch and production (dried fish) of a land- based skipjack fishing operation in Bor- neo; fishing gear and boats described and figured; yellowfin and bigeye catch, catch rates, weather, and oceanographic data for 19 longline stations; observations of surface schools of skipjack and small yel- lowfin. 1934. Report of the soutbem fisheries investiga- tion for 1932. Bur. Fish. Min. Agr. and For. 196 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Japanese Bureau of fisheries. — Continued (1932) 1934: 347 p. [J.P] Results of tuna longlining and purse sein- ing by boats of the factory ship Haruna Maru (1,500 tons) off the N. coast of Borneo and the W. coast of Sumatra; com- plete operational data; results of shipboard canning and freezing experiments; des- criptions and figures of fishing gear and boats; daily catches by species by each of 8 boats fishing up to 40 baskets of long- lines; catch rates given, positions of sets plotted; yellowfin stomach contents (non- quantitative) recorded for 66 samples of up to 29 fish, together with notes on plank- ton samples from the same stations; graph plotting yellowfin and bigeye catch rates with transparency and water temperature at 0, 100, and 150 meters. 1935. Report of the southern fisheries investigation for 1933. Bur. Fish. Min. Agr. and For. (1933) 1935: 298 p. [J,P] Results of tuna longlining and purse sein- ing by boats of the factory ship Haruna Maru (1,500 tons) off south coasts of Su- matra, Java, and the lesser Sundas; com- plete operational data; shipboard canning and freezing results; fishing logs and catch data for 67 longlining stations; water tem- perature and specific gravities to 200 m. given as high, low, and average for each of three sections of cruise; total catch rates (all species) given similarly; details of longline and purse seine construction; descriptions of boats used; description of processing techniques; tuna catch com- prised yellowfin and bigeye. 1939. Results of encouragement given to the devel- opment of albacore fishing grounds during 1938. Bur. Fish., Min. Agr. and For. 1939: 298 p. [J.P] Detailed results of exploratory albacore longlining by 11 research ships at 28°N.- 43°N., 165°E.-165°W. from May to Sep- tember; track charts and fishing logs; data on surface and subsurface water tem- peratures and salinities, correlated with catch rates; albacore .stomach contents noted; measurement and sex data; catch records also for bigeye, yellowfin, and skipjack. 1940. Results of encouragement given to the de- velopment of albacore fishing grounds during 1939. Bur. Fish., Min. Agr. and For. 1940: 173 p. (Translated as Spec. sci. Rep.: Fish. U. S. Fish Wildl. 33). [P] Detailed results of exploratory albacore longline fishing by 9 research ships at 30° N.-45°N., 163°E.-175°W. from May Japanese Bureau of fisheries. — Continued through October; track charts and fishing logs; data on surface and subsurface water temperatures and salinities, cor- related with albacore catch rates; stomach contents noted; measurement and sex data; catch records for bigeye also; data on plankton collections (nonquantitative). 1942. Results of encouragement given to the devel- opment of albacore fishing grounds during 1940. Bur. Fish. Min. Agr. and For. 1940: 135 p. [J] Results of exploratory albacore longlining in the central North Pacific. JOUBIN, M. (Ed.) 1934. Faune Ichthyologique de I'Atlantique Nord, No. 15. Copenhagen, Andr. Fred Host and Fils (published for Conseil Permanent International pour I'Exploration de la Mer). [P] Plates including description, synonymy, geographical distribution of: Germo ala- lunga, Atixis thazard, Katsuwonus pelamis, Euthynnus alletteratus, Sarda sarda. June, Fred C. 1950a. Preliminary fisheries survey of the Hawai- ian-Line Islands area. Part 1. The Hawaiian long-line fishery. Comm. Fish. Rev. 12(1) :1- 23. [P] Information on the boat, crew, description of gear, bait, setting the line, fishing areas and depths, amount and efficiency of gear used, catch composition. 1950b. The tuna industry in Hawaii. Pan-Amer. Fish. 4(10) : 11, 19. [P] Brief description of the skipjack (Katsu- wonus pelamis j fishery. 1951a. Preliminary fisheries survey of the Hawai- ian-Line Islands area. Part 2. Notes on the tuna and bait resources of the Hawaiian, Lee- ward, and Line Islands. Comm. Fish. Rev. 13(l):l-22. [P] Includes sea conditions, tuna and bait re- sources, for the Hawaiian Islands, Lee- ward Islands, Line Islands, and Canton Island. 1951b. Preliminary fisheries survey of the Ha- waiian Line Islands area. Part 3. The live- bait skipjack fishery of the Hawaiian Islands. Comm. Fish. Rev. 13(2) : 1-18. [P] Description and notes on biology of skip- jack, development of the fishery, fishing boats and crews, bait, fishing methods, fishing areas and seasons. 1952a. Observations on a specimen of bluefin tuna (Thunnus thynnus) taken in Hawaiian waters. Pacif. Sci. 6(1) :75-76. [P] Comparison with Thunnus orientalis; mor- BIBLIOGRAPHY ON THE TUNAS 197 jLi^NE, Frb:d C. — Continued phometric measurements and meristic counts of the specimen given. 1952b. An "unusual" yellowfin tuna (Neothunnus macroptenis) from the waters of the northern Line Islands in the central Pacific Ocean. Copeial952 (3):210-211. [P] Description of a 24-lb. female which be- cause of its coloration at first appeared to be a bluefin tuna (Thumius thynnus) or Thunnus maccoyi. Meristic counts and measurements indicated that it was a yellowfin tuna. 1953. Spawning of yellowfin tuna in Hawaiian waters. Fish. Bull., U. S., 54(77) :47-64. [P] Collection and treatment of ovary samples, description of the ovaries, development of the ova, relation of ovary size to fish size as a measure of maturity, number of ova spawned, spawning season, spawning and the fishing season. June, Fred C, and J. W. Reintjes. 1953. Common tuna-bait fishes of the central Pa- cific. Res. Rep. U. S. Fish. Wildl. 34:54 p. [P] Keys and descriptions of families and spe- cies of bait fishes: figures; evaluation of tuna bait resources in the central Pacific. K.\FUKU, TAKEICIURO. 1950. On the dark muscle tissue in fishes. (Rep. No. 1.) The dark muscle tissue of the tunas, from the viewpoint of comparative anatomy. Jap. J. Ichth., Tokyo 1(2) :89-100. [Je,P] Tuna: anatomy. KAGOSniMA PREFECTURE FISHERIES EXPERIMENT STATION. 1930a. Experimental skipjack fishing. Kagoshima- ken suisan shikenjo jigyo hokoku (1928) :1-18. [J,P] Results of 11 skipjack livebait fishing cruises off southern Japan, the Ryukyus, and northern Formosa from March to June: fishing logs; surface and subsurface water temperatures at fishing stations. 1930b. Experimental longline fishing for tuna. Kagoshima-ken suisan shikenjo jigyo hokoku (1928) :18-31. [J,P] Results of 3 longlining cruises from south- ern Japan to the Ryukyus between Novem- ber and February; bigeye, yellowfin, and albacore catches, correlated with tides; surface and subsurface water tempera- tures at fishing stations; fishing logs. 1930c. Experimental fishing by small motor ves- sels: experimental longline fishing for alba- core. Kagoshima-ken suisan shikenjo jigyo hokoku (1928): 54-60. KAGOSHIMA PREFECTURE FISHERIES EXPERIMENT Station. — Continued Results of 2 longlining cruises with a 20- ton vessel in Ryukyu waters in March and April; description of gear; albacore, yel- lowfin, and bigeye catch, surface and sub- surface water temperatures at fishing sta- tions; fishing logs, plots of sets. 1931a. Experimental skipjack fishing. Kagoshima- ken suisan shikenjo jigyo hokoku (1929) :1-16. [J.P] Results of 10 livebait skipjack fishing cruises in Ryukyu waters between March and June; average monthly surface water temperatures for 7 years; commercial fishing correlated with surface tempera- tures; a few subsurface temperature data; fishing logs. 1931b. Experimental longline fishing for tuna. Kagoshima-ken suisan shikenjo jigyo hokoku (1929) :16-30. [J,P] Results of 4 exploratory tuna longlining cruises in Ryukyu waters from October to January; average monthly water tempera- tures; catches of yellowfin, albacore, and bigeye tuna recorded with surface and subsurface temperatures at the stations, moon phase and tides, transparency; de- scription of gear; plot of station locations and fishing logs. 1932a. Experimental skipjack fishing. Kagoshima- ken suisan shikenjo jigyo hokoku (1930) :l-20. [J,P] Results of 9 livebait skipjack fishing cruises in Ryukyu waters from March to June; surface water temperature iso- therms plotted; fishing logs and plot of station locations. 1932b. Experimental longline fishing for tuna. Kagoshima-ken suisan shikenjo jigyo hokoku (1930) : 21-28. [J,P] Results of 7 exploratory tuna longlining cruises in Ryukyu waters from October to February; catches of yellowfin, bigeye, and black tuna recorded with surface and subsurface temperatures at fishing sta- tions; fishing logs and plot of station lo- cations. 1932c. Experimental longline fishing for albacore and pole and line fishing for mackerel. Kago- shima-ken suisan shikenjo jigyo hokoku (1930) 54-59. [J,P] Results of 3 exploratory longlining stations in Ryukyu waters in March; catch (a total of 3 albacore) recorded with surface and subsurface temperatures on the stations; fishing logs and plot of station locations. 198 FISHEKY BULLETIN OF THE FISH AND WILDLIFE SERVICE KAGOSHIMA PEEFECTURE FISHERIES EXPERIMENT Station. — Continued 1933a. Investigation of skipjack fishing. Kagoshi- ma-ken suisan shikenjo jigyo hokoku (1931) : 1-16. [J,P] Results of 8 exploratory live-bait skip- jack fishing cruises in Ryukyu, Formosan, and Philippines waters from March to June; fishing logs and plots of station lo- cations; surface temperatures discussed, isotherms plotted. 1933b. Experimental longline fishing for tuna. Ka- goshima-ken suisan shikenjo jigyo hokoku (1931): 16-23. [J,P] Results of 3 exploratory tuna longlining cruises (23 stations) in Ryukyu waters in October-December; fishing logs and plots of station locations; yellowfin, alba- core, and bigeye catch recorded with sur- face and subsurface temperatures and sal- inities, transparencies, on the stations; total catch rates averaged by area. 1934. Investigation of skipjack fishing. Kagoshl- ma-ken suisan shikenjo jigyo hokoku (1932): 1-27. [J,P] Results of 8 exploratory skipjack live-bait fishing cruises in Ryukyu waters from February to June; fishing logs and plot of station locations; water temperature dis- tribution discussed with data on commer- cial landings at local ports. 1935a. Investigation of skipjack fishing. Kagoshi- ma-ken suisan shikenjo jigyo hokoku (1933) : 1-13. [J,P] Results of 9 exploratory skipjack live-bait fishing cruises in Ryukyu waters from March to June; fishing logs and plot of station locations; water temperature dis- tribution discussed with data on commer- cial landings at local ports. 1935b. Cooperative South Seas tuna fishing survey. Kagoshima-ken suisan shikenjo jigyo hokoku (1933) : 15-21. [J,P] Results of 4 combination skipjack live- bait and tuna longlining exploratory cruises in the Sulu and Celebes seas by subsidized commercial vessels; catch rates for total tuna, species not recorded. 1936a. Investigation of skipjack fishing. Kago- shima-ken suisan shikenjo jigyo hokoku (1934):1-16. [J,P] Results of 10 exploratory skipjack live- bait fishing cruises in Ryukyu waters from March to June; discussion of water temperatures and fishing conditions; com- mercial landings at local ports; lengths and weights of 728 skipjack, average con- dition factors of samples. KAGOSHIMA Prefecture Fisheries experiment Station. — Continued 1936b. Cooperative southern skipjack fishing ex- periment. Kagoshima-ken suisan shikenjo jig- yo hokoku (1934) : 17-21. [J,P] Results of 4 exploratory skipjack live-bait fishing cruises in the Sulu and Celebes seas by a subsidized commercial vessel; chart of locations fished; notes on feeding and care of live-bait. 1936c. Investigation of the migration of important fishes. Kagoshima-ken suisan shikenjo jigyo hokoku (1934) :86-87 [J,P] Release records of 45 skipjack tagged on the caudal peduncle in Ryukyu waters. 1937a. Investigation of skipjack fishing. Kago- shima-ken suisan shikenjo jigryo hokoku (1935) 1-8. [J,P] Results of 8 exploratory skipjack live-bait fishing cruises in Ryukyu waters from February to June discussed in relation to surface water temperatures; data on land- ings at local ports by months; average weights, lengths, and condition factors of ten 50-fish samples; distribution of water temperatures and skipjack fishing grounds recorded and plotted for several years; seasonal and annual variations in size composition of the catch; fishing logs and station plots. 1937b. Cooperative southern tuna fishing experi- ment. Kagoshima-ken suisan shikenjo jigyo hokoku (1935) :9-ll. [J,P] Results of 4 combination skipjack live-bait fishing and tuna longline exploratory cruises to the Sulu Sea; skipjack, yellow- fin, and bigeye catches recorded, fishing locations plotted. 1937c. Survey of the present condition of the skip- jack fishing industry. Kagoshima-ken suisan shikenjo jigyo hokoku (1935) : 96-103. [J,P] Numbers of skipjack vessels by size classes in the prefecture, their equipment and operating regime; economic and finan- cial data on the fishery. 1938a. Investigation of skipjack fishing. Kago- shima-ken suisan shikenjo jigyo hokoku (1936) :l-4. [J,P] Results of 9 exploratory skipjack live-bait fishing cruises in Ryukyu waters from February to July; discussion of water tem- perature in relation to fishing conditions; monthly skipjack landings at local ports; average lengths and weights of fifteen 20-fish samples; fishing logs and station plot. BIBLIOGRAPHY ON THE TUNAS 199 KAGOSHIMA PREFECTURE FISHERIES EXPERIMENT Station. — Continued 1938b. Cooperative southern skipjack fishing ex- periment. Kagoshima-ken suisan shikenjo ji- gyohokoku (1936) : 7-10. [J,P] Results of 2 exploratory skipjack fishing cruises in the Celebes and Sulu seas from November to February by a subsidized commercial vessel. 1938c. Investigation of the migration of important fishes. Kagoshima-ken suisan shikenjo jlgyo hokoku (1936) :89. [J,P] Release records for 45 skipjack tagged in Ryukyu waters. 1939a. Investigation of skipjack fishing. Kago- shima-ken suisan shikenjo jigyo hokoku (1937) :l-3. [J,P] Results of 7 exploratory skipjack live-bait fishing cruises in Rjoikyu waters from March to June; fishing logs and plot of stations; average length and weight of 8 samples of approximately 20 fish; month- ly commercial landings at local ports. 1939b. Cooperative southern skipjack fishing ex- periment. Kagoshima-ken suisan shikenjo jigyo hokoku (1937) :7-9. [J,P] Results (not very detailed) of 10 days' exploratory skipjack live-bait fishing in the Sulu Sea by a subsidized commercial vessel; chart of locations fished. 1939c. Investigation of the migration of important fishes. Kagoshima-ken suisan shikenjo jigyo hokoku (1937) :69. [J,P] Relase records of 36 skipjack tagged in Ryukyu waters. 1940a. Experimental skipjack fishing. Kagoshima- ken suisan shikenjo jigyo hokoku (1938) :l-3. [J,P] Results of 9 exploratory skipjack live-bait fishing cruises in Ryukj-u waters from March to July; brief discussion of water temperature and fishing conditions; plot of locations fished; monthly commercial landings at local ports; average lengths and weights of 13 samples of 20 fish each. 1940b. Cooperative southern skipjack fishing ex- periment. Kagoshima-ken suisan shikenjo ji- gyo hokoku (1938) :7-9. [J,P] Results of 4 exploratory skipjack live-bait fishing cruises to the Sulu Sea from Octo- ber to February by a subsidized com- mercial vessel. 1940c. Investigation of the migration of important fishes. Kagoshima-ken suisan shikenjo jigyo hokoku (1938) :43. [J,P] Release records for 20 skipjack tagged in Ryukyu waters. KAGOSHIMA PREFECTURE FISHERIES EXPERIMENT STATION. — Continued 1941a. Investigation of skipjack fishing. Kago- shima-ken suisan shikenjo jigyo hokoku (1939) :l-3. [J,P] Results of 10 exploratory skipjack Uve-bait cruises in Ryukyu waters from March to July; fishing log and station plot; brief discussion of water temperature and fish- ing conditions; average lengths and weights of 8 samples of 20 fish each; monthly commercial landings at local ports. 1941b. Cooperative southern skipjack fishing ex- periment. Kagoshima-ken suisan shikenjo ji- gyo hokoku (1939) :7. [J.P] Fishing logs for 3 exploratory skipjack livebait fishing cruises to the Sulu Sea from October to January by a subsidized commercial vessel. KAKIMOTO, DAIICHI, AKIO KANAZAWA, and KENICHI Kashiwada. 1953. Biochemical studies on skipjack (Katsuwo- nusvagans). IV. Distribution of amino-acid in pyloric caeca. Bull. Jap. Soc. sci. Fish. 19 (6) :729-732. [Je,P] Chemical analysis. Kamimura, Tadao, and MISAO HONMA. 1953. Biology of the big-eyed tuna, Parathunnris mebac;ii(Kishinouye). I. Length frequency of the big-eyed tuna caught in the North Pacific with special reference to biennial frequency. Contrib. Nankai reg. Fish. Res. Lab. 1, Contrib. 46:18 p. [Je,P] Analysis of size composition of bigeye landed by longUnes from 130°E. to 165°W. north of 26°N. from 1948 through 1953; compared for areas and years; discussion of reproduction, growth, and migration. KANAGAWA PREFECTURE FISHERIES EXPERIMENT STATION. 1951a. Report of South Sea tuna fishing experi- ments, 1951. 49 p. [J,P] Detailed results of a longlining cniise to 0°-6°N., 154°-162°E. in January-March: distribution, longline catch rates, relative depth of capture, length frequencies, sex ratios, stomach contents for yellowfin and bigeye tuna. Subsurface water tempera- tures, salinities; notes on plankton. 1951b. Report of work of the Kanagawa Prefecture Fisheries Experiment Station, 1950:1-107. Results of 5 longlining cruises between 5 - 38°N. and 150°-175°E. Yellowfin, bigeye, and albacore catch rates, length frequen- cies, sex ratios, stomach contents (non- quantitative); relative depth of capture; subsurface water temperatures and saU- 200 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Kanagawa Prefecture fisheries experiment Station. — Continued nities, notes on plankton; fishing logs, N. macropteriis and P. mebachi recorded from stomach contents; shipboard refrig- eration experiments. 1952a. Report of experimental tuna fishing on the eastern grounds by the Sagami Maru. Maguro gyogyo shiken hokoku 4:21 p. [J.P] Report of an exploratory longlining cruise in Feb.-Mar. around 32°N., 171°E. Catch rates of albacore and bigeye; average sizes; length frequency distribution of al- bacore; estimated depth of capture; sex ratios; water temperatures and salinities to 300 m. ; fishing correlated with oceano- graphic conditions. Complete fishing logs. 1952b. Report of experimental tuna fishing on the eastern grounds by the Sagami Maru. Maguro gyogyo shiken hokoku 5:26 p. [J,P] Report of an exploratory longlining cruise in Feb.- Apr. around 30 °N., 175 °E. Alba- core and bigeye catch rates, sex ratios, length frequency distribution; water tem- perature and salinities to 300 m., correla- ted with fishing conditions; catch by branch line number. Complete fishing logs. KANAI, MOTO, and NOBORU KASU. 1938. Report of an investigation of the economics of the Okinawan tuna longline and deep-sea handline fisheries. Okinawa-ken suisan shi- kenjo hokoku 1:1-42. [J,P] History of the fishery, charts of grounds showing expansion, fishing seasons; des- cription and figures of gear and vessels; list of vessels, catch statistics; economics of the fishery: costs of gear, provisions, bait, fuel. Kanamura, Masami, and Kakuji Imaizumi. 1936a. Report of experimental tuna longlining east of Formosa. In Report of experimental fish- ing by the Shonan Maru in 1935. Taiwan s6- tokufu suisan shiken jo shuppan 3:165-202. [J,P] Results of fishing at 21°-23°N., 121°-126° E. in Nov.-Dec. Gear dimensions, station plot, fishing log, weather, surface and subsurface temperature and salinity, transparency, water color; yellowfin, big- I eye: catch rates, depth of capture; body temperature, sex, maturity, length, weight of 25 yellowfin and 3 bigeye; shark dam- age rate; tuna catch and water tempera- ture correlated. 1936b. Fishing conditions for tuna longline boats based at Takao. Taiwan sotokufu suisan shi- kenjo shuppan 3: 203-207. [J,P] Operational data (average cruises per kanamura, Masami, and Kakuji Imaizumi. — Cont. season, Oct. -May, average fishing effort, average catch per cruise by species) for 7 commercial boats fishing the Celebes, Sulu, and S. China Seas; plots by 1° squares of catch rates for yellowfin and bigeye. Kanamura, Masami, and Haruo Yazaki. 1940a. Investigation of tuna longline fishing grounds in the East Philippine Sea. In Report of fishing ground investigations by the Sho- nan Maru in 1937. Taiwan sotokufu suisan shiken jo shuppan 21:1-65. [J,P] Results of exploratory longlining at 3°- 20°N., 123°-131°E. in June-Sept. Fishing log, track chart and station plot; setting, hauling, and soaking time; construction of gear. Subsurface temperatures and salini- ties, transparency, water color at stations. Yellowfin, bigeye, skipjack, albacore catch rates; depth of capture; catch rates plotted against latitude; stomach contents (non- quantitative) ; sex and maturity; body temperature (compared with water tem- perature) ; length, weight, length-weight relation, age according to Aikawa's tables, condition factor; correlation of yellowfin catch with water temperature, salinity, currents, and weather examined; shark damage rates. 1940b. Investigation of tuna longline fishing grounds in the South China Sea. In Report of fishing ground investigations by the Shonan Maru in 1937. Taiwan sotokufu suisan shi- kenjo shuppan 21:67-117. [J,P] Results of exploratory longlining at 16°- 22°N., 116°-121°E. in Feb.-May. Fishing log with operational data, track chart and station plot; gear construction. Subsur- face temperatures and salinity, transpar- ency, water color at stations. Yellovirfin, albacore: catch rates; depth of capture; stomach contents (non-quantitative) ; body temperature, compared with water tem- perature; length, weight, age according to Aikawa's tables, condition factor, sex and maturity; correlation of dark or "green" flesh color with kind of feed, con- dition factor, and sexual maturity at- tempted. Kashiwada, Kenichi, Daiichi Kakimoto, and YOSHISHIGE Horiguchi. 1952. Biochemical studies on skipjack (Katsu- wonus vagaiis). I. Chemical components of pyloric caeca and extractive matter. Bull. Jap. Soc. sci. Fish. 18(4) : 147-150. [Je,P] Fat, moisture, ash, protein content; changes in nitrogen compounds by auto- lysis. BIBLIOGRAPHY ON THE TUNAS 201 KaSHIWADA, KENICHI, DAIICHI KAKIMOTO, and TOSHIMORI Yamasaki. 1953. Biochemical studies on skipjack, m. On the nitrogen compounds in skipjack pyloric caeca extract. Bull. Jap. Soc. sci. Fish 19(1) : 15-18. [J,P] Chemical analysis. KatO, Genji. 1940. An account of longline fishing for tuna. Nanyo suisan joho 4(7) : 8-10. [J.P] General account of a longlining trip in Palau waters: brief notes on yellowfin sex ratio and maturity, on soaking time and bait loss, and on working efficiency of the fishermen. KAWAMURA, HYOzO. 1939. Observations on oceanography and fishing conditions in Palau waters. Nanyo suisan joho 3(1) :2-6. [J,P] General discussion of yellowfin and skip- jack fishing in relation to ocean currents in the Palau area. KAWANA, TAKESHI. 1934. On the relation between the tuna fishery and oceanographic conditions. Hokkaido sui- san shikenjo suisan chosa hokoku 31:80 p. (Spec. sci. Rep.: Fish. U. S. Fish. Wildl. 781. [P] Thunnus orieiitalis, northern Japan: catch statistics; fishing conditions related to temperature, currents, abundance of other fishes, sunspots, lunar period, wind direc- tion; monthly average size of fish in com- mercial landings; tag recovery records for 6 fish. 1935. The tuna spawns in the Japan Sea. Suisan kenkyti shi 30:284-286. Black tuna: spawning. 1937. The catch of tunny, Thunnus orientalis T. and S., off Kushiro, Hokkaido, in relation to the vertical difference in water temperature. Bull. Jap. Soc. sci. Fish. 6(2) :73-74. (Pacific Oceanic Fishery Investigations Translation No. 50. In: Spec. sci. Rep.: U. S. Fish Wildl. 52 . [P] Temperature difference between surface and 50 m., and average numbers of large, medium, and small fish taken per trip in 13 years; Pacific Ocean-northwest. 1938. On tuna fishing conditions at Urakawa. Hoku sui shi sui cho ho 43:125-134. [J] Thuntius orientalis -Jisbing conditions. Pa- cific Ocean - northwest. Kawasaki, Tsuyoshi. 1952. On the populations of the skipjack, Katsuwo- nus pelamis (Linnaeus), migrating to the Northeastern Sea Area along the Pacific Kawasaki, Tsin-osm. — Continued coast of Japan. Bull. Tohoku reg. Fish. Res. Lab. 1:1-14. [Je,P] Populations distinguished by size compo- sition, condition factor, and biting qualties; distribution correlated with oceanographic conditions; length-weight relation; age determination (according to Aikawa's ta- bles) ; Pacific Ocean-northwest. KIDA, Takeo. 1936. On the surface temperature of water in the tunny fishing grounds off Kushiro and Ura- kawa in summer. Bull. Jap. Soc. sci. Fish. 5(2):87-90. [Je,P] Thynnus thynnus: fishing condition cor- related wdth water temperature; size com- position of schools; relation to other fishes, birds, plankton; Pacific Ocean-northwest. KIKAWA, ShOji. 1953. Observations on the spawning of the big- eyed tuna (ParathuH7ius mebachi Kishinouye) near the southern Marshall Islands. Contrib. Nankai reg. Fish. Res. Lab. 1, Contrib. 24:10 p. [Je,P] Mature and ripe fish abundant in longline catch in southern Marshalls, June- August; occurrence of ripe fish in relation to area, oceanographic conditions, month, and size composition; ripe eggs described and fig- ured; successful artificial fertilization des- cribed and embryo figured. KIMURA, KINOSUKE. 1932. Growth curves of blue-fin tuna and yellow- fin tuna based on catches near Sigedera, on the west coast of Prov. Izu. Bull. Jap. Soc. sci. Fish. 1(1) :l-4. (Pacific Oceanic Fishery Investigations Translation No. 37. In: Spec, sci. Rep.: Fish. U. S. Fish Wildl. 22). [P] Neothunnus macropterus, Thunnus orien- talis: g^rowth rates determined from size g^roups; Pacific Ocean-northwest. 1933. Statistical analysis of the catch of yoimg bluefin tuna and yellowtail in Suruga Bay. Bull. Jap. Soc. sci. Fish. 2(4) : 183-194. [Je,P] Thunnus orientalis: trap catches correlated with area and with season; Pacific Ocean- northwest. 1935. Statistical analysis of the catch by keddle nets, along the coast of Suruga Bay. Rec. oceanogr. Wks. Jap. 7(1) :l-36. Neothunnus macropterus, Thunnus orien- talis: Pacific Ocean-northwest; seasonal distribution of trap catches, 1927-32; weight frequencies; age-growth curves. 1941. Skipjack fishing conditions. Suisan seizo kogaku koza 1 :36 p. [J] Northwest Pacific Ocean : distribution, mi- 202 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE KIMURA, KINOSUKE. — Continued gration, catch correlated with water tem- perature; age and size composition of commercial catches. 1942a. Tuna and spearfish fishing conditions. Sul- san seizo kogaku koza 5:122 p. [J,P] Albacore, yellowfin, bigeye tuna; north- west Pacific, Indonesian waters; landings from various longlining grounds (1936- 40), fishing seasons, catch per day fished, relation of water temperature to fishing conditions; age (from Aikawa's tables) and rough size composition of albacore and yellowfin catches. 1942b. High seas fisheries. Kaiyo no kagaku 2(3) : 142-7. [J,P] Albacore, black tuna, skipjack, bigeye, yellowfin; popular accoimt of livebait fish- ing and longlining grounds and seasons, relation to water temperature; proportion of skipjack landings from various areas, 1937-39. 1949. Atlas of skipjack fishing groimds — with data on the albacore grounds. Tokyo. Kuroshio Publ. Co., 44 p. [J,P] Japan: catches of albacore and skipjack in relation to surface water temperature; chart of fishing situation and isotherms for each 10-day period of the year. KiMURA, KINOSUKE, and Kazumi Ishii. 1931. Fishing conditions in the northeastern part of Suruga Bay (Part 1). Tunas and spear- fishes and their young. Suisan butsuri dan- wakai kaiho 33:526-538. [J] Pacific Ocean - northwest: tuna, young. 1932. Fishing conditions in the northeastern part of Suruga Bay (Part 2) , with the growth rates of black tuna and yellovvrfin based on catches from the Shigedera fishing grounds. Suisan butsuri danwakai kaiho 38:562-580. [J] Thunnus orientalis and Neothunntis ma- cropterus: growth, fishing conditions; Pa- cific Ocean-northwest. 1933a. Statistical analysis of the catch at the northeastern end of Suruga Bay. I. Bluefin tuna (Thunnus orientalis T. and S.). Bull. Jap. Soc. sci. Fish. 1(5) :221-229. [Je,P] Size composition, fishing conditions cor- related with season; Pacific Ocean-north- west; annual and seasonal variations in catch and sizes of fish in Japanese tuna traps. 1933b. Statistical analysis of the catch at the northeastern end of Suruga Bay. II. Yellowfin tuna, swordfish, yellowtail, etc. Bull. Jap. Soc. sci. Fish. 2(2) :69-79. [Je,P] KiMURA, KiNOSUKEj and Kazumi Ishii. — Continued Seasonal distribution of trap catches; cor- relation with water temperature; Pacific Ocean-northwest. KIMURA, KINOSUKE, MITSUO IWASHITA, and ToSHmO Hattori. 1952. Image of skipjack and tuna recorded on echo sounding machine. Bull. Tohoku reg. Fish. Res. Lab. 1:15-19. [Je,P] Depth recorder traces of skipjack, alba- core, and bigeye schools; notes on verticaj distribution and schooling habits. KODAMA, Masaharu, K. Iizuka, and T. Harada. 1934. Weight ratio of various body parts and anal- yses of the normal constituents of fresh flesh of important South Sea fish. Taiwan sotokufu suisan shikenjo jigyo hokoku (1932), Technol. Sect. 1-6. [J,P] Skipjack and tuna exanained. KOYASU, ShOzO. 1931. On the skipjack fishing conditions off eastern Honshu. Suisankai 579:2-25; 580-24-30. [J] Katsuwonus pelamis: fishing conditions. Pacific Ocean-northwest. Kkeutzer, Conradin. 1951a. Ein elektrisches ThunfischangelgerSt. Arch. Fischereiwiss. 3(1/2). [ Fishing methods and gear; North Sea. 1951b. Thune werden elektrisch geangelt. Fisch- ereiwelt 3(10) : 160-1. [P] Description and figure of electrified hook, line, and accessories for stunning tuna. KUMAMOTo Prefecture fisheries experiment Station. 1946. Experimental pole and line fishing for skip- jack. Kumamoto-ken suisan shikenjo jigyo ho- koku (1942, 1943, 1944) : 3-5. [J,P] Summary results of 4 cruises off southern Japan in July-Nov. Surface water tem- peratures discussed. KUMATA, TOSHIRO, ET AL. 1941. Illustrated atlas of edible marine animals and plants of the South Seas. Nissan Fish. Res. Sta., p. 62, 65. [J] Katsuwonus vagans, Neothtmnus niacrop- terus, Parathunnus mebachi: distribution; English and Japanese common names, fig- ures, Dutch and Malay common names of N. viacropterus, and P. mehachi. KURONUMA, KATSUZO. 1940. A young of ocean sunfish, Mola mola, taken from the stomach of Germo germo, and a specimen of Masturus lanceolatus as the sec- ond record from Japanese waters. Bull, bio- geogr. Soc. Japan 10(2):25-28. [JE] Stomach contents of Germo germo noted. BIBLIOGRAPHY ON THE TUNAS 203 KURONUMA, KATSUZO, TAKEICHIRO KAFUKU, and ShOji Kikawa. 1949. Report of investigations of skipjack £md tuna resources in 1947 by the Nakamura research staff. Cent. Fish. Exp. Sta. Rep. 1:7 p. (Pa- cific Oceanic Fishery Investigations Transla- tion No. 33. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 17). [P] Katsuwonus pelamis, central and south Japan; morphometric data, merlstic counts; size composition and sex ratio of commercial catch; sexual maturity, gonad weights, egg counts; stomach contents; catch statistics; fishing grounds and sea- sons. Le Danois, E. 1933. Les transgressions oc^aniques. Bull. Inst. oc6anogT. Monaco 613:1-16. Germo alalunga, Thunniis thynnus: oc- currence in relation to water temperature and salinity; Atlantic Ocean. 1938. L'influence des transgressions sur la biologie et la peche. In: L'Atlantique (Histoire et Vie d'un Oc^an), p. 246-286. Paris, Albin Michel. Germo alalunga: races, migrations. 1951. Sur la presence du thon blanc ou germon sur les cotes du Venezuela et sur le lieu de ponte de cette esp^ce dans I'Atlantique Nord. C. R. Acad. Sci. Paris 232:1029-1030. Thunmts (Germo) alalunga; Atlantic Ocean, Caribbean Sea: spawning, migra- tion; Sargasso Sea as probable spawning ground. Le Gall, J. 1934a. Auxis thazard. In: Faune Ichthyologique de I'Atlantique Nord, No. 15. [P] Plate including description, synonymy, dis- tribution. 1934b. Euthynnus alletteratus. In: Faune Ichthy- ologique de I'Atlantique Nord, No. 15. [P] Plate including description, distribution, synonymy. 1934c. Germo alalunga. In: Faune Ichthyolo- gique de I'Atlantique Nord, No. 15. [P] Plate including description, distribution, synonjrmy. 1934d. Katsuxvonus pelamis. In: Faune Ichthy- ologique de I'Atlantique Nord, No. 15. [P] Plate including description, distribution, synonymy. 1949. Germon. R6sum6 des connaissances acquises sur la biologie du germon. Rev. Trav. Off. Pfiches marit. 15:1-42. [P] Germo alalunga: synonymy, common and scientific names; systematics, figure, measurement data; digestive, respiratory, Le Gall, J. — Continued and circulatory apparatus; reproduction, distribution (geographical and vertical), food, parasites, age and growlh, migration, Isirveie; bibliography. 1951. Ichthyom^trie des Thonidds. De I'emploi d'une technique intemationale. J. Cons. Int. Explor. Mer 17(3) :267-273. [P] Attempt to standardize techniques in taking morphometric measurements. LEGENDRE, R. 1932. La nourriture du germon (Germo alalunga Bonnaterre). Arch. Zool. exp. g6n. 74:531-540. [P] Stomach contents of Germo alalunga. 1933. Presence d'Anotopterus pharao Zugmayer dans I'estomac des germons. C. R. Acad. Sci., Paris 197:1261-1752. Germo alalunga, Atlantic Ocean: food. 1934. La faune p^lag^que de I'Atlantique au large du Golfe de Gascogne recueillie dans des esto- macs de germons. Premifere Partie: Poissons. Ann. Inst. Oc^anogr. 14:249-418. Germo alalunga, Atlantic Ocean: food, bio- nomics, bibliography. 1936. La peche du germon, son int^rfit scientifique. Rev. Sci. 74(9) : 266-273. Biology of Germo, methods of capture. 1937. Presence de Parathutuius obesus dans le Golfe de Gascogne. C. R. Soc. Biog^ographie 13:66-67. P.atlanticus: distribution; Atlantic Ocean. 1940. La faune p^lagique de I'Atlantique au large du Golfe de Gascogne recueillie dans des esto- macs de germons. Troisifeme Partie: Inver- t^br^s (C^phalopodes exclus.). Parasites du germon. Ann. Inst. Oc6eaogrr. 20:127-310. Germo alalunga, Atlantic Ocean: food; parasites. Letaconnoux, R. 1950. Le germon (Germo alalunga Gmelin). Blx- amen du nombre de branchiospines. Ann. biol., Copenhague 7:63. [P] Meristic counts. Lozano, Fernando. 1950. Notas sobre el bonito del norte o albacora (Germo alalunga Gmelin) de Galicia, Bol. Inst. esp. Oceanogr. 36:1-13. [P] Morphometries, notes on sexual maturity, food, age, fishing season and catch, rela- tion of albacore and sardine fisheries. LUling, Karl H. 1950. Thunfischanlandungen am Hamburger Fisch- markt in der Schleppnetz- Heringssaison 1949. Fischereiwelt 2(4):57-9. [P] 204 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE LOling, Karl H. — Continued Seasonal changes in frequency distribution of fish weights in German bluefin land- ings. 1951. Die Thunfischanlandungen in der Schlepp- netz-Heringssaison 1949 und 1950. Fischerei- welt 3(6) : 90-2. [P] Weight of landings and frequency distri- bution of fish weights in North Sea blue- fin catch. 1952a. Thunfischbeobachtungen und Thunfisch- fang. Fischereiwelt 4(2) :10-2. [P] Thunnus thynnus: habits; fishing methods and gear; echo sounder record of tima figured. 1952b. Die Thunfischanlandungen in der Schlepp- netz-Heringssaison 1951. Fischereiwelt 4(3): 42-3. [P] Seasonal changes in weight of landings and frequency distribution of fish weights in North Sea bluefin fishery. McHuGH, John l. 1952. The food of albacore (Germo alalunga) off California and Baja California. Bull. Scripps Instn. Oceanogr. tech. 6(4) : 161-172. Also Contributions from the Scripps Institution of Oceanography, New Series, No. 562. [P] Quantitative analysis of contents of 321 stomachs collected from northern Cali- fornia: variations in composition with latitude, year, season; compared with re- ports from other areas; indications of vertical distribution. McKernan, Donald l. 1953. Pioneer longlining for tuna along the equa- tor. Pacif. Fisherm. 51(8) :19, 21, 23. [P] Results of semi-commercial fishing by a West Coast vessel at 33 stations between 140° and 153°W. Yellowfin catch rates, size of fish, cannery rejection rates. Maldura, Cablo M. 1946. Gli olii di tonno. Natura, lavorazione, ap- plicazioni nell' industria e in terapia. Boll. Pesca Piscic. Idrobiol. 1 (n. s.) Fasc. 2:105-130. [P] Thynnus thynnus L., Thynnus alalonga L., Thynmis thunnina Cuv. and Val., Thyn- nus pelamys L., Pelamys unicolor Geoff: Mediterranean; distribution, synonymy; Italian and foreign names; chemical anal- ysis of oils. MANTER, HAROLD W. 1940. Digenetic trematodes of fishes from the Gala- pagos Islands and the neighbouring Pacific. A. Hancock Pacif Exped. 2(14) :448. Gymnosarda alletterata and G. pelamis listed as hosts. Makr, John C. 1948. Observations on the spawning of oceanic skipjack (Katsuwonus pelamis) and yellowfin tuna (Neothunnus macropterus) in the north- ern Marshall Islands. Fish. Bull., U. S. Fish Wildl. 51 ( 44 ) : 201-206. [ P] K. pelamis: records and descriptions of juveniles, ova measurements. K. pelamis, N. macropterus: spawning, length, sex, maturity. M.A.RR, John C, and M. B. Schaefer. 1949. Definitions of body dimensions used in de- scribing tunas. Fish. Bull., U. S. Fish Wildl. 51(47) :241-244. [P] Tunas: measuring methods, meristic counts; determination of sex and degree of maturity. MARTIN, CLARO. 1938. Tuna fishing and longline fishing in Davao Gulf, Philippines. Philipp. J. Sci. 67(2) :189. Katsuwonus pelamis, Neothunnus itosibi, Neothunnu^ macropterus: listed in com- mercial catch. MARUKAWA, HISATOSHI. 1939a. Bait fishes for tuna and skipjack. Part 4 of Fisheries of the South Sea Islands. Nsmyo suisan 5(5):4-10. [J,P] Scientific and common names of baitfish species of Micronesia; description; habitat; methods of capture ; value as bait. 1939b. Fisheries of the South Sea Islands : natural food of skipjack and tunas. Nanyo suisan 5(7):12-14. [J,P] Yellowfin tuna, stomach contents: juve- niles of Arixis thazard, Katsuwonus pel- amis, and Parathunnus sibi mentioned as food of tunas. 1939c. Tunas and the tuna fishery. Parts 1, 2, and 3. Appendix to Fisheries of the South Sea Islands. Nanyo suisan 53:14-22; 54:16-30; 55: 34-42. [J,P] Germo alalunga, Neothunnus albacora, Parathunnus obestis, Thunnus thynnus: general discussion of European tuna fish- eries as compared with Japanese and Am- erican; distribution, description, spawning, feeding habits, migrations, catch statis- tics, fishing gear and methods. (Trans- lation of a German article by Hans Thiele in Die Fischeboote; no date given, but see Thiel, 1938). M.\THER, FRANK J., Ill, and C. G. Day. 1954. Observations of pelagic fishes of the tropical Atlantic. Copeia 1954(3) :179-188. [P] Records of capture by trolling of a few Katsuwonus pela7nis, EiUhynnus allettera- tus, Neothunnus albacores, Parathunnus BIBLIOGRAPHY ON THE TUNAS 205 Mather, Frank J. HI, and C. G. day.— Continued atlanticus, Thunmis thynnus: distribution, length measurements. Matsubara, Kiyomatsu. 1943. Southern fishes. Part 2. Taiwan suisan zas- shi 334:11-14. [J,P] Brief general description of tuna longline and skipjack livebait fisheries, with re- marks on location of fishing grounds in relation to oceanographic conditions. Matsui, KizO. 1942a. Growth, water and fat content of the brain of skipjack and tuna. Kagaku nanyo 5(1) :106- 116. [J.P] Skipjack, yellowfin tuna from Palau waters: maturity of gonads, weight of brain in relation to body length; brain anatomy, regressions of weights of parts of brain on total brain weight; compari- son with other fishes and rat. 1942b. On the gonads of skipjack from the adja- cent waters of Palau. Kagaku nanyo 5(1): 117-122. (Pacific Oceanic Fishery Investiga- tions Translation No. 19. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 20) . [P] Gonad weight used as a criterion of sexual maturity; gonads figured. Matsui, Y. 1938. La pesca del atiin en la Baja California. Bol. Dep. for. Mex. 3:101-106. Tuna, Pacific Ocean - northeast; purse- seining, livebait fishing. Matsumoto, Takeshi. 1937. An investigation of the skipjack fishery in the waters of Woleai Island. Nanyo suisan joho 3:2-4. (Pacific Oceanic Fishery Investi- gations Translation No. 42. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 46). [P] Results of experimental livebait fishing for skipjack at Woleai, Caroline Is.; ob- servations on the scarcity of bait fish at Lamotrek and Puluwat; some oceanogra- phic data; release of tagged skipjack re- corded. Matsumoto, Walter M. 1952. Experimental surface gill net fishing for skipjack (Katsuwonus pelamis) in Hawaiian waters. Spec. sci. Rep. Fish. U. S. Fish Wildl. 90:20 p. [P] Description of gear, history of gill netting for tuna, fishing operations, fishing groimds and sesisons. Mazzarelli, G. 1935. Programmi vecchi e nuovi per lo studio del tonno {Thunnus thynnus L.) . Mem. Biol. mar. 3. Appendix: 1-9. T. thynnus, Mediterranean Sea. Mead, Giles W. 1951. Postlarval Neothunnus macropterus, Auxis thazard, and Euthynnus lineatus from the Pa- cific coast of Central America. Fish. Bull., U. S. Fish Wildl. 52(63) :121-127. [P] Observations on spawning seaison; key. Meyer, P. F. 1951. Erfahrungen mit der elektrischen Thunfisch- angel. Fischereiwelt 3(11) :176. Method of using an electrified tuna hook and its effect in stunning tuna. MIE PREFECTURE FISHERIES EXPERIMENT STATION. 1930a. Investigation of skipjack fishing grounds and guidance in fishing. Mie-ken suisan shi- kenjo jigyo hokoku (1927): 1-15. [J,P] Fishing logs; skipjack, bigeye, albacore catches recorded with surface water tem- perature and specific gravity, Japanese waters. 1930b. Skipjack fishing and oceanographic condi- tions. Mie-ken suisan shikenjo hokoku ( 1927 ) : 15-17. [J,P] Skipjack fishing conditions discussed in relation to water temperature and specific gravity; Japanese waters, May-August; plots of two 100-mile oceanographic sec- tions. 1930c. Investigation of tuna fishing grounds and guidance in fishing. Mie-ken suisan shikenjo jigyo hokoku (1927) : 18-33. [J,P] Longline fishing logs and station plots, Japanese waters; albacore, bigeye, black tuna, yellowfin catches recorded with sur- face water temperature and specific grav- ity; plots of two 100-mile oceanographic sections; fishing discussed in relation to oceanographic conditions. 1930d. Investigation of skipjack fishing grounds and guidance in fishing. Mie-ken suisan shi- kenjo jigyo hokoku (1928) :1-18. [J,P] Livebait fishing logs and station plots, Japanese waters; skipjack, albacore, big- eye, yellowfin catches recorded ■5\ith sur- face water temperature and specific grav- ity; plots of two 100-mile oceanographic sections; fishing discussed in relation to oceanographic conditions. 1930e. Investigation of tuna fishing grounds and guidance in fishing. Mie-ken suisan shikenjo jigyo hokoku (1928) : 19-33. [J,P] Longline fishing logs and station plots, Japanese waters; albacore, bigeye, black tuna, yellowfin catches recorded with sur- face water temperature and specific grav- ity; plots of two 100-mile oceanographic 206 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE MiE Prefecture Fisheries Experiment Station. — Continued sections; fishing discussed in relation to oceanographic conditions. 1950a. No. 1 Taiyo Maru's 2nd exploratory tuna fishing cruise in 1950. Mie sui shi jih6 165:8- 12. [J] Fishing conditions; Pacific Ocean - north- west. 1950b. Shinro Maru's investigation of the coastal tuna longline fishery. Mie sui shi jih5 165: 15-16. [J] Fishing conditions; Pacific Ocean — north- west. 1950c. Outline of the skipjack fishery in 1950. Mie sui shi jiho 165:24-25. [J] Fishing conditions; Pacific Ocean — ^north- west. MiGlTA, M., and K. Arakawa. 1948. Melanophorhonnone of fishes. Contrib. cent. Fish. Sta. Japan (1946-48) 39:241-244. [Je] Frigate mackerel, skipjack, Thunnus ori- entalis, yellowfin tuna: melanophorhor- mone content of pituitary glands tabula- ted; proportional weight of various pEirts of T. orientalia brain; brain of yellowfin figured. MiLiC, N. 1937. Tunolov. Ribarski kalendar: 53-57. Fishing methods and gear, Adriatic Sea. MINE., TATSUZ6, and Satoru Iehisa. 1940. Homogeneity of the groups of black tuna in the Satsunan Sea area. Bull. Jap. Soc. sci. Fish. 8(6) : 292-294. (Pacific Oceanic Fishery Investigations Translation No. 53. In: Spec, sci. Rep: Fish. U. S. Fish WUdl. 52). [P] Size composition of conuuerclal longline catch, seasonal changes; age composition (by Aikawa's tables) ; seasonal changes in catch per trip; southern Japan. MIURA, T. 1941. Fishes of South Seas. Tokyo, Unebi Book Co., Ltd. 416 p. [J] Narrative of trip through various fishing grounds (skipjack, etc.). Includes: (1) In search of skipjack, p. 3-32; (2) South Sea skipjack, p. 35-66; (3) Bait for South Sea skipjack, p. 67-72; (4) South Sea Neothun- nus macropterus, p. 73-87. MIYAMA, YOSHIMICHI, and I. OSAKABE. 1938. On the character of the fats obtained from the various bodily parts of fishes. Bull. Jap. Soc. sci. Fish. 7(2) :105-106. [Je.P] Katsuwonus vagans, Parathunnus aibi, Thunnus orientalis: chemical analysis of fats. MiYAMAj YOSHIMICHI, and I. OSAKABE. — Continued 1940. Note on the vitamin oil contained in the liver of fishes. BuU. Jap. Soc. sci. Fish. 9(1) : 16-20. [Je,P] Bigeye, black tuna, skipjack, yellowfin: chemical analysis of liver and liver oil. MIYAMA, YOSHIMICHI, K. Saruya, and T. Hasegawa. 1939. On the characters of the fats obtained from various body parts of fish. Bull. Jap. Soc. sci. Fish. 8(4): 185-186. [ Je,P] Thunnus macropterus: South Seas; chemi- cal analysis of various body parts; length- weight data, sex, and stomach contents of specimens recorded. MiYAZAKi Prefecture High-Seas Fishery Guidance Center. 1953. Tuna longline fishery guidance. Miyazakl- ken enyo gyogo shidosho gyOmu gaiyo: 3-41. [J,P] Reports of 7 longlining cruises to South China Sea, Philippine, and New Guinea waters; complete fishing logs and opera- tional information; catch rates, size com- position, sex ratios and maturity for yel- lowfin and bigeye; water temperature at surface, 50 m., and 100 m. MIZUSHIMA^ KOichirO, et al. 1951. Studies on green meat in albacore. In: Itsumi sho tosen rombvm : 1-81. Publ. by Shi- zuoka kanzume kyokai gijutsubu. [J,P] Green discoloration of albacore flesh stu- died from physiological, histological, chemical, and technological points of view to determine its cause. Molteno, C. J. 1948. The South African tunas. Cape Town, South African Fishing Industry Research Institute, 34 p. [P] Economics of the tuna fishing industry; how to recognize a tuna; commercial tuna . , fishing methods; synonymy, description, •"C' distribution, utilization of Thunnus thyn- nus L., Germo albacora (Lowe), Sarda sarda (Bloch), Euthynnus pelamis L., Euthynnus alletteratus (Rafinesque), Ger- mo alalunga (Bonnaterre), Auxis thazard (Lac^pSde), Germo obesus (Lowe), Neo- thunnus itosibi (Jordan and Evermann). Mooke, Harvey L. 1951a. Estimation of age and growth of yellowfin tuna (Neothunnus macropterus Temminck and Schlegel) in Hawaiian waters. (Unpublished thesis submitted for the degree of Master of Science, University of Hawaii, Honolulu). [P] Scale and vertebra reading, and weight frequency analysis. BIBLJOGRAPHY ON THE TUNAS 207 Moore, Harvey L. — Continued 1951b. Estimation of age and growth of yellow- fin tuna (Neothunnus macropterus) in Ha- waiian waters by size frequencies. Fish. Bull., U. S. Fish Wildl. 52(65) : 133-149. [P] Scale and vertebra reading, and weight frequency analysis. MORiCE, Jean 1953a. Essai syst^matique sur les families des Cybiidae, Thunnidae, et Katsuwonidae, pois- sons scombroides. Rev. Trav. Inst. sci. tech. P6ches marit. 18(l):35-63. [P] Discussion of anatomy and systematics of the genera Acanthocybiuvi, Cybium, Grammatorcynus, Sarda, Orcynopsis, Gymnosarda, Thunnus, Germo, Parathyn- nus, Neothunnus, Katsutvojius, Euthynnus, and Auxis; keys to the genera; brief biblio- graphies; figures of A. solandri. C. ma- culatum, C. regale. C. cavalla, S. sarda, S. orientalis, G.alalonga, K. pelamis, E. aUet- teratus, and A. thazard. 1953b. Un caractfere syst^matique pouvant servir k s^parer les espfeces de Thunnidae atlantiques. Rev. Trav. Inst. sci. tech. Pgches marit. 18 (l):65-74. [P] Descriptions and figures of livers of T. thynnus, G. alalonga, P. obesus, Neothun- nus albacora, and S. sarda; liver morph- ology as a systematic character. MOEOVIC, DINKO. 1950. Prilog bibliografiji jadranskog rlbarstva. Split, Jugoslavia, Institut Oceanograflju i Rib- arstvo u Splitu, 142 p. [P] Bibliography of material on Adriatic (and other) fisheries published in the Croatian language from 1869 to 1949. Morrow, James e. 1954. Data on dolphin, yellowfin tuna, and little tuna from East Africa. Copeia 1954 (1) :14- 16. Morphometric measurements and sex of 11 Euthynnus af finis and 29 Thunnus alba- cora from the Mombasa area. MOWBRAY, LOUIS L. 1935. Description of the Bermuda large-eyed tuna, Parathunnu^ ambigutis, n. sp. Government Aquarium, Bermuda. (Three-page, imnum- bered, privately printed pamphlet.) MURAYAMA, BiNZO, and ShirO OKtmA. 1950. A study of experimental AmericEin-style purse seining (III). J. Fish. Res. Inst. 3:233- 257. [J,P] Catches of purse seiners fishing skipjack and black tuna off the Japanese coast; construction details of a 50-ton purse seiner. MURAYAMA, BiNZO, and ShirO Okura. — Continued 1952. A study of experimental American-style purse seining (IV). J. Fish. Res. Inst. 4:381- 394. [J,P] Catches and details of operations of purse seiners fishing for skipjack and black tuna off the Japanese coast; details of seine construction. Murphy, Garth, I., and E. L. Niska. 1953. Experimental tuna purse seining in the cen- tral Pacific. Comm. Fish. Rev. U. S. Fish Wildl. 15(4) :1-12. [P] Description of gear, results of purse sein- ing in vicinity of Phoenix, Line and Ha- waiian islands; factors affecting success: weather, clarity of water, vertical thermal structure, behavior of surface schools of skipjack and yellowfin; use of livebait in skipjack purse seining. Murphy, Garth I., and Richard S. Shomxhia. 1952. New tuna source : Fish and Wildlife Service's investigation reveals potential new grounds. Pan-Amer. Fish. 6(10): 14-16. [P] Results of experimental longlining in equatorial waters between 150° and 165° W. Yellowfin, bigeye, albacore, skipjack catches, total catch rates, geographical distribution; relation of upwelling to tuna abundance; possibility of commercial ex- ploitation. 1953a. Longline fishing for deep-swimming timas in the central Pacific, 1950-51. Spec. sci. Rep: Fish. U. S. Fish Wildl. 98:47 p. [P] Description of longline fishing; horizontal and vertical distribution of deep-swimming tunas in equatorial waters; sex ratios and size composition of yellowfin, bigeye, al- bacore, skipjack; relation of upwelling to zooplankton and tuna abundance; com- mercial possibilities. 1953b. Longline fishing for deep-swimming tunas in the central Pacific, January-June 1951. Spec. sci. Rep: Fish. U. S. Fish. Wildl. 108:32 p. [P] Results of experimental longlining in equa- torial waters from 120°W. to 180°. Ver- tical and horizontal distribution of yellow- fin, bigeye, albacore, skipjack; size com- positions and sex ratios. Operational data and comparison of gear with shallow and deep float lines. Effect of -wind on up- welling and tuna abundance. Murray, J. J. 1952. Report on 1951 exploratory blue-fin tuna fishing in the Gulf of Maine. Comm. F^sh. Rev. 14(3) : 1-19. [P] Results of experimental purse seining; de- scription of equipment tmd methods; log of operations. 208 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Nakamura, Hiroshi. 1935. tjber intersexualitat bei Katsuwonus pelamis (Linn.) Trans. Nat. Hist. Soc. Formosa 25 (141) :197-198. [J] Example of hermaphroditism recorded and described. 1936. On the food habits of yellowfin tuna, Neo- thunnus macropterus (Schlegel), from the Celebes Sea. Trans. Nat. Hist. Soc. Formosa 26(148) :l-8. (Pacific Oceanic Fishery Inves- tigations Translation No. 17. In: Spec. sci. Rept: Fish. U. S. Fish Wildl. 23). [P] Analyses of stomach contents of 57 long- line-caught fish; length -weight data. 1938. Preliminary report on the habits of the black tuna, Thumius orientalis (Schlegel). Zool. Mag. 50(5) : 279-281. [J,P] Description and figure of mature egg; gonads figured, distribution, sexual ma- turity, spawning areas and seasons. 1939a. Summary of an investigation of scombroids of Formosan waters. Taiwan suisan zasshi 288:22-26. [J,P] N. macropterus, T. orientalis, P. mebachi, T. germo, K. pelamis, Euthynnus yaito, Kishinoella rara, Gymnosarda nuda, Auxis hira, A. maru: listed as occurring in For- mosan waters; Japanese common names. 1939b. Report on the investigation of Thunnidae in Formosan waters. Taiwan sotokufu suisan shikenjo shuppan 13 : 1-15. [Je,P] Auxis hira, A. maru, Euthynnus yaito: classification; Japanese common names; synonymy. Katstiioonus pelamis, Kishin- oella rara, N. macropterus, Parathunnus mebachi, Thumius germo, T. orientalis: classification; description; distribution; synonymy; Japanese common names; fig- ures of all except N. macropterus, P. me- bachi, and T. germo. N. macropterus: spawning, morphometric data; compared externally with N. itosibi and Semathun- nus guildi. K. rara compared externally with K. zacalles; spawning of T. orientalis. 1939c. Notes on differences between Neothunnus ma- cropterus and Neothunnus itosibi. Taiwan suisan zasshi 288:27-32. [J,P] N. macropterus : classification, morpho- metric data, synonymy; Neothunnus com- pared with 8em,athunnus. 1941. On the body temperatures of some species of Thunnidae and Istiophoridae. Suisan gakkwai ho 8(3) :256-263. [J,P] Yellowfin tuna, bigeye tuna, Formosan and Philippine waters; body temperatures compared with water temperatures at pre- sumed depth of capture. Nakamura, Hiroshi. — Continued 1943. Tunas and spearfishes. Kaiyo no kagaku 3 (10) : 445-459. (Pacific Oceanic Fishery In- vestigations Translation No. 47). [P] Neothunnus macropterus, N. rarus, Para- thunnus mebachi, Thunnus germo, T. ori- entalis: classification, distribution, food, Japanese common names, spawning. 1949. The tunas and their fisheries. Tokyo, Take- uchi Shobo. 118 p. (Translated as Spec. sci. Rep: Fish. U. S. Fish Wildl. 82.) [P] Thunnus orientalis, Germ,o germo, Para- thunnus mebachi, Neothunnus macrop- terus, N. varus : anatomy, description, fig- ures, classification, keys; general account of food, spawning, growth, migration, dis- tribution of longline catch rates in western Pacific; fisheries, relation of fishing grounds to topography and oceanography, fishing seasons; bibliography. Similar ma- terial on spearfishes. 1951. Tuna longline fishery and fishing grounds. Tokyo, Takashima Shoten pub. 144 p. (Also published as Nankai Regional Fisheries Re- search Laboratory Rept. No. 1. Translated as Spec. sci. Rep: Fish. U. S. Wildl. 112) [P] Compilation of research vessel longline catch rates for approximately 20 prewar years. Geographical and seasonal distri- bution of N. macropterus, P. sibi, G. ger- mo, T. orientalis, and spearfishes in the western Pacific and Indonesian waters; average catch rates for each species plot- ted by 1° squares. Relation of fishing grounds and seasons to meteorological and oceanographical phenomena. Nakamura, Hiroshi, tadao kamimura, and YOichi YABUTA. 1953. Size composition of the albacore and bigeyed tuna caught in the North Pacific area. Con- trib. Nankai reg. Fish. Res. Lab. 1, Contrib. 12:6 p. [Je,P] Length composition of longline-caught albacore and bigeye from the central North Pacific; discussion of age and growth, and of annual variations in size composition. Nakayama, TakuzO. 1948. Calculation of the cost price in the tuna fishery. Suisankai 770:10-16. (In: Spec. sci. Rep: Fish: U. S. Fish Wildl. 79). [P] Japanese tuna longline fishery; economic statistics. Nankai Regional fisheries Research Laboratory. 1951a. Report No. 1. 144 p. [J,P] See Nakamura, Hiroshi, 1951. Contents are identical, but this Report has the text and the charts bound separately. BIBLIOGRAPHY ON THE TXINAS 209 Nankai Regional fisheries research laboratory. — Continued 1951b. Supplementary report no. 1:194 p. [J,P] Distribution of albacore longline catch rates in the northwest Pacific by 1° squares, 1948-51; albacore length frequen- cies by month and area; survey of con- struction and dimensions of longline gear of numerous vessels, with area of employ- ment and catch rates by species. Navarro, Francisco de p., and F. lozano. 1950. Carta de pesca de la costa del Sahara, desde el Cabo Juby al Cabo Barbas. Trab. Inst. esp. OceanogT. 21:24 p. [P] Atun (Thiouuis thynnus L.), patudo (Parathunnus obesus Lowe), rabil (Neo- thiuums albacora Lowe), bonito de altura {Katsuwomis pelamys L.), bonito del norte [Germo alalunga Gmelin), melva (Auxis thazard Lacep.) : occurrence and fishing methods. Navaz, Jose M. 1950. Contrlbuci6n al estudio de los esc6mbridos de la costa vasca (atiin, bonitos, y melva). Bol. Inst. esp. Oceanogr. 31:21 p. [P] Morphometries, fishing seasons, catch of Thunnus tliynniis, Germo alalunga, Katsu- womis pelamis, and Auxis thazard in the Bay of Biscay; bibliography; Spanish and Basque names. Nichols, John t., and F. R. La Monte. 1941. Yellowfin, Allison's, and related tunas. Ichth. Contr. Int. Game Fish Assn. 1(3) : 27-32 [P] Neothunniis albacora, N. allisoni, N. cata- linae, N. rarus: classification, description, English common names, key, synonymy. Provisional subspecies: Neothunnus alba- cora macropterus, N. allisoni allisoni, N. allisoni itosibi, and N. rarus zacalles, pro- posed. NiGRELLi, Ross F., and H. W. Stunkard. 1947. Studies on the genus Hirudinella, giant tre- matodes of scombriform fishes. Zoologica, N. Y., 31(4) :185-196. (Contribution No. 747, Dept. of Tropical Research, N. Y. Zoological Society). Table 4: Hirudinella from scombriform fishes other than Acanthocybium: Para- thunniis atlanticus, Katsutconus pelayjiis, Euthynnus alletteratus, Neothunnus ma- cropterus, Thunnus thynnics. Lists name of collector, host, locality and figure. Nishikawa, Sadaichi. 1934. On the future of the high-seas skipjack and tuna fisheries and standards for their opera- ting methods. Rakusui 29(4) : 20-22. [J] • Fishing methods and gear. NiSKA, EDWIN L. 1953. Construction details of tuna long-line gear used by Pacific Oceanic Fishery Investiga- tions. Comm. Fish. Rev. U. S. Fish Wildl. 15(6) :l-6. Also Separate No. 351. [P] Description of gear construction. NnVA, HITOMARO. 1937. On the pigments in the muscles of fishes. Rep. No. 1, Pigments of tuna muscle. Suisan kenkyu shi 32(6) :306-313. [J] Chemical analysis. NOGUCHI, SADAMI. 1938. The Ogasawara Is., and the future of our tuna fishery. Suisankai 666:44-46. [J] Pacific Ocean - northwest. Nomura, ToshizO, et al. 1952-3. Survey of the high-seas tuna fisheries based on landings at Misaki, Tokyo, and Yaizu. Reports of the following months' operations are in the indicated numbers of the Kanagawa suishi geppo: April and May 1952, 1:3-20; June 1952, 2:1-7; July 1952, 3:1-6; Aug. 1952, 4:1-10; Sept. 1952, 5:1-8; Oct. and Nov. 1952, 6:1-10; Dec. 1952 and Jan. 1953, 7:1-15; April 1953, 9:1-10; June 1953, 11:1-20; July 1953. 12:1-18; the reports of February and March 1953 operations were published as Maguro-rui mizuagechi chosa (imnumbered, undated) by the Kanagawa Prefecture Fishesies Experi- ment Station; the report for May 1953 opera- tions is in Report of survey for tuna fishing 1:1-22; the report for August 1953 operations is in Tuna fishing 4 : 17-32. [ J,P] Average catch rates for yellowfin, big- eye, albacore, bluefin, skipjack, and spear- fishes reported by Japanese longline ves- sels from various areas of the western and central Pacific, Indonesian waters, and the Indian Ocean by months; plots of lo- cations fished by the vessels investigated; discussion of fishing conditions in each area for the month; length frequencies of each species by area. oiTA Prefecture fisheries experiment St.ation. 1930. Report of experimental tuna fishing in the Kanto region (1927). oita-ken suisan shiken- j6 jigyo hokoku (1927-28) :l-40. [J,P] Longline catches of yellowfin, bigeye, and black tuna off central Japan : morphome- tric data, body temperatures compared with water temperatures; gear construc- tion, general account of tuna fisheries and bases of the region. OKADA, YAICHIRO, and KlY0M.\TSU MATSUBARA. 1938. Keys to fishes and fish-like animals of Japan. Tokyo, Sanseido Co., Ltd., p. 146-150. [J] Axixis hira, A. tapeinosoma, Euthynnus alletteratus, E. yaito, Germo germo, Kat- 210 FISHERY BULLETIN OP THE FISH AND WILDLIFE SERVICE Okada, Yaichiro, and Kiyomatsu Matsubara. — Con- tinued suwomis vagans, Kishinoella rata, Neo- thunnus itosibi, N. macropterus, Parathun- nus orientalis: classification, description, key, Japanese common names. 1953. Bibliography of fishes in Japan (1612-1950). Mie Prefecture, Faculty of Fisheries, Prefec- tural University of Mie, 228 p. [P] General bibliography of Japanese and foreign literature on fishes which occur in Japanese waters; arranged by years, not annotated. Okada, YaichirO, et al. 1935. Illustrated atlas of Japanese fishes. Tokyo, The Sanseido Co., Ltd. Scomber tapeinocephalus, Auxis tapeino- soma, Sarda oi-ientalis, Katsuwonus va- gans, Euthynnus yaito, Thunnus orientalis, Germo germo, Parathunnus sibi, Neothun- nus macropterus. All figured, briefly de- scribed, brief notes on distribution, habits and utilization; spawning. OKAJIMA, KIYOSHI. 1939. Tuna fisheries of Kanagawa and Shizuoka prefectures. Nanyo suisan joho 3(1) :7-23. [J.P] General description of ports of Misaki and Omaezaki; accounts of vessels, gear, fish- ing methods, fishing grounds, and sample catches of longline, livebait, and mother- ship-type handlining fisheries. Okamoto, GorozO. 1940. On the composition of shoals of "katsuo," Euthynnus vagans (Lesson), in the north- eastern Japanese waters as analyzed by the body weight. Bull. Jap. Soc. sci. Fish. 9(3) : 100-102. (Pacific Oceanic Fishery Investiga- tions Translation No. 52. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 51) . [P] Size composition of commercial catch from various groimds, seasonal changes; age (according to Aikawa's tables). Okinawa prefecture fisheries experiment Station. 1931. Investigation of the maturity of skipjack. Okinawa-ken suisan shikenjo jigyo hokoku (1930:106-107. [J,P] Skipjack length-weight data; gonad weight and maturity. 1936a. Experimental skipjack fishing. Okinawa-ken suisan shikenjo jigryo hokoku (1934) :l-28. [J,P] Results of livebait fishing in Ryukyu waters ;fishing logs of 15 trips, with opera- tional data, catch, weather, surface tem- perature. Okinawa Prefecture fisheries experiment Station. — Continued 1936b. Experimental longline fishing for tuna. Okin- awa-ken suisan shikenjo jigyo hokoku (1934) : 29-35. [J,P] Results of two cruises to Philippines and Hainan I. waters; construction of gear; fishing logs with catch, weather, surface temperature, operational data; yellowfin, bigeye: catch per imit of effort. 1936c. Experiment on holding livebait for skipjack. Okinawa-ken suisan shikenjo jigyo hokoku (1934) :36-46. [J,P] Attempt to correlate captures of gatsun {Trachurus sp.) with air and water tem- perature, atmospheric pressure, specific gravity of water. 1940a. Experimental skipjack fishing. Okinawa-ken suisan shikenjo jigyo hokoku (1939) :3-5. [J,P] Ryukjru Islands: skipjack catch recorded with air and water temperature. 1940b. Experimental tuna fishing. Okinawa-ken suisan shikenjo jigyo hokoku (1939) :6-8. [J,P] Bigeye tuna, black tima: Bonin Islands; catches recorded with water temperature. 1943. Elxperimental skipjack fishing. Okinawa- ken suisan shikenjo jigyo hokoku (1941) :4-14. [J,P] Ryiikyu Islands: distribution of skipjack; catch recorded with air and water tem- peratures. OKUMA, YASUMICHI, ET AL. 1935. Investigation of South Sea fisheries by the Shonan Maru; investigation of tuna fishing grounds. Taiwan sotokufu suisan shikenjo jigyo hokoku (1933) :120-123. [J,P] Yellowfin tuna: Indo-Pacific region; dis- tribution, stomach contents, length-weight data, sexual maturity, fishing conditions in relation to oceanography and weather, catch per unit of effort. OKUMURA, ISABURO. 1943. Management of the southern tuna fisheries. Suisankai 728:67-72. [J] OMORi, Kageyu, and T. Fujimoto. 1940. Experimental longline fishing for tuna. Na- gasaki-ken suisan shikenjo jigyo hokoku (1938) :175-214. [J,P] Bigeye tuna, black tuna: Japan; catches in relation to water temperature and spe- cific gravity. OMORI, KAGEYU, and M. Fukuda. 1938. Experimental longline fishing for tuna. Na- gasaki-ken suisan shikenjo jigyo hokoku (1936) : 47-48. [J,P] BIBLIOGRAPHY ON THE TUNAS 211 OMORI, KAGETi'U, and M. FUKUDA. — Continued Bigeye tuna, black tuna: Japan; catches in relation to water temperature and spe- cific gravity. 1940. Experimental longline fishing for tuna. Na- gasaki-ken suisan shikenjo jigyo hokoku (1937):45-92. [J,P] Bigeye tuna, black tuna: Japan; catches in relation to water temperature and spe- cific gravity. Onodera, Matsuji. 1941. The relation of freshness and condition fac- tor of Palau Islands skipjack to the ratio of finished products. Nanyo suisan joho 5(2): 7-17. [J,P] Skipjack leng^-weight data, body con- dition of fish. OYA, Takeo, and T. Takahashi. 1936. On the growth acceleration substance in the liver of the marine animals. Bull. Jap. Soc. sci. Fish. 5(3) :192-194. [Je,P] Growth hormones in skipjack livers. Partlo, J. M. 1950. A report on the 1949 albacore fishery (Thun- nus alalunga) . Circ. Pac. biol. Sta., Nanaimo 20:37 p. [P] Thumtus alalunga: age and growth, catch in relation to water temperature, tagging, food, fishing gear. 1951. A report on the 1950 albacore fishery of Brit- ish Columbia {Thunnus alalunga). Circ. Fac. biol. Sta., Nanaimo 23:7 p. [P] Thunnus alalunga: distribution of catch for temperature intervals, tagging, size groups in commercial catch, offshore water temperatures, log records. Phillipps, W. J. 1932. Notes on new fishes from New Zealand. N. Z. J. Sci. Tech. 13(4) :226-234. Description of Pacific yellow-finned alba- core, Neothunmis itosibi, as new to N. Z. PoissoN, R., and E. Postel. 1951. Sur la presence d'une vessie natatoire chez certains individus d'Euthynmis alliteratus (Rafn.) (poisson t616ost6en). C. R. Acad. Sci. Paris 233:201-203. Anatomy of air bladder; description. POSTEL, E. 1949. Les thonidds d'Afrique Occidentale Francaise. Bull. Serv. £lev. Industr. anim. A. O. F. 2(4) : 39-46. 1950. Pfiche sur les cfltes d'Afrique occidentale. n. Rapport et note sur quelques poissons de surface de la presqu'ile de Cap-Vert. Inspec- tion g6n6rale de I'^levage, Dakar, French West Africa. 77 p. [P] Postel, E. — Continued Euthynnus alletteratus : morphology, ana- tomy, ecology; fishing methods. Powell, a. w. b. 1937. Marine fishes new to New Zealand; includ- ing the description of a new species of Halie- utaea. Trans, roy. Soc. N. Z. 67(1) :80. Neothunnus itosibi: recorded, synonymy, description, figured. POWELL, Donald E. 1950. Preliminary report on 1950 North Pacific albacore tuna explorations of the John N. Cobb. Comm. Fish. Rev. 12(12) : 1-7. [P] Results of exploratory trolling, gillnetting, and longlining, Oregon to Alaska; water temperatures related to fishing; stomach contents, tagging, plankton; size range of catch. Powell, Donald e., and H. A. Hildebrand. 1950. Albacore tuna exploration in Alaskan and adjacent waters, 1949. Fish. Leafl., U. S. Fish Wildl., 376:33 p. [P] HUstory of West Coast fishery; abundance and location as shown by exploratory surface and deep trolling catches; water temperatures related to fishing; length composition of catch, stomach contents (non-quantitative) ; potential livebait sources evaluated; plankton, saury, blue shark, birds, cetaceans as indicators of tuna. Powell, Donald E., D. L. Al\'erson, and R. Living- stone, JR. 1952. North Pacific albacore tuna exploration-1950. Fish. Leafl., U. S. Fish Wildl., 402:56 p. [P] Trolling, gillnet, longline gear and meth- ods; migration, tagging, vertical distribu- tion; water temperature related to fish- ing; diurnal fishing trends; length com- position; stomach contents; shark and other damage to catch. PRIOL, E. p. 1944. Observations sur les germons et les thons rouges captures par les pgcheurs bretons. Rev. Trav. Off. Pfiches marit. 13:387-439. [P] Thynnus (Germo) alalonga Gmelin: ecol- ogy, enemies, parasites; anatomy; figures of otoliths, vertebrae, measuring tech- nique; reproduction, food. Thynnus (Or- cynus) thynnus L. : color, food, measure- ment data. Rasalan, Santos B. 1950. New methods of fish capture in the Philip- pines. Bull. Fish. Soc. PhiUpp. 1:57-66. [P] Diagram of tuna longline gear Eind esti- mated cost. 212 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Rawlings, John e. 1953. A report on the Cuban tuna fishery. Comm. Fish. Rev. 15(1) :8-21. [P] History, fishing methods and gear, bait, catch, seasons; Parathunmis atlanticus, Katsuwonus pelamis, Neothunnus argenti- vittatus recorded in catch. Reintjes, JOHN W. 1952. Food and feeding habits of the yeUowfin tuna, Neothunnus •macroptertis, in relation to its distribution in the central Pacific region. (Unpublished thesis submitted for the degree of Master of Science, University of Hawaii, Honolulu.) [P] Kinds and sizes of organisms taken; gor- ging; diurnal variation; variation in vol- ume of stomach contents related to lo- cality, habitat, fish size. ReintjeSj John W., and J. E. King. 1953. Food of yeUowfin tuna in the central Paci- fic. Fish. Bull., U. S. 54(81) : 91-110. [P] Kinds and sizes of organisms taken; gor- ging ;diurnal variation; variation in vol- ume of stomach contents related to lo- cality, habitat, fish size. Reiss, p., and E. Vellinger. 1929. Mesures du pH de I'eau de mer aux environs de Tunis en vue d'une application a I'^tude des migrations du thon. Bull. Sta. oc^anogr. Sal- ammbo 15:1-19. Bluefin tuna; migrations related to pH of sea water. RIVAS, LUIS R. 1951. A preliminary review of the western North Atlantic fishes of the family Scombridae. Bull, mar. Sci. Gulf Caribb. 1(3) :209-230. [P] Key :Auxis thcuzard, Katsuwonus pelamis, Euthynnus alletteratus, Thunnus thynnus, Thunnus atlanticus, Thunnus argentivit- tatus, Thunnus alalunga, Sarda sarda. 1953. The pineal apparatus of tunas and related scombrid fishes as a possible light receptor controlling phototactic movements. Bull. mar. Sci. Gulf Caribb. 3(3) : 168-180. [P] Thunnus thynnus, Germo, Parathunnus, Neothunnus, Auxis, Katsuwonus, Euthyn- nus, Sarda: anatomy; photo taxis. Robins, J. P. 1952. Further observations on the distribution of striped tuna, Katsuwonus pelamis L., in east- ern Australian waters, and its relation to sur- face temperature. Aust. J. Mar. Freshw. Res. 3(2):101-110. [P] ROEDEL, Phil M. 1948a. Common marine fishes of California. Fish. Bull., Sacramento. 68:59-63. [P] Katstiwonus pelamis, Neothunnus niacrop- terus, Thunnus germo, T. thynnus: clas- ROEDEL, Phil M. — Continued sification, description, key, distribution, English common names; anatomical dif- ferences between Parathunnus niebachi and T. germo and between P. mebachi and N. macropterus noted. 1948b. Occurrence of the black skipjack (Euthyn- nus lineatus) off southern California. Calif. Fish Game 34(1) :38-39. [P] Distributional note. RONQUILLO, INOCENCIO A. 1953. Food habits of tunas and dolphins bsised upon the examination of their stomach contents. Philipp. J. Fish. 2(1) : 71-83. Neothunnus macropterus, Katsuwonus pel- amis, Euthynnus yaito: stomach contents of troll-caught fish from Philippine waters identified. Rosa, Horacio, Jr. 1950. Scientific and common names applied to tunas, mackerels, and spearfishes of the world with notes on their geographic distribution. Washington, D. C, Food and Agriculture Or- ganization of the United Nations, 235 p. [P] Rosen, Nils. 1943. Tonfiskens upptradande i vara farvatten. Fauna och Flora 1943:23-26. [P] Orcynnus thynnus: migrations, trends in catch, seasons; North Sea and Baltic. ROYCE, William f. 1953. Preliminary report on a comparison of the stocks of yellowftn tuna. Proc. Indo-Pacif. Fish. Coun. 4 (Section H) :130-145. [P] Neothunnus macropterus : morphometric data; distribution; analysis, comparison, and evaluation of data for defining pop- ulations. Russell, f. S. 1933a. Tunny in the North Sea. Nature 132(3342) : 786. Thunnus thynnus: distributional note. 1933b. Tunny in the North Sea. Nature 132(3344) : 860. Thunnus thynnus: distributional note. 1934a. Tunny investigations made in the North Sea on Col. E. T. Peel's yacht St. George, summer 1933. Part 1: Biometric data. J. Mar. biol. Assoc. U. K. 19(2) : 503-522. Thunnus thynnus: tuna hooks marked, measurements of tuna described. 1934b. The tunny, Thunnus thynnus, Linnaeus — An account of its distribution and biology. Sci. Progr. Twent. Cent. 28(112) : 634-649. Chart showing distribution, where caught by hooks and where by nets; habits; de- scriptions of madrague or thonnaire (nets BIBLIOGRAPHY ON THE TUNAS 213 RUSSELL, F. S. — Continued used in Mediterranean and on coasts of Spain, Portugal, and Africa). 1936. Submarine illumination in relation to ani- mal life. Rapp. Cons. Explor. Mer Copen- hague 101(2) :l-8. [P] Thiouius thymius: effect of illumination on migration. S.\itO, Munek.\zu. 1937. Oceanographic investigations and tuna fish- ing conditions in the Solomon Islands. Suisan kenkyu shi 32(5) :260-271. [J] Oceanographic conditions in relation to fishing; Pacific Ocean-southwest. Sakai, MorisabukO, and Michio Uno. 1940. Tuna (maguro) fisheries and boats in Japan. J. Imp. Fish. Exp. Sta. 10:1-37. [Je,P] Statistical study of 80 Japanese tuna long- line boats : equipment, fishing gear, crews, finances, fishing seasons and groimds. Samson, v. J. 1940. Notes on the occurrence of albacore Germo alalunga in the North Pacific. Copeia 1940 (4):271. Distributional note. Sanzo, L. 1932. Uova e primi stadi larvali di tonne (Orcy- nus thynmis Ltkn.). Mem. R. Com. Talass. Ital. 189:16 p. [P] Spawning season in Mediterranean de- duced from egg and larva collections; de- scription and figures of mature ovarian eggs and pelagic eggs; compared with eggs of T. germo and Auxis; development, measurements, description, and figures of larvae at various stages. 1933. Uova e primi stadi larvali di alalonga (Or- cynus germo Ltkn.). Mem. R. Com. Talass. Ital. 198:9 p. Spawning season in the Mediterranean deduced from collections of eggs and lar- vae; descriptions and figures of mature eggs and developing embryos and larvae. Sasaki, takeo. 1932. A consideration of albacore fishing condi- tions and oceanographic conditions north of Zunan. Rakusui 27(6) :9-ll. [J] Germo alalunga: fishing conditions in re- lation to oceanography; Pacific Ocean- northwest. 1939a. Oceanographic conditions and the skipjack fishing grounds of the Northeastern Sea Area. Miyagi Pref. Fish. Exp. Sta., 12 p. (In: Spec, sci. Rep: Fish. U. S. Fish Wildl. 83). [P] Katsuwoyius pelamis: migrations; fishing conditions in relation to water tempera- ture; Pacific Ocean-northwest. Sasaki, Takeo. — Continued 1939b. Oceanographic conditions and the albacore grounds east of Cape Nojima. Miyagi Pref. Fish. Exp. Sta. 14 p. (In: Spec. sci. Rep: Fish. U. S. Fish Wildl. 77) . [P] Germo alalunga: fishing conditions re- lated to water temperature; Pacific Ocean — northwest; catch statistics; sizes and putative age composition of fish in long- line and livebait catch; migrations. Sasaki, takeo, and Isaku Takehisa. 1932. A consideration of the skipjack fishery in the Northeastern Sea Area in 1931. Rakusui 27(4):1-10. [J] Pacific Ocean — northwest; Katsuwonus pelamis. Scagel, R. F. 1949. Report on the investigation of albacore. Circ. biol. stas. Nanaimo and Prince Rupert 17:23 p. [P] Thunnus alalunga: catch in relation to oceanographic conditions in the northeast Pacific; size composition of commercial catch; stomach contents; tagging; body temperature. SCHAEFER, MILNEE B. 1948a. Morphometric characteristics and relative growth of yellowfin tuna {Neothunnus macrop- terus) from the Central Pacific. Pacif. Sci. 2(2):114-120. [P] Morphometric data; length-weight rela- tion; growth; classification based on va- riations in dorsal and anal fin length. 1948b. Size composition of catches of yellowfin tuna {NeothMunus macropterus) from Central America, and their significance in the deter- mination of growth, age, and schooling habits. Fish. Bull., U. S. Fish Wildl. 51 (44) :197-200. [P] Size composition of commercial catch; sexual maturity; age, growth; schoohng habits; length-frequency data from mixed school of skipjack and yellowfin tuna. 1948c. Spawning of Pacific tunas and its implica- tions to the welfare of the Pacific tuna fish- eries. Trans. N. Amer. Wildl. Conf. 13:366- 371. [P] Auxis sp., Euthynnus Uneatus, E. yaxto, Katsuwonus pelamis, Neothunnus viacrop- terus, Thunnus germo: Pacific Ocean, dis- tribution; review of records and observa- tions on spawning and juveniles; manage- ment problem. 1951. Some recent advances in the study of the biology and racial division of Pacific tunas. Proc. Indo-Pacif. Fish. Coun. 2(2/3) :63-69. [P] Neothunnus macropterus, Thunnus germo, 214 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE SCHAEFERj MiLNEE B. — Continued Katsuwonus pelamis, Parathunnus sibi, Thunnus thxmnina, Euthynnus lineatus, Auxis thMzard: general review of work on age and growth, morphometries and racial studies, size composition, spawning. 1952. Comparison of yellowfin tuna of Hawaiian waters and of the American west coast. Fish. Bull., U. S. 52(72) :353-373. [P] Neothunnus macropterus : morphometries, meristic coimts, relative growth, definition of populations. SCHAEFER, MILNEK B., and J. C. MARK. 1948a. Juvenile Euthynnus lineatus and Auxis thazard from the Pacific Ocean off Central America. Pacif. Sci. 2(4) :262-271. [P] Anatomy, meristic counts, description, fig- ures, and records of capture of juveniles. 1948b. Spawning of yellowfin tuna (Neothunnus macropterus) and skipjack {Katsuwonus pela- •mis) in the Pacific Ocean off Central America, with descriptions of juveniles. Fish. Bull. U. S. Fish Wildl. 51(44) : 187-195. [P] Neothunnus macropterus: size composi- tion in commercial catch ; sexual maturity, spawning. Katsuwonus pelamis: sexual maturity, spawning; records of capture, descriptions and figpures of N. macropterus and Katsuwonus pelam,is juveniles. SCHAEFEK, MILNER B., and L. A. WALFORD. 1950. Biometric comparison between yellowfin tunas (Neothunnus) of Angola and of the Pacific coast of Central America. Fish. Bull., U. S. 51(56) : 425-443. [P] Neothunnus albacora, N. macropterus: morphometries, meristic counts, definition of populations; taxonomy. SCHAEFERS, EDWARD A. 1952. North Pacific albacore tuna exploration, 1951. Comm. Fish. Rev. 14 (5): 1-12. [P] Thunnus germo: results of exploratory trolling, longlining, gillnetting; fishing conditions in relation to water tempera- ture and season; stomach contents noted; gear described ;tagging. 1953. North Pacific albacore tuna exploration, 1952. Comm. Fish. Rev. 15(9) : 1-6. [P] Thunnus germo: results of exploratory trolling and gillnetting; fishing conditions in relation to water temperature; descrip- tion of gear; notes on size composition, stomach contents, tagging. SCHMIDT, p. 1930. A check-list of the fishes of the Riu-Kiu Islands. J. Pan-Pacif. Res. Instn. 5(4) :3. [P] Euthynnus aUetteratus: listed. SCHMITT, WALDO L., and L. P. SCHULTZ. 1940. List of fishes taken on the Presidential Cruise of 1938. Smithson. misc. Coll. 98(25) :3. Euthynnus aUetteratus, E. lineatus: listed and compared. SCHUCK, H. A. 1951- Northern record for the little tuna, Euthyn- nus aUetteratus. Copeia 1951 (1):98. [P] Distributional note, eastern Atlantic Ocean. SCHUCK, H. A., and J. F. MATHER, in. 1951. A blackfin tuna (Parathunnus atlanticus) from North Carolina waters. Copeia 1951 (3):248. [P] Distributional note. SCHULTz, Leonard p. 1949. A further contribution to the ichthyology of Venezuela. Proe. U. S. nat. Mus. 99:1-211. [P] Thunnus thynnus: listed; synonymy, Eng- lish and Venezuelan common names. SCHULTZ, Leonard p. and A. C. De Lacy. 1936. A catalogTie of the fishes of Washington £ind Oregon with distributional records and a bibliography. Mid-Pacif. Mag. 49(1) :70-71. [P] Germo alalunga, Thunnus thynnus: syn- onymy, distribution. SCHWEIGGER, ERWIN. 1943. Pesqueria y oceanografia del Peril y propo- siciones para su desarrollo futuro. Lima, Perti, Compaflla Administradora del Guano. 356 p. Bonito (Sarda chilensis) : distribution; Neothunnus macropterus: description and distribution. 1949. El attin frente a la costa peruana. Bol. Comp. Admin. Guano 25(8) :27 p. [P] Thunnus macropterus: length frequency; length-weight relationship; distribution In Peruvian waters in relation to season, water temperature, currents; growth; food, sex ratios. Scordia, C. 1930-1939. Per la biologia del tonno. Mem. Biol, mar. I, 1, 3; H, 1, 2; in, 1, 2, 3; IV, 2, 4, 5; V, 4, 5, 6, 7, 8; VI, 3. Thunnus thynnus, Mediterranean Sea. 1939a. La biologia del tonno secondo Le Danois. Mem. Biol. mar. 6(4) :l-4. Thunnus thynnus, Mediterranean Sea. 1939b. Notizie sulle migrazioni dei tonni del basso Adriatico. Mem. Biol. mar. 6(1) :l-7. Thunnus thynnus: migrations. 1940. Le migrazioni dei tonni tirreno-jonici e I'entrata di essi in tonnara. Atti Conv. biol. mar. 2. Thunnus thynnus: migrations, Tyrrhenian and Ionian seas; trap fishery. BIBLIOGRAPHY ON THE TXJNAS 215 SCORDIA, C. — Continued 1943. Prime indagini sul valore quantitativo delle concentrazioni gamiche del tonno {Thunnus thynnus L.) Boll. Zool. agr. Bachic. 14(1/3): 93-103. Thunnus thynnus: age, spawning; Medi- teiranesm Sea. Seale, Alxtn. 1940. Report on fishes from Allan Hancock Ex- peditions in the California Academy of Sciences. A. Hancock Pacif. Exped. 9(1) : 17-18. [P] Euthynnus lineatus, Katsuwoniis pelamis, Neothunnus macropterus : descriptions, records of capture; Mexico to Galapagos. Sexla, M. 1930. Distribution and migrations of the tuna (Thunnus thynnus L.) studied by the method of hooks and other observations. Int. Rev. Hydrobiol. 24 : 446-466. See (SeUal952). ' 1931. The tuna (Thunnus thynnus L.) of the western Atlantic. An appeal to fishermen for the collection of hooks found in tuna fish. Int. Rev. Hydrobiol. 25(1/2) :46-47. [P] Thunnus thynnus L. compared with T. J secundodoTsalis Storer; harpooning of tuna; distribution in western Atlantic, mi- grations. 1932. Studio sul tonno. Conferenza sulla pesca del tonno. Boll. Pesca Piscic. Idrobiol. 8(1): 68-73. Thunnus thynnus. 1952. Migrations and habitat of the tima (Thun- nus thynnus L.), studied by the method of the hooks, with observations on growth, on the operation of the fisheries, etc. Spec. scl. Rep: Fish. U. S. Fish WUdl. 76:20 p. (Translation of R. Com. talass. It. Mem. 156, Venice, 1929) . [P] Migration, distribution, growth and age from vertebral annuli, spawning, photo- tropism of young, food, influence of salin- ity, cyclical fluctuations in abundance. Serventy, D. L. 1941a. The Australian tunas. Pamphl. Coun. scl. ind. Res. Aust. 104:48 p. [P] Auxis thazard, Euthynnus alletteratus, Katsuwonus pelamis, Kishinoella tonggol, Neothunnus macropterus, Thunnus germo, T. maccoyi: distribution, description, key, figures, Australian common names, size groups, migration and spawning of T. maccoyi; length-weight relationship and internal and external differences of K. tonggol and T. maccoyi compared; livers of K. tonggol and T. m,accoyi figured. Serveinty, D. L. — Continued 1941b. Victorian tunas and some recent records. Vict. Nat., Melb. 58:51-55. Southern bluefin (T. maccoyii) , albacore (T. germo), yeUowfin (Neothunnus m/i- cropterus), striped tuna (Katsuwonus pelamis), bonito (Sarda australis) : re- corded; description of N. macropterus. 1942a. Notes on the economics of the northern tuna (Kishinoella tonggol). J. Coun. scl. ind. Res. Aust. 15(2) : 94-100. Distribution, feeding habits, stomach con- tents, spawning. 1942b. The tuna Kishinoella tonggol Bleeker in Australia. J. Coun. sci. ind. Res. Aust. 15(2) : 101-112. Distribution, description, ratios of various body proportions, internal anatomy, syn- onymy; compared with K. zacalles, Neo- thunnus rarus, Thunnus tiicolsoni, Thynnus tonggol; figures of T. tonggol and K. tonggol, cranium of K. tonggol figured. 1947. A report on commercial tuna trolling tests in southeastern Australia. J. Coun. sci. ind. Res. Aust. 20(1): 1-16. Katsuwonus pelamis, Thunnus germ,o, T. maccoyi: catch per unit of effort; size composition of T. maccoyi. 1948. Allothunnus faUai, a new genus and species of tuna from New Zealand. Rec. Canterbury (N. Z.) Mus. 5(3) :131-135. Classification, description, morphometries; internal anatomy of type specimen; records of specimens and occurrences; compared with Katsuwonidae. Sette, Oscar E. 1954. Progress in Pacific Oceanic Fishery Inves- tigations 1950-53. Spec. sci. Rep: Fish. U. S. Fish Wildl. 116:75 p. [P] Brief summary of POFI tuna research In the central Pacific. Yellowfin: catch per unit of effort, fishing conditions in rela- tion to area; size composition; spawning; growth; distribution of larvae; morpho- metries, definition of populations. Skip- jack: distribution, fishing conditions in relation to season and oceanog:raphic fea- tures. Fishing methods: results of troll- ing, gillnetting, purse seining, livebfiit fishing, longlinlng. Shapiro, Sidney. 1948a. Aquatic resources of the Ryukjoi area. SCAP Nat. Resour. Sect. Rep. 117:54 p. Also Fish. Leaf!., U. S. Fish Wildl., Wash. 333. [P] Auxis hira, A. tapeinosoma, E. yaito, Kat- suwonus pelamis, Neothunnus macrop- terus, Parathunnus sibi, Thunnus germo. 216 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Shapiro, Sidney. — Continued T. orientalis: bibliography, distribution, Ryukyuan common names, migration, spawning of K. pelaniis. 1948b. The Japanese tuna fisheries. SCAP Nat. Resour. Sect. Rep. 104:60 p. [P] Katsuivonus pelaniis, Neothunnus macrop- terus, Thunnus germo, T. orientalis: bibli- ography, classification, description, distri- bution, migration, food, habits, fishing conditions in relation to oceanography, figures, Japanese common names. 1950. The Japanese long-line fishery for timas. Comm. Fish. Rev. U. S. Fish Wildl. 12(4) :l-26. [P] Details of construction, operation of gear. Shim ADA, Bell M. 1951a. An annotated bibliography on the biology of Pacific tunas. Fish Bull., U. S. Fish Wildl. 52(58): 1-58. [P] Arranged alphabetically by author, and papers by the same author chronologically. Detailed subject index. 1951b. Contributions to the biology of tunas from the western equatorial Pacific. Fish. Bull., U. S. Fish Wildl. 52(62) :111-119. [P] Spawning of Neothunnus viacropterus and Parathunnus sibi; records of juvenile Kat- suwonus pelamis; occurrence of Thunnus orientalis. 1951c. Japanese tuna-mothership operations in the western equatorial Pacific Ocean. Comm. Fish. Rev. Fish Wildl. 13(6) :l-26. Also Sepa- rate No. 284. [P] Description and operation of longline gear with diagrams; total landings; composi- tion of catch; mothership vessel opera- tions. 1951d. Juvenile oceanic skipjack from the Phoe- nix Islands. Fish. Bull., U. S. Fish Wildl. 52 (64) :129-131. [P] Descriptions of 5 specimens of Katsuwo- nus pelamis; table of published records of juvenile K. pelamis from Pacific Ocean. 1954. On the distribution of the big-eyed tuna, Parathunnus sibi, in the tropical eastern Pacif- ic Ocean. Pacif. Sci. 8(2) :234-5. [P] Distributional records; anatomical differ- ences from Neothunnus macropterus. Shimizu, Wataeu. 1947. Seasonal changes in the composition of tuna flesh. Bull. Jap. Soc. sci. Fish. 13(1): 27-28. [Je,P] Analysis of black tuna flesh for water, ash, protein, and fat content. Shimoda, Mokuichi. 1937. Southern fisheries. Kaiyo gyogyo 7:1-136. [J,P] General description and history of the ex- ploration and the expansion of the Japa- nese longline and livebait fisheries into the Mandated Islands and Indonesian waters; plan for mothership-type opera- tions. Shizuoka Prefecture Fisheries Experiment Station. 1936. Investigation of skipjack fishing. Shizuoka- ken suisan shikenjo jigyo hokoku (1934) : 1-19. [J,P] Skipjack catch in relation to water temp- erature in Japanese waters. 1937a. Investigation of skipjack fishing. Shizuo- ka-ken suisan shikenjo jigyo hokoku (1935) : 259-285. [J,P] Skipjack catch in relation to water tem- perature in Japanese waters. 1937b. Albacore grounds of the central Pacific. Kaiyo gyogyo 2(15) : 50-55. [J] Thunnus germo, Pacific Ocean — northwest. Sigma. 1941. Spunti biologichi: perche non debba rite- nersi che i nostri tonni ci vengano dall' Atlan- tico. Corr. Pesca 15(21) : 1-2. Thunnus thynnus: migrations; Atlantic, Mediterranean. Smith, J. L. B. 1935. New and little known fishes from South Africa. Rec. Albany Mus. 4:358-364. Neothunnus itosibi: figure. Smith, Osgood R., and Milner B. Schaefe:r. 1949. Fishery exploration in the western Pacific (January to June 1948, by vessels of the Pacific Exploration Company) . Comm. Fish. Rev. 11(3) :1-18. [P] Bigeye tuna, Euthynnus yaito, Katsuwo- nus pelamis, yellowfin tuna; Hawaiian, Line, former Japanese mandated islands, occurrence recorded; E. yaito and K. pela- 7nis figured. Smith, Robert O. 1947. Survey of the fisheries of the former Japan- ese mandated islands. Fish. Leafl. U. S. Fish Wildl. 273:89-95. [P] Katsuwonus pelamis, Neothunnus macrop- terus, Thunnus germo: Hawaiian and Micronesian common names; livebait fish- ing for K. pelamis. Society for the Promotion of oceanic Fisheries. 1936. Tunas. Kaiyo gyogyo 4:1-62. [J,P] Germo alalunga: differences in quality of summer and winter fish from the north- west Pacific in relation to sexual maturity, body condition, season; spawning, migra- BIBLIOGRAPHY ON THE TXJNAS 217 Society for the promotion of oceanic fisheries. — Continued tion, location of fishing' grounds in rela- tion to oceanography; compared with Thu7inus orientalis. 1937a. An investigation of the summer albacore fishing grounds. Kaiyo gyogyo 2(15) : 1-4. [J] Germo alalunga; Pacific Ocean — northwest. 1937b. Tuna fishery from the canning point of view. Kaiyo gyogyo 2(10) :1-120. [J] Japanese tuna fisheries. SOLDATOVj V. K., and G. J. Lindberg. 1930. A review of the fishes of the seas of the Far East. Bull. Pacif. sci. Inst. Fish., Vladi- vostok 5:103-109. Auxis hira, A. maru, Germo alalunga, Katsuwonus pelamis, Neothunnus macrop- terus, Thunnus thynmi^: classification, description, distribution, key, synonymy; keys to Parathunmis and Kishinoella; Thunnus thynnus figured. SoLJAN, T. 1930. Tunolov u hrvatskim primorju. Ribarski list 1-2:7-12; 3-4:25-29. Fishing gear and methods, Adriatic Sea. South Seas Go\'ernment-general fisheries Experiment Station. 1937a. Investigation of the fisheries of the Mari- ana Islands, 1924. Nanyo-cho suisan shikenjo jigyo hokoku 1 (1923-35) :6-9. [J,P] Distribution of yellowfin tuna and skip- jack. 1937b. Investigation of the fisheries of the Mar- shall Islands, 1926-27. Nanyo-cho suisan shi- kenjo jigyo hokoku 1 (1923-35) :14-24. (Pacific Oceanic Fishery Investigations Translation No. 31. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 47). [P] Results of exploratory trolling and long- lining for yellowfin and bigeye at several atolls; distribution. 1937c. Survey of fishing grounds and channels in Palau waters, 1925-26. Nanyo-cho suisan shikenjo jigyo hokoku 1 (1923-35) : 25-37. (Pacific Oceanic Fishery Investigations Trans- lation No. 5. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 42). [P] Results of exploratory trolling and livebait fishing close to the islands; distribution of dogtooth tuna, frigate mackerel, skip- jack; bait survey; fishing seasons. 1937d. An investigation of the waters adjacent to Ponape. Nanyo-cho suisan shikenjo jigyo hokoku 1 (1923-35) : 78-83. (Pacific Oceanic Fishery Investigations Translation No. 12. South Seas Government-General fisheries Experiment Station. — Continued In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 46). [P] Results of exploratory trolling and long- lining; skipjack livebait survey. 1938. Report of oceanographical investigations, 1927-37. Palau, March 1938, 124 p. [J,P] Skipjack, yellowfin tuna: catch in rela- tion to oceanographic factors, especially water temperature; catch statistics for mandated islands. 1941a. Experimental tuna fishing in waters adja- cent to Woleai. Nanyo-cho suisan shikenjo jigyo hokoku (1936) :8-12. [J,P] Yellowfin, bigeye tuna, skipjack: distri- bution. 1941b. Investigation of skipjack fishing in the central Caroline Islands. Nanyo-cho suisan shikenjo jigyo hokoku (1937) : 63-68. [J,P] Release of tagged skipjack recorded. 1941c. Investigation of skipjack fishing in Yap waters. Nanyo-cho suisan shikenjo jigyo ho- koku (1937) :69-75. [J,P] Skipjack length-weight data, body condi- tion. 1941d. Investigation of tuna fishing in waters adjacent to foreign territories. Nanyo-cho suisan shikenjo jigyo hokoku (1936) : 17-26. [J,P] Indonesian waters: yellowfin tuna catch in relation to oceanographic conditions. 1942. Report of a survey of the tuna fishery in Palau waters. Nanyo suisan joho 6(1) : 10-13. (Pacific Oceanic Fishery Investigations Trans- lation No. 4. In: Spec. sci. Rep.: Fish. U. S. Fish WUdl. 42). [P] Yellowfin, bigeye tuna, skipjack: records of catches at 11 longlining stations off Palau In relation to surface and subsur- face water temperatures and currents. 1943a. Investigation of albacore fishing in the central Pacific Ocean. Nanyo-cho suisan shikenjo jigyo hokoku (1939) :1-13. [J,P] Results of 3 exploratory longlining cruises around Saipan, Marcus, and Wake Islands; yellowfin and albacore catches, average, maximum, and minimum length, weight, body depth for each station; subsurface temperatures; gear construction, opera- tional data, station plots. 1943b. Report of the investigation of tuna fishing in waters adjacent to Palau. Nanyo suisan joho 7(1) :2-4. [J,P] Skipjack, yellowfin: fishing conditions in relation to currents and water tempera- ture. 218 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE SUDA, AKIEA. 1953. Juvenile skipjack from the stomach con- tents of tunas and marlins. Bull. Jap. Soc. sci. Fish. 19(4) :319-340. [Je.P] Records of juvenile skipjack collected from stomachs of longline-caught yellow- fin, bigeye, and marlin in Bonln, Caroline, Marshall, and Midway areas; summary of past records; figures and measurements of juveniles; notes on seasonal occurrence; seasonal changes in size composition of juveniles; rough analysis of average stom- ach contents of yellowfin and bigeye in various areas. SUGHJRA, YASUKICHI. 1932. The Kushiro tima fishery under the appli- cation of catch restrictions. Suisankai 594: 15-21. [J] Thunnus orientalis: laws and regulations, Pacific Ocean — northwest. SUYEHIRO, YASUO. 1936. The reasons why the bonito does not take to baits. Fish. Invest. (Suppl. Rep.) Imp. Fish. Exp. Sta. 3(31): 14-16. [J,P] Skipjack: stomach anatomy, stomach con- tents; leng:th, weight, body condition; fish from different schools compared. 1938. The study of finding the reasons why the bonito does not take to the angling baits. J. Imp. Fish. Exp. Sta. 9(69) :87-101. [Je,P] Skipjack: stomach anatomy, stomach contents, food organisms figured. 1941. On the islets of Langerhans of teleost fishes. J. Imp. Fish. Exp. Sta. 11(82) : 121-138. [Je,P] Germo germo, Katsuwonus vagans, Neo- thunnus macropterus, Parathunniis sibi. 1942. A study of the digestive system and feeding habits of fish. Jap. J. Zool. 10(1): 1-303. Katsuwonus vagans, Neothunnus macrop- terus, Parathunnus sibi, Thunnus orienta- lis: digestive system described, analysis of stomach contents; digestive system of K. vagans, N. macropterus, T. orientalis figured. 1950. On the pituitary body of the skipjack. Bull, physiogr. Sci. Res. Inst. Tokyo Univ. 6(9?): 25-27. [J] Katsuwoniis pelamis: anatomy. TACHIKAWA, TAKUITSU. 1932. The rise and decline of the skipjack fishery in Ryukyu waters. Suisankai 590:93-103; 591:18-26. [J] Katsuwonus pelamis; Pacific Ocean — northwest TAIHOKU PROVINCE FISHERIES EXPERIMENT STATION. 1932. Report of the investigation of skipjack fish- ing grounds. Taihoku-shu suisan shikenjo gyomu hokoku 7(1930)1932:1-17. [J,P] Ryukyu Islands: skipjack fishing condi- tions in relation to water temperature and color. Takayama, I., and S. AndO. 1934. A study of the "magro" (Thunnus) fishing in 1930. J. Imp. Fish. Exp. Sta. 5(38):1-21. [Je,P] Neothunnus macropterus, Parathunnus mebachi, Thunnus germo, T. orientalis: fishing conditions in relation to surface water temperature; optimum tempera- tures for occurrence of fish recorded. Takayama, I., et al. 1934. A study of the "katsuwo" (Katsuwonus pelamis) fishing in 1930. J. Imp. Fish. Exp. Sta. 5(39) :23-56. [Je,P] Fishing conditions in relation to surface water temperature; temperature range and optimum temperature for occurrence of fish recorded. Tanaka, Shigeho. 1931. On the distribution of fishes in Japanese waters. J. Fac. Sci. Tokyo Univ. 3(1, Sec. 4) : 22, 50. Auxis hira, Euthynnus alletteratus, E. pelamys, Germo macropterus, Parathun- nus sibi, Thynnus alalunga, T. thynnus: distribution, synonjnny. 1935. A new theory concerning the habits of the snapper and the tuna. Shokubutsu oyobi dobutsu 3(11) :1929-1930. [J] Habits. 1936. Tuna of the Japan Sea. Suisankai 639: 51-54. [J] 1939. The black tuna problem. Suisankai 684: 44-47. [J] Thunnus orientalis, Pacific Ocean — north- west. tapiador, Domingo d. 1951. A report on deep-sea longline fishing for tima in the Bull. Fish. Soc. Philipp. 2:3-27. Yellowfin tuna: fishing conditions in re- lation to water depth, temperature, salinity. taranetz, a. R. 1937. Handbook for the identification of fishes of Soviet Far East and adjacent waters. Bull. Pacif. sci. Inst. Fish., Vladivostok 11:86-87. Auxis hira, A. viaru, Germo alalunga, Katsuwonus pelamis, Neothunnus macrop- terus, Thunnus thynnus: classification, distribution, key. BIBLIOGRAPHY ON THE TUNAS 219 Tauchi, Morisabubo. 1940a. On the stock of Thunnus orieyitalis (T. and S.). Bull. Jap. Soc. sci. Fish. 9(4):133- 135. (Pacific Oceanic Fishery Investigations Translation No. 35. In: Spec. sci. Rep.: Fish. U. S. FishWildl. 16). [P] Survival rates, rates of exploitation, age and size composition. 1940b. On the stock of Neothunntis macropterus (T. and S.). Bull. Jap. Soc. sci. Fish. 9(4): 136-138. (Pacific Oceanic Fishery Investiga- tions Translation No. 35. In : Spec. sci. Rep. : Fish. U. S. FishWildl. 16). [P] Survival rates, rate of exploitation, age and size composition, migration, popula- tion analysis. 1940c. On the stock of Germo germo (Lac^pSde). Bull. Jap. Soc. sci. Fish. 9(4) :139-141. (Paci- fic Oceanic Fishery Investigations Translation No. 35. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 16). [P] Survival rates, rate of exploitation, age and size composition. 1943. On the stock of Euthynnus vagans (Les- son). Bull. Jap. Soc. sci. Fish. 11(5/6): 179-183. (Translation in Spec. sci. Rep.: Fish. U. S. FishWildl. 83). [P] Definition of populations based on age (according to Aikawa's method) and size composition of landings from Japanese and Palau waters. TESTE2S, ALBERT L., ET AL. 1952. Establishing tuna and other pelagic fishes in ponds and tanks. Spec. sci. Rep: Fish. U. S. Fish Wildl. 71 :20 p. [P] Neothunnus macropterus, Katsuwonus pelamis, Euthynnus yaito, Auxis thaaard: trolling catch per unit of effort in Hawaiian waters; behavior in captivity. 1952. Reaction of tuna and other fish to stimuli, 1951. Spec. sci. Rep.: Fish. U. S. Fish Wildl. 91:83 p. [P] Neothunnus macropterus, Euthynnus yai- to: behavior in captivity; responses to flesh extracts and chemicals, light, and sound; sound production. Thiel, Hans. 1938. Thunf isch und Thunfischf ang. Fischmarkt 6. Fishing methods and gear, tuna. Tinker, Spencer w. 1944. Hawaiian fishes. Honolulu, Tongg Publ. Co., p. 154-160. tP] Auxis thazard, Euthynnus alletteratus, E. pelamis, Gerino alalunga, Neothunnus macropterus, Parathunnus sihi, Sema- thunnus itosibi, Thunnus orientalis, T. Tinker, Spencer W. — Continued thynnus: description, distribution; figures for all except S. itosibi; English, Hawai- ian, Japanese, and European common names. Tominaga, SeijirO. 1943. Bonitos. Kaiyo no kagaku 3(10) : 460-465. [J] Frigate mackerels, black skipjack: de- scriptions, distribution, habits, food, Jap- anese common names; migration and population analysis of Japanese skipjack stock. tomiyama, Tetsuo. 1933. The chemical composition of tunny liver oil. Bull. Jap. Soc. sci. Fish. 2(1) :l-7 [Je,P] Black tuna: chemical analysis. TOyama, YuzO, et al. 1941. Studies on fish in Japan as a source of insulin. Bull. Jap. Soc. sci. Fish. 10(4) : 153- 155. [P] Germo germo, Katsuwonus pelamis, Neo- thunnus macropterus, Parathunnus sibi: yields of insulin. TUBE, J. A. 1948. Whale sharks and devil rays in North Borneo. Copeia 1948 (3) :222. Euthynnus macropterus: recorded. UCHIDA, KEITARO. 1937. On flotation structures seen in planktonic larvae of fishes, I, II. Kagaku 7:540-546, 591-595. [J] Tunas: young. UCHIHASHI, KIYOSHI. 1953. Ecological study of Japanese teleosts in relation to the brain morphology. Bull. Japan Sea reg. Fish Res. Lab. 2:166 p. [Je,P] Morphology of brain of Katsuivonus pela- mis and Auxis tapeinosoma: described, figured, and discussed in relation to the habits and ecology of the fish. UDA, MICHITAKA. 1931a. A consideration of Zunan albacore fishing conditions and oceanographic conditions. Suisan butsuri danwakai kaiho 27:413-416. [J] Thunnus germo: fishing conditiona in re- lation to oceanography; Pacific Ocean — northwest. 1931b. Studies of skipjack fishing conditions north of Zunan in 1930. Suisan butsuri d£in- wakai kaiho 2. [J] Katsuwonus pelamis, Pacific Ocean — northwest. 1932a. On the body weight of some scombroid fishes of Japan. Bull. Jap. Soc. sci. Fish. 220 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Uda, Micmitaka. — Continued 1(3) :124-129. [Je,P] Thunnus thymms, Euthymms vat/ans : sea- sonal variations in size composition. 1932b. A study of the body weight of yellowtail, young black tuna, cybiids, and skipjack. Sui- san butsuri danwakal kaiho 35:610-619. [J] Thunnus orientalis, Katsuivonus pelamis: size composition. 1933. The shoals of "Katuwo" and their angling. Bull. Jap. Soc. sci. Fish 2(3) : 107-111. (Trans- lation in Spec. sci. Rep. : Fish. U. S. Fish Wildl. 83). [P] Euthynnus vagans: distribution, density, biting qualities of schools in Japanese waters in relation to salinity, stomach contents, and objects with which asso- ciated; catch per unit of effort in livebait fishery. 1935a. On the estimation of favorable tempera- ture for longline fishing of tunny. Bull. Jap. Soc. sci. Fish. 4(1) :61-65. (Pacific Oceanic Fishery Investigations Translation No. 44. In: Spec. sci. Rep.: Fish. U.S. Fish Wildl. 52). [P] Gernio macropterus, Parathtmnus sibi, Thunnus alalunga, T. thunnus: distribu- tion in relation to water temperature in Japanese waters. 1935b. The results of simultaneous oceanographi- cal investigations in the North Pacific Ocean adjacent to Japan made in August 1933. J. Imp. Fish. Exp. Sta. 6(43) :1-130. [Je,P] Euthynnus vayans: fishing conditions and distribution in relation to oceanographic conditions; size composition of schools and commercial catch; habits. 1935c. Skipjack schools congregating along cur- rent boundaries. Kagaku 5(12) :503-504. [J] Katsuwonus pelamis: fishing conditions in relation to oceanography. 1936a. Locality of fishing centre and shoals of "katuwo" Euthynnus vagans (Lesson) corre- lated with the contact zone of cold and warm currents. Bull. Jap. Soc. sci. Fish. 4(6): 385-390. [Je,P] Euthynnus vagans : occurrence and migra- tion in relation to water temperatures in Japanese waters; recoveries of tagged skipjack noted. 1936b. The winter albacore grounds and the sub- tropical convergence. Kagaku 6:416-417. [J] Gerrtio alalunga: fishing in relation to oceanographic conditions. 1938. Correlation of the catch of "Katuo" in the waters adjacent to Japan. Bull. Jap. Soc. sci. UdA;, Michitaka. — Continued Fish. 7(2) :75-78. [Je,P] Katsuwonus pelamis: amount and distri- bution of catch in Japanese waters in relation to annual variation in water temperature and amount of early season landings. 1939. On the characteristics of the frequency curve for the catch of "katuo" Euthynnus vagans (Lesson) referred to the water tem- perature. Bull. Jap. Soc. sci. Fish. 8(4) :169- 172. [Je,P] Catches correlated with water tempera- ture off Japan. 1940a. A note on the fisheries condition of "katuo" as a function of several oceanogra- phic factors. Bull. Jap. Soc. sci. Fish. 9(4): 145-148. (Pacific Oceanic Fishery Investiga- tions Translation No. 45. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 51 ) . [P] Euthymms vagans: occurrence of fish in relation to oceanographic factors; catch of E. vagans correlated with temperature, salinity, transparency and food supply of sardines and copepods. 1940b. On the recent anomalous hydrographical conditions of the Kurosio in the south waters off Japan proper in relation to the fisheries. J. Imp. Fish. Exp. Sta. 10 ( 74 ) : 231-278. [ Je,P] Albacore, black tuna, skipjack: fishing conditions in relation to movements of water masses. 1940c. The time and duration of angling and the catch of "katuo," Euthynnus vagans (Lesson). Bull. Jap. Soc. sci. Fish. 9(3) : 103-106. (Pacific Oceanic Fishery Investigations Trans- lation No. 46. In: Spec. sci. Rep.: Fish. U.S. Fish Wildl. 51). [P] Diurnal feeding activity and length of fishing time per school as they affect the catch. 1941. The body temperature and bodily features of "katuo" and "sanma." Bull. Jap. Soc. sci. Fish. 9(6) :231-236. (Pacific Oceanic Fishery Investigations Translation No. 51. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 51). [P] Body temperatures of skipjack correlated with water temperature; proportional body measurements analyzed. 1948. On the oceanographic conditions of the summer of 1942 and the abnormal skipjack fishing conditions, oyo kisho 3(1):? [J] Katsuwonus pelamis : fishing conditions in relation to oceanography. 1952. On the relation between the variation of the important fisheries conditions and the BIBLIOGRAPHY ON THE TUNAS 221 Uda, Michitaka. — Continued oceanographical conditions in the adjacent waters of Japan 1. J. Tokyo Univ. Fish. 38(3) :363-389. [Je,P] Relation between fluctuations of Thutmus orie^italis abundance and oceanographic conditions in southern Japanese waters; relation between Neothunnus macropterus fishing grounds in the equatorial Pacific and hydrographic fluctuations. Uda, Michitaka, and Eimatsu Tokunaga. 1937. Fishing of Geinio germo (LacepSde) in re- lation to the hydrography in the North Pacific waters (Rep. 1). Bull. Jap. Soc. sci. Fish. 5(5) :295-300. [Je,P] Annual variations in catch of summer and winter fisheries related to oceanic circu- lation and water temperatures; hypothe- tical relationships of stocks on Japanese and mid-Pacific grounds. Uda, Michitaka, and JiRo Tsukushi. 1934. Local variations in the composition of vari- ous shoals of "Katuwo," Euthynnus vagans (Lesson), in several sea-districts of Japan. Bull. Jap. Soc. sci. Fish. 3(4) : 196-202. (Trans- lation in Spec. sci. Rep.: Fish. U. S. Fish Wildl. 83). [P] Size composition of the catch on various Japanese grounds; association of schools of different character with different ob- jects and organisms; definition of popu- lations in Japanese fishery. Uda, Michitaka, and N. Watamabe. 1938. Autumnal fishing of skipper and bonito influenced by the rapid hydrographical change after the pass of cyclones. Bull. Jap. Soc. sci. Fish. 6(5) :240-242. [Je,P] Effects of cyclones on water temperature and indirectly on location of skipjack fish- ing grounds in Japanese waters. Uehara, TOKUZO. 1941. A survey of tuna grounds in equatorial waters. Nanyo suisan joho 5(3) : 13-17. (Pa- cific Oceanic Fishery Investigations Transla- tion No. 14: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 45). [P] Bigeye tuna, skipjack, yellowfin tuna: catches and catch rates at 6 longlining stations off N. Guinea and Halmahera; water temperatures and currents recorded. U. S. Fish and Wildlife Service, BRANCH OF C0MMERCI.4L FISHERIES. Quarterly outlook for marketing fishery products. (This appears as Fish. Leaf 1., U.S. Fish Wildl., Wash. 336 and began with April 1949). Gives statistics on the general fishery situation, and also on specific marketing situations, such as for tunas. u. s. fish and wildlife service, Branch of Commercial fisheries. Market News Service. Fishery Products Report, San Pedro, Calif. [P] Fishery Products Report, Seattle, Washington. [P] Daily mimeographed reports on various fish- eries, including tuna. Monthly summary — Fishery products. San Pedro, Calif. [P] California cannery receipts and pack of tuna, mackerel and anchovies; U. S. imports into Arizona and California, and other information. U. S. FISH AND Wildlife Service, Branch of Commercial Fisheries, Market News Service, San Pedro, Calif. California fisheries. (Annually). [P] California receipts, comparison by year — also by month comparing 3-year period — of tuna and tuna-like fishes: albacore, skipjack, yel- lowfin, bonito, yellowtail in tons; imports of frozen tuna by tons; California ex-vessel prices for tuna and tuna-like fishes in dollars per ton; canned tuna prices; California pack of tuna and tuna-like fishes with comparative data. UNO, Michio. 1936a. Germo germo (Lacepede) in the waters east of Nozima promontory, Tiba Prefecture (Prelim. Rep. No. 1). Bull. Jap. Soc. sci. Fish. 4(5) :307-309. [Je,P] Size and age composition fby vertebral annuli); length-weight relationship of pole-and-line-caught fish from Japanese waters. 1936b. Germo germo (Lacepede) in the waters east of Nozima promontory, Tiba Prefecture (Prelim. Rep. No. 2). Bull. Jap. Soc. sci. Fish 5(4) :235. [Je,P] Size and age composition (by vertebral annuli); length-weight relationship of pole-and-line-caught fish from Japanese waters. Van Campen, W. G. 1952. Japanese mothership-type tuna fishing operations in the western equatorial Pacific, June-October 1951 (Report on the seventh, eighth, and ninth expeditions). Comm. Fish. Rev. U. S. Fish Wildl. 14(11) : 1-9. [P] Catch rates for each expedition given: Neothunnus macropterus, Parathunnv^ sibi, Thunnus germo, Thunnus orientalis. Katsuwonus pelamis. 1953. Tuna fishing at Tahiti. Comm. Fish. Rev. 15(10) :l-4. Also Separate No. 358. [P] Methods of capture; description of indus- try. 222 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE VAN Campen, W. G., and B. M. Shimada. 1951. The Japanese albacore fishery of the north central Pacific. Fish. Leafl., U. S. Fish. Wildl., Wash. 388. [P] Charts of area covered by Japanese alba- core explorations prior to the war; Japan- ese longline gear used in albacore fishing. Van Cleave, Harley J. 1940. The Acanthocephala collected by the Allan Hancock Expedition, 1934. A. Hancock Pacif. Exped. 2(15) :512. Euthynnus alletteratus and Katsuwonua pelamis listed as hosts. VlTLOV, V. 1949. Ribolov na moru (Razvoj tehnike tunolova kod nas). Ribarski kalendar: 120-123. Fishing gear and methods, Adriatic Sea. WADE, Charles B. 1949. Notes on the Philippine frigate mackerels, family Thunnidae, Genus Auxis. Fish. Bull., U. S. Fish Wildl. 51(46) :229-240. [P] Auxis tapeinosoma, A. thazard: classifi- cation, description, key, synonymy, distri- bution, counts of meristic characters, records and descriptions and figures of juveniles. 1950a. Juvenile forms of Neothunnus macrop- terus, Katsuwonus pelamis, and Euthynnus yaito from Philippine seas. Fish. Bull., U. S. Fish Wildl. 51(53) : 395-404. [P] Description, figures. 1950b. Observations on the spavifning of Philip- pine tuna. Fish. Bull., U. S. 51(55) : 409-423. [P] Classification of gonads; Euthynnus yaito, Katsuwonus pelamis, Neothunnus ma- cropterus: tables showing locality, length, sex, degree of maturity, and gonad weight. 1951. Larvae of tuna and tuna-like fishes from Philippine waters. Fish. Bull., U. S. Fish Wildl. 51(57) : 445-485. [P] Five genera, embracing four known spe- cies, of larvae previously unknown in the western Pacific are described and illus- trated: Grammatorcynus bicarinatus, Neo- thunnus niacropterus, Katsuwonus pela- mis, Euthynnus yaito, and Au,xis sp.; dis- tribution and abundance of larvae, spawn- ing areas, diurnal vertical migration. Walfoed, Lionel a. 1931. Handbook of common commercial and game fishes of California. Fish. Bull., Sacramento 28:74-77. [P] Germo alalunga, Katsuwonus pelamis, Neothunnus 'niacropterus, Thunnys thyn- nus: classification, English common names, description, figures, key. Walfoed, Lionel a. — Continued 1937. Marine game fishes of the Pacific coast — Alaska to the Equator. Berkeley, Univ. of California Press, p. 1-20, 20-30. Euthynnus lineatus, Germo alalunga, Kat- suwonus pelamis, Neothunnus macrop- terus, Thunnus thynnus: description, dis- tribution, food, spawning, key, English common names, figures. Auxis thazard: description, distribution, English common names, figure, key. Migration of G. ala- lunga, K. pelamis, and N. macropterus. N. macropterus compared with Allison's tuna. Wakfel, H. E. 1950. Outlook for development of a tuna industry in the Philippines. Res. Rep. U. S. Fish. Wildl., 28:37 p. [P] History of Philippine tuna fishery; recent exploration for tuna (live bait fishing ex- periments, longline trawl experiments, trolling experiments, traps). Descriptions of Neothunnus macropterus, Katsuwonus pelamis, Euthynnus yaito, Auxis tlutzard, Grammatorcynus bicarinatus, Gymnosar- da nuda, Sarda orientalis. For each species mentioned: distribution, figure, and de- scription. Watanabe, Hajime. 1939. Investigation of albacore. Shizuoka-ken suisan shikenjo jigyo hokoku (1936-38) 1939: 22-23. [J,P] Albacore: mid-Pacific; stomach contents, sexual maturity, body proportions of males and females compared, morphome- tric data, probable spawning season, area, number of eggs. Watanabe, Haruo. 1940. Fishing conditions south of the Marshall Islands. Nanyo suisan 58, March 25, 1940; No. 59, April 25, 1940; No. 60, May 25, 1940. (Pacific Oceanic Fishery Investigations Trans- lation No. 11. In: Spec. sci. Rep.: Fish. U. S. Fish Wildl. 43). [P] Narrative report of an exploratory long- line fishing cruise from Jaluit to the Solo- mons and back. Some oceanographic in- formation and lengths and weights of the fish taken. Bigeye tuna, yellowfin tuna. Watanabe, Nobuo. 1941. Measurements on the bodily density, body temperature and swimming-velocity of "Katu- wo," Euthynnus vagans (Lesson). Bull. Jap. Soc. sci. Fish. 11(4) :146-148. [J,P] Body volume and weight for three speci- mens, temperature immediately after cap- ture for 10, and swimming speeds of 10 as timed from a fishing boat; body tem- peratures compared with water tempera- tures. J BIBLIOGRAPHY ON THE TXnSTAS 223 Welsh, James P. 1949a. A preliminary study of food and feeding habits of Hawaiian kavvakawa, mahimahi, ono, aku, and ahi. Fish. Progr. Rep. 1(2) :26 p. [P] Quantitative analyses of stomach contents of Euthynnus yaito, Neothiinnus macrop- terus, and Katsuwonus pelamis from Hawaiian waters; food organisms listed, numbers and volumes given for each. 1949b. Range extension of the file fish Monocan- thus Dielanocephalus. Pacif. Sci. 3(1) :100. [P] Hawaii: specimens recovered from Eu- thynnus yaito. 1949c. A trolling survey of Hawaiian waters. Fish. Res. Progr. Rep. 1(4) : 30 p. [P] Trolling catch per unit of effort for Euthynnus yaito, Neothunnus macrop- terus, and Kaisuw&jius pelamis; fishing areas and lures compared. Westman, J. R., and P. W. Gilbert. 1941. Notes on age determination and growth of the Atlantic bluefin tuna, Thunnus thynnus L. Copeia 1941:70-72. Thunnus thynnus: scale reading, age- length relationship; measurements taken from 100 tuna off Long Island in 1938. Westman, James R., and Willlam c. Neville. 1942. The tuna fishery of Long Island. New York, Board of Supervisors, Nassau County. 30 pp. Thunnus thynnus: scale reading, age- fishing methods described — trolling, hand- lining; catch per unit of effort; lengrth frequency; length-weight relationship; age and growth; scale reading, maturity, tagging; Atlantic. Whitehead, S. S. 1930. California bluefin tima. Calif. Fish Game 16(3) : 231-233. Thunnus thynnus: distribution. 1931. Fishing methods for the bluefin tuna (.Thunnus thynnus) and an analysis of the catches. Fish. Bull. Sacramento 33:32 p. [P] Classification, distribution, figure, migra- tion, spawning, catch per unit of effort. Whitley, Gilbert P. 1937. The Middleton and Elizabeth Reefs, South Pacific Ocean. Aust. Zool. 8(4) :229-231. Wanderer wallisi proposed as new genus and species; synonymy, description, food; compared with E. yaito and E. allettera- tus. 1947. New sharks and fishes from Western Australia. Aust. Zool. 11(2) : 129-150. Whitley, Gilbert P. — Continued Euthynnus alletteratus, Katsuwonus pela- mis, Kishinoella tonggol, Neothunnus ma- cropterus, Thunnus maccoyi: recorded, Australian common names. Wilson, Robert C. 1953. Tuna marking, a progress report. Calif. Fish Game 39(4) : 429-442. [P] History of tagging of tuna, present tag- ging methods, tag application methods. Figures of various types of tags. WoLi-E Murray, D. K. 1932. Tunny (Thunmis thynnus L.) in the North Sea. J. Cons. int. Explor. Mer 7(2) : 251-254. [P] Some observations made by the author during his voyages as to occurrence, habits, and food of T. thynnus. A table is given showing the first and last ap- pearances in the seasons of 1923-31. YABE, HIROSHI. 1953. Juveniles collected from South Seas by Tenyo Maru at her second tuna research voy- age (preliminary report). Contrib. Nankal reg. Fish. Res. Lab. 1, Contrib. 25:14 p. [J,P] Surface trawl and night-light collections in the Carolines area include 2 Katsu- loonus pelamis and 4 unidentified scom- briform larvae; description, measure- ments, and figure of a 5.2-mm. Katsu- wonus pelamis. Yabe, Hiroshi, Noboru Anraku, and Tokumi mori. 1953. Scombroid youngs found in the coastal seas of Aburatsu, Kyushu, in summer. Contrib. Nankai reg. Fish. Res. Lab. 1, Contrib. 11 :10 p. [J,P] Young Katsuivonus pelamis, Euthynnus yaito, Auxis tapeinosoma, TMmnus orien- talis, Neothunnus macropterus, Sarda orientalis, and Auxis hira taken by various gear in coastal waters; fishing conditions correlated with oceanography; habits, food, size composition, morphometries. Yabe, Hiroshi, and Tokumi Mori. 1948. Report of skipjack investigations for 1947. Cent. Fish. Exp. Sta. Rep. 30. [P] Ryukyu Islands: length -weight data, stomach contents, catch correlated with water temperature; maturity of gonads, age analysis, spawning season, estimated number of eggs, past records of Juveniles captured. 1950. An observation on the habit of bonito, Katsuwonus vagans, and yellowfin, Neothun- nus macropterus, school under the drifting timber on the surface of ocean. Bull. Jap. Soc. sci. Fish. 16(2):35-39. [Je.P] 224 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE YABE, HiROSHi, and TOKUMI MORI. — Continued Katsuwonus pelamis, Neothiinnus macrop- terus: ecologry; live-bait fishing; yellowfin length frequencies. Yabuta, YOichi. 1953. On the stomach contents of the tuna and marlin from the adjacent seas of Bonin Islands. Contrib. Nankai reg. Fish. Lab. 1, Contrib. 15:6 p. [J,P] Lists and gives numbers of organisms found in 76 longline-caught albacore, big- eye, and yellowfin tuna stomachs (no volumes); describes food organisms; dis- cusses seasonal differences in stomach contents, relation to oceanographic condi- tions; young skipjack as food. Yabuta, YOichi and Shoji ueyanagi. 1953a. The distributions of tunas in the equatorial region. I. Contrib. Nankai reg. Fish. Res. Lab. 1, Contrib. 28:8 p. [Je,P] Distribution and migration of yellowfin, bigeye, and black marlin in southern Mar- shalls area as shown by longline catch rates of mothership fleet; seasonal changes in size composition of yellowfin; fishing conditions correlated with oceano- graphic conditions. 1953b. The distribution of tunas in the equatorial region. II. Hooked-rate of yellowfin tuna. Contrib. Nankai reg. Fish. Res. Lab. 1, Con- trib. 29:6 p. [J,P] Longline catch rates of yellowfin tuna from mothership fleet operations in south- em Carolines waters related to locality, hydrography, time, and size composition. Yamamoto,, Shigeo. 1933. Points of information for the skipjack fish- Yamamoto, Shigeo. — Continued ery gained from the study of fish's eyes. Rakusui 28(11) : 927-930. [J] Skipjack: anatomy. 1934. Points of information for the skipjack fish- ery gained from the study of fish's eyes. Rigakkai 32(1):28. [J] Skipjack: anatomy. YAMAMOTOj Shokichi. 1940. Views on increasing the commercial value of dried fish sticks from the South Seas. Nanyo suisan 3(11) : 21-35. [J,P] Skipjack: Japan, Formosa, South Seas; proportional weights of various body parts. yamanakAj Ichiro. 1950. On the size composition of skipjack in the Northeastern Sea area. Nippon kaiyogakkai Shi 5(214). [J] Yonezavva, Matsunosuke. 1950. Skipjack fishing experiences. Kaiyo no kagaku 6(1) :47-49. [J] Skipjack, Japanese waters. Yoshihara, Tomokichi. 1951-52. Distribution of fishes caught by the long- line. I. Horizontal distribution. II. Vertical distribution. III. Determination of the swim- ming depth. Bull. Jap. Soc. sci. Fish. 16(8) : 367-369; 16(8) :370-374; 18(5) :187-190. [Je,P] Statistical study comparing catch rates on different parts of longline sets; rela- tive swimming depths of Germo germo, Parathunnus sibi, and spearfishes deduced from estimated depths of hooks on which captured. Zei, M. 1948. zivot naseg Jadrana. 220 p. Tuna, Adriatic Sea. INDEX BY SUBJECTS Adriatic Sea Fortunie, 1930. Hadzi, 1934. Hirtz, 1933. Mili6, 1937. Morovie, 1950. Scordia, 1939b. soljan, 1930. Vitlov, 1949. Zei, 1948. Age Brock, 1943. Conseil International pour 1' Exploration de la Mer, 1933. Heldt, 1950. Higashi, 1941. Ikebe, 1939a, 1939b, 1940a, 1940b, 1940c, 1941a, 1941b. Kanamura and Yazaki, 1940a, 1940b. Kawasaki, 1952. Kimura, 1935, 1941, 1942a. LeGall, 1949. Lozano, 1950. Mine and lehisa, 1940. Moore, 1951a, 1951b. Nakamura, Kamimura, and Yabuta, 1953. Partlo, 1950. Sasaki, 1939b. Schaefer, 1948b, 1951 Scordia, 1943. Sella, 1952. Uno, 1936a, 1936b. Westman and Gilbert, 1941. Westman and Neville, 1942. Aku. See Katsuwonus. Albacore. See Germo. Allison's tuna. See Neothunnus itosibi. AUothunnus Fraser-Brunner, 1950. Serventy, 1948. Anatomy Berg, 1947. Chabanaud, 1930. Conrad, 1937. Eckles, 1949. 1 . Frade, 1930a, 1930b, 1931. Frade and DeBuen, 1932. Godsil and Byers, 1944. Greenbood, 1952. Higashi, 1941b. Imamura, 1949. Kafuku, 1950. LeGall, 1949. Letaconnoux, 1950. Nakamura, 1935, 1949. Anatomy — Continued Poisson and Postel, 1951. Priol, 1944. Rivas, 1953. Roedel, 1948a. Schaefer and Marr, 1948a. Serventy, 1941a, 1942b, 1948. Shimada, 1954. Sueyhiro, 1936, 1938, 1941, 1942, 1950. Uchihashi, 1953. Yamamoto, 1933, 1934. As food of tunas Asano, 1939. Marukawa, 1939b. Suda, 1953. Yabuta, 1953. Atlantic Ocean Alaejos, 1931. Beebe, 1936. Beebe and Tee-Van, 1936. BelWn and Bardan de BellOn, 1949. Bigelow and Schroeder, 1953. Bini, 1931. Bouxin and Legendre, 1936. Carlson, 1951. Chilton, 1949. Crane, 1936. DeBuen, 1930, 1931, 1932, 1935, 1937. Ehrenbaum, 1934. Farina, 1931. Ferreira, 1932. Frade, 1930, 1931a, 1931b, 1937. Godsil and Holmberg, 1950. Heldt, 1931, 1932a, 1932b. LeDanois, 1933, 1938, 1951. LeGall, 1934a, 1934b, 1934c, 1934d, 1949. Legendre, 1932, 1933, 1934, 1937, 1940. Letaconnoux, 1950. Lozano, 1950. Mather, 1954. Molteno, 1948. Morice, 1953b. Mowbray, 1935. Murray, 1952. Navarro and Lozano, 1950. Navaz, 1950. Postel, 1949, 1950. Priol, 1944. Rivas, 1951, 1953. Russell, F. S., 1934b. Schaefer and Walford, 1950. Schuck, 1951. Schuck and Mather, 1951. Schultz, 1949. Sella, 1930. 225- 226 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Atlantic Ocean — Continued Westman and Gilbert, 1941. Westman and Neville, 1942. Australian waters Robins, 1952. Serventy, 1941a, 1941b, 1942a, 1942b, 1947, 1948. Auocis Asano, 1939. Bamhart, 1936. Brock, 1949. Chu, 1931. DeBeaufort and Chapman, 1951. DeBuen, 1930, 1932, 1935. Domantay, 1940. Fish, 1948. Fisheries Society of Japan, 1931. Food and Agr. Organ., U. N., 1949b. Fowler, 1936, 1938, 1944, 1949. Graham, 1938. Herald, 1951. Herre, 1940, 1943. Herre and Umali, 1948. Imamura, 1949. Joubin, 1934. LeGall, 1934a. Marukawa, 1939b. Mead, 1951. Molteno, 1948. Morice, 1953a. Nakamura, 1939a, 1939b. Navarro and Lozano, 1950. Navaz, 1950. Okada and Matsubara, 1938. Rivas, 1951, 1953. Sanzo, 1932. Schaefer, 1948c, 1951. Schaef er and Marr, 1948a. Serventy, 1941a. Shapiro, 1948a. Soldatov and Lindberg, 1930. Tanaka, 1931. Taranetz, 1937. Tester, 1952. Tinker, 1944. Tominaga, 1943. Uchihashi, 1953. Wade, 1949, 1951. Walford, 1937. Warfel, 1950. Yabe et al., 1953. Aiixis hira Imamura, 1949. Nakamura, 1939b. Okada and Matsubara, 1938. Shapiro, 1948a. Soldatov and Lindberg, 1930. Tanaka, 1931. Taranetz, 1937. Yabe et al., 1953. Auxis inaru Imamura, 1949. Nakamura, 1939a, 1939b. Soldatov and Lindberg, 1930. Taranetz, 1937. Auxis tapeinosoma Okada and Matsubara, 1938. Okada et al., 1935. Shapiro, 1948a. Uchihashi, 1953. Wade, 1949. Yabe et al., 1953. Auxis thasard Bamhart, 1936. Brock, 1949. DeBeaufort and Chapman, 1951. DeBuen, 1930, 1932, 1935. Domantay, 1940. Fish, 1948. Fisheries Society of Japan, 1931. Food and Agr. Organ., U. N., 1949b. Fowler, 1936, 1938, 1944, 1949. Fraser-Brunner, 1950. Graham, 1938. Herald, 1951. Herre, 1940. Herre and Umali, 1948. Joubin, 1934. LeGall, 1934a. Marukawa, 1939b. Mead, 1951. Molteno, 1948. Morice, 1953a. Navarro and Lozano, 1950. Navaz, 1950. Rivas, 1951. Schaefer, 1951. Schaefer and Marr, 1948a. Serventy, 1941a. Tester, 1952. Tinker, 1944. Wade, 1949. Walford, 1937. Warfel, 1950. Baltic Sea Ros^n, 1943. Behavior Anonymous, 1953b. Bigelow and Schroeder, 1953. Hiatt and Brock, 1948. Imamura, 1949. Kida, 1936. Kimura, Iwashita, and Hattori, 1952. LUling, 1952a. Murphy and Niska, 1953. Nakamura, 1949. Russell, F. S. 1934b, 1936. Schaefer, 1948b. Serventy, 1942a. BIBLIOGRAPHY ON THE TUNAS 227 Behavior — Continued Tanaka, 1935. Tester, 1952. Tester et al., 1952. Tominaga, 1943. Uda, 1940c. Uda and Tsukushi, 1934. Watanabe, 1941. Yabe and Mori, 1950. Yoshihara, 1951, 1952. Bibliography Bini, 1952. Corwin, 1930. Heldt, 1930, 1931, 1932, 1934. LeGall, 1949. Legendre, 1934. Morovie, 1950. Nakamura, 1949. Navaz, 1950. Okada and Matsubara, 1953. Rosa, 1950. Shapiro, 1948a. Shimada, 1951a. Bigeye tuna. See Parathunnus spp. Black tuna. See Thunnus orientalis. Blackfin tuna. See Parathunnus atlanticits. Bluefin tuna. See Thunntts thynnus. Body condition Aikawa and Kato, 1938. Ikebe and Matsumoto, 1937. Kanamura and Yazaki, 1940a, 1940b. Kawasaki, 1952. Mizushima et al., 1951. Morice, 1953a. Onodera, 1941. Schweigger, 1949. South Seas Gov't., 1941c. Suyehiro, 1936. Body temperature Kanamura and Imaizumi, 1936a. Kanamura and Yazaki, 1940a, 1940b. Nakamura, 1941. oita Pref . Fish. Expt. Sta., 1930. Scagel, 1949. Society for the Promotion . . . 1936. Uda, 1941. Watanabe, N., 1941. Bonito. See Katsuwonus pelamis. Caribbean Sea LeDanois, 1951. Rawlings, 1953. Catch per unit of effort Bates, 1950. Chiba Pref. Fish. Expt. Sta., Katsuura Br., 1938, 1941. Formosa Gov't.-Gen. Fish. Expt. Sta., 1933. Heldt, 1930, 1932. Hiratsuka and Imaizumi, 1934. Hiratsuka and Ito, 1934. Catch per unit of effort — Continued Ikebe, 1940, 1941, 1942. Imaizumi, 1937. Inoue, 1953. Jap. Bur. Fish., 1939, 1940. Kanagawa Pref. Fish. Expt. Sta. 1951b. Kanamura and Imaizumi, 1936a. Kanamura and Yazaki, 1940a, 1940b. Kimura, 1942a. Marukawa, 1939c. Mine and lehisa, 1940. Murphy and Shomura, 1953b. Nakamura, 1949. Nomura et al., 1952-3. Okinawa Pref. Fish. Expt. Sta., 1936b. Okuma et al., 1935. Serventy, 1947. Sette, 1954. Tester, 1952. Van Campen, 1952. Westman and Neville, 1942. Whitehead, 1931. Yabuta and Ueyanagi, 1953. Chemical analysis Asaka, Noguchi, and Mimoto, 1953. Dontcheff and Legendre, 1938. Higashi and Hirai, 1948. Horiguchi, Kakimoto, and Kashiwada, 1950. Horiguichi, Kashiwada, and Kakimoto, 1953. Kakimoto, Kanazawa, and Kashiwada, 1953. Kashiwada, Kakimoto, and Horiguchi, 1952. Kashiwada, Kakimoto, and Yamasaki, 1953. Kodama, lizuka, and Harada, 1934. Maldura, 1946. Matsui, K., 1942b. Migita and Arakawa, 1948. Miyama and Osfikabe, 1938, 1940. Miyama, Saruya, and Hasegawa, 1939. Mizushima et al., 1951. Niwa, 1937. 6ya and Takahashi, 1936. Shimizu, 1947. Tomiyama, T., 1933. Tomiyama, Y., et al., 1941. Classification Berg, 1947. DeBuen and Frade, 1932. Dieuzeide, 1930. Frade and DeBuen, 1932. Fraser-Brunner, 1949, 1950. Godsil and Byers, 1944. Herre, 1953. Nakamura, 1939b, 1939c, 1943, 1949. Nichols and LaMonte, 1941. Okada and Matsubara, 1938. Roedel, 1948a. Serventy, 1948. Shapiro, 1948b. Soldatov and Lindberg, 1930. Taranetz, 1937. 228 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Classification — Continued Wade, 1949. Walford, 1931. Whitehead, 1931. Color, water. See Oceanographic conditions. Common names Ancieta C, 1952. Bamhart, 1936. Chevey, 1932. DeBuen, 1930. Delsman and Hardenburg, 1934. Fish, 1948. Food and Agr. Organ. U. N., 1949b. Ikebe and Matsumoto, 1938. LeGall, 1949. Marukawa, 1939a. Nakamura, 1939b, 1943. Navaz, 1950. Nichols and LaMonte, 1941. Okada and Matsubara, 1938. Roedel, 1948a. Rosa, 1950. Schultz, 1949. Serventy, 1941a. Shapiro, 1948a, 1948b. Smith, 1947. Tinker, 1944. Tominaga, 1943. Walford, 1931, 1937. Whitley, 1947. Condition, body. See Body condition. Contents, stomach. See Food. Currents. See Oceanographic conditions. Description Ancieta C, 1952. Barnard, 1948. Bamhart, 1936. Beebe and Tee-Van, 1936. Bigelow and Schroeder, 1953. Boeseman, 1947. Brock, 1949. Chabanaud, 1930. Chevey, 1932. Chilton, 1949. Clemens and Wilby, 1946. Crane, 1936. DeBeaufort and Chapman, 1951. DeBuen, 1932. Delsman and Hardenburg, 1934. Eckles, 1949. Fish, 1948. Fisheries Society of Japan, 1931. Fowler, 1933, 1936, 1938. Frade, 1931. Fraser-Brunner, 1950. Godsil and Byers, 1944. Graham, 1938. Heldt, 1931, 1932. Hildebrand, 1946. Description — Continued Ikebe and Matsumoto, 1938. Imamura, 1949. Joubin, 1934. June, 1952b. Kanamura and Yazaki, 1940a. LeGall, 1934a, 1934b, 1934c, 1934d. Marukawa, 1939a, 1939c. Molteno, 1948. Mowbray, 1935. Nakamura, 1939b, 1949. Nichols and LaMonte, 1941. Okada sind Matsubara, 1938. Okada et al., 1935. Poisson and Postel, 1951. Powell, A. W. B., 1937. Roedel, 1948a. Schaefer and Marr, 1948a, 1948b. Schweigger, 1943. Seale, 1940. Serventy, 1941a, 1941b, 1942b, 1948. Shapiro, 1948b. Soldatov and Lindberg, 1930. Tinker, 1944. Wade, 1949, 1950a. Walford, 1931, 1937. Whitley, 1937. Distribution Ancieta C, 1952. Anonymous, 1941b, 1953a, 1953b. Bahr, 1952. Bamhart, 1936. Bigelow and Schroeder, 1953. Bini, 1952. Brock, 1939. Carlson, 1951. Chapman, 1946. Chevey, 1932a, 1932b. Chu, 1931. Clemens and Wilby, 1946. Cowan, 1938. DeBuen, 1930. Delsman, 1933. Delsman and Hardenburg, 1934. Fish, 1948. Fitch, 1953. Food and Agr. Orgsui. U. N., 1949a, 1949b. Formosa Gov't.-Gen. Fish. Expt. Sta., 1933. Fowler, 1938, 1944. Fraser-Brunner, 1949, 1950. Godsil, 1949. Godsil and Greenhood, 1948. Hasegawa, 1937. Heldt, 1931a, 1931b, 1932. Herre, 1932, 1933, 1935, 1936, 1940 Hildebrand, 1946. Hiratsuka and Imaizumi, 1934. Hiratsuka and Ito, 1934. Imamura, 1949. Isawa, 1935. BIBLIOGRAPHY ON THE TUNAS 229 Distribution — Continued Jap. Bur. Fish., 1940. Joubin, 1934. Kanamura and Yazaki, 1940a, 1940b. Kimura, 1941, 1942b. Kumata, 1941. LeDanois, 1933, 1938. LeGall, 1934b, 1934c, 1934d. Leterdre, 1937. Maidura, 1946. Martin, 1938. Mather, 1954. Molteno, 1948. Murphy and Shomura, ^953a, 1953b. Nakamura, 1939b, 1943. 1949, 1951. Navarro and Lozano, 1950. Okada and Matsubara, 1938. Okada et al., 1935. Okinawa Pref. Fish. Expt. Sta., 1943. 6kuma et al., 1935. Powell, D., 1950. Powell and Hildebrand, 1950. Powell et al., 1952. Robins, 1952. Roedel, 1948a, 1948b. Rosa, 1950. Royce, 1953. Russell, F. S., 1933a, 1933b, 1934b. Samson, 1940. Schaefer, 1948c. Schuck and Mather, 1951. Schultz and DeLacy, 1936. Schweigger, 1943, 1949. Sella, 1930. Serventy, 1941a, 1942a, 1942b, 1948. Sette, 1954. Shapiro, 1948a, 1948b. Shimada, 1954. Smith and Schaefer, 1949. Soldatov and Lindberg, 1930. South Seas Gov't. . . . 1937a, 1937b, 1941a. Tanaka, 1931. Taranetz, 1937. Tinker, 1944. Uda, 1935a. Wade, 1949. Walford, 1937. Whitehead, 1930, 1931. Wolfe Murray, 1932. Yabuta and Ueyanag:!, 1953. Ecology (other than oceanog:raphic conditions or food) DeBeaufort and Chapman, 1951. Lozano, 1950. Marukawa, 1939a. Nakamura, 1949. Powell, D. E., 1950. Powell and Hildebrand, 1950. Powell et al., 1952. Priol, 1944. Ecology — Continued Uchihashi, 1953. Yabe and Mori, 1950. Eggs DeJong, 1940. Delsman, 1931. Delsman and Hardenburg, 1934. Hatai et al., 1941. Kikawa, 1953. Marr, 1948. Nakamura, 1938. Sanzo, 1932, 1933. Watanabe, Hajime, 1939. Yabe and Mori, 1948. Euthynnus Bigelow and Schroeder, 1953. Boeseman, 1947. Bonham, 1946. Brock, 1949. Carlson, 1951. Chabanaud, 1930. Chapman, 1946. Chevey, 1932a, 1932b, 1934. Chiba Pref. Fish. Expt. Sta., 1936. Chilton, 1949. Conrad, 1937. DeBeaufort and Chapman, 1951. DeBuen, 1930, 1935. DeJong, 1940. Delsman, 1931. Delsman and Hardenburg, 1934. Domantay, 1940. Dung and Royce, 1953. Eckles, 1949. Fish, 1948. Fisheries Society of Japan, 1931. Fitch, 1953. Food and Agri. Organ. U. N., 1949b. Fowler, 1931, 1938, 1944, 1949. Frade, 1932. Fraser-Brunner, 1949, 1950. Fukuda and lizuka, 1940. Godsil and Greenhood, 1948. Hatai et al., 1941. Herald, 1949. Herre, 1932, 1933, 1940. Herre and Umali, 1948. Hiatt and Brock, 1948. Higashi, 1940. Hildebrand, 1946. Imamura, 1949. Joubin, 1934. Kagoshima Pref. Fish. Expt. Sta., 1930c. LeGall, 1934b. Maidura, 1946. Mather, 1954. Mead, 1951. Mie Pref. Fish. Expt. Sta., 1930c, 1930e. Molteno, 1948. Morice, 1953a. 230 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Euthynnus — Continued Morrow, 1954. Nakamura, 1939a, 1939b. Nigrelli and Stunkard, 1947. Okada and Matsubara, 1938. Okada et al., 1935. Poisson and Postel, 1951. Postel, 1950. Rivas, 1951, 1953. Roedel, 1948b. Ronquillo, 1953. Schaefer, 1948c, 1951. Schmidt, 1930. Schmitt and Schultz, 1940. Schuck, 1951. Scale, 1940. Serventy, 1941a. Shapiro, 1948a. Smith and Schaefer, 1949. Tanaka, 1931. Tester, 1952. Tester et al., 1952. Tinker, 1944. Tominaga, 1943. Tubb, 1948. Van Cleave, 1940. Wade, 1950a, 1950b, 1951. Walford, 1937. Warfel, 1950. Welsh, 1949a, 1949b. Whitley, 1937, 1947. Yabe, et al., 1953. Euthynnus af finis. See also E. cUletteratus. Dung and Royce, 1953. Fraser-Bnmner, 1949. Euthynnus alleterata. See E. alletteratus. Euthynnus alleteratu^. See E. alletteratus. Euthynnus alletteratus Bigelow and Schroder, 1953. Boeseman, 1947. Carlson, 1951. Chabanaud, 1930. Chapman, 1946. Chilton, 1949. DeBeaufort and Chapman, 1951. DeBuen, 1930, 1935. DeJong, 1940. Delsman, 1931. Delsman and Hardenburg, 1934. Food and Agr. Organ. U. N., 1949b. Fowler, 1931. Frade, 1932. Herre, 1932, 1940. Hildebrand, 1946. Joubin, 1934. LeGall, 1934b. Maldura, 1946. Manter, 1940. i ;."■ -■- Mather, 1954. ■^• Molteno, 1948. Euthynnus alletteratus — Continued Morice, 1953a. Nigrelli and Stunkard, 1947. Okada and Matsubara, 1938. Poisson and Postel, 1951. Postel, 1950. Rivas, 1951. Schmidt, 1930. Schmitt and Schultz, 1940. Schuck, 1951. Serventy, 1941a. Smith and Schaefer, 1949. Tanaka, 1931. Tinker, 1944. Van Cleave, 1940. Whitley, 1937, 1947. Euthynnus allitteratus. See E. alletteratus. Euthynnus lineatus Fowler, 1938, 1944. Fraser-Brunner, 1949. Mead, 1951. Roedel, 1948b. Schaefer, 1948c, 1951. Seale, 1940. Walford, 1937. Euthynnus pelamis. See Katsuwonus. Euthynnus pelamys. See Katsuwonus. Euthynnus vagans. See Katsuwonus. Euthynnus yaito Bonham, 1946. Brock, 1949. Chabanaud, 1930. Chevey, 1932a, 1932b, 1934. Chiba Pref. Fish. Expt. Sta., 1936. Domantay, 1940. Eckles, 1949. Fisheries Society of Japan, 1931. Fitch, 1953. Fraser-Brunner, 1949. Fukuda and lizuka, 1940. Godsil and Greenhood, 1948. Hatai et al., 1941. Herald, 1949. Herre, 1933, 1940. Herre and Umali, 1948. Hiatt and Brock, 1948. Higashi, 1940. Imamura, 1949. Kagoshima Pref. Fish. Expt. Sta., 1930c. Mie Pref. Fish. Expt. Sta., 1930c, 1930e. Nakamura, 1939a, 1939b. Okada and Matsubara, 1938. Okada et al., 1935. Ronquillo, 1953. Schaefer, 1948c. Shapiro, 1948a. Tester, 1952. Tester et al., 1952. Tominaga, 1943. Wade, 1950a, 1950b, 1951. ■\-:'rK^ bibliography: ON' THE TUNAS T, ■!;?>..;• 201 Suthynntts yaito — Continued Warfel, 1950. Welsh, 1949a, 1949b. WhiUey, 1937. Yabe et al., 1953. Euthymis alletteratus. See Euthynnus allettei-atus. Figures Anonymous, 1938. Barnard, 1948. Barnhart, 1936. Bigelow and Schroeder, 1953. Bini, 1952. Chevey, 1932. DeBuen, 1930. Delsman and Hardenburg, 1934. Eckles, 1949a, 1949b. Fisheries Society of Japan, 1931. > Fowler, 1944. Frade, 1931. Fraser-Brunner, 1949, 1950. Godsil and Byers, 1944. Heldt, 1931a, 1931b, 1932b. Joubin, 1934. LeGall, 1934a, 1934b, 1934c, 1934d, 1949. Morice, 1953a. Nakamura, 1939b, 1949. Okada et al., 1935. Powell, A. W. B., 1937. Schaefer and Marr, 1948a, 1948b. Serventy, 1941a, 1942. Smith, 1935. Smith and Schaefer, 1949. Suyehiro, 1936, 1942. Tinker, 1944. Tominaga, 1943. Wade, 1949. Walford, 1937. Whitehead, 1931. Fishing conditions correlated with area Schweigger, 1949. Fishing conditions correlated with season Clemens and Wilby, 1946. Kimura, 1933, 1942a. Kimura and Ishii, 1933. Lozano, 1950. Mine and lehisa, 1940. Nakamura, 1949. Navaz, 1950. Ros^n, 1943. Schaefers, 1952. Schweigger, 1949. Sette, 1954. Uda and Watanabe, 1938. Wade, 1950a. Walford, 1931. Fishing methods and gear (other than purse seining, longlining, and llvebait) Anonymous, 1937b, 1953b. Bates, 1950. Fishing methods and gear — Continued Bini, 1931, 1933. Carlson, 1951. Cleaver and Shimada, 1950. De La Tourrasse, 1951. Dieuzeide, 1931. Domantay, 1940. Farina, 1931. Ferreira, 1932. Fortunia, 1930. Godsil, 1938. Heldt, 1931a, 1932. Hirtz, 1933. Imamura, 1953. June, 1951b. Kimura, Iwashita, and Hattori, 1952. Kreutzer, 1951a, 1951b. Legendre, 1936. LUling, 1952a. Markukawa, 1939a, 1939c. Matsumoto, W., 1952. Meyer, 1951. Mili(5, 1937. Navarro and Lozano, 1950. Nishikawa, 1934. Postel, 1950. Powell, D. E., 1950. Powell and Hildebrand, 1950. Powell et al., 1952. Rawlings, 1953. Russell, F. S., 1934b. Schaefers, 1952, 1953. Schweigger, 1949. ,.'>;i--: Scordia, 1940. Sella, 1931. Sette, 1954. S" ' goljan, 1930. South Seas Gov't. . . . 1937. Tester, 1952. Thiel, 1938. Van Campen, 1953. Vitlov, 1949. Welsh, 1949c. Westman and Neville, 1942. Yabe et al., 1953. Food Anonymous, 1938. Asano, 1939. Ban, 1941. Beebe, 1936. Bouxin and Legendre, 1936. Carlson, 1951. Chapman, 1946. Clemens and Wilby, 1946. Crane, 1936. Delsman and Hardenburg, 1934. Eckles, 1949. Fitch, 1950. Formosa Gov't.-Gen. Fish. Expt. Sta., 1933. Hart and HoUister, 1947. ^c j ■ 232 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Food — Continued Hart et al., 1948. Hatai et al., 1941. Heldt, 1934. Herald, 1949. Hildebrand, 1946. Imai, 1950. Imamura, 1949. Iwate Pref. Fish. Expt. Sta., 1953. Jap. Bur. Fish., 1933, 1934, 1935, 1939, 1940. Kanagawa Pref. Fish. Expt. Sta., 1951a. Kiironuma, 1940. LeGall, 1949. Legendre, 1932, 1933, 1934, 1940. Lozano, 1950. McHugh, 1952. Marukawa, 1939b, 1939c. Miyama et al., 1939. Nakamura, 1936, 1943, 1949. okuma et al., 1935. Partlo, 1950. Powell, D. E., 1950. Powell and Hildebrand, 1950. Powell et al., 1952. Priol, 1944. Relntjes, 1952. Reintjes and King, 1953. Ronquillo, 1953. Scagel, 1949. Schaefers, 1952, 1953. Schweigger, 1949. Sella, 1952. Serventy, 1942a. Shapiro, 1948b. Suda, 1953. Suyehiro, 1938. Tominaga, 1943. Uda, 1940a. Walford, 1937. Watanabe, Hajime, 1939. Welsh, 1949a, 1949b. Whitley, 1937. Yabe et al., 1953. Yabe and Mori, 1948. Yabuta, 1953. Germo Aikawa, 1933. Aikawa and Kato, 1938. Alverson and Chenowith, 1951. Anonymous, 1938, 1953a. Arcidiacono, 1935. Asano, 1939. Bamhart, 1936. Belloc, 1935. Bini, 1952. Bouxin and Legendre, 1936. Brock, 1939, 1943, 1949. Chiba Pref. Fish. Expt. Sta., 1936a, 1936b. Chiba Pref. Fish. Expt Sta., Germo — Continued Katsuura Branch, 1937, 1938b, 1941a, 1941b, 1941c, 1941d, 1941e, 1941f. Clemens and Wilby, 1946. Conseil Int'l pour 1' Exploration de la Mer, 1933. Cowan, 1938. DeBuen, 1930, 1935. De La Tourrasse, 1951. Dontcheff and Legendre, 1938. Dung and Royce, 1953. Ferreira, 1932. Fish, 1948. Fisheries Society of Japan, 1931. Food and Agr. Organ. U. N., 1949b. Fowler, 1938, 1944. Frade, 1953. Fraser-Brunner, 1950. Ganssle and Clemens, 1953. Godsil, 1945, 1948, 1949a, 1949b. Godsil and Byers, 1944. Godsil and Greenhood, 1948. Hart and HoUister, 1947. Hart et al., 1948. Hasegawa, 1938. Heldt, 1950. Herre, 1940. Herre and Umali, 1948. Hildebrand, 1946. Ikebe, 1939a, 1940. Inanami. 1942. Inoue, 1953. Iwate Pref. Fish. Expt. Sta., 1953. Jap. Bur. Fish., 1939, 1940, 1942. Joubin, 1934. Kagoshima Pref. Fish. Expt. Sta., 1930b, 1930c, 1931b, 1932c, 1933b. Kanagawa Pref. Fish. Expt. Sta., 1952a, 1952b. Kanamura and Yazaki, 1940, 1940b. Kimura, 1942a, 1942b, 1949. Kimura, Iwaahita and Hattori, 1952. Kuronuma, 1940. LeDanois, 1933, 1938, 1951. LeGall, 1934c, 1949. Legendre, 1932, 1933, 1934, 1936, 1940. Letaconnoux, 1950. Lozano, 1950. McHugh, 1952. Maldura, 1946. Marukawa. 1939c. Mie Pref. Fish. Expt. Sta., 1930c, 1930e. Mizushima et al., 1951. Molteno, 1948. Morice, 1953a, 1953b. Murphy and Shomura, 1953a, 1953b. Nakamura, 1939b, 1949, 1951. Nakamura et al., 1953. Nankai Reg. Fish. Res. Lab., 1951a, 1951b. Navarro and Lozano, 1950. Navaz, 1950. BIBLIOGRAPHY ON THE TUNAS 233 Germo — Continued Okada and Mataubara, 1938. Okada et al., 1935. Partlo, 1950, 1951. Powell, D. E. 1950. Powell and Hildebrand, 1950. Powell et al., 1952. Priol, 1944. Rivas, 1951, 1953. Roedel, 1948a. Samson, 1940. Sanzo, 1932, 1933. Sasaki, 1932, 1939b. Scagel, 1949. Schaefer, 1948c. Schaefers, 1952, 1953. Schultz and DeLacy, 1936. Serventy, 1941a, 1941b, 1947. Shapiro, 1948a, 1948b. Shizuoka Pref . Fish. Expt. Sta. 1937b. Smith, 1947. Society for the Promotion . . . 1936, 1937a. Soldatov and Lindberg, 1930. South Seas Gov't. . . . 1943a. Suyehiro, 1951. >' Takayama and Ando, 1934. Tanaka, 1931. Taranetz, 1937. Tauchi, 1940c. Tinker, 1944. Toyama, Y., et al., 1941. Uda, 1931a, 1935a, 1936b, 1940b. Uda and Tokunaga, 1937. Uno, 1936a, 1936b. Van Campen, 1952. Van Campen and Shimada, 1951. Walford, 1931, 1937. Watanabe, Hajime, 1939. Yabuta, 1953. Yoshihara, 1951-52. Germo alalunga. See Germo. Germo aVbacores. See Neothunntts itosibi, Germo argentivittatus. See Neothunnus argentivittatus. Germo germo. See Germo. Germ.0 germon. See Germo. Germo macropterus. See Neothunnus -macropterus. Germo obesus. See Parathunnus atlanticus. Germo sihi. See Parathunnus sibi. Germon. See Germ^o. Growth Aikawa and Kato, 1938. Brock, 1943. Conseil Int'l pour 1' Exploration de la Mer, 1933. Frade, 1937a, 1937b. Galtsoff, 1952. Heldt, 1930, 1931a, 1943, 1950. Kamimura and Honma, 1953. Kimura, 1932, 1935. Kimura and Ishii, 1932. Matsui, K., 1942b. Growth — Continued Moore, 1951a, 1951b. Nakamura, 1949. Nakamura et al., 1953. Partlo, 1950. Schaefer, 1948b, 1951, 1952. Schaefer emd Walford, 1950. Schweigger, 1949. Sella, 1952. Sette, 1954. Westman and Gilbert, 1941. Westman and Neville, 1942. Gymnosarda Manter, 1940. Nakamura, 1939a. Warfel, 1950. Gymnosarda af finis. See Katsuwonus. Gymnosarda alletterata. See Euthynnus alletteratus. Gymnosarda pelamis. See Katsuwonus. Habits. See Behavior. Honmaguro. See Thunnus orientalia. Indiem Ocetin Molteno, 1948. Morrow, 1954. Nomura et al., 1952-53. Smith, 1935. Indonesian waters Anonymous, 1941b. Japanese Bureau of Fisheries, 1933, 1934, 1935. Kimura, 1942a. Matsubara, 1943. Nakfimura, 1936, 1951. Nomura et al., 1952-53. Okuma et al., 1935. Shimada, 1937. South Seas Gov't. . . . 1941d. Juveniles. See Young. Katsuo. See Katsuwon%is. Katsuwonidae Serventy, 1948. Katsuwonus Abe, 1939. Aikawa, 1933, 1937. Aikawa and Kato, 1938. Anonymous, 1937a, 1937b, 1939a, 1939b, 1941b. Auffret, 1931. Bini, 1952. Blackburn and Rajoier, 1951. Brock, 1949. Chapman, 1946. Chiba Pref. Fish. Expt. Sta., 1936a, 1936b. Chiba Pref. Fish. Expt. Sta., Katsuura Branch, 1937, 1938, 1941c, 1941d. Cleaver and Shimada, 1950. Clemens and Wilby, 1946. DeBuen, 1930, 1935. 234 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Katsuwonus — Continued Delsman and Hardenburg, 1934. Domantay, 1940. Dung and Royce, 1953. Eckles, 1949a, 1949b. Fish, 1948. Fisheries Society of Japan ,1931. Food and Agr. Organ. U. N., 1949b. Formosa Gov't.-Gen. Fish. Expt. Sta., 1930, 1931, 1932, 1933, 1934. Fowler, 1931, 1934, 1938, 1944, 1949. Fukuda and lizuka, 1940. Godsil, 1936, 1938a, 1938b, 1945, 1949. Godsil and Byers, 1944. Herald, 1951. Herre, 1932, 1933, 1935, 1936, 1940, 1953. Herre and Umali, 1948. Higashi, 1940a, 1940b, 1941, 1942. Higashi and Hirai, 1948. Hildebrand, 1946. Horiguchi, Kakimoto, and Kashiwada, 1950. Horiguchi, Kashiwada, and Kakimoto, 1953. Ikebe, 1938. Ikebe and Matsumoto, 1937, 1938. Ikeda, 1932. Ikeda and Ando, 1933. Inanami, 1942b, 1942c, 1942d. Joubin, 1934. June, 1950b, 1951b. ■ '■•■• ■■-'' Kagoshima Pref. Fish. Expt. Sta., 1930a, 1931a, 1932a, 1933a, 1934, 1935a, 1935b, 1936a, 1936b, 1936c, 1937a, 1937b, 1937c, 1938a, 1938b, 1938c, 1939a, 1939b, 1939c, 1940a, 1940b, 1940c, 1941a, 1941b. Kakimoto, Kanazawa, and Kashiwada, 1953. Kashiwada, Kakimoto, and Horiguchi, 1952. Kashiwada, Kakimoto, and Yamasaki, 1953. Kimura, 1941, 1942b, 1949. Kimura, Iwashita, and Hattori, 1952. Kodama, lizuka, and Harada, 1934. Koyasu, 1931. Kumamoto Pref. Fish. Expt. Sta., 1946. Kumata et al., 1941. Kuronuma et al., 1949. LeGall, 1934d. Maldura, 1946. Manter, 1940. Marr, 1948. Martin, 1938. Marukawa, 1939a, 1939b. Mather, 1954. Matsui, K., 1942a, 1942b. Matsumoto, T., 1937. Matsumoto, W., 1952. Mie Pref. Fish. Expt. Sta., 1930a, 1930b, 1930d, 1950. Migita and Arakawa, 1948. Miura, 1941. Miyama and Osakabe, 1938, 1940. Molteno, 1948. Morice, 1953a. Katsuwonus — Continued Murayama and Okura, 1950, 1952. Murphy and Niska, 1953. Murphy and Shomura, 1953a, 1953b. Nakamura, 1935, 1939a, 1939b. Navarro and Lozano, 1950. Navaz, 1950. Nigrelli and Stunkard, 1947. Nishikawa, 1934. Nomura et al., 1952-53. Okada and Matsubara, 1938. Okada et al., 1935. Okamoto, 1940. Okinawa Pref. Fish. Expt. Sta., 1931, 1936a, 1940a, 1943. Onodera, 1941. oya and Takahashi, 1936. Rawlings, 1953. Rivas, 1951, 1953. Robins, 1952. Roedel, 1948a. Ronquillo, 1953. Sasaki, 1939a. Sasaki and Takehisa, 1932. Schaefer, 1948b, 1948c, 1951. Schaefer and Marr, 1948b. Seale, 1940. Serventy, 1941a, 1941b. Sette, 1954. Shapiro, 1948a, 1948b. Shizuoka Pref. Fish. Expt. Sta., 1936, 1937a. Smith and Schaefer, 1949. Smith, 1947. Soldatov and Lindberg, 1930. South Seas Gov't. . . . 1937a, 1937c, 1938, 1941a, 1941b, 1941c, 1942, 1943b. Suda, 1953. Suyehiro, 1936, 1938, 1941, 1942, 1950. Tachikawa, 1932. Taihoku Prov. Fish. Expt. Sta., 1932. Takayama, 1934. Tanaka, 1931. Taranetz, 1937. Tauchi, 1943. Tester, 1952. Tinker, 1944. Tominaga, 1943. Toyama, Y., et al., 1941. Uda, 1932b, 1933, 1935b, 1935c, 1936a, 1938, 1939, 1940a, 1940b, 1940c, 1941, 1948. Uda and Tsukushi, 1934. Van Campen, 1952. Wade, 1950a, 1950b, 1951. Walford, 1931, 1937. Warfel, 1950. Watanabe, N., 1941. Welsh, 1949a. Whitley, 1947. Yabe, 1953. Yabe and Mori, 1948, 1950. BIBLIOGRAPHY ON THE TtJNAS 235 Katsuwoniis — Continued Yamamoto, S., 1933, 1934. Yamamoto, Shokichi, 1940. Yamanaka, 1950. Yonezawa, 1950. Katsuwcmris pelamis. See Katsuwonus. Keys Brock, 1949. DeBeaufort and Chapman, 1951. DeBuen, 1930. DeBuen and Frade, 1932. Delsman and Hardenburg, 1934. Frade and DeBuen, 1932. Fraser-Brunner, 1949, 1950. Godsil and Byers, 1944. Hildebrand, 1946. Mead, 1951. Morice, 1953a. Nakamura, 1949. Nichols and LaMonte, 1941. Okada and Matsubara, 1938. Rivas, 1951. Roedel, 1948a. Soldatov and Ldndberg, 1930. Taranetz, 1937. Uda, 1933. Wade, 1949. Walford, 1931, 1937. Kihada. See Neothunnua macropterus. Kishinoella Brock, 1949. Dung and Royce, 1953. Fraser-Brunner, 1950. Nakamura, 1939a, 1939b. Soldatov and Lindberg, 1930. Whitley, 1947. Kishinoella rara Brock, 1949. Nakamura, 1939a, 1939b. Kishiyioella tonygol Dung and Royce, 1953. Serventy, 1941a, 1942a, 1942b. Whitley, 1947. Kishinoellu zacalles Serventy, 1942b. Kuromaguro. See Thunnus orientalis. Larvae. See Young. Laws and regulations Anonymous, 1952. Cerquetelli, 1936. Sugiura, 1932. Length-weight data. See Morphometries. Length-weight relationship Anonymous, 1938. Bahr, 1952. Formosa Gov't.-Gen. Fish. Expt. Sta. ,1933. Hiratsuka and Imaizumi, 1934a, 1934b. Hiratsuka and Morita, 1935, 1936. J- Ikebe, 1940b, 1940c, 1941. Length-weight relationship — Continued Inanami, 1940. Kagoshima Pref. Fish. Expt. Sta., 1936a, 1937a. 1938a, 1939a, 1940a, 1941a. Kanamura and Imaizumi, 1936a. Kanamura and Yazakl, 1940b. Kawasaki, 1952. Miyama et al., 1939. Nakamura, 1936. Schaefer, 1948a. Schweigger, 1949. Serventy, 1941a. South Seas Gov't. . . . 1941c, 1943a, Uno, 1936b. Westman and Neville, 1942. Yabe and Mori, 1948. Little tuna (tunny). See Euthynnus atletteratus. Livebait fishing Anonymous, 1937b. Chapman, 1946. Chiba Pref. Fish. Expt. Sta., Katsuura Branch, 1941c. De La Tourrasse, 1951. Domantay, 1940a, 1940b. Flett, 1944. Higashi, 1941b. Ikebe and Matsumoto, 1938. Ikeda, 1932. Ikehara, 1953. Imamura, 1953. June, 1951b. June and Reintjes, 1953. Kagoshima Pref. Fish. Expt Sta., 1935b, 1936b, 1937c, 1938b, 1939b, 1940b, 1941a. Kanai and Kasu, 1938. Matsubara, 1943. Matsumoto, T., 1937. Miura, 1941. Murphy and Niska, 1953. Okajima, 1939. Powell. D. E. and Hildebrand, 1950. Sette, 1954. Shimoda, 1937. South Seas Gov't. . . . 1937. Yabe and Mori, 1950. Longline fishing Anonj-mous, 1937a, 1938, 1941a, 1941c. Chiba Pref. Fish. Expt. Sta., 1936b. Chiba Pref. Fish. Expt. Sta., Katsuura Branch, 1937b. 1938b, 1941a, 1941b, 1941f. Ikebe, 1941a. Imaizumi, 1937. Iwate Pref. Fish. Expt. Sta., 1953a, 1953b. June. 1950a, 1950b. Kagoshima Pref. Fish. Expt. Sta., 1930b, 1930c, 1931a, 1931b, 1932a, 1932b, 1932c, 1933b, 1935b. Kanagawa Pref. Fish. Expt. Sta., 1951b, 1952a, 1952b. Kanamura and Imaizumi, 1936a, 1936b. Kanamura and Yazaki, 1940a, 1940b. 236 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Longline fishing — Continued Kato, 1940. McKeman, 1953. Matsubara, 1943. Mie Pref. Fish. Expt. Sta., 1950. MiyazEiki Pref. High-Seas Fish. Guidance Center, 1953. Murphy and Shomura, 1952. 1953a, 1953b. Nakamura, 1951. Nakayama, 1948. Nankai Reg. Fish. Res. Lab., 1951a, 1951b. Niska, 1953. Nomura at al., 1952-53. Okajima, 1939. omori and Fujimoto, 1940. omori and Fukuda, 1938, 1940. Powell, D. E., 1950. Powell et al., 1952. Rasalan, 1950. Sakai and Uno, 1940. Schaefers, 1952. Sette, 1954. Shapiro, 1950. Shimada, 1951c. Shimoda, 1937. South Seas Govt'. . . . 1937, 1942, 1943a. Tapiador, 1951. Uda, 1935a. Van Campen and Shimada, 1951. Watanabe, Haruo, 1940. Yoshihara, 1951-52. Mackerel, frigate. See Auicis spp. Management Okumura, 1943. Schaefer, 1948c, 1951. Measurement data. See also Morphometries. Kagoshima, 1936a, 1937a, 1938a, 1939a, 1940a, 1941a. Marr, 1948. Russell, F. S., 1934a. Schaefer, 1952. Schaefer and Walford, 1950. Measuring methods LeGall, 1951. Marr and Schaefer, 1949. Priol, 1944. Russell, F. S., 1934a. Mebachi. See Parathunnus mebachi. Mediterranean Sea and Strait of Gibraltar Anonymous, 1932. Arcidiacono, 1935. Aric6 and Genovese, 1953. Auffret, 1931. Bonamico, 1933. Cerequetelli, 1936. DeBuen, 1931. Dieuzeide, 1930. Farina, 1931a, 1931b. Frade, 1937b, 1953. Genovese, 1952. 1953. Mediterranean Sea and Strsiit of Gibraltar — Continued Heldt, 1932a, 1934, 1937, 1938, 1943. Maldura, 1946. Reiss and Vellinger, 1929. Russell, F. S., 1934b. Sanzo, 1932, 1933. Scordia, 1930, 1939a, 1940, 1943. Sigma, 1941. Meristic counts Conseil Int'l pour 1' EJxploration de la Mer, 1933. Godsil and Byers, 1944, Heldt, 1931a, 1932b. June, 1952a, 1952b. Letaconnoux, 1950. Marr and Schaefer, 1949. Schaefer and Marr, 1948a. Schaefer and Walford, 1950. Wade, 1949. Migration Bini, 1952. DeBuen, 1931. Hatai et al., 1941. Heldt 1930, 1931a, 1932a, 1932b, 1934, 1943. Kagoshima Pref. Fish. Expt. Sta., 1936c. Kajnimura and Honma, 1953. Kawasaki, 1952. Kimura, 1941, 1942b. LeDanois, 1938, 1951. Marukawa, 1939c. Nakamura, 1949. Powell, D. E., et al., 1952. Reiss and Vellinger, 1929. Ros6n, 1943. Russell, F. S., 1936. Sasaki, 1939a. Schaefers, 1953. Scordia, 1940. Sella, 1930, 1931, 1952. Serventy, 1941a. Shapiro, 1948a, 1948b. Sigma, 1941. Tauchi, 1940b. Tominaga, 1943. Uda. 1936a. Uda and Tokunaga, 1937. Wade, 1951. Walford, 1937. Whitehead, 1931. Yabuta and Ueyanagi, 1953. Miscellaneous species (Auxis to Neothunnus) Manter, 1940. Mather, 1954. Rivas, 1951. Tubb, 1948. Warfel, 1950. Miscellaneous species (Orcynus to Wanderer) Boeseman, 1947. Chu, 1931. DeBeaufort and Chapman, 1951. Fowler, 1949. ,-, BXBLJOGRAPHY ON THE TUNAS 237 Miscellaneous Species — Continued Ginsburg, 1953. Rivas, 1951. Schaefer, 1951. Schaefers, 1952, 1953. SeUa, 1931. Morphometries Aikawa and Kato, 1938. Aric6 and Genovese, 1953. Bell6n and BardAn de Bell6n, 1949. Bini, 1931. Bonliam, 1946. Conseil Int'l pour 1' Exploration de la Mer, 1933. DeBuen, 1932. Dung and Royce, 1953. Frade, 1931a, 1931b. Godsil, 1948, 1949. Godsil and Byers, 1944. Greenhood, 1952. Heldt, 1937, 1938. Higashi, 1942. Hiratsuka eind Morita, 1935, 1936. Ikebe and Matsumoto, 1937. Inanami, 1942d. Jap. Bur. Fish., 1939, 1940. June, 1952a, 1952b. LeGall, 1949, 1951. Legendre, 1934. Letaconnoux, 1950. Marr and Schaefer, 1949. Mather, 1954. Nakamura, 1939b, 1939c. Navaz, 1950. Oita Pref. Fish. Expt. Sta., 1930. Priol, 1944. Royce, 1953. Russell, F. S., 1934a. Schaefer, 1948a, 1951, 1952. Schaefer and Walford, 1950. Serventy, 1948. Uda, 1941. Watanabe, Hajime, 1939. Yabe et al., 1953. Neothunnus (Neothynnus) Abe, 1939. Aikawa, 1933. Aikawa and Kato, 1938. Ancieta C, 1952. Arai and Matsumoto, 1953. Ban, 1941. Barnard, 1948. Bamhart, 1936. Bates, 1950. Beebe, 1936. Beebe and Tee-Van, 1936. Bini, 1931, 1952. Boeseman, 1947. Bonham, 1946. Brock, 1949. Neothunnus (Neothynnus) — Continued Chapman, 1946. Chiba Pref. Fish .Expt. Sta., 1936b. Chiba Pref. F^sh. Expt. Sta., Katsuura Branch, 1941f. Chu, 1931. Copley, 1947. DeBeaufort and Chapman, 1951. DeBuen, 1930, 1935. Delsman and Hardenburg, 1934. Domantay, 1940. Dung and Royce, 1953. Eckles, 1949a. Fish, 1948. Fisheries Society of Japan, 1931. Fitch, 1950. Food and Agr. Organ. U. N., 1949b. Formosa Gov't. -Gen. Fish. Expt. Sta., 1933a, 1933b. Fowler, 1931, 1936, 1949. Frade, 1931b, 1931c. Ginsburg, 1953. Godsil, 1936, 1938a, 1938b, 1945, 1948, 1949a, 1949b. Godsil and Byers, 1944. Godsil and Greenhood, 1948, 1951, 1952. Greenhood, 1952. Hatai et al., 1941. Herald, 1949. Herre, 1932, 1935, 1936, 1940. Herre and Umali, 1948. Higashi, 1940, 1941, 1942. Higashi £ind Hirai, 1948. Hildebrand, 1946. Hiratsuka emd Imaizumi, 1934. Hiratsuka and Ito, 1934. Hiratsuka and Morita, 1935, 1936. Ikebe, 1939a, 1939b, 1940a, 1940b, 1940c, 1940d, 1941b, 1941c, 1942. Inanami, 1940a, 1940b, 1940c, 1942a, 1942b. Iwate Pref. Fish. Expt. Sta., 1953a, 1953b. Jap. Bur. Fish., 1933, 1934, 1935. June, 1952b, 1953. Kagoshima Pref. Fish. Expt. Sta., 1930b, 1930c, 1931b, 1933b. Kanamura and Imaizumi, 1936a. Kanamura and Yazaki, 1940a, 1940b. Kato, 1940. Kawamura, 1939. Kimura, 1932, 1935, 1942a, 1942b. Kimura and Ishii, 1932, 1933b. Kumata et al., 1941. Marr, 1948. Martin, 1938. Marukawa, 1939b, 1939c. Mather, 1954. Mead, 1951. Mie Pref. Fish. Expt. Sta., 1930c, 19306. Migita and Arakawa, 1948. Miura, 1941. Miyama and Osakabe, 1940. Miyama et al., 1939. 288 FISHERY BTJLIjETINOm THE FISH AND WILDLIFE SERVICE Neothunntis (Neothy^mnsJ— Continued ■ ■.-■ :m Miyazaki Pref., High-Seas Fish. Guidance Center, 1953. Molteno, 1948. Moore, 1951a, 1951b. Morice, 1953a, 1953b. Morrow, 1954. :», ;, Murphy and Niska, 1953. ■■ ' : "":i'" Murphy and Shomura, 1953a, 1953T3. Nakamura, 1936, 1939a, 1939b, 1939c,. 1941, 1943, -■ 1949, 1951. :- Nankai Reg. Fish. Res. Lab., 19f51a. .': ^i.'.vC; Nichols and LaMonte, 1941. ; "'.' ::-'^ Nigrelli and Stunkard, 1947. ►- Nomura et al., 1952-53. Oita Pref. Fish. Expt. Sta., 1930. Okada and Matsubara, 1938. Okada et al., 1935. Okinawa Pref. Fish. Expt. Sta., 1936b. okuma et al., 1935. Phillipps, 1932. Powell, A. W. B., 1937. Rawlings, 1953. Reintjes, 1952 Reintjes and King, 1953. Rivas, 1951, 1953. Roedel, 1948a. Ronquillo, 1953. Royce, 1953. Schaefer, 1948a, 1948b, 1948c, 1951, 1952. Schaefer and Marr, 1948b. Schaefer and Walford, 1950. Schweigger, 1943, 1949. Seale, 1940. Serventy, 1941a, 1941b. Sette, 1954. Shapiro, 1948, 1948b. Shimada, 1951, 1954. Smith and Schaefer, 1949. Smith, 1947. Soldatov and Lindberg, 1930. South Seas Gov't'. . . . 1937a, 1938, 1941a, 1941d, 1942, 1943a, 1943b. Suda, 1953. Suyehiro, 1941, 1942. Tanaka, 1931. Tapiador, 1951. Taranetz, 1937. Tauchi, 1940b. Tester, 1952. Tester et al., 1952. Tinker, 1944. Toyama, Y., et al., 1941. Uda, 1935a, 1952. Uehara, 1941. Van Campen, 1952. Wade, 1950a, 1950b, 1951. Walford, 1931, 1937. Warfel, 1950. Watanabe, Haruo, 1940. Neothunnus {Neothynnus)-^.Contixiued. Tj!i!i':.:-£.\\;'. Welsh, 1949a, 1949c. • ty ; • Whitley, 1947. : i. Yabe et al., 1953. Yabe and Mori, 1950. Yabuta, 1953. Yabuta and Ueyanagi, 1953a, 1953b. Neothunnus albacora Barnard, 1948. Blni, 1931. DeBuen, 1930, 1935. Frade, 1931b, 1931c. Marukawa, 1939c. Navarro and Lozano, 1950. Nichols and LaMonte, 1941. Schaefer and Walford, 1950. Neothunnus albacares. See N. macropterus. N. albacora albacora. See N. macropterus. N. albacora macropterus. See N. macropterus, N. allisoni Nichols and LaMonte, 1941. Walford, 1937. N. allisoni allisoni. See N. allisoni. N. allisoni itosibi. See N. itosibi. N. argentivitattus Beebe, 1936. Beebe and Tee-Van, 1936. Fowler, 1944. Rawlings, 1953. N. catalinae Nichols and LaMonte, 1941. N. itosibi Domantay, 1940b. Martin, 1938. Molteno, 1948. Nakamura, 1939c. Okada and Matsubara, 1938. Phillipps, 1932. Powell, A. W. B., 1937. N. m,acropterus Abe, 1939. Aikawa, 1933. Aikawa and Kato, 1938. Ancieta C, 1952. Anonymous, 1938. Aral and Matsumoto, 1953. Asakawa, Noguchl, and Mimoto, 1953. Ban, 1941. Barnhart, 1936. Bates, 1950. Bini, 1952. Boeseman, 1947. Bonham, 1946. Brock, 1949. Chapman, 1946. Chiba Pref. Fish. Expt. Sta., 1936b. Chiba Pref. Fish. Expt. Sta., Katsuura Branch, 1941. Chu, 1931. Copley, 1947. ■JBIBLilOGRAPHYiON! THE TUNAS 239 N. macropterus — Continued p^eaufort and Chapman, 1951. DeBuen, 1935. Delsman and Hardenburg, 1934. Domantay, 1940. Dung and Royce, 1953. Eckles, 1949a. Fish, 1948. Fisheries Society of Japan, 1931. Fitch, 1950. Food and Agr. Organ. U. N., 1949b. Formosa Gov't.-Gen. Fish. Expt. Sta., 1933a. Fowler, 1931, 1936, 1949. Ginsburg, 1953. Godsil, 1936, 1938a, 1938b, 1945, 1948, 1949a, 1949b. Godsil and Byers, 1944. Godsil and Greenhood, 1948, 1951, 1952. Greenhood, 1952. Hatai et al., 1941. Herald, 1949. Herre, 1932, 1935, 1936, 1940. Herre and Umali, 1948. Higashi, 1940a, 1941a, 1941b, 1942. Higashi and Hirai, 1948. Hildebrand, 1946. Hiratsuka and Imaizumi, 1934. Hiratsuka and Ito, 1934. Hiratsuka and Morita, 1935, 1936. Ikebe, 1939a, 1939b, 1940a, 1940b, 1940d, 1941b, 1941c, 1942. Ikehara, 1953. Imaizumi, 1937. Inanami, 1940a, 1940b, 1940c, 1942a, 1942b. Iwate Pref. Fish. Expt. Sta., 1953a, 1953b. Jap. Bur. Fish., 1933, 1934, 1935. June, 1952b, 1953. Kagoshima Pref. Fish. Expt. Sta., 1930b, 1930c, 1931b, 1933b. Kanagawa Pref. Fish. Expt. Sta. 1951a. Kanamura and Imaizumi, 1936a. Kanamura and Yazaki, 1940a, 1940b. Kato, 1940. Kawamura, 1939. Kimura, 1932, 1935, 1942a, 1942b. Kimura and Ishii, 1932, 1933b. Kumata et al., 1941. Marr, 1948. Martin, 1938. Marukawa, 1939b. Mather, 1954. Mead, 1951. Mie Pref. Fish. Expt. Sta., 1930c, 1930d, 1930e. Migita and Arakawa, 1948. Miura, 1941. Miyama and Osakabe, 1940. Miyama et al., 1939. Miyazaki Pref. High-Seas Fish. Guidance Center, 1953. Moore, 1951a, 1951b. Morrow, 1954. N. macropterus — Continued Murphy and Niska, 1953. Murphy and Shorn ura, 1952, 1953a, 1953b. Nakamura, 1936, 1939a, 1939b, 1939c, 1941, 1943, . 1949, 1951. .f Nigrelli and Stunkard, 1947. ■ ••■ . Nomura et al., 1952-53. Oita Pref. Fish. Expt. Sta., 1930. Okada and Matsubara, 1938. Okada et al., 1935. Okinawa Pref. Fish. Expt. Sta., 1936b. okuma et al., 1935. Reintjes, 1952. Reintjes and King, 1953. Roedel, 1948a. '^''''' ', Ronquillo, 1953. Royce, 1953. Schaefer, 1948a, 1948b, 1948c, 1951, 1952. Schaefer and Marr, 1948b. Schaefer and Walford, 1950. Schweigger, 1943, 1949. Seale, 1940. Serventy, 1941a, 1941b. Sette, 1954. Shapiro, 1948a, 1948b. Shimada, 1951b, 1954. Smith and Schaefer, 1949. Smith, 1947. Soldatov and Lindberg, 1930. South Seas Gov't. . . . 1937a, 1938, 1941a, 1941d, 1942, 1943a, 1943b. Suda, 1953. Suyehiro, 1941, 1942. Takayama and Ando, 1934. Tanaka, 1931. Tapiador, 1951. Taranetz, 1937. Tauchi, 1940b. Tester, 1952. Tester et al., 1952. Tinker, 1944. Toyama, Y., et al., 1941. Uda, 1952. Uehara, 1941. Van Campen, 1952. Wade, 1950a, 1950b, 1951. Walford, 1931, 1937. Warfel, 1950. Watanabe, Haruo, 1940. Welsh, 1949a, 1949c. Whitley, 1947. Yabe et al., 1953. Yabe and Mori, 1950. Yabuta, 1953. Yabuta and Ueyanagi, 1953a, 1953b. N. rarus Delsman and Hardenburg, 1934. Herre, 1940. Nakamura, 1943, 1949. Nichols and LaMonte, 1941. 240 FISHERT BULLETIN OF THE BISH AND WILDLIFE SERVICE N. rarua — Continued Serventy, 1942b. N. rarus zacalles. See Kishinoella zacallea. North Sea and English Channel Bahr, 1952. Delsman, 1933. Flck, 1937. Kreutzer, 1951a. LUUng, 1950, 1951, 1952b. Ros6n, 1943. Russell, F. S., 1933a, 1934a. Oceanographic conditions correlated with fishing or distribution Aikawa, 1933. Anonymous, 1937a, 1941c, 1942. Ban, 1941. Bini, 1952. Chiba Pref. Fish. Expt. Sta., 1936a. Chiba Pref. Fish. Expt. Sta., Katsuura Branch, 1937a, 1937b, 1938a, 1938b, 1941a, 1941b, 1941c, 1941d, 1941f. Formosa Gov't.-Gen. Fish. Expt. Sta., 1930, 1931, 1932, 1933a, 1933b, 1934. Fujii, 1932. Fukuda and lizuka, 1940. Genovese, 1952, 1953. Hart and HoUister, 1947. Hiratsuka and Imaizumi, 1934. Hiratsuka and Ito, 1934. lehisa, 1939. Ikebe, 1940b, 1941a, 1942. Ikebe and Matsumoto, 1937. Imamura, 1949. Inanami, 1938, 1940b, 1940c, 1941, 1942a, 1942b. Inoue, 1953. Jap. Bur. Fish. 1933, 1934, 1935, 1939, 1940. Kagoshima Pref. Fish. Expt. Sta., 1930a, 1930b, 1930c, 1931a, 1931b, 1932a, 1932c, 1933a, 1933b, 1934, 1935a, 1936a, 1937a. Kanagawa Pref. Fish. Expt. Sta., 1951a, 1951b, 1952a, 1952b. Kanamura and Imaizumi, 1936a. Kawamura, 1939. Kawana, 1934, 1937. Kawasaki, 1952. Kida, 1936. Kimura, 1941, 1942a, 1949. Kimura £ind Ishii, 1933b. Kumamoto Pref. Fish. Expt. Sta., 1946. LeDanois, 1933, 1938. Matsubara, 1943. Matsumoto, T., 1937. Mie Pref. Fish. Expt. Sta., 1930a, 1930b, 1930c, 1930d, 19306. Miyazaki Pref. High-Seas Fish. Guidance Center, 1953. Murphy and Niska, 1953. Murphy and Shomura, 1953a, 1953b. Nakamura, 1949. oita Pref. Fish. Expt. Sta., 1930. Oceanographic conditions — Continued Okinawa Pref. Fish. Expt Sta., 1940a, 1940b, 1943. Okuma, 1935. Omori and Fujimoto, 1940. Omori and Fukuda, 1940. Partlo, 1950, 1951. Powell, D. E. 1950. Powell and Hildebrand, 1950. Powell et al., 1952. Reiss and Vellinger, 1929. Sait6, 1937. Sasaki, 1932, 1939a, 1939b. Scagel, 1949. Schaefers, 1952, 1953. Shapiro, 1948b. Shizuoka Pref. Fish. Expt. Sta., 1936, 1937a. Society for the Promotion . . . 1936. South Seas Government . . . 1937c, 1938, 1941d 1942, 1943a, 1943b. Taihoku Prov. Fish. Expt. Sta., 1932. Takayama and And5, 1934. Takayama et al., 1934. Tapiador, 1951. Uda, 1931a, 1933, 1935a, 1935b, 1935c, 1936a, 1936b, 1938, 1939, 1940a, 1940b, 1941, 1948, 1952. Uda and Tokunaga, 1937. Uehara, 1941. Yabe et al., 1953. Yabuta, 1953. Yabuta and Ueyanagi, 1953a, 1953b. Orcynus Priol, 1944. Sanzo, 1932, 1933. Orcynus alalonga. See Germo. Orcynus germo. See Germo. Orcynus pacificus. See Germo. Orcynus thynnus. See Thunnus thynnus. Pacific Ocean, NE Bamhart, 1936. Bates, 1950. Brock, 1938, 1939, 1943, 1949, 1954. Chapman, 1946. Chiba Pref. Fish. Expt. Sta., Katsuura Branch, 1941e, 1941f. Clemens and Wilby, 1946. Cowan, 1938. Eckles, 1949a, 1949b. Fish, 1948. Fitch, 1950, 1953. Godsil, 1936, 1937, 1938a, 1938b, 1938c. Godsil and Greenhood, 1948, 1951, 1952. Godsil and Holmberg, 1950. Hart and HoUister, 1947. Hart et al., 1948. Ikehara, 1953. Inanami, 1941. June, 1950a, 1953. McHugh, 1952. McKeman, 1953. Matsui, Y., 1938. ',^ BIBLIOGRAPHY ON THE TUNAS 241 Pacific Ocean, NE — Continued Moore, 1951a, 1951b. Murphy and Niska, 1953. Murphy and Shomura, 1953a, 1953b. Nakamura et al., 1953. Niska, 1953. Partlo, 1950, 1951. PoweU, D. E., 1950. Powell and Hildebrand, 1950. PoweU et al., 1952. Reintjes and King, 1953. Roedel, 1948a, 1948b. Samson, 1940. Schaefer, 1948a, 1948b, 1952. Schaefer and Marr, 1948a, 1948b. Schaefer and Walford, 1950. Schaefers, 1952, 1953. Schultz and DeLacy, 1936. Sette, 1954. Smith and Schaefer, 1949. South Seas Gov't. . . . 1943£L Tester, 1952. Tinker, 1944. Van Campen and Shimada, 1951. Walford, 1931, 1937. Welsh, 1949a, 1949b, 1949c. Whitehead, 1930, 1931. Pacific Ocean, NW Abe, 1939. Aikawa, 1932, 1933, 1937. Anonymous, 1939, 1941b, 1941c. Boeseman, 1947. Chapman, 1946. Chevey, 1932a, 1934. Chiba Pref. Fish. Expt. Sta., 1936a, 1936b. Chiba Pref. Fish. Expt. Sta., Katsuura Branch, 1937a, 1937b, 1938a, 1938b, 1941a, 1941b, 1941d, 1941e. Cleaver and Shimada, 1950. DeJong, 1940. Delsman, 1931. Domantay, 1940a, 1940b. Ego and Otsu, 1952. Espenshade, 1948. Fish, 1948. Fisheries Society of Japan, 1931. Formosa Gov't.-Gen. Fish. Expt. Sta., 1930, 1931, 1932, 1933a, 1933b, 1934. Fujii, 1932. Fukuda Eind lizuka, 1940a, 1940b. Hasegawa, Kiichl, 1937. Hasegawa, Kimpei, 1938. Hatai et al., 1941. Herre, 1933, 1935. Herre and Umali, 1948. Hiatt and Brock, 1948. Hiratsuka and It6, 1934. Hiratsuka and Morita, 1935, 1936. lehisEi, 1939. Ikebe, 1938, 1939a, 1939b, 1940a, 1940b, 1940c, Pacific Ocean, NE — Continued 1940d, 1941c. Ikebe and Matsumoto, 1937a. Ikeda, 1932. Ikeda and Ando, 1933. Imamura, 1949. Inanami, 1940a, 1940b, 1942a, 1942b, 1942c, 1942d. Jap. Bur. Fish. 1933, 1934, 1935, 1940, 1942. June, 1951b, 1952a, 1952b. Kagoshima Pref. Fish. Expt. Sta., 1930a, 1930b. 1930c, 1931a, 1932a, 1932c, 1 1933a, 1933b, 1934. 1935a, 1935b, 1936b, 1937a, 1937b, 937c. 1938a, 1938b, 1938c, 1939a, 1939b, 1939c, 1940a, 1940b. 1940c, 1941a, 1941b. Kanagawa Pref. Fish. Expt. Sta., 1951a, 1951b. 1952a, 1952b. Kanai, Moto and Kasu, 1938. Kanamura and Imaizumi, 1936a. 1936b. Kanamura and Yazaki, 1940b. Kato, 1940. Kawamura, 1939. Kawana, 1934, 1935, 1938. Kida, 1936. Kikawa, 1953. Kimura, 1933, 1935, 1941, 1942a, 1942b, 1949. Kimura and Ishii, 1931, 1932, 1933b. Koyasu, 1931. Kumamoto Pref. Fish. Expt. Sta., 1946. McKeman, 1953. Manter, 1940. Matsubara, 1943. Matsui, K., 1942b. Mie Pref. Fish. Expt. Sta., 1930a, 1930b, 1930c. 1930d, 1930e, 1950a, 1950b, 1950c. Mine and lehisa, 1940. Miura, 1941. Murayama and Okura, 1950, 1952. Nakamura, 1938, 1939b, 1939c, 1941, 1943, 1949, 1951. Nakamura, et al., 1953. Nankai Reg. Fish. Res. Lab., 1951a, 1951b. Noguchi, 1938. Nomura et aJ., 1952-53. 6ita Pref. Fish. Expt. Sta., 1930. Okada and Matsubara, 1938. Okajima, 1939. Okamoto, 1940. Okinawa Pref. Fish. Expt. Sta., 1931, 1936a, 1940a, 1940b. okuma et al., 1935. Omori and Fujimoto, 1940. 6mori and Fukuda, 1938, 1940. Onodera, 1941. Rasalan, 1950. Ronquillo, 1953. Sakai and Uno, 1940. Sasaki, 1932, 1939a, 1939b. Sasaki, and Takehisa, 1932. Scagel, 1949. Schmidt, 1930. 242 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Pacific Ocean, NW — Continued Shapiro, 1948a, 1948b. Shimada, 1951b, 1951c. Shizuoka Pref. Fish. Expt. Sta., 1936, 1937a, 1937b. Smith and Schaefer, 1949. Smith, 1947. Society for the Promotion . . . 1936, 1937a. Soldatov and Lindberg, 1930. South Seas Gov't. . . . 1937a, 1937b, 1937c, 1937d, 1938, 1941a, 1941b, 1941c, 1941d, 1942, 1943a, 1943b. Suda, 1953. Sugiura, 1932. Tachikawa, 1932. Taihoku Prov. Fish. Expt. Sta., 1932. Takayama and Ando, 1934. Takayama at al., 1934. Tanaka, 1931, 1936, 1939. Tapiador, 1951. Taranetz, 1937. Tauchi, 1940a, 1940b, 1940c, 1943. Tominaga, 1943. Uda, 1931a, 1931b, 1932a, 1933, 1935a, 1935b, 1935c, 1936a, 1938, 1939, 1940b, 1940c, 1941, 1948, 1952. Uda and Tokunaga, 1937. Uda and Tsukushi, 1934. Uehara, 1941. Uno, 1936a. Van Campen and Shimada, 1951. Wade, 1950a. Warfel, 1950. Watanabe, Hajime, 1939. Watanabe, Haruo, 1940. Watanabe, Nobuo, 1941. Yabe, 1953. Yabe et al., 1953. Yabe and Mori, 1950. Yabuta, 1953. Yabuta and Ueyanagi, 1953a. Yamamoto, Shokichi, 1940. Pacific Ocean, SE Ancieta C, 1952. Bini, 1952. Chapman, 1946. Fish, 1948. Fowler, 1938. Schweigger, 1943, 1949. Scale, 1940. Shimada, 1951d, 1954. Van Campen, 1953. Pacific Ocean, SW Anonymous, 1953a. Ban, 1941. Chapman, 1946. Fish, 1948. Flett, 1944. Formosa Gov't. Gen. Fish. Expt. Sta., 1933. Fowler, 1938. Godsil and Holmberg, 1950. Pacific Ocean SW — Continued Ikebe, 1941b. McKeman, 1953. Nomura et al., 1952-53. PhilUpps, 1932. Robins, 1952. Saito, 1937. Serventy, 1941a, 1941b, 1942a, 1942b, 1947, 1948. Whitley, 1937, 1947. Parasites Arai and Matsumoto, 1953. Crane, 1936. LeGall, 1949. Legendre, 1940. Manter, 1940. Nigrelli and Stunkard, 1947. Priol, 1944. Van Cleave, 1940. Parathunnus (Parathynnus) Aikawa, 1933. Asakawa, Noguchi, and Mimoto, 1953. Beebe, 1936. Beebe and Tee-Van, 1936. Bell6n and Barddn de Bell6n, 1949. Bini, 1931. Brock, 1949. Chiba Pref. Fish. Expt. Sta., 1936b. Chiba Pref. Fish. Expt. Sta., Katsuura Branch, 1941e, 1941f. DeBeaufort and Chapman, 1951. DeBuen, 1930. Domantay, 1940b. Dung and Royce, 1953. Fish, 1948. Fisheries Society of Japan, 1931. Fowler, 1931, 1938, 1949. Frade, 1931b. Fukuda and lizuka, 1940. Godsil, 1945. Godsil and Byers, 1944. Hatai et al., 1941. Herre, 1940. Higashi, 1940a, 1941b. Ikebe, 1939a, 1940a, 1942. Inanami, 1940b, 1940c. Iwate Pref. Fish. Expt. Sta., 1953a, 1953b. Jap. Bur. Fish., 1933, 1934, 1939. Kagoshima Pref. Fish. Expt. Sta., 1930b, 1930c, 1931b, 1933b. Kamimura and Honma, 1953. Kanagawa Pref. Fish. Expt. Sta., 1951a, 1952a. 1952b. Kanamura and Imaizumi, 1936a. Kanamura and Yazaki, 1940a. Kikawa, 1953. Kimura, 1942a, 1942b. Kimura, Iwashita, and Hattori, 1952. Kumata et al., 1941. Mie Pref. Fish. Expt. Sta., 1930c, 1930e. Miyama and Osakabe, 1938, 1340. BIBLIOGRAPHY ON THE TUNAS 243 Parathunntis (Parathynnus) — Continued Miyazaki Pref. High-Seas Fish., Guidance Center, 1953. Molteno, 1948. Morice, 1953b. Mowbray, 1935. Murphy and Shomura, 1953a, 1953b. Nakamura, 1939a, 1939b, 1941, 1943, 1949, 1951, 1953. Nankai Reg. Fish. Res. Lab., 1951a. Navarro and Lozano, 1950. Nomura et al., 1952-53. Oita Pref. Fish. Expt. Sta., 1930. Okada and Matsubara, 1938. Okada et al., 1935. Okinawa Pref. Fish. Expt. Sta., 1940b. omori and Fujimoto, 1940. omori and Fukuda, 1938, 1940. Rivas, 1953. Roedel, 1948a. Schaefer, 1951. Schuck and Mather, 1951. Shapiro, 1948a. Shimada, 1951b, 1954. Smith and Schaefer, 1949. Soldatov and Lindberg, 1930. South Seas Government . . . 1941a, 1942. Suda, 1953. Suyehiro, 1941, 1942. Takayama and Ando, 1934. Toyama, Y., et al., 1941. Uda, 1935a. Uehara, 1941. Van Campen, 1952. Watanabe, Haruo, 1940. Yabuta, 1953. Yabuta and Ueyanagi, 1953a. Parathunnus atlanticus Beebe, 1936. Beebe and Tee-Van, 1936. Bell6n and Bard^n de Bell6n, 1949. Bini, 1931. DeBuen, 1930. Frade, 1931b. Legendre, 1937. Marukawa, 1939b, 1939c. Mather, 1954. Molteno, 1948. Morice, 1953b. Mowbray, 1935. Navarro and Lozano, 1950. Nigrelli and Stunkard, 1947. Rawlings, 1953. Schuck and Mather, 1951. Paiathunnus mebachi Asakawa, Noguchi, and Mimoto, 1953. Fish, 1948. Godsil, 1945. Godsil and Byers, 1944. Ikebe, 1939a. Kamimura and Honma, 1953. Parathunnus (Parathynnus J — Continued Kikawa, 1953. Kumata et al., 1941. Mie Pref. Fish. Expt. Sta., 1930a, 1930d. Nakamura, 1939a, 1939b, 1943, 1949. Nakamura et al., 1953. Roedel, 1948a. South Seas Government . . . 1941a. Takayama and Ando, 1934. Parathunnus obesus. See P. atlanticus. Parathunnus sibi Aikawa, 1933. Brock, 1949. Chiba Pref. Fish. Expt. Sta., 1936b. Chiba Pref. Expt. Sta., Katsuura Branch, 1941e, 1941f. DeBeaufort and Chapman, 1951. Domantay, 1940b. Dung and Royce, 1953. Fisheries Society of Japan, 1931. Fowler, 1931, 1938, 1949. Fukuda and lizuka, 1940a. Hatai et al., 1941. Herre, 1940. Higashi, 1940a, 1941b. Ikebe, 1940a, 1942. Inanami, 1940b, 1940c. Iwate Pref. Fish. Expt. Sta., 1953a, 1953b. Jap. Bur. Fish., 1934, 1935, 1939. Kagoshima Pref. Fish. Expt. Sta., 1930b, 1930c, 1931b, 1933b. Kanagawa Pref. Fish. Expt. Sta., 1951a, 1952a, 1952b. Kanamura and Imaizumi, 1936a. Kanamura and Yazaki, 1940a. Kimura, 1942a, 1942b. Kimura, Iwashita, and Hattori, 1952. Kumata et al., 1941. Marukawa, 1939b. Mie Pref. Fish. Expt. Sta., 1930c, 1930e. Miyama and Osakabe, 1938, 1940. Miyazaki Pref. High-Seas Fish. Guidance Center, 1953. Murphy and Shomura, 1953a, 1953b. Nakamura, 1941, 1951. Nankai Reg. Fish. Res. Lab. 1951a. Nomura et al., 1952-53. oita Pref. Fish. Expt. Sta., 1930. Okada and Matsubara, 1938. Okada et al., 1935. Okinawa Pref. Fish. Expt. Sta., 1940b. omori and Fujimoto, 1940. omori and Fukuda, 1938, 1940. Schaefer, 1951. Shapiro, 1948a. Shimada, 1951b, 1954. Smith and Schaefer, 1949. South Seas Gov't 1942. Suda, 1953. Suyehiro, 1941, 1942. Toyama, Y., et al., 1941. 244 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Parathunnus sibi — Continued Uda, 1935a. Uehara, 1941. Van Campen, 1952. Watanabe, Haruo, 1940. Yabuta, 1953. Yabuta and Ueyanagi, 1953a. Yoshihara, 1951-52. Parathynnus sibi. See Parathunnus sibi. Pelamys (Pelamis) Maldura, 1946. Pelamys affine. See Euthynnus alletteratus. Pelamys macropterus. See Neothunnus macropterus. Pelamys pelamys. See Katsuwonus. Population dynamics Aikawa, 1937. Bini, 1952. Brock, 1943. Tauchi, 1940a, 1940b, 1940c, 1943. Tominaga, 1943. Populations, definition of Godsil and Byers, 1944. Royce, 1953. Schaefer, 1951, 1952. Schaefer and Walford, 1950. Sette, 1954. Uda and Tokunaga, 1937. Uda and Tsukushi, 1934. Purse-seining Carlson, 1951. Heldt, 1932b. Imamura, 1953. Murayama and Okura, 1950, 1952. Murphy and Niska, 1953. Murray, 1952. Sette, 1954. Reactions to stimuli Tester et al., 1952. Red Sea Copley, 1947. Red tuna. See Thunnus thynnus. Reproduction Anonymous, 1941b. Bini, 1952. Brock, 1943. DeBuen, 1931, 1937. DeJong, 1940. Eckles, 1949b. Frade and Managas, 1933. Genovese, 1953. Hatai et al., 1941. Heldt, 1934. Ikebe, 1941b. June, 1953. Kamimura and Honma, 1953. Kawana, 1935. Kikawa, 1953. Kuronuma et al., 1949. LeDanois, 1951. LeGall, 1949. Reproduction — Continued Lozano, 1950. Marr, 1948. Marukawa, 1939c. Mead, 1951. Nakamura, 1938, 1939b, 1943, 1949. Okada et al., 1935. Priol, 1944. Sanzo, 1933. Schaefer, 1948c, 1951. Schaefer and Marr, 1948b. Sella, 1952. Serventy, 1941a, 1942a. Sette, 1954. Shapiro, 1948a. Society for the Promotion . . . 1936. Wade, 1950b. Walford, 1937. Watanabe, Hajime, 1939. Whitehead, 1931. Yabe and Mori, 1948. Rote Thun. See Thunnus thynnus. Salinity. See Oceanographic conditions and related subjects. Scomber Conrad, 1937. Scouting methods Heldt, 1932a. SewMthunnus Fowler, 1933, 1934. Nakamura, 1939c. Tinker, 1944. Sex. See Morphometries Sex ratios Brock, 1943, 1954. Crane, 1936. Iwate Pref. Fish. Expt. Sta., 1953a, 1953b. Kanagawa Pref. Fish. Expt. Sta., 1952a, 1952b. Marr, 1948. Miyazaki Pref. High-Seas Fish., Guidance Center, 1953. Murphy and Shomura, 1953a, 1953b. Schweigger, 1949. Size composition Aikawa, 1937. Aikawa and Kato, 1938. Bonham, 1946. Brock, 1943. Hart et al., 1948. Inanami, 1942c. Kagoshima Pref. Fish. Expt. Sta., 1937a. Kamimura and Honma, 1953. Kanagawa Pref. Fish. Expt. Sta., 1951a, 1951b, 1952b. Kawana, 1934. Kawasaki, 1952. Kida, 1936. Kikawa, 1953. Kimura, 1935, 1941, 1942a. Kimura and Ishii, 1933a. Mine and lehisa, 1940. BIBLJOGRAPHT ON THE TUNAS 245 Size composition — Continued Miyazaki Pref . High-Seas Fish. Guidance Center, 1953. Murphy and Shomura, 1953a, 1953b. Nakamura et al., 1953. Nankai Reg. Fish. Res. Lab., 1951b. Nomura et al., 1952-53. Okamoto, 1940. Okinawa Pref. Fish. Expt. Sta., 1931. okuma et al., 1935. Onodera, 1941. Partlo, 1951. Powell, D. E., 1950. Powell and Hildebrand, 1950. Powell et al., 1952. Sasaki, 1939a, 1939b. Scagel, 1949. Schaefer, 1948b, 1951. Schaefer and Marr, 1948b. Schaefers, 1953. Schweigger, 1949. Serventy, 1947. Sette, 1954. Tauchi, 1940a, 1940b, 1940c. Uda, 1932b. Uda and Tsukushi, 1934. Westman and Neville, 1942. Yabe and Mori, 1950. Yabuta and Ueyanagi, 1953a, 1953b. Yamanaka, 1950. Skipjack. See Katsuioonus. Skipjack, black. See Euthynnus spp. S6dagatsuwo. See Auxis spp. South China Sea Anonymous, 1938. Kanamura and Imaizumi, 1936b. Southern bluef in tuna. See Thunnus maccoyi Spawning. See Reproduction. Statistics Alaejos, 1931. Anderson, Stolting, et al., 1953. Anonymous, (1), (2), (3), 1932, 1945, 1947, 1949a, 1949b, 1952. California. Dept. Fish and Game California. Dept. Fish and Game, Marine Fish. Br. Chiba Pref. Fish. Expt. Sta., 1936b. Chiba Pref. Fish. Expt. Sta., Katsuura Branch, 1938b, 1941e, 1941f. Conseil Int'l'. pour 1' Exploration de la Mer, 1933. Ego and Otsu, 1952. Ehrenbaum, 1934. Espenshade, 1948. Farina, 1931a. Federation of Japan Tuna . . . 1951a, 1951b, 1952, 1953a, 1953b. Food and Agr. Organ. U. N., 1949a. Godsil, 1937, 1949. Kanai, Moto and Kasu, 1938. LUling, 1951, 1952b. Nakayama, 1948. Statistics — Continued Navaz, 1950. South Seas Gov't 1938. U. S. Fish and Wildlife Service. Stomach contents. See Food. Striped tuna. See Katsuwonus. Sumagatsuo. See Euthynnits yaito. Synonymy Barnard, 1948. Beebe and Tee-Van, 1936. Boeseman, 1947. Chevey, 1932b. Chu, 1931. DeBeaufort and Chapman, 1951. DeBuen, 1935. Fish, 1948. Food and Agr. Organ. U. N., 1949b. Fowler, 1931, 1934, 1936, 1949. Frade, 1931c. Fraser-Brunner, 1949, 1950. Ginsburg, 1953. Heldt, 1930, 1931a, 1931b. Herre, 1936. Hildebrand, 1946. Joubin, 1934. LeGall, 1934a, 1934b, 1934c, 1934d. Maldura, 1946. Molteno, 1948. Nakamura, 1939b, 1939c. Nichols and LaMonte, 1941. Powell, A. W. B., 1937. Rosa, 1950. Schaefer and Walford, 1950. Schultz, 1949. Schultz and DeLacy, 1936. Serventy, 1942b. Soldatov and Lindberg, 1930. Tanaka, 1931. Wade, 1949. Whitley, 1937. Tagging Alverson and Chenowith, 1951. Anonymous, 1939a. Conseil Int'l pour 1' Exploration de la Mer, 1933. Fukuda and lizuka, 1940b. Ganssle and Clemens, 1953. Godsil, 1936, 1938b, 1938c. Heldt, 1932b. Kagoshima Pref. Fish. Expt. Sta., 1936c, 1938c, 1939c, 1940c. Kawana, 1934. Matsumoto, T., 1937. Partlo, 1950, 1951. Powell, D. E., 1950. Powell et al., 1952. Russell, F. S., 1934a. Scagel, 1949. Schaefers, 1952, 1953. South Seas Gov't. . . . liJ41b. 246 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Tagg-ing — Continued Uda, 1936a. Wilson, 1953. Temperature. See Body temperature, oceanographic conditions. Thon blanc. See Germo. Thon rouge. See Thunnus thynnus. Thunnidae Bellbn and Bardan de Bell6n, 1949. Nakamura, 1939b, 1941. Thunniformes Berg, 1947. Thunniinae Frade, 1932. Thunmis Aikawa, 1933. Aikawa and Kato, 1938. Alaejos, 1931. Aricb and Genovese, 1953. Bahr, 1952. Barnhart, 1936. Bigelow and Schroeder, 1953. Blackburn and Rayner, 1951. Boeseman, 1947. Bonamico, 1933. Brock, 1938, 1949. Cerquetelli, 1936. Chabanaud, 1930. Conrad, 1937. Conseil Int'l pour 1' Exploration de la Mer, 1933. Crane, 1936. DeBeaufort and Chapman, 1951. DeBuen, 1930, 1931, 1932, 1935, 1937. De La Tourrasse, 1951. Delsman, 1933. Dieuzeide, 1930, 1931. Dung and Royce, 1953. Ehrenbaum, 1934. Fick, 1937. Fish, 1948. Fisheries Society of Japan, 1931. Food and Agr. Organ. U. N., 1949b. Fowler, 1931, 1934, 1936, 1938, 1944. Frade, 1930a, 1930b, 1931a, 1931b, 1931d, 1935, 1937a, 1937b, 1953. Frade and Managas, 1933. Fraser-Brunner, 1950. Fujii, 1932. Galtsoff, 1952. Genovese, 1952, 1953. Ginsburg, 1953. Godsil, 1945, 1949b. Godsil and Byers, 1944. Godsil and Holmberg, 1950. Heldt, 1930, 1931a, 1931b, 1932b, 1934, 1937, 1938, 1943. Herre, 1936, 1940. Hildebrand, 1946. lehisa, 1939. June, 1952a. Thunnus — Continued Kawana, 1934, 1935, 1937, 1938. Kida, 1936. Kimura, 1932, 1933, 1935. Kimura and Ishii, 1932, 1933a. LeDanois, 1933. LUling, 1950, 1951, 1952a, 1952b. Maldura, 1946. Marukawa, 1939c. Mather, 1954. Mazzarelli, 1935. Migita and Arakawa, 1948. Mine and lehisa, 1940. Miyama and Osakabe, 1938, 1940. Molteno, 1948. Morice, 1953a, 1953b. Murayama and Okura, 1950, 1952. Murray, 1952. Nakamura, 1938, 1939a, 1939b, 1943, 1949, 1951. Nankai Reg. Fish. Res. Lab., 1951a. Navarro and Lozano, 1950. Navaz, 1950. Nigrelli and Stunkard, 1947. Nomura et al., 1952-53. Oita Pref. Fish. Expt. Sta., 1930. Okada and Matsubara, 1938. Okada et al., 1935. Okinawa Pref. Fish. Expt. Sta., 1940b. omori and Fujimoto, 1940. omori and Fukuda, 1938, 1940. Priol, 1944. Reiss and Vellinger, 1929. Rivas, 1951, 1953. Roedel, 1948a. Ros6n, 1943. Russell, F. S., 1933a, 1933b, 1934a, 1934b. Sanzo, 1932. Schaefer, 1948c, 1951. Schaeffers, 1952, 1953. Schultz, 1949. Schultz and DeLacy, 1936. Schweigger, 1949. Scordia, 1930, 1939a, 1939b, 1940, 1943. Sella, 1930, 1931, 1952. Serventy, 1941a. 1941b, 1942b, 1947. Shapiro, 1948a, 1948b. Shimada, 1951b. Shimizu, 1947. Society for the Promotion . . . 1936. Soldatov and Lindberg, 1930. Sugiura, 1932. Suyehiro, 1942. Takayama and Ando, 1934. Tanaka, 1931, 1939. Taranetz, 1937. Tauchi, 1940a. Tinker, 1944. Uda, 1932a, 1932b, 1935a, 1940b, 1952. Van Campen, 1952. Walford, 1931, 1937. BIBLIOGRAPHY ON THE TUNAS 247 Til iin nus — Continued Westman and Gilbert, 1941. Westman and Neville, 1942. Whitehead, 1930, 1931. Whitley, 1947. Wolfe Murray, 1932. Yabe et al., 1953. Thuntui^ alalunga. See Germo. Thunnus albacora. See N. macropterus. Thutinus germo. See Germo. Thunnus maccoyi Blackburn and Rayner, 1951. Boeseman, 1947. Dung and Royce, 1953. Godsil and HoUnberg, 1950. Nomura et al., 1952-53. Serventy, 1941a, 1941b, 1947. WhiUey, 1947. Thunnus macropterus. See N. macropterus. Thtmnus mebachi. See Parathunnus mebachi. Thunnus nicolsoyii Serventy, 1942b. Thunnus obesus. Fraser-Brunner, 1950. Thunytus orientalis Aikawa, 1933. Aikawa and Kato, 1938. Dung and Royce, 1953. Fisheries Society of Japan, 1931. Fowler, 1934. lehisa, 1939. June, 1952a. Kawana, 1934, 1935, 1937, 1938. Kida, 1936. Kimura, 1932. 1933, 1935. Kimura and Ishii, 1932, 1933a. Migita and Arakawa, 1948. Mine and lehisa, 1940. Miyama and Osakabe, 1938, 1940. Murayama and Okura, 1950, 1952. Nakamura, 1938a, 1939a, 1939b, 1943, 1949, 1951. Nankai Reg. Fish. Res. Lab., 1951a. Nomura et al., 1952-53. Oita Pref. Fish. Expt. Sta., 1930. Okada and Matsubara, 1938. Okada et al., 1935. Okinawa Pref. Fish. Expt. Sta., 1940b. omori and Ftijimoto, 1940. Omori and Fukuda, 1938, 1940. Shapiro, 1948a, 1948b. Shimada, 1951b. Shimizu, 1947. Society for the Promotion . . . 1936. Sugiura, 1932. Suyehiro, 1942. Takayama and Ando, 1934. Tanaka, 1939. Tauchi, 1940a. Tinker, 1944. Uda, 1932b, 1940b, 1952. Thunnus orientalis — Continued Van Campen, 1952. Yabe et al., 1953. Thuimus rarus. See Neothunnics varus. Thunnus schlegeli. See Thunnus orientalis Thunnus sibi. See Pai-athunnus sibi. Thunnus thunniiia. See Euthynnus alletteratus. Thunnus thunmis. See Thunnus thynnus. Thunnus thynnus Alaejos, 1931. Aricb and Genovese, 1953. Bahr, 1952. Barnhart, 1936. Bigelow and Schroeder, 1953. Bonamico, 1933. Brock, 1938, 1949. Cerquetelli, 1936. Chabanaud, 1930. Conseil Int'l pour 1' Exploration de la Mer, 1933. Crane, 1936. DeBuen, 1930, 1931, 1932, 1935, 1937. De La Tourrasse, 1951. Delsman, 1933. Dieuzeide, 1930, 1931. Dung and Royce, 1953. Ehrenbaum, 1934. Fick, 1937. Fish, 1948. Food and Agr. Organ. U. N., 1949b. Fowler, 1931, 1936, 1938, 1944. Frade, 1930a, 1930b, 1931a, 1931b, 1935, 1937a, 1937b. Frade and Managas, 1933. Fraser-Brunner, 1950. Fujii, 1932. Galtsoff, 1952. Genovese, 1952, 1953. Ginsburg, 1953. Godsil, 1949b. Godsil and Byers, 1944. Heldt, 1930, 1931a, 1931b, 1932b, 1934, 1937, 1938, 1943. Herre, 1936, 1940. June, 1952a. Kida, 1936. Kimura, 1932. LeDanois, 1933. Luling, 1950, 1951, 1952a, 1952b. Marukawa, 1939c. Mather, 1954. Mazzarelli, 1935. Molteno, 1948. Morice, 1953b. Murray, 1952. Navarro and Lozano, 1950. Navaz, 1950. Nigrelli and Stunkard, 1947. Priol, 1944. Reiss and Vellinger, 1929. Rivas, 1951, 1953. Roedel, 1948a. 248 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Thunnus thynnus — Continued Ros6n, 1943. Russell, F. S., 1933a, 1933b, 1934a, 1934b, 1936. Sanzo, 1932. Schultz, 1949. Schultz and DeLacy, 1936. Scordia, 1930, 1939a, 1939b, 1940, 1943. Sella, 1930, 1931, 1932, 1952. Society for the Promotion . . . 1936. Soldatov and Lindberg, 1930. Tanaka, 1931. Taranetz, 1937, 1944. Uda, 1932a, 1935a. Walford, 1931, 1937. Westman and Gilbert, 1941. Westman and Neville, 1942. Whitehead, 1930, 1931. Wolfe Murray, 1932. Thunnus tonggol DeBeaufort and Chapman, 1951. Fraser-Brunner, 1950. Serventy, 1942b. Thunnus zacalles. See Kishinoella zacalles. Thynnus af finis. See Euthynnus allettei-atus. Thynnus alalonga. See Germo. Thynnus germo. See Germo. Thynnus maccoyi. See Thunnus maccoyi. Thynnus macroptertis. See N. macropterus. Thynnus orientalis. See Thunmis orientalis. Thynnus pacificus. See Germo. Thynnus pelamys. See KO'tsuwonus. Thynnus sibi. See Parathunnus sibi; also Germo. Thynnus Vumina. See Euthynnus alletteratus. Thynnus thunnina. See Euthynnus alletteratus. Thynnus thynnus. See Thunnus thynnus. Thynnus tonggol. See Thunnus tonggol. Tonno. See Thunnus thynnus. Tuna (otherwise unspecified) Aikawa, 1932. Anonymous, 1939b. Auffret, 1931. Bini, 1931, 1933. Chiba Pref. Fish. Expt. Sta., Katsuura Branch, 1941e. Corwin, 1930. DeBuen and Frade, 1932. Domantay, 1940. Farina, 1931b. Federation of Japan Tuna . . . 1951a, 1951b, 1952, 1953a, 1953b. Flett, 1944. Food and Agr. Organ. U. N., 1949a. Frade, 1932. Godsil, 1938c. Had2i, 1934. Hasegawa, 1937. Heldt, 1932a. Hirtz, 1933. Imai, 1950. Imaizumi, 1937. Tuna (otherwise unspecified) — Continued Imamura, 1953. Isawa, 1935. June, 1951a. Kafuku, 1950. Kanagawa Pref. Fish. Expt. Sta., 1951b. Kawana, 1935. Kimura and Ishii, 1931. Kodama, lizuka, £ind Harada, 1934. Kreutzer, 1951b. LeGall, 1934e, 1951. McKeman, 1953. Marr and Schaefer, 1949. Marukawa, 1939a. Matsui, K., 1942a. Matsui, Y., 1938. Meyer, 1951. Mie Pref. Fish. Expt. Sta., 1950a, 1950b. Morovid, 1950. Murphy and Niska, 1953. Murphy and Shomura, 1952, 1953a, 1953b. Nishikawa, 1934. Niska, 1953. Niwa, 1937. Noguchi, 1938. Okumura, 1943. Postel, 1949. Rasalan, 1950. Rawlings, 1953. Ronquillo, 1953. Rosa, 1950. Saito, 1937. Sakai and Uno, 1940. Scordia, 1940. Sella, 1932. Sette, 1954. Shapiro, 1950. Shimada, 1951a. Society for the Promotion . . . 1937b. goljan, 1930. South Seas Gov't. . . . 1937b, 1941a. Tanaka, 1935, 1936. Tester et al., 1952. Thiel, 1938. Tomiyama, 1933. U. S. Fish and Wildlife Service Vitlov, 1949. Wilson, 1953. Zei, 1948. Wanderer Whitley, 1937. Weather correlated with fishing or distribution Murphy and Niska, 1953. Murphy and Shomura, 1953b. Yellowfin tuna. See Neothunnus macropterus. Young Bini, 1952. BIBLIOGRAPHY ON THE TUNAS 249 Young — Continued Delsman, 1931. Delsman and Hardenburg, 1934. Eckles, 1949b. Greenhood, 1952. Hatai et al., 1941. Herald. 1951. Inanami, 1942d. Kimura and Ishii, 1931. LeDanois, 1951. LeGall, 1949. Marr, 1948. Marukawa, 1939b. Young — Continued Sanzo, 1932, 1933. Schaefer, 1948c. Schaefer and Marr, 1948a, 1948b. Sette, 1954. Shimada, 1951b, 1951d. Suda, 1953. Uchida, 1937. Wade, 1949, 1950a, 1951. Yabe, 1953. Yabe et al., 1953. Yabe and Mori, 1948. 02' S. 2°07' N. 1°19' S. 2-54' N, 1°04' N. 4°04' N. 2°56' S. 3°49' N. 4°04' N. 4°04' N. 3°49' N. 6°08' S. 4°04' N. 4°C4' N, 4°04' N. 4°04' N. 4°04' N. 1°20' S, 1°59' N. 1°20' S. 1°20' S, 1°20' N. r20' S, 1°69' N, 1°69' N. 1°69' N. 1°67' N. r59'N. 1°20' N. 1°59' N. 1°20' S. 1°57' N. 1020' S. 1°59' N, 1°59' N. 1°20' S, 2°46' N. 1°59' N. X'iT N. 1°59' N. 3°51' N. 1°59' N. 2°42' N, 1°57' N. 1°69' N. 2°42' N. l-S?' N. 1°69' N, 3''52' N. 3°6r N. 1°59' N. 0°22' S. 154°56' W. 154°59' VV. 154°59' VV. 154°56' W. 156''41' VV. 156°41' VV. 162°19' VV. 162°19' VV. 162°Q6' VV. 162°06' VV. 162°06' VV. 168°17' VV. 168°21' VV. 161°15' VV. I71°05' VV. 171°31' VV. 171°05' VV. 171°05' VV. 171°16' VV. 150°00' VV. 150°09' VV. 1.54°,57'VV. 150°36' VV. 150°11' VV. 150°09' VV. 150°11' VV. 150°24' VV. 150°09' VV. 150°09' VV. 150°11' VV. 150°11' VV. 150°24' VV. 154°57' VV. 160°19' VV. 150°24' VV. 154°57' VV. 150°09' VV. 150°24' VV . 150°19' VV . 151°06' VV. 154°66' VV. 150°08' VV. 150°07' VV. 154°56' VV. 154°66' VV. 150°07' VV . 150°09' VV. 154''66'VV. 164°56' VV. 154°56' VV. 164°56' VV. 154°56' VV. 156°06' VV. 167°31' VV. 165°06' VV. 165°06' VV. 155°03' VV. 155°0C' VV. 157°31' VV. 157°31' VV . 157°31'VV. 167°32' VV. 167°31' VV. 155°03' VV. 167°31' VV. 166°06'W. 167°32' VV. 156°06' VV. 157»31' VV. 157°31' VV. 155°06' VV. 155°10' W. 157°31' W. 157°32' VV. 157°31' W. 159°26' VV. 157°31' VV. 165°05' W. 157°32' VV. 167°31' W. 165°05' VV. 167°32'W. 157°31' W. 159°20' VV. 159°26' VV. 157°3I' VV. leo-or VV. M Im Im Im M In In In In Im Im Im Im Im In In Im Im Im M M M M M M M M M M M M M M In In In In In In In In In In In In Im Im Im Im Im Im Im M M M M M M M M M M M M M M In In In In In In In In In In In In In In In In In In Im Im Im Cm. 113 104 96 78 143 146 135 132 127 133 119 85 83 66 98 93 91 78 68 149 148 147 146 145 143 143 143 139 138 138 138 137 135 152 150 148 or 140 142 142 141 141 140 138 136 129 128 140 136 126 123 lis 114 108 148 148 147 143 142 139 139 138 134 133 130 126 124 76 142 140 140 138 137 136 136 135 136 134 133 132 131 131 130 130 130 98 123 120 112 Feb. 10, 1951--- 0°22' S. 3°51' N. 3°51' N. 1°69' N. 0°22' S. 5°53' N . 0°22' S. 5°63' N. 3°52' N. 5°53' N. rei'N. 3°62' NT. 3°62' N. 5°63' N. 3°52' N. 1°51' N. 4°30' S. 3°07' S. 2°60' S. 3°15' S. 4°30' S. 3°07' S. 0°02' N. 0°02' N. 2''39' .?. 4°03' S. 2°39' S. 4°03' S. .3°00' N. 1°18' S. 1°I8' S. 6°47' S. 8°00' S. 1°00' N. 1°00' N. 1°51' S. I°48' S. TOO' N. l^OO' S. 1°00' N. 4°12' N. 0°09' N. 0°09' N. 6°16' S. 1°00' S. 1°00' S. 3°07' N. 2°12' N. 3°25' S. 0°09' N. 2°12' N. 3°07' X. 1°48' S. 1°00' S. l°0O' N. 1°00' N. 4°12'N. 4°12' N. 2°I2' N. 2°12' .M. 3°07' N. C°09' N. 6°16' S. 1°0()' S. I°51' S. 3°07' N. 2°12' N. 4°12' N. 13°31'S. 13''31' S. 4''10'N. 6°40' S. 6°40' S. 1°20' S. 6°18' S. 1°20' N. o'or N. l'=20' .S. 2°50' N. 2°50' N. 6°40' S. 0°01' N. 2°43' S. l'>20' N. 2°50' N. 4''02' S. 1°61' N. lecoi' W. 159°26' W. i59°26' v^^ 157°31' W. i6o°or w. 162°05' VV. 160°01' VV. 162°05' VV. 169°20' VV. 162°05' VV . 167°20' VV. 159°20' W. 169°20' VV. 162°06' W. 159°20'VV. 157°20' W. 172°10' W. 171°05' VV. 171°40' W. 171°30' VV. 172''10' W. I71°05' VV. 179°48' E. 179°48' E. 179°54' E. 179°58' E. 179°64' E. 179°68' E. I80°00' 180°00' 180°00' 179°59' VV. I79°56' E. 140°00' VV. 140°00' W. 140°11' VV. 139°69' VV. 140°00'VV. 140°05' VV. 140°0O' VV, 140°20' W. 139°47' VV. 139''47' VV. 141°32' W. HOBOS' VV. 140°05' W . 140°07' VV. 140°18' W. 140°03' W. 139<'47' VV. 140°18' W. I40°07' VV. 139°59' VV. 140°05' W. 140°00' VV. 140°00' W. 140°20' W. 140°20' VV. 140°18' VV. 140°18' VV. 140°07' W. 139°47'VV. 141°32' VV. 140°05' VV. 140°11' VV. 140°07'VV. 140°18' VV. 140°20' VV. 147°08' VV. 147°OS' VV. 168°30' VV. 169°03'W. 169°03' VV. 169°00' VV. 169°03' W. 169°05' W. 169°04' VV. 169°00' VV. 169°07' VV. 169°07' VV. 169°03' VV. 169<'04' W. 169°00' W. 169''05' VV. 169°07' VV. 169°04' W. 157°2n' \V. Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im M M Im Im Im Im M M M M M M M M M In In R R R M M M M M M M M M M M M M M -M M M M M M M M M M M M M M M M M In M Im R M M M M M M M M M M M M M In In M Cm. IK Jan. 27, 1951 Feb. 5, 1953 Do 10" Do --- lOi Jan 26, 1951 Feb. 3, 1963.. 10^ Jan. 31, 1953 Feb. 16, 1961 Feb. 21, 1951 Feb. 11, 1951... 9 Do Jan. 23, 1953 9( 8f Do _-. Feb. 21, 1961 .- 8> Jan. 25, 1953 Feb. 4, 1951.. 8; Do.--- -- Do Feb. 21, 1951...- -.- Feb. 16, 1951 Feb. 17, 1951 8 Jan. 21, 1951 7 Jan. 19, 1951 Feb. 19, 1961 Feb. 22, 1961. Jan. 27, 1953 Jan. 26, 1951- Jan. 31, 1951 Feb. 12, 1951 Feb. 13, 1951. 7 7 Jan. 26, 1951 Do Feb. 8, 1950 Feb. 5, 1951 Feb. n, I960. 12( 10 Jan. 29, 1961 . 12 Feb. 11, 1953.- - Feb. 10,1962 Feb. 8, 1950 -.- Feb. 5, 1951 Feb. 18, 1952 Do Feb. 20, 1952 Feb. 21,1952-- - Feb. 20, 1952 Feb. 21,1962 Feb. If), 1952 Feb. 19, 1952 Do -.. Feb. 23, 1962 - Feb. 24, 1962 Mar. 15, 1953 Do Mar. 12, 1963--- Mar. 11, 1963 _. Mar 15, 1953 8( Feb. 4, 1953 Feb. 4, 1952-.- Feb. 14, 1953 - Feb. 12, 1953 7 6 14 14 Feb. 4, 1953 Feb. 12, 1953 14 I4( Feb. 13, 1953-- 13 Feb. 4, 1953 Do 13 13 Feb. 12, 1953-.- Do 13 12 Feb. 13,1963 Feb. 4, 1952 Feb. 3, 1953 13 13 14 Feb. 13, 1953 14 Feb. 4, 1952 14 Feb. 4, 1953 16 Feb. 13, 1953 15 Feb 3, 1963 Mar. 13, 1953 Mar. 16, 1953-- Mar 18, 1963 16 Feb. 10, 1953 . . 14 Feb. 1, 1962 14 Feb. 17, 1953 Mar 14 1963 14 Feb. 2, 1953 Do Mar. 8, 1953 14 Feb. 1, 1952 14 Do Mar 13 1953 14 Feb. 2, 1953 Do 14 Feb. 19, 1963 Mar 17 1963 14 Feb. 1, 1952 Mar 16, 1953 14 Do Mar 10 1963 14 Do - 14 Do Mar 16 1953 14 Do Mar. 17, 1953 Mar. 11,1963.-- --- Mar. 13, 1963 14 Feb. 6, 1962 14 Feb. 3, 1963 14 Feb. 5, 1952 14 Do Do 14 Feb. 3, 1962 Mar. 18, 1953 Do Mar. 16, 1953 Do 14( Feb. 5, 1952 141 Feb. 3, 1953 13 Do --_ -- 13 Do Mar 17 1953 13 Feb. 1, 1953 _-- 13 Feb. 3, 1953 Mar 8 1953 13 Feb. 3, 1952 __- 13 Feb. 3, 1953 Mar 12 1953 13 Feb. 5, 1952 13 Feb. 1, 1953 Mar 16 1953 13 Feb. 5. 1952 Mar. IS, 1953 Mar 4, 1953 . - - 14 Feb. 3, 1953 ,-- 14 Do -„ Do 7 Feb. 5, 1952 14 Feb. 2, 1962 Mar 4 1952 14 Feb. 3, 19.53 Do 14 Feb. 1, 1953 Mar 8 1952 14 Feb. 3, 1953 14 Feb. 5. 1963 -- Mar 10 1952 14 Feb. 3, 1953-- 14 Feb. 6, 1952 Mar 8 1952 14 Feb, 1, 1963-- 14 Feb. 3, 1963-- -- Do 13 Feb. 6, 1952- - 13 Feb. 1, 1953 Mar 9 1952 13 Feb. 3, 1963 - -... 13 Feb. 4, 1961 Mar 10 1952 13 Feb. 5, 1953- Mar. 11, 1952 - Mar fi 1952 14 Feb. 3, 1953 12 Feb. 11,1951 Apr. 30, 1951 - 11 SPAWNING OF YELLOWFIX TUNA 253 T.VBLE 1. — Daln on 7/,0 i/ellowfin tuna specimens from the central equatorial Pacific for which matiirilij determinalions were made in the laboratory — Continued Date Position Stage of maturity Fish length Cm. 98 108 96 96 81 110 104 98 98 98 92 90 80 80 80 75 70 148 146 146 135 151 146 146 143 140 140 136 136 134 133 133 130 129 129 127 123 122 121 117 114 106 101 100 92 92 86 128 122 121 118 117 115 115 114 110 109 108 107 106 104 104 100 lOO 98 98 94 92 90 90 89 88 88 88 85 82 149 115 112 111 111 108 104 103 102 101 101 101 100 96 Date Position Stage of maturity Fish Latitude Longitude Latitude Longitude length Apr.26.1951. Anr 27 1 951 1°51' N. I'sr X, \'sy X. I'si'x. I'Sl' X. s-ss' X. 5°53' X. 0°22' S. 5''53' X. 5°53' N. 5°5.3'X, 5°53' X. psr N. 1''51' X. 1°51'X. i°5rx. S-SS'X. 7°09' X. 4''18' X. 4°18' X. 4°18' X. 4°55' X. 4°02' X. 4°02' X. 4°02' X. 4°55' X. 3''58' -V. 3''52' X. 1°51'X. 4°45'X. 4°55' X. 4'>52'X. 6'>02'X. 5<>26'X. 4°.55' X. 6''25'X. 4°56' X. 4°45' X. 4°17' X. 4°02' X. 4°17' X. 6°25' X. 6-25' X. 6°25' X. 6°25' X. 6'>25' X. 1°51' X. 4°52' X. 4°02' X. 4°45' X. 5°,53'X. 3''58' X. 4°42'X. 6''25'X. 3°58' X . 4<'17'X. 6°25'X. 3''58' X. 1°51'X. 5°58' X. 5-53' X. 3° 58' X. 1°51' X. 6°25' X. 1°51' X. 3°58' N. 6°25' X. fi°25'X. 6°25'X. 6°25' X. 6°25' X. 6''25' X. 6°25'X. 6°25'X, 6°25' X. 6''25' X. S'SS' X. 6°25' N . 5'>53' X . 4''42' X. 5°58' X. 4°42' X, 4°42' X. 3°58' N. 4''02' X. 5'>S8' X. 4°02' X. 3"'68' X. 6°58' X. 6''25' X. 157°20' W. 157020' W. 157°20' W. 157°2fl' W. 157''20' W. 162°05' W. 162''05' W. 160°01' \V. 162°05' W. 162°05' W. 162°05' \V. 162°05' W. 157°20' W. 157°20' W. 157-20' W. 1.57°20' W. 162°05' W. 119°00' W. 119"'35'W. 119''35' W. 119°35' \V. 161''19' W. 159''34' W. 159°34' W. 169''34' \V. 161°19' W. 159°04' W. 159°20' W. 157°20'\V, 160°11' W. 161°19' W. 1.59°35' W. 162°28' W. 161°37' W. 16ri9' W. 162''26' W. 160°32' W. 160°11' W. 160°28' W. 159-34' W. 160-28' W. 162-26' W. 162-26' W. 162-26' \V. 162-26' W. 126-26' W. 157-20' W. 159-35' W. 159-34' VV. 160-11' W. 162-05' W. 159-04' W. 160-24' W. 162-26' W. 159-04' W. 160-28' W. 162-26' W. 159-04' W . 157-20' \V. 162-52' \V. 162-05' W. 159-04' W . 157-20' W. 162-26' W. 157-20' W. 159-04' W. 162-26' W. 162-26' W. 162-26' W. 162-26' W. 162-26' W. 162-26' \V. 162-26' \V. 162-26' VV. 162-26' \V. 162-26' W. 162-52' W. 162-26' W. 162-05' W. 160-24' W. 162-52' W. 160-24' W. 160-24' \V. 159-04' W. 1.59-34' W. 162-52' W. 159-34' W. 159-04' W. 162-52' W. 162-26' W. M In In In In Im Im Im Im Im Im Im Im Im Im Im Im R M M M M M M M M M M M M M M M M M M M M M M M M M M M M M In In In In In In In In In In In In In In In In In In In In In In In In In In In In In Im Im Im Im Im Im Im Im Im Im Im Im Im Ira Mav 28. 1954 Mav 30, 1951 4''02' N. 6-25' N. 6-26' X. 4-56' X. 6-25' X. 6-25' X. 6-25' X. 6-25' X. 6-25' N. 1-51' X. 1-51' X. 1-61' X. 1-51' N. rsi' X. 3-52' X. 4-42' N. 6-25' X. 6-25' X. 5-53' X. 6-25' X. 4-26' S. 3-27' S. 6-51' S. 4-26' S. 4-26' S. 4-26' S. 4-26' S. 6"'41' S. 4-26' S. 4°26' R. 0-19' X. 6-06' X. 2-19' X. 6-06' X. 8-00' X. 6-06' N. 8-00' X. i-or s. 0-21' X. 2-19' X. 2-19' X. 2-19' X. 6=25' X. 1°,52' .X. 2-29' X. 3-04' X. 1-47' X. 3-04' X. 2-29' N. 1-52' X. 6-25' X. 4-42' X. 6-25' X. 6-53' N. 5-53' X. 5-53' N. 1-47' X. 3-04' N. 1-52' N. 2-03' X. 2-03' X. 3-04' X. 4-42' N. 2-or N. 2-03' X. 2-or X. 1-52' N. 4-42' X. 2-01 ' X. 3-52' X. 3-52' X. 3?52' X. 4-42' X. 3-52' X. 3-52' X . 3-52' X. 5-53' X. 5-53' X. 5-53' X. 5-53' X. 0-30' S. 2-14' S. 0-30' S. 0-30' S. 0-30' S. 2-14' S. 0-30' S. 2-14' S. 2-50' S. 2-50' S. 159-34' W. 162-26' W. 162-26' \V. 160-32' W. 162-26' VV. 162-26' VV. 162-26' VV, 162-26' VV. 162-26' VV. 157-20' VV. 157-20' VV. 157-20' VV. 157-20' VV. 157-20' VV. 159-20' VV. 160-24' VV. 162-26' VV. 162-26' VV. 162-05' VV. 162-26' VV. 170-09' VV. 170-12' VV. 170-02' VV. 170-09' VV. 170-09' VV. 170-09' VV. 170-09' VV. 169-44' VV. 170-09' VV. 170-09' VV. 119-58' VV. 129-55' VV. 130-07' VV. 129-55' VV. 130-24' VV. 129-55' VV. 130-24' VV . 125-56' VV. 129-23' VV. 130-07' VV. 130-07' VV. 130-07' VV. 162-26' VV. 156-47' VV. 158-22' VV. 1.59-13' VV. 158-16' VV. 1,59-13' VV. 158-22' VV. 156-47' VV. 162-26' VV. 160-24' VV. 162-26' VV. 162-05' VV. 162-05' VV. 162-05' VV. 158-16' VV. 159-13' VV. 156-47' VV. 157-40' VV. 157-40' VV. 159-13' VV. 160-24' VV. 157-09' VV. 157-40' VV. 157-09' VV. 156-47' VV. 160-24' VV. 1.57-09' VV. 1.59-20' VV. 159-20' VV. 159-20' VV. 160-24' VV. 1,59-20' VV. 159-20' VV. 1,59-20' VV. 162-05' VV. 162-05' VV. 162-05' VV. 162-05' VV. 169-52' VV. 170-00' VV. 169-52' VV. 169- .52' VV . 169-.52'VV. 170-00' VV. 169-52' VV. 170-00' VV. 171-40' VV. 171-40' VV. Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im R M M M M M M M t'" Im M R R R R R R M M M M Im R M M M M M M M M M M M M M In In In In In In In In In Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im R M M M M M M M Im Im Cm. 93 90 Do Mav 20, I95I 88 Apr 30, 1951 Mav 24, 1954 May 30, 1951 88 Anr 26 1 951 88 Do 87 Anr 16 195! Do - 86 Apr. 11, 1951 Do - May 31, 1951.. 86 85 Apr. 16. 1950 - \pr 15, 1951 May 1,1951 Do Do 83 82 \nr 16 1951 81 Do 80 Apr. 26. 1951 Apr 27 1951 Do 80 May 7, 1951 79 Anr 26 1951 May 3, 1950 -- . 75 \Dr 13 1951 May 11, 1951 75 May 29. 1952 Do May 27, 1950 Do 75 Mav 31, 1952 74 Do 74 Do Mav 30, 1953 134 Mav 23, 1954 May 31,1953 May 28, 1953 149 Mav 28, 1954 148 Do -.- Mav 30, 1953 144 Do "Do 136 Mav 23, 1954 Do 136 Mav 30 1954 Do 133 Mav 7, 1951 May 29, 1953 103 Mav 13, 1950 -- Mav 30, 1953 130 Mav 25, 1954 Do 94 Mav 23, 1954 June 3, 1952 - June 12 1952 157 Mav 26, 1954 156 Mav 18, 1954 June 9, 1952 150 Mav 22, 1954 June 12, 1952 - 142 Mav 23, 1954 June 13, 1952 139 Mav 31, 1951 - June 12, 1952 129 May 24, 1954 June 13 1952 94 Mav 25, 1954 --- 153 Mav 27, 1954. June 8, 1952 - 153 Mav 28, 1954 June 9 1952 146 May 27, 1954 Do 135 Mav 31, 1951 Do 142 Do. ..:.: 90 Do June 7 1954 152 Do. June 2 1954 147 Do June 1 1954 146 Mav 1, 1951 Tune 9 1954 142 May 26, 1954. 140 Mav 28, 1954 --- 138 Mav 25, 1954... 126 May 28, 1950. June I. 1951 118 Mav 30, 1954 , June 5, 1950 " -.- 99 May 3, 1950. .- 90 Mav 31, 1950 Do 87 May 30, 1954 Do. 84 Mav 27, 1954-- Do 83 Mav 29, 1950 128 Mav 30, 1954 Tune 1 1954 118 May 12, 1950 June 7, 1954 115 May 17, 1954 June 4, 1954... Do 114 May 11, 1951 112 May 30. 1954 no May 1. 1951 June 4 1950 no Mav 31, 1951 Junes, 1954 108 Mav 1. 1951 __.. June 4, 1954 IOC May 30, 1954 June 8 1954 12C Mav 30, 1951 _.. June 7, 1954 no May 12, 1951 June 4, 1950 IOC May 30, 1951 Junes. 1954 94 Mav 31. 1951 Tune 6 1951 85 May 30. 1951. Do 83 Mav 31. 1951 Do 83 Do June 4. 19.50 82 Do 82 May 28, 1950 Do 82 May31,I951 Do 8C May 17, 1954 June 1 1951 77 May 31, 1950 Do 76 May 27, 1950. May 3, 1950 Do 76 June 3 1950 75 May 17, 1954 June 2. 19,53.-- June 1 1953 138 May 3. 1950 142 1)0 Mav 30, 1954.. May 28, 1954. Mav 17, 19.54. May 28. 19.54 May 30, 19.54 May 17. 19.54 Mav 3(1, 1961 ... June 2. 1953.-- Do Do June 1. 1953 141 133 131 129 June 2. 1953 June 1. 1953 121 102 June 12. 1951 135 June 17. 19.51 122 254 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 1. — Data on 740 yellowfin tuna specimens from the central equatorial Pacific for ivhich maturity determinations were made in the laboratory — Continued Date June 17, 1951- June 1. 1953- . Julv 17, 1950- July 15, 1950- July 19, 1950-. Do Aug. 24, 1952- Aug. 23, 1952. Aug. 24, 1952- Aug. 31, 1952- Aug. 2(;, 1952, Aug. 31, 1952- Aug. 2«, 1952- Aug. 24, 1952- Aug. 21, 1952- Aug. 23, 1952. Aug. 28, 1952- Aug. 29, 1962- Aug. 27, 1952- Aug. 31, 1952. Aug. 2fi, 1952- Aug. 29, 1962- Aug. 13, 1953- Do Do. Aug. 27, 1961. Aug. 13, 1953. Do Do Aug. 26, 1953. Aug. 16, 1953. Aug, 12, 1953. Aug. 19, 1953. Aug. 7, 1953.. Do Aug. 21, 1953. Aug. 25, 1953. Aug. 7, 19,i3-. Aug. 20, 1963- Aug. 21, 1953 Do Do Do Aug. 19, 1953- Aug. 25, 1953- Aug. 7, 1953- - Aug. 12, 1953- Do Aug. 19, 1963- Aug. 20, 1953- Aug. 7, 1953.- Do Aug. 12, 1953- Aug. 14, 1953. Aug. 19, 1953- Aug. 23, 1953. Aug. 7. 1953.- Do- Aug. 26, 1960- Aug. 14, 1953. Aug. 19, 1963- Aug. 14, 1963- Aug. 16, 1953- Aug. 14, 19.53- Aug. 18, 1953. Aug. 19, 1953. Do Aug. 14, 19.53. Aug. 18, 1953. Aug. 14, 1953. Aug. 21, 1953. Aug. 19, 1953. Aug. 21, 1953. Aug. 16, 1953- Aug. 18, 1953. Aug. 20, 1953. Aug, 21, 1963. Aug. 14, 1953. Aug. 25, 1953. Aug. 18, 1953. Aug. 16, 1960 Do Aug. 17, I960. Do- Do Sept. fi, 1952. Sept. 9, 19.52. Sept. 7, 1952. Sept. 9, 1952. Do Latitude Longitude a'sn' R. 2"! 4' s. 2",50' s. 2''.50' s. 2°.5n' s. 2"50' s. 4-28' N. ,6"16' N 4°28' N. 3"45' N. 2"23' N. 3''45' N. ron' N. 4°2R' N. 7''02' N. ,5-16' N i°on' N. 2"no' N. r33' N. 3"45' N. 2°23' N. 2''no' N. 0"flS' N. n"08' N. 0"0S' N. 9",36' N. flans' N. o"n8' N. n"08' N. 7''.5()' N. 4°10' ,S, 1»21' N 1°31' S. 2"n5' N. 2''n6' NT. rir N. B-IO' NT. 2"06' N. cor N, rii' N. i"ii' N. i"ii' N. i"n' N. r3i' S. 6''l(l' N. 2"05' N. 1-21' N. 1"21' N. 1"31' S. O^fll' N. 2"n5' N. 2"05' N. r2i' N 1°0S' S. r.3i' K. 3"22' N. 2''05' N. 2°06' N. (,"?.!<' N. IW S, rm' S. 1"08' s. 4"10' s. I'-OS' 8. 2''.56' K, 1°3I' S. rsi S, ro8' s. 2".56 s 1°08 R. rii' N. rsi' S. rn N. 2°.33 R 2°,56 R 0°01 N. ni N, rns S. 6°10 N. 2"5B H. a-ss R, a'-as W. 3''07 S, 3"n7 S. 3''07 S. 2-06 N. 2"33 N. 1''42 N. 2°33 N. 2°33 N. 171°40'W. 170°00' W. 171''40' W. 171°40' W. 17r40' W. 171''40' W . 139''51' W. 140°28' W. 139°61' W. 140°10'W. 140° 12' W. 140°10'W. 140°22' W. 139°61' W. 140°46' W. 140°28' W. 140°22' W. 140''40'W. 140°13' W. 140°in' W. 140°12' W. 140°40' W. 164°51' W. 154°51' W. 154°51' W. 150°06' W. 154°61' W. 154°5I' W. !54°61' W. I59°24' W. 1.55°33' W. 165''16' W. 159°63' W. 157°.'i8' W. 157°38' W. 160°08' \V. 160°02' W. 157°38' W . 159°56' W. 160''08' W. 160"'08' W. 160°08' W. 160°08' W. 159°53' W. 160°02' W . 167°38' W. 155°16' W. 155'>16' W. 159''53' W. 169°56' W. 167''38' W. 167°38' \V. 1.55°16' \V. 155°18' W. 169°.53' W. 160°24' \V. 157''38' W. 167''38' W. 162°26' W. 1.55°18' W. 169°53' W. 155°18' W . 1,55''33' W, 156°18' W . 160°14' W. 1.59°53' W. 159°63' W. 165''18' W. 160°14' \V. 15.5°18' W. 160°08' W. 159°53' W. lecos' W. 155''23' W. wo'ir W. 159°66' W. 160°08' \V. 155°18' \V. 160°02' W. 160°14' W. 171°31' W. 171°31' W. 171°0j' W. 171°06' W. 171°06' W. 140°.56' \V. 143°22' W. 141°24' W. 143°22' W. 143°22' W. Stage of maturity Im Im M M In Im M M M M M M M M M M In In In In In In M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M In In In In In In In In In In In In In In In Im Im Im Im Im Im M Im Im Im Im R R M M M Fish length Cm. 118 81 140 128 134 118 166 151 151 151 146 143 139 129 124 121 157 153 or 139 147 142 141 139 168 143 143 143 or 131 142 141 138 153 151 148 144 143 142 142 142 141 141 141 141 140 140 139 139 138 136 136 136 136 135 135 135 134 134 128 126 124 111 147 142 140 139 138 138 138 136 134 132 131 127 125 111 89 139 116 109 104 102 89 118 114 86 66 64 165 136 153 148 147 Date Sept. 3, 1952 Sept. 6, 19.52 Sept. 7, 1952 Do Sept. 2, 1962 Sept. 6, 19.52 Sept. 2, 1952 Sept. 4, 1952 Sept. 7, 1952 Sept. 1, 1952 Sept. 2, 1952 Sept. 4, 19.52- Sept. 5, 1951 Sept. 13, 1952 Sept. 2, 1951 Sept. 13, 1952 Sept. 16, 1952 Sept. 4, 1951 Sept. 16, 1952 Sept. 19, 1951 Sept. 3, 1951 Sept. 19, 1961 Sept. 3, 1961- Sept. 18, 1961 Sept. 19, 1961 Sept. 20, 1951 Sept. 3, 1961 Sept. 19, 1961 Sept. 3, 1961 Do Sept. 4, 1951 Sept. 6, 1951 Sept. 19, 1951 Sept. 4, 1951 Do Sept. 17, 1951 Sept. 18, 1951 Sept. 19, 1951 Sept. 2, 1961 Sept. 3, 1951 Sept. 5, 1951 Do- Sept. 19, 1951 Do Sept. 20, 1951 Sept. 2, 1961 - Sept. ,5, 1951 Sept. 19, 1951 Do Sept. 2, 1951 Do Sept. 3, 1951 Sept. 19, 1951 Do Sept. 20, 1951 Sept. 2, 1951 Do Sept. 3, 1951 Sept. 5, 1951 Sept. 13, 1952 Sept. 17, 1951 Sept. 19, 1961 Sept. 20, 1961 Sept. 4, 1951- Sept. 6, 1951 Do Sept. 18, 1951 Sept. 2, 1951 Do Sept. 5, 1951 Sept. 16, 1962 Sept. 2, 1961 Sept. 3, 1961 Sept. 18, 1951 Sept. 3, 1951 Sept. 6, 1961 - Do Sept. 3, 1951 Sept. 4, 1961- Sept. 5, 1951 Sept. 6, 1951 Sept. 2, 1961 Sept. 17, 1952 Sept. 4, 1961 Sept. 3, 1951 Sept. 18, 1961 Sept. 17, 1951 Sept. 19, 1951 Sept. 17, 1951 Sept. 19, 1951 Latitude Longitude 4°04' N. 2°06' N. 1''42' N. r42' N. 3°05' N. 2°06' N. 3''05' N. 3°20' N. 1°42' N. 3°3r N. 3°05' N. 3''20' N. 2°02' N. 1°22' N. 4°04' .Nf. 1°22' N. 2''28' N. r59' N. 2°05' N. 2°00' N. 2°57' N. 2°00' N. 2°57'N. 2°02' N . 2°00' N. 0''54'N. 2°57' N. 2'>00' N. 2°.57' N. 2°67' N. 1°59' N. 2''03' N. 2°00' N. 1''59' N. 1°59' N. 2°or N. 2°02' N. 2°00' N. 4°04' N. 2° 57' N. 2°02' N. 2°02' N. 2°00' N. 2°00' N. 0''64' N. 4°04' N. 2°02' N. 2°00' N. 2°00' N. 4°04' N. 4°04' N. 2"57'N. 2°00' N. 2°00' N. 0°54' N. 4°04' N. 4°04' N. 2° 67' N. 2°02' N. 1°22' N. 2''01' N. 2°00' N. 0''54' N. 1°59' N. 2°03' N. 2''03' N. 2''02' N. 4°04' N. 4°04' N. 2'>02' N. 2°28' N. 4''04' N. 2°57' N. 2<'02' N. 2°57' N. 2°02' N. 2°02' N. 2°57' N. 1°59' N . 2°02' N. 2°03' N. 4<'04' N. 3°26' N. 1°59' N. 2°57' N. 2°02' N. 2°or N. 2°00' N. 2°01' N. 2''00' N. 140°09' W. 14n°66' W. 141°24''W. 141°24' W. 140°02' W. 140°56' \V. 140°02' W . 140°10' W. 141°24' W. 140"'28' W. 140°02' W. 140°10' W. 151°50' W. 149''.54' W. 150°06' W. 149°54' \V. 150°38' W. 150°12' W. 150°23' W. 15I°24' W. 150°17' W. 151°24' W. ISO'l?' W. 153-12' W. 151°24' W. ISO^OO' W. 150°17' W. 151°24' W. 150°17' W. 150°17' W. 150°12' W. 153°12' W. 151°24' W. 150"'12' W. 150°12' W. 154'>50' W. 153°12' W. 151°24' W. 150°06' W. 150°17' W. 151''60' W. 151°60' W. 151°24' W. 151°24' W. 160°00' W. 160°06' W. 151°60' W. 151°24' W. 15r24' W. 150°06' W. 150°06' W. 150°17' W. 161°24' W. 15r24' W. 150°00' W. 150°06' W. 150°06' W. 150°17' W. 151°50' W. 149°54' W. 154°50' W. !5r24' W. 1,50''00' W. iso'iy W. 153°12' W. 163°12' W. 153°12' W. 150°06' W. 150°06' W. 161°50' W. 160°38' W. 150°06' W. 160°17' W. 153°12' W. 150°17' W. 151°50' W. 151°50' W. 160°17' W. 150°12' W. 161='50' W. 153°12' W. 150°06' W. I51°40' W. 150°12' W. 150°17' W. 153''12' W. 154° 50' W. 151°24' W. 164°50' W. 161°24' W. Stage of maturity M M M M In In In In In In In Im R R M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M M In In In SPAWNING OF YELLOWFIN TUNA 255 Table 1. — Data on 740 yellowfin tuna specimens from the central equatorial Pacific for which maturity determinations were made in the laboratory — Continued Date Position Stage of maturity Fish length Date Position Stage of maturity Fish Latitude Longitude Latitude Longitude length Sept 22, 1951 1°07' S. 2°01' N. 2°01' N. 2°01' N. 2°01' N . 2°00' N. 2°01' N. 2°01' N. 0°54' N. 2°00'N. 2°01' N. 1°59'N. 2°01' N. 4°56' S. 2°01'N. 2°01' N. 2°02' N. 2°or N. 4°04' N. 4°56' S. 4°56' S. r52' N. 1°19' N. 1°52' N. 1°52' N. 1°52' N. 1°59' N. r52'N. 2°02' N. 2°02' N. 1°52' N. 2°02' N. 1°52' N. 2°02' N. 2°02' N. 2°02' N. 1°52' N. 2°02' N. r62' N. 1°19'N. 2°02' N. 1°59' N. 1°59' N. 1=59' N. 1°59' N. 1°59' N. 8°14' N. 3°58' S. 5°36' S. 5°34' N. 3°12' N. 5°34' N. 2°15' N. 5°34' N. 2°15' N. 5°34' N. 4°00' N. 7°17' N. 6°25' N. 6°25' N. 6°25'N. 6°25' N. 7°33' S. 7°33' S. 7°33' S. 7°33' S. 7°33' S. 3°11'S. 3°irs. 1°00' N. 1°00' N. 2°13' N. 6°53' N. 6°13' N. 6°59' N. 7°24' N. 6°25' N. 7°24' N. 3°52' N. 6°13' N. ri2' N. I'Sl'N. 3°52' X. 3°.52' N. 6°13' N. 2''55' N. 2°.55' N. 1°12' N. 150°21' W. 154°50' \V. 154° 50' W. 154°50' W. 154° 50' W. 15r24' W. 154°50' W. 154°50' W. 150°00' W. 151°24' W. 154°50' W. 150°12' W. 154°60' W. 150°I3' W. 154°50' W. 154° 50' W. 153°12' W. 154° 50' W. 150°06' W. 150°13' W. 150°13' W. 155°24' W. 157°30' \V. 155°24' W. 155°24' W. 156°24' W. 157°36' W. 155°24' W. 156°20' W. 156°20' W. 165°24' W. 156°20' W. 156°24' W. 156°20' W. 156°20' W. 156°20' W. 156°24' W. 156°20' W. 156°24' W. 157°30' W. 156°20' W. 157°36' \V. 167°36' W. 157°36' W. 157°36' W. 157°36' W. 120°32' W. 120°14' W. 120°25' W. 152°26' W. 152°05' W. 152°26' W. 151°19' W. 152°26' W. 151°19' W. 152°26' W. 152°20' W. 157°04' W. 162°26' W. I62°26' W. I62°26' W. 162°26' W. 120°21' \V. 120°21' \V. 120°21' \V. 120°21' W. !20°2r W. 130°17' W. 130° 17' W. 151°26' W. 151°26' W. 151 °5r W. 162°05' W. 1B3°05' W. 163°54' \V. 164°23' W. Ifi2°2f)' W. 164°23' W. 1.59°20' W. 158°53' W. 160°21' W. 157°20' W. 159°57' W. 159°57' W. 158°53' W. 160°20' W. 160'20' W. 160°2r W. In In In In In In In In In In In In In In In In In In In In Im M M M M M M In In In In In In In In In In In In In In In In In Im Im R In Im M M M In In In In In M Im Im Im Im M In In In Im M Im In In In M M M M M In In In In In In In In Im Im Im Cm. 142 141 141 140 140 140 139 139 139 138 137 136 134 133 131 131 129 128 127 122 117 144 143 138 137 133 113 144 144 144 142 141 140 136 136 136 136 135 133 124 115 114 109 93 104 96 147 133 159 143 135 127 149 140 137 135 133 143 99 95 95 88 138 139 127 123 153 132 150 ' 145 142 135 146 134 133 133 129 143 143 140 139 136 134 132 122 143 141 138 Nov.20.1950 . Nov. 23, 1950 Nov. 21, 1950 Do.. Nov. 23, 1950... Nov. 16, 1950 2°55' N. 5°04' N. 3°52' N. 3°52' N. 5°04' N. 3°62' N. 3°52' .N. 3°52' N. 6°13' N. 5°53' N. 3°52' N. 4°42' N. 6°13' N. 4°42'N. 6°25' N. 5°53' N. 4°42' N. 4>'42' N. 3"'54' N. 4°42' N. 4''42' N. 6°53' N. 4°42' N. 4°42' X. 4°42' N. 4°42' N. 3°52' N. 4°42' N. 5°53' N. 4°42' N. 5°53' N. 4°42' N. 5»63' N. 4°42' N. 4°42' N. 6°25' N. 6°25' N. 4°42' N. 6°25' N. 5°53' N. 4°42' N. 4°42' N. 4°42' N. 5°53' N. 6°25' N. 5°63' N. 6°25' N. 3°3fi' S. 6°25' N. 1°00' S. 0°04' N. 2°24' N. 5°00' S. 4°14' N. 4°14' N. 4°14' N. 4°14' N. 2°27' N. 3°31' N. 1°59' N. 1°28' S. 1°28' S. 2°14' N. 3°02' S. 1°22' N. 3''31' N. 2°27' N. 2°27' N. 2°27' .v. 4°33' S. 3°02' S. 1°69' N. 2°14' N. 1°59' N. 1°.59' N. 2°I4' N. 2°14' N. 2°14' N. 3°02' S. 2°01' N. 1°28' S. 1°59' N. 2°fll' N. 3°3r N. 3°31' X. 1°28' S. 1°28' S. 3°31' N. 160°20' \V. 159°03' W. 159°57' W. 1.59°57' W. 159°03' W. 1,59°20' W. 1.59°20' \V. l.';9°20' W. 158° 53' \V. 162°05' W. 169°20' W. 160°24' W. 158°53' W. 160°24' \V. 162°26' W. 162°05' W. 160°24' \V. 160°24' W. 159°26' W. 160°24' W. 160°24' \V. 162°05' W. 160°24' W. 160°24' W. 160°24' W. 160' 24' W. 159°20' W. 160°24' W. 162°05' W. 160°24' W. 162°05' W. 160°24' W. 162°05' \V. 160°24' W. 160°24' W. 162°26' W. 162°26' \V. 160°24' W. I62°26' \V. 162°05' W. 160°24' W. 160°24' \V. 160°24' W. 162°05' W. 162°26' W. 162°05' W. 162°26' W. 170°02' W. 167°32' W. 169°27' W. 168°48' VV. 168°44' VV. 170°08' VV. 154°60' VV. 1.54°56' VV. 1,54°56' W. 1.54°.%' VV. 15.5°26' VV. 1!J5°23' VV. 156°09' \V. 155°25' VV. 15.5°25' VV. 1.57°08' VV. I!i5°12'VV. 1.5.5°18'VV. 15.5°23'VV. 155°26' VV. 1.5.5°2fi' VV. 155°2fi' \V. 155°08' VV. 1.5,5° 12' VV. 156°09' VV. 157°08' VV. 156°09' VV. 1.56°09' VV. 1.57°08' W. 157°08' VV. 1.57°08' VV. 15.5°12' VV. 1.58° 15' VV. 155°25' VV. 156°09' VV. 158°15' W. 15,5°23' W. 155°23' VV. !55°25' VV. 156°25' VV. 155 23' VV. Im Im Im Im Im Im Im Im Ira Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im R M M M In In In Im Im Im In In In In In In In In In In In In In In In In In In In In In In In Im Im Im Im Im Im Im Im Cm. 137 Sept. 17. 1951 - Do .. 137 136 Do Do - 135 135 Sept. 19,1951 Sept. 17, 1951... Do 130 Do... Do 130 122 Sept. 20, 1951 Sept. 19, 1951... Sent 17 1951 Nov. 24. 1950 122 Nov. 6, 1950 120 Nov. 16, 1950 120 Nov. 23, 1950-- 120 Spnt 17 1951 Nov. 24, 1950 118 Sept. 25. 1951 --. Sept. 17. 1951 - Do Nov. 23, 1950... 114 Nov 3, 1950 110 Nov. 28, 1950 - 110 Nov. 23, 1950.-- - 107 Nov. 2. 1950 105 Sept. 2. 1951 Nov. 16, 1950 Nov. 24, 1950.- --- 104 103 Do Nov. 2, 19iO 102 Sept. 12.1951 Sept. 15, 1951..- Sept. 12, 1951 Do Nov. 6. 1950 Nov 22, 1950 101 101 Do 99 .Nov. 23, 1960 .Vov. 24, 1950 ■. 99 98 Sept. 14, 1951 Sept. 12, 1951.- Sent 13 1951 Nov. 17. 1950 Nov. 23. 1950 96 94 Nov 27. 1950 93 Do Nov. 23. 1950 Nov. 27. 1950.- -. Nov 24. 1950 92 Sept. 12.1951 Sent 13 1951 88 87 Sept. 16. 1951 Sept 13, 1951 Nov. 28. 1950 86 Nov. 23. 1950 Nov. 24. 1950 Nov. 4, 1950 85 Do 85 Do -- - 83 Sept lli. 1951 Nov. 3. 19->0 Nov. 22. 1950 82 81 Sept 12, 1951 Nov. 2. 1950 Nov 4, 1950 78 Sept 15 1951 77 Nov. 22. 1950 77 Sept 14, 1951 Nov. 24. 1950 76 Do Do 71 Do Nov. 28. 1950 68 Do Nov. 4, 1950 63 Do Nov 27, 1950 . . 63 Oct 19, 1952 Nov. 30, 1950 61 Oct. 30, 1952 - Nov. 20, 1952 Nov. 11. 1950 Nov. 23, 1952 142 Oct. 31, 1952 153 135 Oct 29 1952 Nov 24 1952 .. . 132 Oct 27, 1952 Nov. 26, 1952 139 Oct 30, 1952 Nov. 19, 1952- - 131 Oct 27 1952 Dec. 12, 1953 - 126 Oct 30, 1952 Do 114 Oct. 27, 1952 Do 114 Oct 2H, 1952 Do 103 Oct 25 1950 Dec 6 1953 14$ Oct 31 1950 Dec 11, 1953 -.. 147 Dec. 7, 1953 146 Do Dec 3, 1953 144 Do Do 142 Nov. 1, 1952 Dec. 8, 19.53... 140 Do Dec 2. 1953 138 Do Dec. 5. 1953 137 Do Dec 11 1953 137 Do Dec 6, 1953 136 Nov. fi, 1952 Do 134 Do Do 127 Nov. 3, ig.w Dec 1. 1953 123 Do . - Dec. 2. 1953 122 Nov. 2, 19.52 Dec 7, 1953 . . 118 Nov. fi, 19.50 Dec. 8, 1053 110 Nov. 7, 1950 Dec. 7. 1953 107 Nov. 8, 19.50 Do . 105 Nov. 9, 1950 Dec. 8, 1953 105 Nov. 3(1. 19.50 Do 103 Nov. 9. 19.50 Do 103 Nov. Ifi. 19,50- Dec. 2, 1953 88 Nov. 24. 19.50 Dec. 9. 1953 87 Nov. 19. 1950.. Dec. 3, 1953 144 Nov. 1. ig-V) Dec 7, 1953 144 Nov. 21, 1950 Dec. 9, 1953 144 Do Dec. 11, 1953 144 Nov. 24. 1950 Do . . 144 Nov. 20. 1950 Dec. 3, 1953 Do 142 Do 141 Nov. 19.1950 Dec. 11, 1953 141 ' From length-weight relation. 256 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 1. — Data on 740 yellowjin tuna specimens from the central equatorial Pacific for which maturity determinations were made in the laboratory — Continued Date Position Stage of maturity Fish length Date Position Stage of maturity Fish Latitude Longitude Latitude Longitude length Dec. 7, 1953 1°59' N. 3°31' N. 3°31' N. 1°69' N. 3°31' N. 4°33' S. 1°22' N. 3°02' S. 1=59' N. 3002' S. 2=27' N. 2°27' N. 3°02' S. l°S,9' N. 2°14' N. 156°09' W. 165°23' W. 155''23' W. 156°09' W. 155023' W. 165°08' W. 165°18' W. 155°12' W. 156°09' W. 155''12' \V. 155°26' W, 155°26' W. 155°12' W. 156°09' W. 157''08' W. Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Cm. 138 138 137 136 129 125 123 121 120 119 119 116 115 110 no Dec. 8, 1953... 2°H' N. 2°14' N. 2=14' N. 2°14' .N. 2''01' N. 2°14' N. 2=14' N. 2°14' N. 2°14' N. 2°14' N. 2°14' N. 2°14' N. 2°14' N. 2°01' N. 2'>14' N. l.=i7'08' W. 167°08' W. 1.57'08' W. 1.57°08' W. 158°15' W. 157°08' W. 157°08' W. 157°08' W. 157°08' W. 157°08' W. 157'>08' W. 157°08' W. 157°0S' W. 158°16' W. 157°08' W. Im Im Im Im Im Im Im Im Im Im Im Im Im Im Im Cm. Dec. 11, 1953._ Do. Do Dec. 7, 1963 Do Do Dec. 9, 1963 Dee. 8, 1953 Do.... Do Do Do.... Do Do 104 103 103 102 99 98 98 98 97 97 Dec. 11, 1953.. Dec. 1,1953 --. Dec. 5, 1953 Dec. 2, 1953 Dec. 7, 1953. - Dec. 2, 1953. _ Dec. 6, 1953 Do. Dec. 2, 1953 Do Dec. 9, 1963 Dec. 8, 1953 Dec. 7, 1953 Dec. 8, 1953 94 88 On several cruises, observations were made on the state of maturity of ovaries which, with a few exceptions, were then discarded. Although these field observations were subjective and liable to differences between observers, they were used to supplement the seasonal and areal coverage. After discussion with the various observers, and after comparisons of field observations with labor- atory classifications, we were able to classify most of the ovaries reliably into two groups, "inactive" (immature and intermediate) and "active" (ma- turing and ripe). Field classifications are given in table 2. Questionable observations were not considered, and are not included in the table. Table 2. — Data on yellowfin tuna specimens from the central equatorial Pacific for which maturity determinations were made in the fi,eld [A, active; I, inactive] Date Position Stage of maturity Fish length Date Position Stage of maturity Fish Latitude Longitude Latitude Longitude length May 28, 1954 May 30, 1954 4°02' N. 3°68' N. 4°02' N. 4°02' N. 5058' N. 4°17' N. 4'>02' N. 4''02' N. 4°02' N. 4°02' N. 4''02' N. 3°68' N. 4°45' N. 4°02' N . 4°02' N. 4°02' N. 4°17' N. 3°58' N. 4°45' N. 3°68' N. 3°68' N. 3°58' N. 3°58' N. 3°58' N. 3<=58' N. 3°58' N. 3°68' N. 3°68' N. 3'>58' N. S^SS' N. 4<'45' N. 3°58' N. 3°58' N. 3°58' N. 3°68' N. 4°46' N. 3'>58' N. 0°30' S. 0°13' S. 0°30' S. O'H' S. 169°34' W. 159°04' W. 159''34' W. 169°34' W. 162''52' W. 160''28' W. 159°34' W. 159°34' W. I59''34' W. 169°34' W. 159°34' W. 159°04' W. 160°11' W. l.W°34' W. IIJ9°34' W. 159°34' W. 160°28' W. 169''04' W. IfWIl' W. 159°04' W. 159°04' W. 159°04' W. 159''04' W. 159°04' W. 159°04' W. 159°04' W. 169°04' W. 169°04' W. 159''04' W. 159°04' W, 160°11' W. 169°04' W, 159°04' W. 159°04' W. 169°04' W. im'W W. 1.59°04' W. 160°19' W. 16O°02' W. 160°19' \V. 160°00' W. A A A A A A A I A A A A Cm. 149 140 139 123 120 118 114 118 Ufi 114 109 108 106 106 103 102 101 100 98 97 96 93 93 93 92 92 91 91 91 89 88 88 88 87 86 84 83 151 150 160 147 Junell,19M June 14, 1964 0°26' S. 0°13' S. O^SO' S. l''43'N. 1°52' N. 0°30' S. 0°30' S. 0026' S. 0°13' S. 0°60' N. Q°30' S. 0°30' S. 0°18' S. 0°13' S. CM' S. 0°18' S. 0°13' S. 0°14' S. 0°18' S. 0''50' N. 0°13' S. 0°14' S. 0°14' S. O'W S. 0°14' S. 0°14' S. Q°14' S. 2''01' N, 0°30' S. 2°01' N. 0°14' S. 0°14' S. 0°14' S. 0°18' S. 2°01' N. 0">14' S. 2-03' N. 0°18' S. 2°03' N. 2''03' N. 2''01' N. 158-67' W. 160°02' W. 160°19' W. 168°28' W. 156'>47' W. 160=19' W. 160°19' W. 168°57' W. Ifi0°02' W. 1.68°63' W. 160°19' W. 160°19' W. 160°16' W. Ifi0°02' W. lOO'OO' W. 160°16' W. 160°02' W. lecoo' w. 160°16' W. 168°53' W. 160'>02' W. leo-oo' w. IfiO'OO' w. 160°00' W. leo'oo' w. 160°00' W. 160°00' W. 167°09' W. 160'>19' \V. 1.57°09' W. 160°00' W. 160°00' W. 160°00' W. ll>0°16' W. 1,67<'09' W. lOO-OC W. 157°40' W. 160''lfi' W. 1.67°40' W. 1.57''40'W. 167°09' W. A A A A A A A A A A A A A A A A A A A A A A A A A A A A ■ A I I I I I I I I I I I I Cm. 146 146 May 28,1954 June 16, 1964 146 Do June 3, 1954 144 May 17, 1954 May 27, 1954 June 7, 1954 June 15, 1954 143 143 Mav28, 1954 Do 143 Do Do ... June 11, 1964 June 14, 1954 142 142 Do Juno 10, 1964 June 15, 1964 141 Do 140 May 30, 1954 ... Do 140 May 25, 1954 June 13, 1964 139 Mav 28, 1954 June 14, 1954 137 Do... June 12, 1954 135 Do June 13, 1954 134 May 27,1954.. May 30, 1954 June 14, 1954 June 12, 1964 133 132 May 25, 1954 June 13, 1964 132 May 30, 1954 June 10, 1954 128 Do June 14, 1954 127 Do.... June 12, 1954 126 Do Do 125 Do Do 122 Do Do . 120 Do Do 120 Do.... Do June 8, 1964 120 Do 117 Do June 15, 1954 92 Do June 8, 1954 .- June 12, 1964 Do.. - 122 Mav 25, 1954 122 Mav 30, 1954 121 Do - . Do June 13, 1964 119 Do.. 118 Do Junes, 1964 June 12, 1964 116 Mav 25, 1954 116 May 30, 1964... June 4, 1954 June 13, 1954 June 4, 1954 116 .June 15, 1954 June 14, 1954 108 104 June 16, 1954 Do . Junes, 1954 102 June 12, 1954 94 SPAWNING OF YELLOWFIN TUNA 257 Table 2. — Data on yeltovifin tuna specimens from the central equatorial Pacific for which maturity determinations were made in Ike field — Continued Date Position Stage o( maturity Fish length Date Position Stage of maturity FUh Latitude Longitude Latitude Longitude length Auff 25 1952 3°26' X. l'>33' X. l'>33' N. 1°33' X. 1°33' X. 1''33' X. 3°26' X. 2''23' X. 2°23' X. 1°33' X. 1°00' X. 3°26' X. l'>33' X. 1°33' X. roo' X. 2-23' X. 1''33' X. l'>33' X. l-SS' X. 1''33' X. 2'>23' X. 4°28' X. 1°00' X. 4°28' X. 9°00' X. 6°10' X. 7°60' X. 7°50' X. 1°U'X. 2°08' N. 9°00' X. 2°08' X. 2°08' X. 4°43' X. 2°08' X. 4°43' X. 4°43' X. 3°22' X. 3''22' X. 3°22' X. 140''08' W. 140''13' W. 140°13' W. 140°13' W. 140°13' W. 140°13' W. 14O°08' W. 140°12' W. 140°12' W. 140°13' W. 140°22' W. 140°0S' W. 140°13' W. 140°13' W. 140°22' W. 140'>12' W. IWU' W. 140''13' W. 140°13' W. 140°13' W. I40°12' W. 139°51' W. 140°22' W. 139°51' W. 159°40' W. 160°02' W. 159°24' W. 159''24' W. 160°08' W. 160°24' W. 159''40' W. 160"'24' W. 160°24' W. 160°00' W. 160°24' W. 160°00' W. 160°00' W. 160°24' W. 160°24' W. 160°24' W. A I A A A A A A A A A On. 142 154 152 150 150 149 148 148 148 148 148 147 147 144 144 142 142 141 141 138 132 131 121 112 147 141 140 140 134 128 128 123 107 140 125 112 111 108 104 100 Aug. 24, 1953 4°43' X. 4°43' X. 4°43' X. 4°43' X. 4°43' X. 4'>43' X. 1°42' X. 4°04' X. r42' N. 2''.33' X. 3005' X. 3°05' X. 2°25' X. 3°05' X. 1°42' X. 2''33' X. .3''31' X. 1°42' X. 2°25' X. 2°57' X. 3°39' X. 2°28' X. 3°39' X. 3°49' X. 2005' X. 3°49' X. 3»39' X. 2»28' X. 3°26' X. 2°28' X. 2°57' X. 1°48' X. 2°08' X. 2°57' X. 2''05' X. 1°48' X. 2'>08' X. 2°05' X. 1°22' X. 2°28' N. leo'oo' W. 16O°0O' W. 160"'00' W. 160°00' W. 160°00' W. 160'>00' \V. 141=24' W. 140°09' \V. 141024' W. 143<'22' W. 14O°02' W. 140°02' W. 140-32' \V. I40''02' W. 141°24' W. 143°22' \V. 140°2)t' W. 141''24' W. I40''32' W. 147°22' W. 151°,54' W. 150''38' W. 151-54' W. 152-10' W. 150-23' W. 152-10' W. 151-54' W. 150-38' W. 151-40' \V. 150-38' W. 147-22' W. 150-05' W. 145-21' W. 147-22' W. 150-23' W. 150-05' W. 145-21' W. 150-23' W. 149-54' W. 150"3S' W. A A A A A Cm. 99 Aug. 27, 1952 Do Do Do 96 Do... Do 9fl 95 Do Do 93 Do Do... .Sept. 7, 1952 Sept. 3, 1952 Sept. 7, 1952 Sept. 9, 1952 .Sept. 2, 1952 Do 88 Aug. 25, 1952 156 Aug. 26, 1952 Do 150 149 Aug. 27, 1952 - Aug. 28, 1952 148 147 Aug 25 1952 145 Aug 27, 1952 Sept. 5. 1952 145 Do Sept. 2, 1952 Sept. 7, 1952 144 Aug. 28, 1952 Aug. 26, 1952 143 Sept. 9, 1952 Sept. I, 1952 143 Aug 27 1952 142 Do Sept. 7, 1952 Sept. 5,1962 Sent. 11. 1952 Sept. 18, 1952 Sept. 16, 1952.... Sept. 18, 1952... Sept. 19, 1952 139 Do 136 Do 157 Aug. 26, 1952 1.50 Aug. 24, 19.52 Aug. 28, 1952. Aug. 24, 1952 Aug 27 1953 146 144 143 Sept. 15, 1952 142 .Aug. 25, 1953 Aug. 26,1953 Do Aug. 21, 1953 Sept. 19, 1952 Sept. 18, 1952 142 138 Sept. 16, 1952 Sept. 17, 1952 Sept. 16, 19.52 137 137 Aug. 22. 1953 Aug. 27, 19.53 Aug. 22, 1953.. Do 135 Sept. 11, 1952.... 148 Sept. 14, 19.52- Sept. 10, 1952 144 143 Aug. 24, ig.w .\ug. 22, 1953 Sept. 11, 1952 ... . 143 Sept. 15, 1952 142 Sept. 14, 1952 135 Do Sept. 10, 1952 134 AuE 23 1953 Sept. 15, 1952 134 Do Sept. 13, 1952 133 Do Sept. 16, 1952 117 Additional data (table 3) are available from field observations reported by the Iwate Prefecture Fishery Experiment Station (1953a and 1953b), which were obtained from Japanese longline expeditions into this area. As with the POFI field observations, the stages of maturity were combined into two groups, "active" and "in- active", and the "spent" category was disregarded. Other pertinent information found in these reports is as follows: Tlie fishing area for the April cruise was between latitudes 9° N. and 11° N. and longitudes 170° W. and 173° W. Fish caught on this cruise ranged from 110 to 150 cm. in length, with only one fish measuring less than 120 cm. Fishing during June and July was done at 3° N. to 4° X. and 175° W. to 177° W. Fork lengths ranged fi'om 114 to 173 cm., with only two fish measuring less than 120 cm. Eighty percent of the fish were caught in June, and the rest were caught in July. Table 3. — A^umber of yellowfin tuna in various stages of maturity, according to Iwate Prefecture Fishery Experiment Station Date 3m S £ 03 a TO Location Longitude Latitude Apr. 4-24, 1953... June 22-July 3, 1953. 3 29 4 41 10 97 14 26 4 22 9°N.-11°X... 3-X.-4-X.... 170° W.-173° W. 175° W.-177- W. Note.— For further data see station reports (1953a and 1953b). SIZE OF FISH AT FIRST SPAWNING To determine the size of first-spawTiing fish, the fork lengths collected by POFI were grouped into classes of 10 cm., and the percentage of fish in the "active" category (maturing and ripe stages), as determined by ovary examination, was calculated for each lengtli class. The results, illustrated in figure 1 , show t hat all the fish smaller than 70 cm. were in a nonspawning condition. In the 70-to-79-cm. class (about 15 to 22 lbs.), 6.9 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 60 70 80 90 100 110 120 130 140 150 69 79 89 99 109 119 129 LENGTH (CM) 139 149 159 12 18 26 37 50 65 84 106 131 160 APPROX. WEIGHT(LBS) Figure 1. — Percentages of sexually active fish at different fork lengths. Figures in parentheses indicate the num- bers of individuals on which the percentages are based. percent were in the maturing or ripe stages. The percentage of sexually active fish increased gradu- ally and irregularly from this length class through the llO-to-119-cm. class (about 57 to 72 lbs.), of which 17.4 percent were active. In the next 10-cm. class (about 74 to 92 lbs.), the percentage of reproductively active fish jumped sharply to 47.0 percent. Above this class the percentage of active fish increased steadily with length. Of the fish measuring 150 to 159 cm. (about 145 to 172 lbs.), 66.7 percent were active. These data suggest that, although yellowfin as small as 70 cm. are capable of reproducing, the greater part of the population reaches sexual maturity at about 120 cm. Schaefer and Marr (1948), however, noted that in Costa Rican waters yellowfin ranging from 70 to 100 cm. spawn later in the year than the larger fish. This presents the possibility that larger fish have longer spawn- ing periods than smaller fish, which in turn sug- gests that the smaller percentage of sexually active fish from 70 to 120 cm. in our samples may have resulted from a shorter spawning season rather than from a diff'erence between the propor- tions of sexually mature fish above 120 cm. and those below 120 cm. Although the representation of fish from 70 to 120 cm. for each month is spotty in our samples, an examination of the monthly percent maturing and ripe (table 4) shows the peak percentage to be far less than that reached by the larger fish. This supports our interpre- tation of the results, that is, that the greater part of the population reaches sexual maturity at about 120 cm. Table 4. — Monthly percentages of sexually active yellowfin tuna below 120 cm. fork length Month Fraction active Percent active January 1/12 2/26 0/1 2/18 11/101 11/45 0/1 3/24 4/12 0/4 0/31 0/33 8.3 7.6 0.0 April .. 11.0 10.0 June 24.4 July 0.0 12.5 September .. ... 33.3 0.0 November 0.0 0.0 LOCALITY OF SPAWNING The data were grouped by months and by 10-degree longitudinal sections. The data for those ovaries collected between 115°00' W. and 124°59' W. are shown in table 5 in the 120° W. longitudinal section (the midpoint of that section), those collected between 125°00' and i;34°59' W. in the 130° W. section, and so on, with the excep- tion of the 180° section, which includes 175°00' W. to 180°. Because the percentage of sexually active fish below 120 cm. was so much smaller, only fish above this size were considered in order to get results that could be used for comparison. The percentage of active fish for each month for each 10-degree section was calculated. The percent- ages for June and July along 180° were calculated from summarized Japanese data, which did not separate the catch of those 2 months. To arrive at the monthly totals for these months (in the extreme right-hand column of table 5), the 193 fish caught along 180° were separated into 154 fish for June and 39 fish for July, because 80 per- cent of the catch was made in June. The results (table 5) show that all the sections had at least one month in which 85 percent or more of the fish were sexuallv active. This, coupled SPAWNING OF YELLOWFIN TUNA 259 Table 5. — Fractions of samples of seitialty active yetlowfin tuna (maturing and ripe) at various longitudes, by , IPcrcontage of se.vually mature fish in sample in parentheses) lonlhs Month At longitude- Total ISO" 170° W. 160° W. 150° W. 140° W. 130° W. 120° W. I (16.7%) |(40.6%, 1 (16.7%) 11(49.3%) ^ (81.8%) I (50.0%) ^^(46.7%) March II (87.5%) 1 (100.0%) |(97.1%) ^|(94.2%) \i)ril =2 (90.3%) If (90.3%) Mav 1 (87.5%) § (85.2%) ■| (100.0%) |(87.2%) li '^•''^''' 1 (77.8%) 1 (88.4%) ^ (90.0%) 1 (100.0%) ^(85.7%) |(85.0%) 1 (66.7%) ^ (83.3%) Au ust § (69.8%) 1 (100.0%) 1(28.2%) i <*'■*''<'> ^ fg(27.8%) ^(72.2%) |r73.3%) J|(67.5%) j (100.0%) 1 (37.5%) 1 (33.3%) ^ (41.7%) I (66.7%? 1; (17.9%) 1 (0.0%) \ (50.0%) \ (20.0%) 5 (25.0%) li (0.0%) ■Y (0.0%) ^2(0.0%) with tlie fact that larvae below 10 mm. have been found in all of these sections (Matsumoto') indicates that yellowfin spawning occurs through- out the central equatorial Pacific. The fact that spawning probably occurs throughout the entire equatorial Pacific is indicated by additional records of spawning yellowfin in the western area by Wade (1950 and 1951), Marr (1948), and Shimada (1951), and in the eastern area by Schaefer and Marr (1948) and Mead (1951). TIME OF SPAWNING The percentage of sexually active fish of 120 cm. and longer was calculated for each month of the year and was plotted on a graph (fig. 2). Yellow- fin that had almost reached the spawning state were found in each month except December, and the greatest percentages of active fish occurred from March (94.2%) through July (83.3%). It was only during November, December, and January that the occurrence of maturing and ripe fish dropped below 40 percent. This, however, does not prove that spawning is a year-round activity, inasmuch as the length of time that the I .Miit.suniuto. Walter M.: T">escriptions of four species of tuna larvae and their distribution in central Pacific waters. POFI. (Unpublished MS.) . . cry. 1/11 IV Ar-TiurlBASED ON OBSERVATIONS FROM • •btxuALLT "l-l l''tj-^^|_ SOURCES ^ < EARLY STAGE OF RESORPTION"! BASED ON OVARIES ^ COLLECTED AFTER . -• LATE STAGE OF RESORPTION J 1951 ONLY 80 - 1 1 T" 1 i 1 — 1 1 1 1 1 1 60 - / ^ \ / / 40 < / / A / _ 20 n - / >/ V 1 A- _L \ 1 1 T" ^ • jI: '*- \ JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC Figure 2.— Monthly percentages of yellowfin tuna -■^exnally active or with rcsitUial eggs. fish are in these stages before spawning is not known. To define tiie spawning season further, the occurrence of residual eggs in these larger fish was invc^stigated with respect to time. Tlie results, plotted in figure 2, show that ovaries with early- 260 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE stage residual eggs have the same occurrence pattern as maturing and ripe ovaries. This is true for all the months except January, which is represented by an inadequate sample. The occur- rence of these early-stage residual eggs indicates that spawning actually is a year-round occurrence. Spawning in other equatorial areas of the Pacific likewise seems to be protracted. Schaefer and Marr (1948) found indications of a prolonged spawning season off Costa Rica. Wade (1950) found that the spawning period of yellowfin in the Philippine Islands extended over a considerable period, but that it was most intense during May, June, July, and August. It is probable that the of these stages are based on gross microscopic examination and are intended to aid future workers in recognizing these structures. Immediately after spawning, these residual eggs (fig. 3) generally resemble the ripe eggs, except that they become shrivelled owing to shrinking of the yolk mass and the resulting collapse of the chorion. The oil sac is usually ruptured, and the released oil appears as bright yellow droplets. The eggs at this stage are still loose and translucent. Subsequently the eggs lose their translucence and collect in masses of semiopaque tubules. The eggs are not within the tubule but are entangled in the manv disordered convolutions of the tubule. Figure 3. — Individual residual eggs; O., oil droplet; O. S., oil sac. prolonged spawning season is accompanied by multiple spawning — in other words, there is more than one spawning per fish in a spawning season. June (1953) considered this to be true for yellowfin in Hawaiian waters, after studying the progression of modal groups in egg-diameter frequencies. DESCRIPTION OF STAGES IN RESORP- TION OF RESIDUAL EGGS In the beginning of this study, several structures found in the ovaries could not be readily identified. As more ovaries were examined, it became evident that these structures were the remains of ripe eggs from a previous spawning which were in different stages of resorption. The following descriptions / .-^-^ ^ ^**^^^« k>-v { 11 ^^ o.t. i.t. IMM 1 1 F.IGURE 4. — Piece of tubule teased from ovary with residual eggs; i. t., inner tubule; o. t., outer tubule. SPAWNING OF YELLOWFIN TUNA 261 The tubular diameter is about 0.20 mm. Within this tubuh' lies anotiier tubule with a diameter of 0.0.5 mm. Circular transverse ridges on the wall of the outer tubule give it a striated appearance. Figure 4 shows a short section of a tubule that had been teased from a mass. Histologically, these masses of tubules and eggs are found to be surrounded by a connective tissue stroma (fig. 5). The wall of the outer tubule seems to be composed of reticular connective tissue. The wall of tlie inner tubule is made up of a single layer of closely arranged minute cells (3 M diameter) with relatively large, deep-staining nuclei. The origin and function of these tubules are open to question, but their pro.ximity to the residual eggs suggests that they are involved in the absorption of these eggs. [^DEVELOPING PORTION ^^DEGENERATING PORTION Figure 5. — Above: Part of section through ovary showing residual egg mass in xilir. c, connective tissue capsule; d. e., developing egg; i. I., inner tubule; o. t., outer tubule; r. e., residual egg. Below: Diagram of this section, outlining developing and degenerating portions. 262 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE The masses of eggs, tubules, and connective tissue which are scattered throughout the ovary appear to shrink with the passage of time. An examination of later stages shows that the residual eggs are not arranged entirely haphazardly but are lined up to form indistinct cords (fig. 6). These masses eventually shrink to nondescript particles (fig. 7) before they are finally lost in the ovarv. OCCURRENCE OF NEMATODES IN THE OVARIES While examining the eggs, we observed several ovaries with nematodes, ranging from 0.5 cm. to 4 cm. in length. The specimens were in too poor a state of preservation to identify. Of 25 ovaries examined for nematodes, 22 (88%) were infested. The extent of infestation did not appear to be serious enough to affect the Figure 6. — Above: Residual egg mass teased from ovary. Below: Diagram of this mass, outlining the rows of eggs. SPAWNING OF YELLOWFIN TtJNA 263 Figure 7. — Part of ovary showing shrunken masses of residual eggs; m., mass of residual eggs. functioning of the ovaries. There were seldom more than five worms in a single ovary, and in only one instance did the ovarian tissue seem to be pathological owing to heavy infestation. Fish with infested ovaries were found throughout the central equatorial Pacific. SUMMARY This study is based on data obtained in the field relative to the time and place of spawning and the size of yollowfin tuna at time of spawning, and on laboratory examination of ovaries of yellow- fin tuna obtained on POFI e.xploratorv-fishing trips made in the central equatorial Pacific from February 1950 to June 1954. Study of the ovaries and of the data on tlie size and distribution of the spawning fish led to the following conclusions: (1) The size at sexual maturity may be as small as 70 cm., but usually is greater than 120 cm.; (2) the spawning season extends throughout most of the year, with November, December, and January the months of lowest spawning intensity; (3) the spawning grounds seem to include the entire equatorial Pacific. During the course of this investigation, stages in the resorption of residual eggs were observed and described. Unidentified nematodes were found in 88 per- cent of a sample of 25 ovaries. In most instances, the nematodes did not seem to be present in sufficient numbers to affect egg production seri- ously. LITERATURE CITED BiNi, Giorgio. 1952. Osservazioni suUa fauna marina delle coste del Chile e del Peril con sp<'ciak' riguardo alle specie ittiche in gcnerale ed ai tonni in particolare. BoUet- tino di Pesca, Piscicoltura e Idriobiologia, vol. 7 (n. s.), fasc. 1: 11-60. Roma. IwATE Prefecture Fishery Experiment St.^tio.v. 1953a. South Sea tuna fishing survey. Rept. No. 1. 1953b. South Sea tuna fishing survey. Rept. No. 2. June, Fred C. 1953. Spawning of yellowfin tuna in Hawaiian waters. U. S. Department of the Interior, Fish and Wildlife Service, Fishery Bull., No. 77, vol. 54, pp. 47-6-1. Marr, John C. 1948. Observations on the spawning of oceanic skipjack (Katsuwonus pelamis) and yellowfin tuna {Neothunnus macroptent.i) in the northern Marshall Islands. U. S. Department of the Interior, Fish and Wildlife Service, Fishery Bull., No. 44, vol. 51, pp. 201-206. Mead, Giles W. 1951. Postlarval \eothunnus mricropterux, Auxis (hazard, and Eiithynnus linealus from the Pacific Coast of Central America. U. S. Department of the Interior, Fish and WildUfe Service, Fishery Bull., No. 63, vol. 52, pp. 121-127. 264 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE ScHAEPER, MiLNER B., and J. C. Marr. 1948. Spawning of yellowfin tuna (Neothunnus macrop- terus) and skipjack (Katsuwonus pelamis) in the Pacific Ocean off Central America, with descriptions of juveniles. U. S. Department of the Interior, Fish and Wildlife Service, Fishery Bull., No. 44, vol. 51, pp. 187-196. Shimada, Bell M. 1951. Contributions to the biology of tunas from the western equatorial Pacific. U. S. Department of the Interior, Fish and Wildlife Service, Fishery Bull., No. 62, vol. 52, pp. 111-119. Wade, Charles B. 1950. Observations on the spawning of Philippine tuna. U. S. Department of the Interior, Fish and Wildlife Service, Fishery Bull., No. 55, vol. 51, pp. 409-423. 1951. Larvae of tuna and tuna-like fishes from Philippine waters. U. S. Department of the Interior, Fish and Wildlife Service, Fishery Bull., No. 57, vol. 51, pp. 445-485. U. S. GOVERNMENT PRINTING OFFICE : 1957 O -406090 A METHOD OF ESTIMATING ABUNDANCE OF GROUNDFISH ON GEORGES BANK By George a. Rounsefell, Fishery Research Biologist In studying the fluctuations in iitjundiinc.e of various species tliat comprise the catcli, it is of paramount importance to know how tlie abun- dance of each species usually vaiies from l)ank to bank and from depth to depth. When vessels arc fishing chiefly for a particular species, they seek the grounds and the depths at which that species is most easily taken in abundance. For such a species, the catch per unit of fishing effort will measure the relative abundance with considerable accuracy, since the vessels will shift to grounds and depths yielding the highest catches. For other species, however, the fluctuations in actual abundance cannot be measured without sufficient knowledge of their average relative density in different depths and on different grounds. There- fore, a study of the distribution of these other species by depth and fishing grounds is a necessary preliminary to a study of their annual fluctuations in abundance. Knowledge of the relative density of each species by fishing grounds is of considerable value from other standpoints. What effect is a fishery in any certain area likely to have on the stock of each species? In certain cases the question arises: What effect will a change in the size of the mesh of the trawl have on the catches? Only by knowing the density of each species by areas and depths can these questions be answered. For many species not extensively sought for economic reasons, it is desirable to know whether there is a possibility of the catch being increased, should it become desirable to increase production. There is also the problem whether the range of a species is wholly covered by the fishery or may extend to areas beyond. Tlic methods developed in this paper have been followed by the haddock investigation of the N'oi'th Atlantic Fishery Investigations in computing indexes of abundance from 1931 to 19.'):i. MATERIAL To obtain a measure of the relative density of each species it was necessary to ascertain the quantity caught by certain units of fishing effort. Collection of the data necessary for this study was started in the fall of 1931, at the Boston Fish Pier, and IS continuing. In 1942 this collection was extended to the ports of Gloucester and New- Bedford, Mass.; in 1953 it was extended to Provincetown, Mass., and Rockland and Portland, Maine. A full description of the methods of collection is given by Rounsefell (1948). The essential data collected for each vessel interviewed are as follows: 1. Name of the vessel, and type of gear employed. 2. Day and hour of departure and of arrival at port. 3. Positions fished, by "unit" areas, each unit comprising a rectangle of 10' of latitude and 10' of longitude, or about 10 miles by 7-plus miles. 4. Depth, in fathoms, at each fishing position. 5. E.'itimated amount of the catch, in thousands of pounds, at each fishing position. 6. Estimated proportion of each species taken on different fishing grounds. 7. For line-trawl vessels, the number of tubs of gear set out at each fishing location. 8. For otter-trawl vessels, an estimate of the time spent on each fishing ground. 9. For otter-trawl vessels, the total amoimt of time lost on the trip (other than the usual running time to and from the banks) because of such occurrences as torn nets, engine trouble, or stormy weather. CALCULATION OF CATCH PER FISHING DAY FOR OTTER-TRAWL VESSELS In order to obtain for otter-trawl vessels a measm-e of fishing effort more or less independent of weather, distance traveled . . ., it was found desirable to calculate the amount of time the vessel actually spent in fishing while on the fishing grounds. P'rom the data available, the actuid number of days the vessel was absent from port 265 266 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE was calculated to the nearest tenth of a day. The time spent away from port was consumed partly in the voyage to and from the bank. To discount this, the groups of otter-trawl vessels selected for study were questioned about their average speed under working conditions, and in many cases this was checked for various voyages by means of radio reports. Tables were made to show the time (to the nearest tenth of a day) that it would take a vessel to make the trip from port to each statistical subarea (see fig. 1.) at the average speed of the selected group. The time the vessel was absent from port, minus this running time, and also minus any time lost by bad weather, . . ., gave the calculated number of fishing days for each trip. On the average, these calculated fishing times were found to agree with the esti- mated fishing times obtained from the interviews but they were used instead of the estimated times given in the interviews because they wei-e con- sidered less subject to personal judgment. In the preceding calculations, the distance was measured from port to a point in the subarea empirically selected as being nearest to the avei'age position fished in that subarea (as shown by plots of many fishing positions over several years). Wlien the vessel fished in two subareas that extended in the same direction from port, only the voyage to and from the most distant of the two subareas was used. When more than one subarea was fished and the subareas were not in line, the running time was taken from port to one subarea, then between subareas, and finally from the last subarea back to port. When a vessel fished in more than one subarea, the calcu'ated fishing time was divided between the subareas in the same proportion as the esti- mated fishing time given in the interview, except that when the estimated and calculated times did not agree and the estimated time in a certain sub- area was only 1 day or a fraction of a day, this estimated time was considered correct, and adjust- ment was made in the time for the subarea or sub- areas in which more fishing was conducted. Al- though this approach is not easily susceptible of statistical proof, it is obvious that the estimates of the shorter periods of time are much more apt to be correct than those of the longer periods. A mate may easily be uncertain whether they fished 6 days or 7 days in a subarea, but an estimate of 12 hours is seldom far off. In some cases, the mate did not remember the number of hours spent in a subarea in which the vessel did little fishing but knew the number of tows made by the otter trawl. In these cases, each tow was considered as an estimated one-tenth of a fishing day. This estimate is predicated on the number of tows per day by large otter trawls, as indicated by careful notes and logs kept by several vessels for W. C. Herrington. These data showed that on the average there were 10 tows per day. SELECTION OF OTTER-TRAWL VESSELS FOR DETERMINING RELATIVE ABUNDANCE The first step in obtaining the catch per day was to select two groups of Boston otter trawlers, each group fairly homogeneous with respect to size of vessels. The first group of 12 large (over 150 gross tons) otter trawlers ranged in size from 163 to 173 gross tons, with an average of 167 gross tons. The second group of 13 vessels ranged from 229 to 262 gross tons, with an average of 247 gross tons, or 48 percent larger than the first group in average size. However, after the data on catch per day were tabulated, it was found that the selection of these groups on the basis of gross ton- nage was apparently erroneous. In order to de- cide on the proper basis for selection, all 25 boats were compared for the year 1938. The levels of fish abundance differ considerably between the New England and Nova Scotia banks; therefore the comparison of fishing ability was con- fined to the New England banks, which accounted for 57 percent of the season's catch. In making tliis comparison, it was found that some of these boats did considerable fishing for ocean perch, while others did little or none. As this is a specialized fishery that yields a far greater poundage per unit of fishing effort, it was necessary to eliminate this cause of variability in order to obtain a valid comparison. Tabulation of the 146 trips or portions of trips made in the deep waters (more than 60 fathoms) of Subareas XXII, F, G, and H in wliich ocean perch were taken, showed 72 instances in whicli over 80 percent of the catch consisted of ocean perch; these trips averaged 95 percent ocean perch. Another 29 trips had be- tween 41 and 80 percent ocean perch and averaged 58 percent, and 45 trips had from 1 to 40 percent ocean perch and averaged 16 percent. Obviously, on the trips vvitli a higli percentage of ocean perch KSTIMATING ABUNDANCE OF GROUNDFISH ON GEORGES BANK 267 ^ >, c3 *^ ■/. u *^ X, r^ » V *^ o "" - «s - Zi X ? X -^^•i = 55 S i S "5 =! '5 = o! 5 s 268 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE the vessels had spent all or a large portion of their time seeking this species. Therefore, all trips con- sisting of over 40 percent ocean perch were elim- inated from the comparison. This amounted to less than 10 percent of tlie catcli in Area XXII. The coefficient of correlation between the aver- age catch per day of the 25 vessels and their gross tonnages, -f-0.4033, was not statistically signifi- cant. What correlation exists is undoubtedly due to the linkage of gross tonnage to other factors, treated below. Since these vessels all employ the same type and size of otter trawl, regardless of differences in the sizes of the vessels, the absence of a significant correlation between size of vessel and fishing ability is not surprising. Obviously, more important factors are the amount of sea bottom covered by the net at each tow and the number of tows made each day. The amount of ground covered in a tow will depend largely on the speed and power of the vessel. Therefore, the catch per day was corre- lated with the power of the vessels. Instead of correlating catch per day directly with horsepower, the ratio of horsepower to length was used as the criterion of power, since the horsepower of a vessel depends more on length than on tonnage. Also, the use of horsepower directly, instead of the ratio, does not give a true estimate of towing ability. This correlation gave a statistically significant coefficient of correlation of +0.75. Since the newer vessels take advantage of all improvements in design and usually obtain tlie best crews, it was suspected that age of the vessel might play a part. The correlation of age of vessel and catch per day gave a significant coeffi- cient of —0.6643, showing that the newer vessels were superior. However, as the newer vessels were often better powered than the older vessels, it was necessary to eliminate the effect of the other variable in comparing the catch per day with either horse- power-length ratio or age of vessel. The coefficient of partial correlation of catch per day and horsepower-length ratio, with age of vessel fixed, was +0.686. The coefficient of par- tial correlation of catch per day and age of vessel, with horsepower-length ratio fixed, was —0.497. Squaring the two partial-correlation coefficients shows that 47 percent of the variability in catch per day was due to differences in the horsepower- length ratio of the vessels and an additional 25 percent of the variability was due to dift'erences in age of the vessels, leaving only 28 percent of the variability in catch per day unaccounted for. In obtaining a more accurate method of rating each boat according to its fishing ability, both age of vessel and horsepower-length ratio were taken into account. For each boat, the amount in standard deviations that it varied from the mean of the horsepower-length ratio was obtained. The same was done for age of vessel. The two figures were then combined, but the age ratio was weighted by 0.52, the ratio of its influence on the catch to the influence of power. The correlation of this adjusted rating of the individual boats with their catch per day of fishing gives a correlation coefficient of +0.817. Squar- ing the coefficient shows that the differences in the adjusted ratings of the vessels accounts for 67 percent of the variability in the catch per day. This accounts for all but 33 percent of the vari- ability in catcli per day, agreeing closely with the 28 percent shown by tlie two coefficients of partial correlation. Because such a large proportion of the varia- bility in catch per day is due to the age and power of the vessel, it was obviously incorrect to intro- duce new boats into the calculation. Therefore, it was decided to reject the data from all vessels except 16 that fished continuously from 1932 through 1938. The use of the same boats every year meant that variations due to age and power of vessel could be held to a minimum. Whether the correlation between age of vessel and catch per day was due to obsolescence or to the increased efficiency of the newer boats cannot be deduced from the correlation. It is safe to say, however, that at least a large share of it is due to improve- ments otlier than power in the design of the newer boats. ADJUSTMENT FOR CATCHING ABILITY OF TWO GROUPS OF OTTER TRAWLERS As a preliminary step in analyzing the catch per unit of fishing effort in various areas and at various seasons it was desirable to determine the relative catching ability of the two groups of otter trawlers. This was to make possible the pooling of their catches so that one final curve of abundance could be obtained for each area. ESTIMATING ABUNDANCE OF GROUNDFISH ON GEORGES BANK 269 The date for this comparison were ohteinc fistiing in each unit area really represent the average for the whole unit area and not for any particular point. Tlie metliod finally adopted was to construct frecjuency ])olygons of days of fishing across the entire Area XXII from north to south and from west to east for each column and each row of unit areas. These frequencies then were smoothed so tliat the amount under the curve would average the correct number of days of fish- ing in each unit area. After the fishing-intensity contoiu's luul i)een drawn (see fig. 4), the area of shallow, medium, and deep water within each contour was measured witli a planimeter. Multiplying the midpoint in tlie range of fishing intensity within contour lines by the areas enclosed gave a total of 82,479 fishing (hiys, compared witli 82,127 in tlie original data, an error of only sligiitly over 1 [)ercent.- B\' this 272 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE P3 3 O X O "T xn P ■g CT> t M O G S ^ (/J s^ S ^ _Q CO fa^ h h J +log J [.mj ul [.mJ + +log y = l There might be some question about the ad- visability of using an unweighted average instead of a weighted average in obtaining these geometric- mean ratios. At this point it must be remembered that in such a chronological series, weighting of the data (thus giving much more weight to certain years) may introduce a bias which we cannot measure. Using the logaritlims of the ratios, an imalvsis of variance was made, which showed no 276 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE significant difference between the means of the years; therefore the years were averaged without weighting. From the analysis of the variance of the loga- rithms of the ratios, the least significant mean difference {P=.05) between ratios was calculated for the mean ratios of the species in the shallow zone and in the deep zone (Snedecor 1940, p. 344) ; these are shown in table 2. Examination of table 2 reveals that the difference in availability is usually statistically significant for any two species at the same depth. The relative abundance (availability) at each depth is shown in figure 6. In interpreting this figure it must be borne in mind that the chief object of this otter-trawl fishery by large vessels has been haddock. Thus, the sample of data used in this table comprises a total of 5,437 cor- rected days of fishing with a catch of 92,201,000 pounds. Of this total, 43.4 percent, or 39,955,000 pounds was large haddock, and 29.2 percent, or 26,914,000 pounds was scrod haddock, making a total of 72.6 percent haddock. Since haddock was the principal object of this fishery, the fleet concentrated where haddock could be taken in greatest abundance. Thus, the fact that the fleet spent most of its time in water of medium depth would indicate that the haddock is most often found at that depth. When the haddock move into shallow water for a 100 - _ SHALLOW -| 50 1 -1 J » - 1 - - 1- z UJ 50 — 1 1 MEDIUM 1 1 (r J - H UJ o. - - DEEP 90 — ~ - 0 - o < CD < _l OD O CO z o 2 O 3 m < X < Q* O U (E O o o a o < o o (E O Vi a o o o o o Q < uJ < O o o tu -I o o o o V) o < Ul < X o o o a o < X Figure 6. — Relative abundance of each species of groundflsh in each depth zone in Area XXII South, from otter-trawl catches. ESTIMATING ABUNDANCE OF GROUNDFISH ON GEORGES BANK Table 2. — Dialribution of species of groundfish according to depth zones on Georges Rank. 1932-38 277 Market site (iMunds) As ratio of medium depth As percent at each depth Species Siiallow zone Deep zone Shallow zone Medium sonc Deep lone Blackback 312.6 303.1 187.2 129.1 113.7 91.5 89.1 78.7 65.9 57.6 41.0 31.8 30.9 28.7 15.3 22.4 24.2 12 fi 54.3 ini.2 3.^0 54.0 88.3 58.9 389.9 671.4 93.8 261.2 148. fi 6.4 3,917.0 6.607.0 71.86 70.93 62.44 45.55 36.11 40.40 36.65 29.48 29.31 10. .W 5.05 14.10 7.88 10.35 12.57 22.99 23.40 33.36 35.29 31.76 44.15 41.14 37.45 44.48 18.26 12.31 44.33 25.50 36.06 82.17 2.49 1.49 5.15 5.66 Yellowtail . . 4.20 Halibut - 19.16 Haddock nareel Over2)4 32.14 Cod (market) - ni to 10 15. 45 Haddock (scrod) Under 2H 22.21 Wolffish 33.07 10 to 25 26.21 71.21 82 64 Cod (whale) Over 25 41.58 Hake 66.62 53.59 Cod (scrod) - - IMto2J4 5.26 97.51 Redfish (owan perth) 98.51 106.4 3.7 106.9 4.2 33.96 31.92 34.12 321.2 5.927 4.028.9 65,580 1,087.9 20.694 short time, tlic fleet follows them. Because we get catches from shallow water only for the period that the fleet is there (when haddock are abundant), it appears as thougli haddock are equally abundant in shallow and in medium-depth water, but such is not the case. According to figure 6, the smaller sizes of had- dock tend to be less abundant in deep water, but the true difference between the depth zones for this species may be more pronounced than the data indicate. In the case of pollock, the data are somewiiat misleading. The otter trawlers make a few large catches of pollock in deep water in the fall and winter months, when the pollock are concentrated in dense schools, but these fish are caught only incidentally to the pursuit of haddock during the remainder of the year. Thus, although the data indicate that the pollock is chiefly a deep-water species, pollock are known to frequent all depths. For example, along the Maine coast the pollock school at the surface and are captured by small purse-seine boats. SUMMARY 1. The fishing intensity, by areas and depths, by otter trawlers during a period of 10 years (1928 to 1937, except 1931) was determined for Georges Bank, Subareas XXII, G, H, J, M, \, and O. The information was obtained from i)lo(s of more than 32,000 days of fishing by otter trawlers of more tlian 50 gross tons. 2. During the 10-year period, the otter trawlers fished 23.4 percent of their time in water of 0 to 30 fathoms in depth, 53.0 percent in water of 31 to 60 fathoms in depth, and 23.6 percent in waters of more than 60 fathoms. 3. During the 10-year period, the productive areas amounted to 54.3 percent of 6,531 square miles of shallow area (0-30 fathoms), 51.3 percent of the medium-depth area (31-60 fathoms) of 9,121.5 square miles, and 37 percent of the area of 6,501 square miles of deep area (61 to 125 fathoms). 4. The relative abundance of each species of groundfisli in each depth zone was determined from 5,437 standard otter trawler days of fishing, land- ing 92,201,000 pounds of groundfish from 1932 to 1938 inclusive. 5. The shallow zone was the center of abundance for blackback, lemon sole, and yellowtail flounders. The medium zone was the center of abundance for scrod cod (1}^ to 2K pounds). The deep zone was the center of abundance for ocean perch, cusk, gray sole, pollock, hake, and dabs. Halibut, wolffish, haddock, and cod of more than 2)i pounds did not differ widely in abundance between depth zones. 6. Because of the differences in relative popula- tion densities between depth zones, the catch per unit of fisliing effort cannot be used as a measure of ahundance for most of the species, unless it is tabulated by depth zones. 7. In order to obtain usable indexes of abund- ance for certain of the species, it may first be 278 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE necessary to obtain accurate estimates of the area occupied in each depth zone to permit proper weighting of the index for eacli depth zone accord- ing to the proportion of the population represented. This areal distribution cannot be obtained from the records of the commercial fishery. Therefore, final abundance indexes depend upon surveys of distribution by a research vessel. Such data have been obtained for recent years and are in process of analysis. LITERATURE CITED Herrincton, William C. 1948. Limiting factors for fish populations; some theories and an example. No. 9 in A symposium on fi.sh popu- lations. Yale Univ.. Bull. Bingham Ocean. Coll., vol. 11, art. 4, pp. 229-283. RorN.SEFELL, George A. 1948. Development of fishery statistics in the North Atlantic. U. S. Department of the Interior, Fish and Wildlife Service, Spec. Sci. Rept. 47. 27 pp. SnEUECOR, CiEORfiE W. 1940. Statistical methods. Iowa State College, Ames, Iowa. 422 pp. U- S. GOVERNMENT PRrNTING OFFICE 1957 O — 406253 EFFECTS OF ENVIRONMENT AND HEREDITY ON GROWTH OF THE SOFT CLAM (Mya arenaria) By Harlan S. Spear and John B. Glude, Fishery Research Biologists The relation of the soft, or soft-sliell, clam {Mya arenaria) to its environment is sucli that some flats are favorable for seed-clam production and are not favorable for growth, while the reverse is true of otlier flats. This situation has resulted in the practice of transplanting clams from "seed" areas to "growth" areas, in order to take full advantage of both environments. Obviously it is desirable to know the relative effects of heredity and environment on the growth of the clams; if heredity has the greatest influence it would be desirable to select clams for transplanting from fast-growing stocks, whereas if environment is the dominant factor any convenient source of seed may be used with equal success. The relative efi"ects of stock origin and growth en- vironment, on clam growth, therefore constitute a subject of commercial importance as well as a subject bearing on tlie biological problem of heredity versus environment, or "nature versus nurture." The growth rate of the soft clam varies along the New England coast (Turner 1948); in general, growth is slower in the more northerly and colder areas. In addition, there are local variations in growth rate not obviously caused by water temperatures. The experiment described here was designed to provide information on tlie relative effects of heredity and environment on the growth rate of soft clams. Assistance in the field work of this experiment was provided by Richard E. Tiller, formerly of the Fisli and Wildlife Service, and by Dana Wallace and John Hurst, of the Maine Department of Sea and Shore Fisheries, which cooperated in the experiment. David W. Calhoun, formerly of the Fish and Wikllife Service, assisted in tiie statistical analyses. The Clam Investigations stafl^ of the U. S. Fish and Wildlife Service has been studying the pro- ductivity of Sagadahoc Bay on Georgetown Island, Maine, in terms of the numbers of clams that can be removed annually without causing depletion. The annual clam census, conducted as a part of these studies, has shown that clams in the main part of Sagadahoc Bay grow much faster than those in Bedroom Cove, an adjacent part of the bay (fig. 1). Figure 2 shows comparative growth rates for the main part of Sagadahoc Bay and for Bedroom Cove, as determined by interpretation of rings on the shells.' Tlie reason for the difference in growth rates of clams in the two parts of Sagadahoc Bay must be known for efficient management of the resource. One possible reason is heredity, that is, that the clams in Bedroom Cove are a slow-growing race while those in tlie center of Sagadahoc Bay are a fast-growing race. Another possible reason is that a combination of factors makes the environment in the center of the bay conducive to rapid growth, whereas the environment in Bedroom Cove permits only slow growth. If growtli rates differ because of heredity, a management plan to increase produc- tion might include replacing the slow-growing stock with fast-growing clams; if dift"erences in growth rates are due to environment, the best management plan miglit be to harvest the clams from Bedroom Cove at a smaller size tiian those in the center of the bay, or to transplant them to areas where they would grow faster. Several researchers have discussed tiie causes of variations in the growth rate of soft clams. Mead (1900) observed tliat clam growth depended di- rectly upon the supply of microscopic organisms in the water. Kellogg (1905) indicated tliat dam growth depended on the amount of available food I Speai. llarlnn S., iy.'i4. Rtsulls of poiJUlntlon ci'nsu.s. Sapadahoc Buy. Maine. lin|)iil)lishcil report on file at V. S. Fishery Laboratory. Boothbay Harbor. Maine, 279 280 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE CP I < I I I 0 \A \/i M \ statute Miles P^UiURE 1. — Sagadahoc Bay, Georgetown Island, Maine. ENVIRONMENT AND GROWTH OF THE SOFT CLAM 281 80 2 60 40 I t- o 20 CENTER SAGADAHOC BAY 12 3 4 5 6 AGE IN YEARS FiGORE 2. — Comparat ive growth rates of soft clams from center of Sagadahoc Bay and from Bedroom Cove. ami that etingli()use Cove is a silty area in the Medoniak River system; seed clams are extremely abundant there and the growth rate is low, but better than in Bedroom Cove. 282 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Figure 3. — Location of test areas and origin of test clams along the New England coast. ENVIRONMENT AND GROWTH OF THE SOFT CLAM 283 Table 1. — Average growth of clams in test areas [Values based on samples taken November 1951 to March 1952] Tested in— Soil type Transplanted from- Soil type Date Initial planted length mt Mm. Feb. 28 26.9 Mar. 1 27.5 Mar. 1 33.7 Mar. 1 29.3 Feb. 28 26.6 Mar. 2 27.8 Mar. 2 32.2 Mar. 2 28.9 Mar. 7 33.4 Mar. 7 28.9 Mar. 7 36.1 Mar. 7 32.0 Apr. 2 25.0 Apr. 2 26.8 Apr. 2 21.3 Apr. 5 26.0 Apr. S 25.6 Apr. 5 35.0 Mean growth Proportion recovered alter 1951 growing season Bedroom Cove. Sagadahoc Bay Rohinhood Cove. Falls Cove Plum Island Sound Sandy silt. Sand Silt Gravelly silt . Sandy silt Western Beach Meetinghouse Cove. Sagadahoc Bay Bedroom Cove Western Beach Meetinghouse Cove. Sagadahoc Bay Bedroom Cove Western Beach Meetinghouse Cove.. Sagadahoc Bay Bedroom Cove Western Beach , Meetinghouse Cove. Falls Cove Western Beach Meetinghouse Cove. Plum Island Sound - Sand Silt- Sand-- Sandy silt Sand-. Silt Sand--- - Sandy silt Sand-- - Silt --.. Sand.-- Sandy silt Sand Silt Gravelly silt. Sand Silt -- Sandy silt Mm. 3.99 6.18 2.42 3.55 17.09 20.30 14.48 18.36 14.69 18.03 11.74 16.21 2.26 3.87 2.85 ' 19.03 > 19. 14 ' 19. 69 Percent 56.0 50.7 56.7 50.0 38.7 24.7 67.3 40.7 3.2 7.6 5.4 5.6 31.3 66.7 6.5 0 0 0 ' Growth at Plum Island Sound is based on shell readings of clams that were dead at time of recovery. To increase the geographical scope of the experi- ment, additional experimental plots were estab- lished in Robinhood Cove and Falls Cove in -Maine, and in Plum Island Sound in Massa- chusetts (fig. 3). The design of the experiment is summarized in table 1. EXPERIMENTAL PROCEDURE Soft clams with an approximate lengtii of 25 mm. were used in the experiment. This initial length was chosen because clams near this size were available in all areas and because (the growth rate of small clams being rapid) differences between areas or plots would be greater than if large clams had been used. Another reason for choosing clams about 25 mm. long was that growth rates would be comparable with those listed by Belding (1930). Each clam was marked with Volger's opaque ink to ensure identification of origin. Previous experiments have shown that this ink remains visible on transplanted clams for a period of 2 to 4 years. Origins were designated by symbols in red or black ink so placed as to cover check marks on the shells that niiglit later be confused with the planting check. Care was taken to avoid injury from contact of the ink with the mantle or other soft parts. Clams from each origin were planted in a separate row containing 13 plots spaced 1 yard apart. Twelve of these plots were 1 stjuare foot in area and contained 50 clams each, for monthly .samples. The thirteenth plot in each row was approxinuitely 4 square feet in area and contained a reserve supply of about 200 clams. 409441 O— 57 2 The rows were parallel, 1 yard apart, and so located that all plots were at the same tidal level and were exposed to the same tidal current. All clams were carefully inserted part way into the substrata to prevent them from being washed away before they became established in the sediment. In discussion of the experiment, each row containing clams from a single origin is termed a "group". One plot from each row in Sagadahoc Bay and in Bedroom Cove was dug each month during 1 year. The high mortality of test clams in the other experimental locations prevented adequate sampling during the entire year, but monthly samples were taken as long as survival permitted. At the time of recovery, all clams were meas- ured to the nearest millimeter with vernier calipers for planted length and total length. The planted length was determined by measuring the length of clams at the check mark on the shell caused by transplanting. Mean growth for each plot was computed from the planted and total lengths. Summaries of growth, by montlis. in the five test areas are given in appendix A; the mean growth for each area, based on select "d samples, is shown in table 2. 284 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE T.\BLE 2. — Mean growth of clams in five test areas, based on selected samples Test area Mean powth Number of clams on which mean growth is based Mm. 3.99 16.84 15.49 3.28 19.35 320 276 Robinhood Cove. . 109 Fnlls Cove 160 Plum Island Sound . . 331 ANALYSIS OF RESULTS SURVIVAL Survival of planted clams at Bedroom Cove and Sagadahoc Bay was high enough that sufficient clams remained after the 1951 growing season to provide reliable growth data. The proportion recovered from plots dug in December 1951 and in January, February, and March, 1952, ranged from 24.7 to 67.3 percent, as shown in table 1. The survival at Robinhood Cove was very poor; only 3.2 to 7.6 percent of the planted clams were re- covered after the 1951 growing season. Green crabs, Carcinides maenas, wliich are serious pre- dators of clams, were very abundant in Robinhood Cove and are believed to have been responsible for the poor survival of planted clams. It was necessary to use clams from the supplementary plots to provide enough measurements for analysis of the growth. Clams from Meetinghouse Cove survived satis- factorily when planted at Falls Cove, as indicated by a recovery of 66.7 percent after the 1951 grow- ing season. Of the Western Beach clams planted at Falls Cove, 31.3 percent were recovered during the winter of 1951 and 1952, but survival of native Falls Cove clams replanted in the experimental area was extremely poor. On November 16, 1951, all of the monthly plots and the supplementary plot were dug, and only 13 live clams were re- covered. The poor survival of Falls Cove clams is believed due to their small size, which made them more susceptible to injury from the marking ink used on their shells. If Volger's opaque ink touches the mantle or siphon of the clam it will injui-e the tissues. Since these clams were smaller than those in any other group, the cliances of injury from this source were greater. The growth of Falls Cove clams that survived was intermediate between that of the Western Beach clams and that of the Meetinghouse Cove clams planted at Falls Cove, which had a much higher rate of survival. If the marking ink was the cause of the mortality, it appears that it did not affect the growth rate of the clams that survived. It is likely that the initial size of 21.3 mm. given in table 1 for native clams replanted at Falls Cove is somewhat higli because it is based on shell measurements of the 13 clams recovered at tlie end of the experiment. Many of the clams planted in the spring of 1951 were 12 to 16 mm. long and had the thin shells characteristic of clams of this size. It therefore appears likely that the marking ink was the cause of the poor survival. It was unfortunate that clams closer to the desired planted size of 25 mm. were not available at this location. Each group of clams planted at Plum Island Sound had a mortality of 100 percent during the late summer and autumn of 1951. Before this time, however, these clams had grown at an extremely rapid rate, as shown in table 1. If we can assume that there was no differential mortality among the three groups, the measurement of growth from the shells of dead clams can be used in the analysis. Since the total mortalities of the three groups were identical and since growth rates were nearly identical, varying only from 19.03 to 19.69 mm., it is likely that inclusion of these data will not cause any significant error in the analysis. In fact, the conclusions are the same whether or not this group is included in the analysis. The percentage recovery after the 1951 growing season shown in table 1 is a rough indication of survival, and is based on the number of clams dug from plots during the winter of 1951-52. It is not a true measure of survival, since it does not take into account the clams that moved, or were moved by hydrographic forces, away from the planting location. Frequently, clams planted in one row were recovered in other rows. Sample digging in the vicinity of the test plots also showed that the marked clams had spread over a considerable area. Therefore, the percentage recovery listed in table 1 might be considered as a minimum percentage survival. INITIAL SIZE All clams obtained from each source for use in this experiment were dug at the same time and had a common mean length, regardless of the area to which they were transplanted. At the time of ENVIRONMENT AND GROWTH OF THE SOFT CLAM 285 recovery, however, the initial length, based on measurement of the check-mark on the shells caused by the transplanting, varied among test areas. In each case, the clams recovered from plantings at Robinhood Cove had a greater initial lengtli tlian those from corresponding groups in the other test areas. It is likely that the smaller clams planted at Robinhood Cove were eaten by green crabs (these smaller clams were nearer the surface of the flats), which resulted in a greater initial length of clams recovered in this area. GROWTH Mean growth shown in table 1 is based on the difference between total length and initial, or "planted," length of each clam as determined at the time of recovery. Monthly samples taken during the winter of 1951-52 were combined to provide an adequate sample for statistical analysis. Combining these samples was justified by the fact that there is virtually no growth during this period, as shown by figures 4 and 5. The mean growth rates of test clams ranged from 2.26 mm. for a group at Falls Cove to 20.30 mm. for a group at Sagadahoc Bay. Table 1 and figures 4 and 5 show that there is a tendency for growth rates to vary less within each test area than between test areas. At Bedroom Cove the native clams grew only 3.55 mm., but clams trans- planted from three other origins also grew slowly. At Sagadahoc Bay the clams from the same origins as those planted at Bedroom Cove grew several times as much. At Robinhood Cove all groups grew much faster than did those at Bedroom Cove. Native clams at Falls Cove averaged only 2.85 mm. growth, and those transplanted from Meet- inghouse Cove and Western Beach also grew very slowly. Contrast this with Plum Island Sound, where clams from Meetinghouse Cove and Western Beach grew more than 19 mm. Statistical analyses (described in appendix B) show that the differences in mean growth between test areas are highly significant. It is safe to conclude that clams from a common origin adopt CENTER SAGADAHOC BAY ..a MAR APR MAY JUNE JULY Aue SEPT OCT NOV DEC JAN FEB MAR BEDROOM COVE FinuKE 4. — Growth curves for groups of dams planted in Sagadahoc Bay and in Bedroom Cove, .smoothed by moving averages of three. Origin of clams was as follows: 1, Western Beach; 2, Meetinghouse Cove; 3, Sagadahoc Bay; 4. Bedroom Cove. 286 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE ROBINHOOD COVE —1 1- FEB MAR APR MAY JUNE JULY AUG SEPT 1 OCT NOV 1 D EC JAN FEB MAR FALLS COVE Figure 5. — Growth curves for groups of clams planted in Robinhood Cove and in Falls Cove, smoothed by moving averages of three. Oiigin of clams was as follows: 1, Western Beach; 2, Meetinghouse Cove; 3, Sagadahoc Bay; 4, Bedroom Cove. significantly different growth rates when trans- planted to areas of different growth conditions. The importance of environment as opposed to heredity in affecting the growth rate of clams is emphasized by these results. If heredity were the cause of the differences in growth rates in various areas, we should expect clams that grew fast in their native beds to continue to grow fast when transplanted. Likewise, slow-growing clams would be expected to continue their slow rate of growth after transplanting. Instead, the growth rates of clams in this experiment varied with new environments. For example, Bedroom Cove clams, which grew only 3.55 mm. in their native environment, grew 18.36 mm. in Sagadahoc Bay. At the same time, Sagadahoc Bay clams, which grew 14.48 mm. in their native area, grew only 2.42 mm. when transplanted to Bedroom Cove. EFFECT OF ORIGINS ON GROWTH RATES Analysis of variance tests by areas (see appendix B, table B-2) also show that there are significant differences in the mean growth of groups of clams within each test area. This result might be expected because of the spread in the growth curves (figs. 4 and 5). The growth curves for clams from Meetinghouse Cove were higher than for other groups in each of the four test areas where these clams were planted. The analysis of variance summarized in appendix B, table B-3, shows that the F value was reduced from 13.0 to 6.0 by omitting clams from Meetinghouse Cove. It is also apparent that Sagadahoc Bay clams contributed greatly to the differences within each test area because their grow^th rate was consistently lower than that of the other groups. Since clams from Meetinghouse Cove appeared ENVIRONMENT AND GROWTH OF THE SOFT CLAM 287 to have grown faster than any other group in each test area except Plum Island Sound, the differences in mean growth were analyzed by origins instead of by test areas. Differences between mean growth of groups of clams from different origins were not statistically significant, as shown by table B-4 in appendix B. Although not statistically significant, the ap- parently faster growth of Meetinghouse Cove clams in four test areas suggests another factor in the experiment. Clams in Meetinghouse Cove have a history of slow growth. If this were a hereditary or racial characteristic, we should expect them to grow slowly after being trans- planted to other areas. Instead, the growth rate of Meetinghouse Cove clams was numerically greater than that of clams transplanted from other origins. On the other hand, native clams in the center of Sagadahoc Bay have a record of fast growth (fig. 2), as indicated by a growth of 14.48 mm. during the present experiment (table 1). In the three test areas where these clams were planted, however, their growth was numerically, although not statistically, less than that of any other group. As far as heredity is concerned, these clams would be expected to have grown fast after transplanting. Since they grew slowly, it is likely that a factor other than heredity was responsible. EFFECT OF PREVIOUS ENVIRONMENT A possible explanation for the fast growth of Meetinghouse Cove clams and the slow growth of Sagadahoc Bay clams after transplanting is the effect of their previous environment. Meeting- house Cove is a shallow, silty cove on the west side of the Medomak River estuary. Tidal currents are slow, and this area is protected from current -inducing winds by the surrounding hills. There is a high concentration of slow-growing clams in this area, and competition for food must be extreme. Sagadahoc Bay is a wide, sandy area exposed to the south winds. Both tidal and wind-induced currents are strong. The clam population consists of a few well-scattered, fast-growing individuals. Competition for food is not likely to be a factor influencing growth in this area. Perhaps competition for food causes clams in Meetinghouse Cove to feed more actively or efficiently than those in Sagadahoc Bay which have an abundance of food. If this characteristic persisted after the clams were transplanted to new areas, the Meetinghouse Cove clams might be expected to grow faster and the Sagadahoc Bay clams slower, as was observed in the experiment. SUMMARY 1. The objective of the experiment was to determine whether differences in growth rates of soft clams in two parts of one bay (Sagadahoc Bay) were caused by environment or by heredity. This determination is an economically important consideration in clam transplantation. 2. Test areas were established at five locations along the coast of New England, including the two parts of Sagadahoc Bay. Native clams and clams from two to four other sources were planted in each location. 3. Growth during one growing season was measured by monthly sampling. 4. Good survival resulted at Sagadahoc Bay and Bedroom Cove and in two of the three groups planted at Falls Cove. Survival was poor at Robinhood Cove because of depredation by the green crab, Carcinides maenas. For unknown reasons clams died in Plum Island Sound during the late summer. 5. Mean growth for clams in each test area was as follows: Bedroom Cove, 3.99 mm.; Sagada- hoc Bay, 16.84 mm.; Robinhood Cove, 15.49 mm.; Falls Cove, 3.28 mm.; Plum Island Sound, 19.35 mm. 6. Differences between mean growths in tlie five test areas were highly significant, as shown by analysis of variance. Clams from a single origin grow at significantly different rates when trans- planted to different environments. 7. Differences between growth rates of groups of clams from different origins within each test area were not statistically significant. Therefore, clams from different origins assume similar growth rates when transplanted to a single enviroimient. 8. Although not statistically significant, the numerically faster growth of Meetinghouse Cove clams, and the slower growth of Sagadahoc Bay clams in all cases except one, suggest anotlier factor influencing growth. A tentative explana- tion is the effect of previous environment, which caused clams from a slow-growing area (Meeting- house Cove) to grow fast, and clams from a fast- growing area to grow slowly after transplanting. 288 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Because the observed growth pattern was the opposite from that which would have been expected if heredity were the principal factor determining growth, the conclusions of the experi- ment are not altered. 9. The experiment demonstrated that environ- ment, not heredity, was the important factor in determining growth of the soft clam. LITERATURE CITED Belding, David L. 1930. The soft-shelled clam fishery of Massachusetts. Massachusetts Department of Conservation, Division of Fisheries and Game, Marine Fisheries Series — No. 1. 65 pp. Kellog, James L. 1905. Conditions governing existence and growth of the soft clam (Mya arenaria). U. S. Commission of Fish and Fisheries, Report of the Commissioner for 1903, pp. 195-224. Mead, A. D. 1900. Observations on the soft-shell clam. Rhode Island Commissioners of Inland Fisheries, 30th Annual Report, pp. 20-42. Newcombe, Curtis L. 1935. Growth of Mya arenaria in the Bay of Fundy region. Canadian Journal of R<>search, vol. 13, sec. D, No. 6 (December), pp. 97-137. QUENOtriLLE, M. H. 1950. Introductory statistics. Ltd., London. 248 pp. Butterworth-Springer Snedecor, George W. 1946. Statistical methods. Iowa State College Press, Ames, Iowa. 4th ed. 485 pp. Turner, Harry J., Jr. 1948. Report on investigations of the propagation of the soft-shell clam, Mya arenaria. Massachusetts Department of Conservation, Division of Marine Fisheries (also Woods Hole Oceanographic Institution, Contribution No. 462), pp. 11-42. APPENDIX A— ORIGINAL DATA NUMBER RECOVERED AND AVERAGE GROWTH, BY MONTHS, OF CLAMS TRANSPLANTED IN FIVE TEST AREAS Table A-1. — Bedroom Cove test area [Samples collected on dates marked by asterisks were used in statistical analysis and to obtain mean growth 0(3.99 mm. based on 320 clams] Date sampled Clams transplanted from— Western Beach Meetinghouse Cove Sagadahoc Bay Bedroom Cove Number recovered Average growth Number recovered Average growth Number recovered Average growth Number recovered Average growth April 2 1961 44 37 42 41 33 34 21 19 30 26 28 Mm. 0.0 .3 .2 .9 1.8 2.7 3.8 4.3 3.7 4.3 4.0 22 35 34 31 33 28 20 26 30 21 25 Mm. 0.4 .6 1.1 1.9 3.9 4.3 3.6 5.8 6.0 6.9 5.8 26 31 42 38 32 28 32 30 27 26 32 Mm. 0.0 .2 .4 1.0 1.9 1.9 1.8 2.7 1.6 3.0 2.7 36 36 39 42 38 29 34 25 29 21 25 Mm. 0 0 Mays... _ 1 June6 1.2 July 5 1 9 August 10 ,__ 2.5 2L9 October 1 3 1 Novembers 3.4 2.9 January 20* mt 3.3 February 26* 4.5 Table A-2. — Sagadahoc Bay test area [Samples collected on dates marked by asterisks were used in statistical analysis and to obtain mean growth of 16.84 mm. based on 276 clams] Clams transplanted from— Date sampled Western Beach Meettoghouse Cove Sagadahoc Bay Bedroom Cove Number recovered Average growth Number recovered Average growth Number recovered Average growth Number recovered Average growth I9S1 April 2 27 44 18 28 16 21 22 10 33 19 25 Mm. 0.0 .5 4.6 7.1 13.7 14.6 17.4 19.8 16.9 15.7 18.4 25 41 21 8 15 18 17 10 11 15 11 Mm. 0.0 1.0 5.5 8.5 14.2 18 5 21.1 17.4 20.0 21.1 19.4 31 42 35 35 42 34 19 22 35 27 39 Mm. 0.0 .5 3.3 6l0 9.7 11.3 14.1 11.5 14.6 15.3 ia8 46 27 24 20 33 25 26 25 27 13 21 Mm. 0.0 May3 _ .5 June 6 __ 4.0 June 26 8.5 August 10 13.9 September 12 16.1 October 1 17.5 18.1 December 4* 16.9 wet January30*.. 18.2 March 11* 20.3 Table A-3. — Robinkood Cove test area [Samples collected on date marked by asterisk were used in statistical analysis and to obtain mean growth of 15.49 mm. based on 109 clams] Clams transplanted from- - Date sampled Western Beach Meetinghouse Cove Sagadahoc Bay Bedroom Cove Number recovered Average growth Number recovered Average growth Number recovered A verage growth Number recovered Average growth 1961 April 2 40 41 28 18 10 11 16 Mm. 0.0 .6 2.7 5.5 8.4 17.4 14.7 40 40 27 21 4 10 38 Mm. 0.0 7.0 9.6 18.2 l&O 42 36 22 20 6 10 27 Mm. 0.0 .4 .6 4.1 3.0 12.0 11.7 45 40 » 14 6 12 28 Mm. 0.0 Mav4.. .4 June 4 1.6 July 2 5.2 AuBust9 7.6 September 12 14.6 December 7* 16.2 289 290 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table A-4. — Falls Cove lest area [Samples collected on dates marked by asterisks were used in statistical analysis and to obtain mean growth of 3.28 mm. based on 160 clams] May 2 June 9- July 6 August 7 September 13 - October 15- -.. November 16* . December 11*. . March 19. Date sampled wee Clams transplanted from — Western Beach Number recovered 87 A verage Rrowth Mm. 0.0 1.4 1.4 3.1 2.0 2.3 2.5 2.1 2.6 Meetinghouse Cove Number recovered 62 Average growth Mm. 0.1 1.9 3.1 2.7 3.2 3.1 3.7 3.9 Falls Cove Number recovered Average growth 0.3 1.2 .8 3.8 1.0 2.8 Table A-5. — PUim Island Sotmd test area (Samples collected on date marked by asterisk were all recovered dead but were used in statistical analysis and to obtain mean growth of 19.3.'j ram. based on 321 clamsj Clams transplanted from- - Date sampled Westem Beach Meetinghouse Cove Plum Island Sound Number recovered Average growth Number recovered Average growth Number recovered A verago growth 19S1 April 30 34 30 9 3 1 3 77 Mm. 1.2 8.9 18.3 27.0 20.0 31.7 19.0 41 36 33 26 1 0 95 Mm. 1.3 10.4 18.0 23.6 33.0 45 36 34 10 3 2 149 Mm. 1.7 May 31 July 2 _ __:.. 9.9 17.0 August 3 -■•.-. 22.7 Septen ber 7 ,. . . .... 23.0 28.5 December 4' 19.1 19.7 APPENDIX B.— STATISTICAL ANALYSIS ANALYSIS OF VARIANCE BETWEEN AND WITHIN TEST AREAS Standard deviations for the 18 groups of clams plotted against their means follow a straight line having the formula E= 1 .15 + 0.284A' (fig. B-1 and table B-1). The slope of this line indicates the need for transformation to make variances independent of the means in order that methods for analysis of variance shall become applicable. The fact that standard deviations plotted against means follow a straight line indicates that the log transformation is the one to be used (Quenouille 1950). Figure B-2 shows tiie variance plotted against the mean for each of the 18 groups of clams after the values had been transformed by taking the log of the midpoint of each '2-mm. class plus 1 (table B-1). The very sliglit slope of the least- squares line, as indicated by the formula E— 0.077 — 0.0304A', indicates that the variances have been made virtually independent of the means. Analysis of variance can therefore be completed, using the transformed values. Table B-2 shows the completed analysis of variance of differences in mean growth between and within test areas using transformed values. The F value for a comparison of between and within test areas was 43.9, which is highly signifi- cant. This indicates that differences between growth rates in the various test areas are liighly significant. A comparison of the differences within test areas and between individuals yielded an F value of 13.0, which is also highly significant. This indicates that there is a considerable amount of variation among the groups of clams that were used in the various test areas. It appears likely from examination of the untransformed mean growths in table 1 (in text) and from the growth curves in Hgiires 4 and 5 (in text) that Meeting- house Cove clams are largely responsible for the high F value in this test. ENVIRONMENT AND GROWTH OF THE SOFT CLAM 291 T.\Bi.E h-\. — Original and transformed mean, variance, and standard deviation for growth of 18 groups of clams in i-mm. classes used in figures B-1 and B-2 [Transformation is based on formula; Transformed X = log (class midpoint + U. Total number of clams, 1,186) (Jroup code letter OrtKin Test area Number of clams Original arithmetic mean (mm.) Original standard deviation (mm.) Trans- formed arithmetic mean Trans- formed variance Bedroom Cove 84 76 85 75 61 101 37 77 28 il 16 ICO 13 47 77 149 95 4.02 6.15 2.36 3.57 18.50 14.58 20.01 16.99 16.14 11.76 18.08 14.62 3.96 2.81 2.37 18.94 19.72 19.15 2.87 2.72 2.13 2.38 5.21 ,'j.54 6.49 .5.43 6.97 3.69 4.37 5.87 1.77 1.60 1.78 9.17 6.67 6.63 0.625 .813 .441 .591 1.274 1.159 1.297 1.233 1.172 1.085 1.268 1.138 .666 .534 .463 1.237 1.283 1.289 0.073 do .045 c D E do .075 do .068 do Sagadahoc Bay - .015 do --- .033 Q do -.- .026 H Western Beach do - .021 Robinhood Cove .081 J ...do - .021 do - Oil do - .082 M Falls Cove .028 Falls Cove do .050 do .-. - -- .060 do Plum Island Sound .075 0 _ _. Plum Island Sound -- do --- -- do - .040 s 7 8 9 10 MEAN GROWTH II IN 12 M M 13 14 15 17 18 19 20 21 Figure B-1. — Standard deviation plotted against arithmetic mean growth for 18 groups of clams listed in table B-1. The slope of the trend line fitted b.v least-squares method indicates the need for transformation to make anal.vsis of variance applicable. o UJ z cr < < cr > .2- Less t Squares Line E= 0.077— 0.0304 X —I 1 1 1 1 1 1- .5 .6 .7 .8 .9 1.0 I.I TRANSFORMED MEAN GROWTH 1.2 1.3 1.4 1.5 1.6 Kif:rRE B-2. Variance plotted against aritlinictic mean growth for 18 groups of clams listed in table B-1 after growths were grouped in 2-inni. classes and transformed by taking the log of each class midpoint plus one. The extremely slight slope of the trend line fitted by the least-squares method indicates that this transformation has made variance slight slope virtually independent of the mean, so analysis of variance method can be used. 292 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table B-2. — Completed analysis of variance of differences in mean growth between and within test areas Itased on 18 groups of clams at 5 areas ll'sinB values transformed by the formula: Transformed .Y = log (midpoint of 2-mm. class + 1)] Source of variation Degrees of free- dom Sum of squares Mean squares F Between test areas - Within test areas 4 13 1,168 114.09 8.44 53. 02 28.52 0.65 0.05 •'43.9 Between individuals "13.0 Total 1,185 176. 15 The analysis of variance by test areas was recomputed without the groups of clams that came from Meetinghouse Cove; the results are listed in table B-3. While the differences in mean growths of groups within tlie test areas are still highly significant, the F value has been reduced from 13.0 to 6.0 by exclusion of Meeting- house Cove clams. Therefore, it appears that clams from this origin were responsible for more than half of the F value of 13 listed in table B-2. Also, it appears likely that the clams from Saga- dahoc Bay contributed a large part of the high F value for this test (figs. 4 and 5). Since the growth pattern of both the Meetinghouse Cove Table B-3. — Completed analysis of variance of differences in mean growth between and within te^l areas ttased on IS groups of clams (excluding those from Meetinghouse Cove) at 5 areas (Using values transformed by the formula: Transformed X= log (midpoint of 2-mm. class + 1)] Source of variation Degrees of free- dom Sum of squares Mean squares F Between test areas. 4 8 827 93.32 2.43 42.17 23.33 0.30 0.06 Within test areas. ••77. 77 "6.00 Total 839 137. 92 clams and the Sagadahoc Bay clams was the opposite of that which might be expected had heredity been the cause of growth differences and growth rate, the significance of this F value does not alter the conclusions given here. ANALYSIS OF VARIANCE BETWEEN AND WITHIN ORIGINS The possibility that variation between the se- ries means was caused by the origin of the test clams needed to be explored. Mean growths of clams from the four origins planted at Bedroom Cove, Sagadahoc Bay, and Robinhood Cove were used in this analysis, because only at these three test areas were all four groups planted. Results of the analysis-of-variance tests are shown in table B-4. The F value of 8.42 for a comparison between origins and within origins was not significant at the 5-percent level. There- fore, the effect of the origin of the clams on their growth rate after transplanting was not significant. The F value of 166.8 for a comparison within origins and between individuals was highly sig- nificant, indicating (as would be expected) that the differences in mean growth of clams from each origin planted in the three test areas were highly significant. T.'^BLE B-4. — Completed analysis of variance of differences in mean growth between and within origins of clams based on 12 groups of clams from 4 origins [Using values transformed by the formula: Transformed X = log (midpoint of 2-mm. class + D) Source of variation Degrees of free- dom Sum of squares Mean squares F Between origins Within origins 3 8 693 2.97 66.74 31.92 .99 8.34 0.05 8.42 ••166.8 Total 704 101. 63 U.S. GOVERNMEt.T PRINTING OFFICE 1957 O — 409441 CLIMATIC TRENDS AND THE DISTRIBUTION OF MARINE ANIMALS IN NEW ENGLAND By Clyde C. Taylor, Fishery Research Biologist, Henry B. Bigelow, Oceanographer, and Herbert W. Graham, Fishery Research Biologist For many years Americans have commented on an apparent warming of their climate; older people have referred to the "old-fashioned winters" they once knew. Climatologists long shrugged off the idea as unfounded, but a melioration in climate is no longer confined to the popular mind : a decided trend toward warmer winters during the past 50 years is now well-documented. Air temperatures in winter, particularly since 1910, are definitely higher in higher latitudes of the Northern Hemi- sphere and probably throughout the world gen- erally. Glaciers have been receding and in far northern latitudes, plants and land animals fol- lowing the retreating ice have extended their ranges northward and to higher altitudes. P'or a bibliography dealing with responses of plants and animals to climatic changes, the reader is referred to Rapports et Proces-Verbaux des Reunions, vol. 125, pp. 42-52, C'onseil Permanent Inter- national I'Exploration de la Aler, Part 1, 1949. Warming of the oceans during periods of higher air temperatures is difficult to demonstrate because of the paucity of observations of sea-water temperatures. Evidence shows, however, as Smed (1949, 1953b) points out, that the Arctic Ocean has warmed appreciably since 1921. This author also presents evidence of increased water temper- atures beginning in the 1920's in the North Sea and in the North Atlantic from the British Isles to the west coast of Greeidand. The warming of northern waters has been ac- companied by the northward extension of many marine vertebrates to the region of Iceland (Fridriksson 1949) and by profound changes in the fish populations around Greeidand (T;\ning 1949). The development of tlie cod fisliery on tlie west coast of Gicenland has been spectacular. As the Note.- Clyde C. Taylor ami llcrbiTl W. Graham, United States Fish and Wildlite Service, Woods Hole. Mass.; Henry B. BiRelow, Museum of Compi',rativ<' Zoology, Harvard I'niversity. Approved for publication, November 1, 1956. waters warmed in this area, an offshoot from the Icelandic stock of cod became established and now supports a substantial fishery (Jensen and Hansen, 1931). In the years 1911 to 1921, the West Greenland cod fishery produced less than 500 tons a year. In 1925, the catch doubled and thereafter continued to increase. In 1952, some 252,758 metric tons of cod were landed from the West Greenland area (International Gommission for the Northwest Atlantic Fisheries, 1954, table 2, p. 28). The fishery now reaches 300 nautical miles farther north than formerly. The Eskimos in some areas who had never seen a cod in 1924 are now busily occupied in the cod fishery, whereas they formerly were seal hunters. Although temperature is only one factor in the ecological complex determining the presence or abundance of a species, in high latitudes temper- ature may in some instances be the sole limiting factor and have a direct effect on distribution. Thus, cod show a definite response to low tempera- tures and their northward extension is probably determined by temperature alone. The abundance of cod in the Greenland area may be related to temperature in a somewhat less-direct manner. Hermann (1953) has shown that the strength of cod year-classes in Greenlaiul waters has a very high correlation witli bottom temperatures in June. Thus, temperature in some way affects survival of whole populations of young fish, per- liaps through affecting their food supply or rate of growth and, conseciuently, their resistance to adverse enviromnental conditions. The warming of arctic areas and the accompany- ing ecological changes have been so marked and so well-documented tliat it seems reasonable to suppose that similar changes liave taken place, although perliaps on a smaller scale, in more southern latitudes. It is the purpose of this paper to examine temperature fluctuations in recent years, and to explore the relations which may 293 294 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE exist between these fluctuations and the abundance and distribution of marine animals along the eastern coast of the United States and in the New England area, in particular. In the following pages we present some of the available data on trends in air and sea tempera- tures and trends in the distribution of certain species of marine fish and invertebrates. We are aware that, in some instances, we may be mis- interpreting the causes of observed changes, or even may be misled in believing that some of the changes have occurred. It is hoped that the pres- entation of these relationships will stimulate others, especially specialists in particular fields, to examine more critically the data tliey may have at hand. A great deal of the theory of fishery science is based on the premise that the environ- ment is unchanging and that the fluctuations which do occur take place within certain limits on either side of a stable norm. We find, therefore, that changes in abundance of fishes are frequently attributed to the effects of overexploitation. If the premise of a stable environment is not valid, it will be necessary, at least, to reexamine the overfishing explanation of such fluctuations. The observations on which this paper is based end with the winter of 1953-54. The authors emphasize that their purpose is to docum.ent the events of a period of warm.ing of air and sea tem- peratures. No prediction of future temperature trends is offered. TRENDS IN AIR TEMPERATURES Extensive evidence of an upward trend in air temperatures over the United States and Canada was presented by Kincer as early as 19.33. In addition to a general upward trend in annual means, Kincer's (1933) analysis showed that winters, springs, and falls were becoming milder, while summer temperatures were remaining about the same. Similarities in trends as well as in the patterns of fluctuations are evident in Kincer's diagrams at representative stations throughout the United States and Canada. Kincer also showed similar trends for other stations through- out the Northern Hemisphere and for a few in the Southern Hemisphere. His data encourage one to believe that air-temperature records at any one point on the eastern seaboard will reflect the general trend of air temperatures for latitudes north and south, although, of course, not in level or in magnitude of fluctuations. It is now generally conceded that a significant warming has occurred throughout the Northern Hemisphere. As Mitchell (1953, p. 244) states — It apparently has taken the relatively severe temperature changes of more recent years, coupled with many kinds of climatological, glaciological, oceanographical and biological evidence, to establish the unmistakable reality of important climatic trends in secular time. The pattern of these changes is examined in the following section to establish a background to which changes in the abundance and distribution of marine species may be related. The longest series of air temperature records for North America was taken at New HaVen, Conn., beginning in June 1778 (Loomis and Newton, 1866). Monthly means of these tem- peratures are available (Clayton, 1927, 1934; Clayton and Clayton, 1947; U. S. Weather Bureau, 1941 to 1953). Many years ago, Loomis and Newton made a comprehensive study of the New Haven series. These authors not only provide an account of the early observers and circumstances of observation, but also reduce the observations to a standard series corrected for the time of day the observations were made. After comparing the temperatures for tlie first and second halves of the period of study, the authors state (p. 246), "We conclude therefore, finally, that during the past 86 years there has been no permanent change at New Haven either in the mean temperature of the year, or in that of any of the separate months; * * *." This conclusion is the more striking because of the subsequent upward trend in New Haven tem- peratures. Annual deviations from the mean computed for the period 1780-1950 (49.3° F.) are presented in figure 1 and a curve of 5-year moving averages has been drawn through them to indicate the trends. The data have several interesting aspects. Considering deviations from the grand mean by 20-year periods beginning with 1781, the minimum deviations occurred over the periods 1821-40 and 1841-60. Since 1900, 44 of the 51 annual means show positive deviations, and over half of the positive deviations are greater than 1° F. Equally warm periods occurred prior to 1900. The out- standing feature revealed by figure 1 is that equally cold periods have not occurred. The CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 295 trend in minimji lias been upward since about tlie first third of the 19th century. Annual means may not indicate all of the significant changes that may be occurring in an area. Mild winters and cool summers may, for example, result in tlic same annual means as severe winters and very warm summers— the climate in the two situations being quite different. In figures 2 and 3, curves for January and July deviations from their respective means are shown for New Haven for the period 1780-1953. The curves are smoothed by 15-year moving averages. The January temperature deviations show a pro- nounced trend from a low in the 1810's. The mean for the period 1780-1900 is 26.72° F., while that for the period 1901-53 is 29.66°, a difference of nearly 3°. The mean since 1930 is 30.57°, an increase of 3.85°, even tliough the trend was gen- erally downward over the latter part of this period. It will be observed, on the other hand, that aside > UJ 4.0 3.0 • • • • 2.0 / • • • • A • • • • • \yV\A 1 .0 n • *• / 'V* •/ \ • * /• ^ •s • 'A '^' ••a/^ • • « - :• % • 1 .0 V ^\ 1 • ^ \ L * \ J * ■/t^ • • 2.0 3.0 • — • • • • • • • ■ • • •• • • • • • • - laio 1830 1930 1850 1870 YEAR Figure 1. — Annual deviations from the mean air temperature, 1781 to 1950, at New Haven, Conn. The solid line is a 5-year moving average. 1790 1810 1830 1880 1870 1890 1910 1930 YEAR Fkure 2. — January air temperature deviations from the mean, 1780 to 1953, at New Haven, Conn. (Curve smoothed by 15-year moving average.) 296 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 1 1 1 1 1 1 1 1 MM III! TTT-r 1 M 1 Mil 1 1 1 1 MM II 11 1 II 1 IT IT MM ■^ - - 3.0 z -_ u: 2.0 '- 'A. -_ z 1.0 - . V / ^ V /^ ■'V" -_ o 0 ; / \ ..W^ ./ \ Ai I^K ^ / - : V \ /«^A -1.0 — ^A \ J - UJ - v ^i/- Y - o -2.0 -3.0 ~ - nil a±xj 1 1 1 1 1 1 1 1 1 1 1 r 1 1 1 1 1 1 1 1 1 1 1 1. MM Mil 1 1 M 1 1 1 1 nil nil ini IMl null 1790 1810 1830 1850 1870 1890 1910 1930 1950 YEAR Figure 3. — July air temperature deviations from the mean, 1780 to 1953, at New Haven, Conn. (Curve smoothed by 15-year moving average.) 1780 1800 1840 I860 1880 YEAR 1920 19 40 Figure 4. — January air temperature deviations at New Haven, Conn., 1780 to 1953 (solid line), compared with those for the winter months of December, January, and February, 1849-1950 (broken line), at the Bluehill Meteorological Observatory. (Curves smoothed by 5-year moving averages.) CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 297 from showing roughly the same periochc fluctua- tions, the July temperatures have remained very nearly tlie same over the entire period. The imperfections of the New Haven tempeia- ture record have been pointed out by Mitchell (1953). He considers tiie most serious of these to be tlie effect of urban development which may have contributed as much as 1° F. to the average temperature during tlie winter montlis. He cites the temperature records at the Bluehill Meteoro- logical Observatory of Harvard University, lo- cated 10 miles south of Boston, as being quite free from urban effects. Conover (1951) has pub- lisiied mean winter temperatures (Decemt)er, January', and February) for the Blueliill Observa- tory covering the period 1849-50 to 1949-50. In figure 4, we have compared the January devia- tions at New Haven with those at Bluehill for tlie winter months. The difference between the January means at New Haven for the periods 1780-1900 and 1901- 53 is 2.94° F. The standard error of this differ- ence is 0.6873. The difference is then iiighly significant (/ = 4.28, 172 d. f.). If the difference were only 1.8°, it would still be highly significant, and if only 1.35°, it would be significant. The reality of the increase can hardh' be denied even if one allows a maximum value for the effect of urban development. With regard to tempera- tures at the Bluehill Observatory over the period 1849-1950, Conover (1951, p. 9) states, "The amount of warming up in general winter temper- atuie over tiie last 100-year perioil has been about 3M° Fahrenheit." To analyze further the pattern of temperature changes, mean temperatures for each month liave been averaged for the 20-year period 1780-1800 and by 25-year periods thereafter to 1950 for New Haven, Conn., and for comparison, from 1876 to 1950 for Eastport, Maine (fig. 5). The New Haven temperatures were adjusted to the level of Eastport temperatures by subtracting from them the mean difference between tempera- tures at New Haven and at Eastport by months for the period 1874-1923. The means for this period for Eastport and New Haven are given by Clayton (1927). Upward trends in temperature are noted fo 22 20 I 8 40 38 h 36 62 lU 60 I- 58 < tt UJ Q. APRIL -8=- JULY 24 - 22 - 20 48r 46 44 62 60 58 36 34 1 1 r FEBRUARY MAY AUGUST 30 28 26 56 54 52 56 54, 52 26 24 h JUNE SEPTEMBER 1780 1801 1826 1S51 1876 1901 1926 1780 1601 1826 1851 1876 1901 1926 1780 1801 1826 1851 1876 1901 1926 TO TO TO TO TO TO TO TO TO TO TO TO TO TO TO TO TO TO TO TO TO 1800 1825 1850 1875 1900 1925 1950 1800 1825 1850 1875 1900 1925 1950 1800 1825 1850 1875 1900 1925 1950 PERIOD Figure !y. — \ comparison of trends in monthly temperatures avcruKed over 21-year (1780-1800) and 25-ycar period.s; New Haven, Conn., 1780-1950 (.solid line), Ea.stport, Maine, 1876-1950 (broken line). •JI4171 O— .")- 2 298 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE the months of January, February, October, No- vember, and December. March and September also show rather pronounced upward trends for the latter part of the period. It will be noted that temperatures at New Haven were, in general, relatively cooler than at Eastport for the period 1875-1900 and relatively warmer for the periods 1901-25 and 1926-50. These differences in the later years may reflect urban development at New Haven. Figure 5 shows clearly that the important changes in temperature occurred during the fall and winter months. SEA SURFACE TEMPERATURES Trends in sea temperatures in the North Atlantic north of latitude 55° have been reported by Smed (1949, 1953b). Some records covering periods of time sufficient to indicate significant trends are available for the Atlantic coast. Hachey and McLellan (1948) have publislied data for St. Andrews, New Brunswick, for the period 1921-47. Annual means, with 12-month moving averages of temperatures, are shown by Lauzier (1952) for St. Andrews and for Sambro Lightship, Nova Scotia, for the period 1936-51. Daily readings of surface temperature were made at Boothbay Harbor, Maine, for the period 1905-49. These readings were made at 8 a. m., 12 noon, and 4 p. m., in conjunction witii the oper- ations of the fish hatchery there during that period. The monthly means are presented in appendix talkie 1, p. 344. Examination of the original records indicates that these temperatures (I'ccorded to the nearest degree) were carefully taken for the most part; however, some tempera- tures below the freezing point of sea water were recorded. The trends and fluctuations in these temperatures are in good agreement with those appearing in the St. Andrews data (cf. figs. 6 and 7). Annual deviations from the 43-year mean at Boothbay Harbor (1906-48) are shown in figure 6. There is only a slight increase in the annual means over the period. For the 25 years, 1906-30, the annual means average 45.9° F.; for the 18 years, 1931-48, the average is 46.4°. A striking melioration of winter conditions in recent years is shown when the January-July difference in water temperature at Boothbay Harbor is plotted as a deviation from the mean difference for the period of record (fig. 8). The average January and July temperatures for various periods are as follows; Period January July 1906-20 -. - 33.3 33.0 34.6 35.2 61. 1 1921-30 -.- 61.6 1931-40 60.7 1941-49 - 60.5 J r < > . -2 -3 I I I I I I I I I I I 1"""^- I ' ' ' I I X_L_ I I I I I I I I I I I I I I I I I 1904 1910 I 920 I 930 YEAR I 940 950 FidiRE 6. — .\iiiiual dt'viatioii.s from the mean surface temperatures, 1906 to 1948, at Boothbay Harbor, Maine. (The solid line is a 5-year movius average.) CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 299 Since 1930, January tcniporaturcs liavc increased about 2° while July temperatures have deereased about 1°. These opposite trends account, of course, for tlie magnitude of the phenomenon to be observed in fifjure 8. Recordsof surface water temperatures, smoothetl by 5-year moving averages, for each month over the period 1905-49 (fig. 9) show significant upward trends for the months of January, February, No- vember, and December. To establish the degree and significance of these trends, we have (livi(U'd < > 1920 1928 1936 1944 1952 YEAR FiGfRE 7. — Annual deviations from the mean surface water temperatures, 1921 to 1951, at St. Andrews, New Brunswick. (The solid line is a 5-year moving average.) 1925 1930 1935 YEAR FifiiRE 8. — .Ianu:iry-.luly differences in surface water temperatures as deviations from the mean difference, 1906 to 1949. at Boothbay Harbor, Maine. (The curve is a 5- .year moving average.) tile records into two approximately equal periods of 22 and 23 years, calculated tlic mean for each month for each period, tiie standard error of the difference between these means, and Student's t (table 1). January, February, and December show highly significant increases in mean tem- perature, while a statistically significant increase has occiu-red for the November period. March, July, August, and September siiow slight down- ward trends, but these trends are not statistically significant. Table 1. — Coinparison of mean monthly nurface tempera- tures at Boothbay Harbor, Maine, for the first and second halves of the period 190o-J,9 [Temperatures In °F.] Month Mean Diflerence ( Period A > Period B ' 33.00 31.26 35.40 39.50 47.04 .54.83 01.21 61.26 56. 98 50.10 43 38 37.00 34.89 33.33 34.95 39.98 47.44 55.01 60. .54 eo. 95 56,90 .50. 73 44.70 38.15 + 1.89 +2.07 -0.45 +0.48 +0.40 +0. 18 -0.67 -0.30 -0.08 +0.63 + 1.32 +1.15 9.27" 3.58" March - 0.70 April 0.93 0.49 June -~ - 0.27 Julv 1.12 August 0.61 September 0.16 1.02 November 2. 14* 5.08" I Periods A and B for January and February are 1906 to 1927 and 1928 to 1949; for March to July, 1905 to 1927 and 1928 to 1949; for -August to December. 1905 to 1926 and 1927 to 1948. respectively. See also appendi.x table 1. p. 344. •5-percent level of significance. '•1-percent level of significance. Water-temperature records for Woods Hole, Mass., for the period 1881-1914, 1932-42, and 1945-52 are given in appendi.x table 2. The means for various periods are as follows : Period January July Annual mean 1885-94 - 34.2 32.9 33.8 34.0 35.2 69.2 68.3 69.6 67.6 70.3 50.9 1896-1904 - - 50.7 51.1 1933-41 50.3 194^51 - - ' 52.3 The trend in surface water temperatures for January at Woods Hole is similar to that for Boothbay Harbor for comparable periods. Except for the period 1933-41, liowever, there does not appear to be any trend at Woods Hole toward declining July temperatures comparable to that for Boothbay Harbor. T1h> annual mean for the period since 1945 is considerably higlier than tlie annual means for the earlier periods. This in- crease is statistically significant, 'i'lie data sliow 300 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE _l I I I I I I M I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I l_ UJ UJ 57 in 3 55 1- 53 < a: 6? ijj Q_ 60 920 1930 YEAR 1940 Fkuire 9. — Trends in surface water temperatures at Boothbay Harbor, Maine, for each month of the year, 1905-49. (Curves are 5-year moving averages.) CLIMATE AND THE DISTKIBUTION OF MARINE ANIMALS 301 that the winters have been generally milder since 1932 and that a significant warming has occurred since 1945. RELATION BETWEEN AIR AND SEA TEMPERATURES Many sources of evidence indicate that tlie warming of the air began perhaps as early as 1850 and that air temperatures have been at a generally high level since 1900. Data on sea temperatures cover, for tiie most part, periods since 1900. Thus, we may be observing fluctuations in sea temperatures at a higher level than occurred in the earlier period. By e.xamining the correlation between air and sea temperatures, the likelihood of this possibility may be determined. The relation between Eastport air temperatures and Boothbay Harbor water temperatures for the month of January during the period 1906-49 is illustrated in figure 10. The correlation coeffi- cient is 0.658, which is highly significant. The same relation using New Haven air temperatures is shown in figure 1 1 . The correlation coefficient is 0.610, also highly significant. The relationships are not (juite good enough to predict with accuracy the surface temperature from a given air tempera- ture; however, the correlation is sufficiently high to conclude that these air temperatures are a rough index of the general level of surface water tem- peratures. On this basis we can assume that win- ter surface water temperatures prior to 1900 were generally lower than after that date. Further confirmation of our conclusion that changes in air temperatures for the period prior to 1900 are generally indicative of changes in sur- face water temperatures is to be found in a series of water temperature records for Eastport, Maine, for the period 1878-87 (Moore 1898). These data are compared with air temperatures for the same period and with air and water temperatures at Eastport for the period 1941-50 (table 2). The general increase in air temperatures for each month between the two periods is faithfully re- flected in a corresponding increase in water tem- peratures. o UJ D= 36 < CC 34 LU CL i^ 32 DC LU 30 I- < ^ 28 5r^ °K- J I 1 I L J L J L J I I L J I L 12 14 16 8 20 22 24 26 28 30 AIR TEMPERATURE F. FiciKE 10. — The relation between Eastport, .Maine, air temperatures and Boothbay Harbor. .Maine, water temperatures for the month of January, 1906-49. 302 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE UJ 01 3 UJ cc 38 36 < a: ^ 34 Q- 32 I- 30 < 28 I I I I I r T I I I \ 1 r "> 1 T X _L ± _L 20 22 24 26 28 30 32 34 AIR TEMPERATURE°F 36 38 40 Figure U. — The relation between New Haven, Conn., air temperatures and Boothlaay Harbor, Maine, surface water temperatures for the month of January, 1906-49. Table 2. — Comparison of mean monthhj water and air temperatures at Eastport, Maine, for the periods 1878-87 and 1941-50 [Temperatures in °F.) Month January... February.. March April May June July Aupust September October. . . November. December- Mean. Air tempera- Water t empera- tures Differ- tures ence 1878-87 1941-50 1878-87 1941-60 20.30 21.62 -t-1.32 36.91 37.73 22.26 22.48 -fO.22 34.08 34.68 28.08 31.20 -t-3. 12 33.82 34.64 38.61 39.42 -1-0.91 36.58 36.77 47.44 48.28 -t-0.84 39.99 40.06 64.78 65.35 -1-0 57 44.02 44.16 60.67 61.13 -fO. 97 47.98 48.32 60.50 61.64 -fl.14 50.53 51.33 66. 89 56.80 4-0.91 51.23 52.25 47.34 48.87 -1-1.63 49.98 51.03 37.01 39.40 -1-2.39 46.33 47.82 25.82 29.39 -1-3.67 41.48 43.02 41.56 42.96 -fl.41 42.74 43.48 Difler- ence -1-0.82 -1-0.60 -1-0.72 -1-0.19 4-0.07 -f-0. 14 -1-0.34 -1-0.80 -1-1.02 -fl.05 -1-1.49 -1-1.54 -1-0.74 TEMPERATURE TRENDS IN OFFSHORE WATERS OF GULF OF MAINE Except for an initial pelagic period common to most marine fish, the more-important commercial species spend the larger part of their existence on the sea bottom, and most pelagic species are found at some time or otlier throughout the water column. If the trends shown in air temperatures and sea surface temperatures are superficial phenomena producing no important changes in the depths where the various species are foimd, any changes in distribution and abundance of species inhabiting these depths must be ascribed to causes other than temperature changes. In the following sections, hydrographic data collected at various Gulf of Maine stations between 1912 and 1954 are examined. GULF OF MAINE TEMPERATURES, 1912-26 The observations recorded during the cruises of 1912 to 1926 (Bigelow 1927, pp. 522-701 and tables, pp. 978-1014) provide the only detailed information that has yet been published on the temperature of the offshore parts of the Gulf of Maine at diff'erent seasons and depths. These observations, therefore, must serve as the basis CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 303 for a comparison between the temperatures of the water of the Gulf in recent j-ears and during the period 1912-26. It seems appropriate, then, to commence tliis survey with a brief summary of the temperatures of the Gulf for the earlier period. Depths to 150 meters Ijile mnter minimum. — The chief cause for the winter chilling that is so conspicuous a phase of the seasonal temperature cycle in the Gulf — reach- ing its clima.x some time in February — is the loss of heat by radiation from the surface of the sea during tlie part of the year when the overlaying air is colder than the water. Since the coldest winds in winter blow out over the Gulf from the quarter between west and north, surface tempera- tures fall as much as 4° to 5° lower along the western and northern margins of the Gulf than over the central basin or Georges Bank offshore (Bigelow 1927, p. 523, fig. 1). This is equally true whether the season is severe or mild. A few days of near-zero weather, with high west-to-north winds, at any time in late December, January, or February are enough to chill the water in enclosed situations all around the coastline of the Gulf to the freezing point of salt water, i. e., to about 28.9° F. at a sahnity of 32°/oo. This is about the lowest temperature to be expected around the shoreline of the Gulf, except close to the mouths of rivers. How much ice actually forms under these conditions varies greatly from place to place, depending on local topography, on strength of tidal currents, on rapidity of interchange with the waters outside, and on the thoroughness with which the channels are kept open by passing vessels. For details in this respect concerning harbors, bays, and rivermouths along the coast from Provincetown, Mass., at the tip of Cape Cod, to tlie mouth of Passamaquoddy Bay, the reader is referred to the Coast Pilot, Atlantic Coast, section A (l^. S. Coast and Geodetic Survey, 1918). Ice has never been known to mass in any amount outside the outer islands and headlands north of Boston. But cases are on record of ice from neighboring harbors and from the shallower parts of Cape Cod Bay massing in heavy fields or windrows, sometimes as mucli as 10 feet tliick, out in Cape Cod Bay (U. S. Coast and Geodetic Survey, 1918; p. 277). It was not unusual in Februarv for the surface to chill below 30° F. there, and along the coast as far north as the offing of Boston. Two such areas with negative temperatures developed in February 1925: one area off the Scituate-Marshficld shore (31°-32° F.), the other in the central part of Cape Cod Bay (30.9° at the surface on the 9th, and 29.2° at 17 meters) doubt- less resulted from the ice that had extended a mile or more offshore south of Wellfleet on December 29, 1924 (Bigelow 1927: p. 655). The area occu- pied by water colder than 32° may have been more extensive still in Massachusetts Bay in 1934. That year the mean water temperature in Boston Harbor was onh' 29.8° for February, and a photo- graph taken from the open Cohasset coast on February 10 shows the pack ice reaching at least 1)2 miles seaward. The severe winter of 1920 was of this same general type in the Gulf — to judge from readings of 33.4° at the surface and 32.6° at 50 meters off Boston as late as March 5 — also the winter of 1923, when a mean water temperature of only about 30.2° in Boston Harbor for February was followed by vernal warming so tardj- that the water off the Scituate shore was still only about 37° at the surface and 32.6° at 80 meters on the 18th of April. In short, winters of the general order of severity represented by 1925 were not exceptional during the 2d to 4th decades of the 20th century, though they did not recur regularly. But the available record makes it most unlikely that the surface temperature ever falls below 32° to 33° F. for more than a few hundred yards out from tide line off the open coast of the Gulf anywhere to the north of Boston, even during the most severe wnnters. On February 7, 1925, for example, when the sur- face temperature was 31° inside Gloucester Har- bor, it was 35° only 1 mile outside. Readings of 32.9° to 34.8° F. taken in March 1920 (a tardy spring) at the 40-meter level — to which vernal warming had not yet penetrated — in the trough between Jeffrey's Ledge and the New Hampshire coast, and near Wood Island, Seguin Island, Great Duck Island, and Petit Manan Is- land off the Maine coast, point to a seasonal mini- mum of 32° to 35° all around the periphery of tlie Gulf northward and eastward from Cape Ann. In the more severe winters, this minimum includes the outer part of the open Bay of Fundy to judge from Mavor's (1923, p. 375) record of 34.6° at the surface and 34.2° at 10 metei-s, on February 7. 304 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 1917. A 40-meter reading by the Albatross of 34.8° some 12 miles off Yarmouth, Nova Scotia, on March 23, 1920, suggests similar temperatures along the west coast of Nova Scotia outside the headlands, except for a brief period in the spring when the cold drift from the east passing Cape Sable mav temporarily chill the surface there to 32° (p. 311). At the coldest time of the year, the water in the Gulf not only is nearly uniform in temperature vertically to a depth of 100 meters or so, but the underlying strata at equal depths (like the surface) out over the deep east-central basin and over Georges Bank offshore are some 3° to 6° warmer than the water closer in around the periphery of the Gulf — a contrast illustrated by the 40-meter and 100-meter charts for 1920 (Bigelow 1927, figs. 12 and 13). Consequently, the temperature of the bottom water is progressively higher passing off- shore along any line normal to the general trend of the coast as the depth increases, reversing the situation characteristic of summer and autumn. Thus, any bottom-dwelling animals capable of active motion need only move a short distance down the slope into deeper water to escape the rigors of winter; and many do just this. While any animal at tide line may be in water soupy with ice crystals and at freezing point, as at Sandy Neck, Barnstable, on February 7, 1901, the minimum temperatures on bottom farther out characteristic of winters as severe as those of 1920 and 1925 were about 32° F. in Cape Cod Bay and 32° to 33° northward and eastward from Boston along the 20-meter zone; about 33° to 36° along the 40- to 50-meter zone; and about 35.4° to 39.7° along the 100-meter zone all around the periphery of the Gulf. The only important exception is the sink off Cape Ann, where interchange with the warmer water of the open basin is barred by the surrounding sill, as evidenced by a 100-meter reading there of 34.7° on March 1, 1920, con- trasted with 37.5° at this same depth on the slope offshore on February 23, and 38.6° a few miles to the southward on March 24. At 150 meters, the usual winter temperature was between 35° and 39° in the Cape Ann sink and between 42° and 43° in the open basin of the Gulf. Readings of 36.6° to 37° F. at 70 to 90 meters on the eastern part of Georges Bank, March 11-13, 1920, show that the bottom water chills to about as low a value there as at equal depths around the basin to the nortiiward. But it is doubtful whether bottom temperatures lower than about 40° ever spread to the southwestern part of the Bank, for the Albatross recorded a bottom reading of 46.5° there at 70 meters on February 22, 1920. During the milder winters included in the period 1913-25, the surface probably did not chill below 35° anywhere along the open coast to the north of Boston, unless close in to tide line, to judge from water temperatures of 35° to 39° F., March 4-5, 1921, at stations scattered between the offings of Seguin Island and of Boston; and perhaps the temperature did not fall below 32° to 33° even in Cape Cod Bay. Readings of 35.1° to 35.4° from the 40-meter level downward in the 180-meter sink 15 miles off Gloucester on March 1, 1920, when contrasted with readings of 38.8° to 39° taken there on March 5, 1921, illustrate the temperature differ- ence that is to be expected at comparable localities from winter to winter in the western side of the Gulf. Yearly maximum. — The surface water is at its warmest sometime in August generally throughout the southern and western parts of the Gulf of Maine (as early as July in some years, particularly in enclosed situations, such as Boston Harbor), but not until sometime in September, or even early October in the northeastern part and in the Bay of Fundy region, where vertical mixing by tidal currents is more active. The regional differences in temperature, also, are much wider in the warm season than in the cold, often within short dis- tances, for reasons the discussion of which would lead us too far afield.' Because of regional differences in the rate of vertical mixing of water by tidal currents, the sur- face warms much more rapidly in the southwestern part of the Gulf, where the cold of the preceding winter is preserved in the underlying strata into autumn, than in the northeastern part, where the heat taken in at the surface is distributed more evenly downward as the season advances. As a consequence of this widespread contrast, combined with local differences caused by smaller-scale mix- ings and upwellings, the surface water temperature is some 3.5° to 7° higher in enclosed situations along the western coast of the Gulf, over an iso- lated pool in Cape Cod Bay, and over the western I Sci an earlier paper in the Geographical Review (Bigelow 1928), for a general summary and a more extended account of the physical oceanography of the Oulf based on the cruises of 1912-26 (Bigelow 19271. CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 305 sido of the gciipral basin of the Gulf tliaii it is dose in along the western coastline; 14° to 15° higher tlian in the northeastern part of the Gulf in gen- eral; and 7° to 1 1° higher than over the shoaler and most tideswept parts of Georges Bank and of Nan- tucket Shoals. In the western part of the Gulf the 20- to 40-meter level is not at its warmest until well into September, the 50- to 70-meter level not until November, the 100-meter level not until late Novertiber or early December, and the 150-meter level not until December or even later. On the other hand, in the northeastern part of the Gulf the entire column of water warms at an almost uniform rate, down to at least 100 meters. In August, when the surface is at its warmest in the western part of the Gulf, the 40-meter level is 5° to 6° colder and the 100-meter level about 4° colder there than in the northeastern part of the Gulf (Bigelow 1927, figs. 53 and 56), reversing the temperature relationship at the surface that is characteristic of that time of year. Since the surface begins to cool in early September in the western part of the Gulf, but not until a month or more later in the eastern, the regional differences in surface temperature decrease as the season advances, until by midautumn the surface is nearly uniform in temperature throughout the Gulf (varying only a degree or two from place to place); and so it continues until the following spring. At the warmest season during the summers cov- ered by the cruises of 1912 to 1926, surface temper- atures ranged from about 64° to 68° F. (occasion- ally 1° to 2° higher) in Cape Cod Bay and over the western part of the basin; from 61° to 64° close in along the western coastline; from 50° to 53.5° in the northeastern part of tlie open Gulf in general and in the lower part of the Bay of Fundy; from 53° to 57° over the parts of Georges Bank where tides run the strongest; and from 57° to 61° over the shallower parts of Nantucket Shoals. The seasonal maxima at deeper levels were as follows in the western part of the Gulf in general and in the lower Bay of Fundy, which cover the regional extremes (table 3). Depths greater than 150 meters In the 180-meter sink oft' Cape Ann, tlie sill depth of which is about 75 meters, the alternate processes of vernal warming and of autumnal 414171 0—57. ;! Table 3. — Seasonal temperature maxima in the western Gulf of Maine, 1912-26, and in lower Bay of Fundy, 1916-17 [Temperatures In °F.] I'epth Western Oulf of Maine Lower Bay of Fundy ' (meters) Month Tem|)era- ture range Month Tempera- ture range 20-40 Sept 50-55.4 46. 4-48. 2 42. 8-46. 4 42. 8-45. .5 Oct 47-50 .W-'O Oct.-N'ov Nov.-Pec - Dec Oct 45. 3^7. 8 100 Nov.' . . .. 44. 8-45. 3 150 . Nov.-Dec.' 44-44.8 1 Scaled from Mavor (1923, p. 375, table 8). 2 Xo November data for 1917. cooling affect the entire column of water from surface to bottom, though in lessening degree at increasing depths. During the winter of 1920-21, for example, the temperature at 150 meters fell from 44.6° F. on December 29, which could not have been much below the maximum for the year, to 39° on March 5, which certainly was close to the minimum for that winter. In the same year, almost as great a seasonal range (4.7°) was regis- tered at 150 meters, between December 30 (44.5°) and March 5 (39.8°) in the 180- to 190-meter trench between Jeffreys Ledge and the coast, where free interchange over the bottom is hindered by an enclosing sill rising to within 120 to 125 meters of the surface. The seasonal difference was about twice as great as this (8.8°) between December 2, 1916 (43.9°) and April 9, 1917 (35°) at 175 meters in the 180- to 208-meter bowl at the mouth of the Bay of Fundy, where the depth of the sill is about 128 to 140 meters (Mavor 1923, p. 375, table 8). The picture is not so clear for the deep bottom water of the open basin of the Gulf. The changes in temperature that take place there from time to time at depths greater than 150 meters are the combined result of such slight influence as the climatic cycle may exert from above modified or intensified, according to the season, by indrafts of water from the continental slope that enter via the deep channel between Georges and Browns Banks. (For a discussion of this general matter, see Bigelow 1927, p. 690.) High salinities, for example, made it clear that the higher mean temperatures recorded in late summer and autumn than were recorded in late winter or spring near the bottom in the eastern part of the basin during our caily cruises (table 4) were the direct result of recent indrafts of this water from offshore. Reatlings of 47.1° F. at 200 meters and of 46.3° at 306 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 250 meters at a station in the southeastern part of the basin in July 1914 and of 46.7° and 47° at 190 meters in the northeastern part in August show that the temperature of the slope water moving inward at some intermediate level via the channel between Georges and Browns Banks is often higher than the reading recorded near the bottom. Readings near the bottom have ranged from 42.9° to 45° (average, 44.2°) at 4 stations in March and April 1920, July 1915, and July 1914. Table 4. — Mean temperatures recorded in Gulf of Maine at depths of 160 to 330 meters, by season (Recorded in 1912-16, 1919-21, 1923, and 1926. Temperatures in °F.] Month Feb.-Mar. Apr.-May June-July- Aug.-Sept Nov .-Jan. Western part of basin Mean tempera- ture 41.2 41.8 42.0 43.1 41.4 Number of readings Eastern part of basin Mean tempera- ture 41.0 41.4 45.9 43.7 44.5 Number of readings R 7 11 1 Since it is the actual temperature with which we are primarily concerned in the present discussion, it is sufficient to add that the 71 readings taken at 160 to 330 meters in various parts of the basin, summarized in the preceding table, ranged be- tween 38.4° and 47° F., with 54 of the 64 readings falling between about 40° and 44°. There is nothing in the record to suggest that the tempera- ture of the deep bottom water within the basin of our Gulf has fallen appreciably below these limits, or risen above them, in any year since Gulf tem- peratures have been recorded. TREND IN WATER TEMPERATURES BETWEEN 1912-26 AND 1953-54 For the period 1926 to 1953, our most instruc- tive sources of information as to ups and downs of temperature around the shoreline of the Gulf of Maine are the mean montlily temperatures that have been reported by the United States Coast and Geodetic Survey for Boston since 1922 and for Eastport since 1930 (U. S. Coast and Geodetic Survey, 1951; with subsequent data contributed in advance of publication); the mean monthly temperatures for St. Andrews, Now Brunswick, 1921 to 1947, tabulated by Hachey and McLcllan (1948, p. 357); and the mean annual temperatures for that same port, 1921 to 1953, published by Lauzier (1954, p. 8, fig. 1). Lauzier (1952, p. 6), has already pointed out that "St. Andrews water temperatures reflect general water conditions over a large section of the Atlantic coast" of Canada. On geographic grounds, a shift in either direction at Boston that persists through several years may be expected to prove an equally reliable index to any upward or downward shifts that may have taken place in the western part of the Gulf in general. At Boston, the mean temperature for the coldest month of the year did not show any long-term trend, either upward or downward, from 1922 through 1936. Wiiile it was higher for the winters of 1029 to 1933 (33°+) than for those of 1922 to 1926 (below 32°), it was again lower than 32° in 1934, 1935, and 1936, with means as low as 29.8° for 1934 and 1936. But the mean for the coldest month, whicli had averaged 31.7° for the 8-year period 1922-29, averaged 33.1° for the period 1930 to 1937, 32.6° for 1938 to 1945, 33.8° for 1946 to 1948, 36.7° for 1949 to 1951, and 36.5° for 1952 to 1954. Furthermore, there has not been a winter since 1944 when the mean water temperature for the coldest month has fallen below 32° in Boston Harbor, though this happened in 11 of the preceding 21 years. Summer air temperatures, also, liave clearly tended upward in Boston Harbor during the past 30 years (1922-53), whetlier expressed as the departure from the mean for July-September or as the mean for tlie warmest month for successive 4-year periods (table 5). Indeed tliere has not been a summer during the 13-year period 1941-53 when the mean temperature for the warmest month was not higher than it was in 8 of the 18 years from 1922 to 1940; while there has been only one summer (1948) since 1941 when the mean temperature was not at least as high as 66.6°, a level equalled only 6 times during the 19-year period 1922-40. Similarly, the mean tempera- ture for the 2 warmest months combined, which was not above 65.3° for any 2 consecutive years between 1922 and 1941 (average, 64.1°) has been 66.5° or higher (mean, 67.1°) in every subsequent year, with the sole exception of 1948. The mean water temperature for the year as a whole also has tended upward at Boston and by about the same amount (table 6). And while mean annual temperature is of little ecological significance in regions wiiere water temperatures CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 307 differ widely from season to season, it does have tlie advantage, as an index to long-term fluctua- tions, of smoothing out the short-period ups and downs that loom large when individual months are compared for different years. Table 5. — Mean air temperatures for warmest month of year at Boston to nearest degree, by periods, 1922-53 (Temperatures in °F.] Period Mean air temper- ature Period Mean air temper- ature 1922-25 54 65 65 66 1938-41 65 1926-29 1942-45 67 1930-33 1934-37 194649. 1950-53 68 67 Table 6. — Mean water temperatures in Boston Harbor to nearest degree, by periods, 1922-53 (Temperatures in °F.| Period Mean water temper- ature Period Mean water temper- ature 1922-25 48 49 50 49 1938-41 49 1926-29 1942-45 - 194649.. 1950-53 1 50 1930-33 1934-37 62 52 I Mean for period 1942-45 based on 3 years only, lacking data for 1944. Water temperatures, similarly, have trended upward unmistakably since a cold period in the early 1940's at Eastport, a short distance within Passamaquoddy Bay (tributary to the lower Bay of Fundy), both for the warmest season and for the coldest (table 7), by about the same amount as at Boston (3°-4°). A mean of 51.9° (50.1°- 53.8°) for the 2 warmest months combined, for the period 1930-41, contrasts with a mean of 53.7° (52.1°-54.6°) for 1949-53. We also read in the annual report of the Fisheries Research Board of Canada for 1951 (p. 34) that Table 1 .■ — Mean water temperatures for coldest and warmest months at Eastport, Maine, by 2-year periods, 1930-53 (Temperatures in "F.] Mean temperature Mean temperature ii — in- Period Period Warmest Coldest Warmest Coldest month month month month 1930-31 63.4 34.0 1942-43 50.9 32.4 1932-33 52.5 35.6 1944-45 52.6 34.4 1934-35 51.8 31.2 1946-47 53.7 34.5 1936-37 52.6 35.2 1948-49 62.9 35. 1 1938-39 51.9 33.9 1950-51 53.3 37.4 1940-41 50.3 33.3 1952-53 53.3 37.1 surface water temperatures at St. Andrews, farther up the Passamaquoddy Bay "have shown a fairly definite upward trend for the past ten years, with temperatures in 1951 the highest in thirty years' records." Lauzier (1954, p. 7) also has recently remarked "that the waters of the (St. Andrews) area are undergoing a general warming since 1940." The ups and downs from year to year have not been great enough to mask this warming on Lauzier's graph of mean annual temperatures (p. 8, fig. 1), although the difference from one year to the next has been about as wide in extreme cases (up to about 3.5°) as the general upward trend has been from the early 1940's to the early 1950's. That this warming trend has extended eastward and northward to the waters along the outer coast of Nova Scotia and to the Magdalen shallows in the southern side of the Gulf of St. Lawrence is apparent from Lauzier's graph of mean yearly temperatures at Sambro Lightship, off Halifax, Nova Scotia, and from his tables of quarterly averages of surface tem- peratures there and at the Magdalen Islands, during the periods 1940-44 and 1949-53 (Lauzier 1952, p. 6, fig. 1; 1954, pp. 9 and 10, tables II and III). COMPARISON OF TEMPERATURES FOR 1953-54 AND 1912-26 March-April Since a colder period during the early 1940's, water temperatures have so clearly averaged some 2° to 4° higher at Boston Harbor, at Eastport, Maine, and at St. Andrews, N. B., than they had previous to about 1940-42 (p. 306) as to make the assumption reasonable that a corresponding tem- perature change had taken place offshore. Com- parison of the bathythermograph records taken by the Albatross III during March 1953 with the temperatures recorded during our late winter and early spring cruises of 1913 to 1925 (p. 302) corrob- orate this assumption, at least for the upper 150 meters. The water may not have been significantly warmer at the mouth of Massachusetts Bay and out in the basin in the early spring of 1953 than at tluit season in the warmer of the years included within the earlier period (table 8). The difference between readings — 1.5° to 3° higher at neighboring localities on March 30-31, 1953, than on March 5, 1921 — may be no greater than can be charged 308 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 71' 44' 43< 42' 41 40" ■ CAPE. A PORTLAND^ 44' 43' 42' 4 I 40 71' 70' 69' 68' 67' 66' FictRE 12. — Areas where comparable records of water temperatures were taken in March and May 1953 (dots) and in March of the earlier years of 1913, 1920, and 1921 (crosses). CLIMATE .•\ND THE DISTRIBUTION OF MARINE ANIMALS 309 against the warming to be expected during a 3- week period at tliat season. But tlie temperature off Gloueester was some 2° warmer from surface to bottom on March 31, 1953, than on April 3, 1913 — a spring that was more fairly representative of temjx'rature conditions prevailing 30 to 40 years ago than was 1921 — and 3° to 5° higher than on either March 1 or April 9 of 1920, a relationship the reverse of what was to be expected on seasonal grounds, other things being equal. T.\BLE 8. — Water temperatures at various depths at the mouth (ff Massachusetts Bay [Area A, flg. 12 . Temperature in "F.) Depth 1913 1920 1921 1953 (meters) Mar. 4 Apr. 3 Mar. 1 Apr. 9 Mar. 5 Mar. 31 0 37.2 37.2 37.4 39.3 39.3 39.3 36.5 36.5 35.4 34.7 34.8 38.0 36.5 38.6 38.8 38 9 38.5 38.5 38.6 38.8 38 9 41.4 20 41.2 40 41.4 100 41.0 120 41. 1 Contrasts of the same order between 1953 and 1920 are demonstrated in table 9 for the stations shown in figure 13. In some instances, the dif- ference in the dates would point to a relationship of the opposite order, other things being equal. In summary, observations taken at comparable dates and localities in March and April of 1920 and of 1953 show that the water in the open basin of the Gulf, outside the 100-meter line north of Georges Bank and in the channel between Georges and Browns Banks averaged some 2.7° to 4.4° warmer, in general, in March 1953 than in April 1920, at all depths from the surface down to 150 meters. The comparison between early spring tempera- tures for 1953 with those for 1920 cannot be ex- tended to Browns Bank (area L, fig. 12), for the Bank was not visited duiing the March cruise of 1953. Mean temperatures, it is true, were be- tween 2° and 4° higher there, from surface to bot- Table 9. — Comparison of temperatures in the Gulf of Maine in 1920 and l,9t)3, by areas |See flg. 12 (or location of areas. Temperature in °F.J Temperature in depths of— 0 meters 20 meters 40 meters 80 meters 100 meters 135 meters 150 meters Mouth of Massachusetts Bay (area A): April 9. 1920 38.0 41.4 37.5 41.0 36.5 41.1 36.5 40.7 37.5 40.9 38.5 38. 7-41. 6 38.5 40.3-41.0 39. 0-39. 5 41. 8-42. 3 .39.5 38.6 33.3-33.3 38.2-42.2 36.5 41.2 36.4 41.0 36.0 40.5 36.0 40.5 36.2 41.0 38.1 39.0-41.3 38.1 39.9-41.5 38. 4-38. 5 42. 0-42. 5 38.5 39.5 37.2-37.6 3S. 5-42.0 38.6 41.4 36.1 41.0 36.0 40.5 136.5 40.5 36.6 41.0 38.4 41. 0-42. 5 37.9 40. 5-41. 5 37. 0-38. 2 42. 2^2. 5 38.4 41.5 37. 7-37. 8 39.5-42.0 38.8 41.0 38.9 41.1 38.5 42.5 38.5 41.5 38.5 41.0 38.4 43.2 39.7 43. 5-44. 0 37.7 40. 9-42. 5 March 31. 1953 Trough west of Jeffreys Ledge (area B): April 9, 1920 39. S 44.0 March 22 1953 Oft northern Cape Cod (area C): March ''4 1^53 Western side of basin, off Cape Ann (area D): March 24 1920 ■40.2 March 23 1953 44.0 Off Mount Desert Island (area E): April 11 1920 40.8 44.5 East-central part of basin (area F): March 23 1920 40.8 » 44. 5-1.1. 5 Southeastern part of basin (area O): 40.0 M6.5 Eastern part of Ocorpt-s Bank (area H): 38. 5-38. 8 ' 42. 2-42. 4 March 29-April 1 1953' Range 4 stations 4:f. 7 47.0 Southeastern slore of Georges Bank (area J): April Ifi. 1920 39.6 44.0 42. 6-43. 7 41.S-44.5 March 27. 1953 . Channel between Georges and Browns Banks (area K): March 13 and April 16, 1920: Range 2 stations . --. ».37.8 ■42.0 42. cV-(4. 3 March 23-30, 1953: Range 4 stations __ .-... 46. 5-47. 5 North of (U-oru'i-s Bank from the 100 meter line, and between Oeorges ancl Bnnvns Banks: March 11-April Hi, 1920: 33.3-39.7 34.9-38.6 42.6-43.7 R;iii(;<' 11 "Stations 42.6-44.3 March l!t-April 1. 1953: 38.1-42.6 39.0-42.2 40.5-46.0 Rancf lit stations 1 44..V47.5 Browns Banli (area L): .\pril Ifi. I920' Mean 4 stations 38.5 39.4-42.0 38.3 39. 8-42. 0 38.0 41.5-42.1 May 3-.'i. 19.W: RanRc 4 .stations I Scaled from vertical graph for this station. ' 2 str.tions only. ' 1 station only. 310 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 7 44° H 43' 42' 41 40 PORTLAND^ CAPE. ANN \ ^t.. _j\00 ^ y \ I \ I ^. / ef of Gf^ e^^' • • •• X /' I 44' 43' 42 ' ^_'- 4 I 40 71* 70^ 69 68' 67' 66' Figure 13. — Stations where the temperature was recorded on March I'J-April 1, 1953 (dots) and on March 1 1-April 16, 1920 (crosses). CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 311 torn, on May 3-5, 1953, than on April 16, 1920 (table 9). But the difl'erence between the two sets of observations is seemingly no wider than can be credited to the warming to be expected to take place over the Bank, in any given spring, during the 3-week interval between the dates when the observations were made in the 2 3'ears. The dif- ference, also, between slightly higher readings in the deep channel north of Georges Bank (area M, fig. 12) on March 23, 1953, than on the 20th of that month in 1920 (table 9), is no greater than is to be expected there from one spring to another during any run of years, depending on the relative sever- ity of the preceding winter, and on the seasonal schedule and temperature of the Nova Scotian drift from the east. But bottom temperatures reported from Browns Bank in 1952, of about 41° to 43° F. between February 15 and March 18 and of about 39° to 46° between April 29 and May 11 (McLellan 1954, p. 408, figs. 2 and 3), contrasted with bottom readings there in 1920 of 38.1° on March 13 and of 37.8° on April 16, point to slightly warmer bottom water on Browns Bank in early spring in 1952 than in 1920. The surface, how- ever, was about as cool in late March of 1953 as in 1920 over the east-central part of the basin, and about 1° colder than in 1920 over the south- eastern slope of Georges Bank. The existence of an area occupied by water as cold as 36° to 40° F., as outlined in the surface- temperature charts for the eastern part of the Gulf of Maine (fig. 14), shows that the responsible factor was the cold Nova Scotian drift, which had spread westward past Cape Sable before the last week of March in 1953 but did not do so until April in 1920, or, seemingly, in 1915 (Bigelow 1927: p. 554, fig. 25; p. 578). In the seasonal schedule of the drift, 1953 paralleled 1919, when a U. S. Coast Guard cutter on ice patrol reported a surface water temperature of 32° (but 38.7° to 41° in the underlying water) on MaiTJi 29, 1919, in the east-central part of the basin of the Gulf (Bigelow 1927, p. 553) where the temperature was 38° to 40° at about the same date (March 23) in 1953. The contrast in this connection between 1953 (a warm spring in the sea), on tlie one hand, and 1919 (a cold spring) and 1920 (a cold and tardy spring), on the other, is good evidence that the date on which the cold Nova Scotian drift affects the temperature of the surface beyond Cape Sable and the extent to which it does so are not correlated with the minimum to which water has chilled during the preceding winter in the Gulf of Maine, or with the relative forwardness of vernal warming there, but that they are governed by events along the Nova Scotian shelf to the east- ward and within the Gulf of St. Lawrence. In 1953, surface water slightly colder than 40° still occupied the area off Cape Sable (fig. 15) as late as the first week in May. The westward drift must have ceased soon thereafter, however, if it had not done so already, for tiie temperature of the surface water on Browns Bank had risen to 46.5°-48.5° by the last of May. It appears, thus, that any tempering effect mild air tempera- tures over tlie land mass in winter may have on water temperatures off southern Nova Scotia early in the following spring is counteracted more or less completel3' by the chilling effect of the drift from the east. August-September It is clear, from the evidence we have presented, that the waters of the Gulf from the surface down to 150 meters entered the season of vernal warming between 1° to 2° and 4° to 4.5° warmer in 1953 than in the period 1913-25, though perhaps no warmer than in the warmest of the years included within that period. The data also show that the temperatures of the Gulf continued about that much warmer during the summer of 1953, though with certain regional exceptions. In Boston Harbor, the mean temperature of the warmest month of 1953 (August, 67.2° F.) was about 2.6° higher than the corresponding mean (64.6°) for the 5-year period 1922-26. Similarly, Frank J. Mather III reported surface readings of 70.5° to 71.5° in the eastern side of Cape Cod Bay, with 70° in the central part, as early in 1953 as July 19. This contrasts with 64.6° recorded by the Halcyon for tlie center of tlie Bay on August 24, 1922, when the surface there was no doubt near its warmest for the year. In 1953, as late as the first week of September, the surface was still as warm as 68° to 70° over the southwestern part of the open basin of the Gulf, where 62.8° had been reported on August 4, 1913, 67.2° to 68.0° August 2.3-24, 1914, and 66.9° on August 12. 1926. Similarly the surface readings in tiie general oflling of Cape Ann (area A, fig. 16) averaged about 2.7° liigher at stations in August 1953 than at 15 stations in that same montii in 1912, 1913, 1914, 312 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 44' 43' 42' 41° 40' 44' 43' - 42' 40' 71' 70' 69' 68' 67' 66 FlGi'RE 14. — Surface temperatures recorded in the Gulf of Maine, March 19 to April 1, 1953. Dots indicate location of observations. (Temperatures in degrees Fahrenheit.) CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 313 71' 44' 43' 42" ■ 41 40' PORTLAND, CAPE. ANN • • ^ • •• 43-45 • • . ' 4^ — 45 I \ I ^ / • • V • • • • •^-^ 4 5°+ / -s_,. ,-'4' X z. 44' 40' 71' 70' 69' 68' 67' 66' Figure 15.— Surface-water temperatures in the eastern part of the Gulf of Maine, April 24-May 8, 1<»53. Dots indicate location of observations. (Temperatures in degrees Fahrenheit.) 414171 0—57^ 4 314 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 44' 43° 42' 41 40 7 I — r PORTLAND^ ® CAPE. ANN 44' - 43" ■■'U.-' I ^_-. 42 40 71" 70' 69' 68' 67' 66' Figure 16. — Areas where comparable observations were taken in August and early September 1953 (crosses), and in late summer of earlier years (dots). CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 315 1915, 1923, and 1926. While the maximum temperature recorded in this particular area was as high (68°) for 1915 as for 1953, the average for the summer of 1953 not only was higher than any individual August reading for the earlier years, with three exceptions, but was about as much higher as comparable data for Boston Harbor. Below-surface readings in the general offing of Cape Ann in August show that, while those for 1953 ^ overlap somewhat those for 1912 to 1926 both in the upper 40 to 50 meters and at the greatest depths sampled, the means for successive strata were consistently higher for 1953, with an overall difference averaging about 3.5° for the water column as a whole (table 10). Table 10. — August temperatures off Cape Ann, in 1963 and in earlier years, to nearest degree [Area A, flg. 16. Temperature In °F.) Deptb (meters) 0 15-20-.. 40-SO... 70-100. 110-130 150-170 180-274 Number of Temperature Mean In— stations In— range In— 1912-26 1953 1912-261 1953 1912-26 1953 15 8 60-68 65-68 63 66 15 7 49-58 61-60 52 55 15 7 41-46 46-47 44 47 15 7 37-43 44-46 41 45 15 5 38-41 44-45 39 44 8 1 37-42 44-45 40 44 5 ' 41-45 43-45 43 44 Mean depar- ture In 1953 -1-3 -1-3 +3 -1-4 -1-5 -1-4 + 1 ' Partly from direct readings; partly as scaled from graphs for the individual stations. It is especially instructive that the water in the western side of the Gulf was between 1.5° and 5.7° warmer in 1953 than in 1926 (table 11), as the mean water temperature for the warmest month was almost precisely the same (67.2° and 67.3° F.) in both summers in Boston Harbor. Evidently the temperature at Boston for any one summer is not a reliable index to conditions in the open Gulf at that time of year, whether for the surface or for the underlying watei-s. Consecjuently, mean water temperatures for the warmest month as high as 67.3° to 68.9° in Boston Harbor in 1926, 1928, 1941, or 1944 do not necessarily mean that the open Gulf was as warm in any one of these sum- mers as it was in 1953, following 4 years when the mean for the warmest month at Boston rose to 67.2° to 69.2° and the mean for the coldest month did not fall below 35.9° to 37.8°. On the other hand, it seems that the residual effects > Readings for 1912 to 1926 Uken tiy reversing thermometer, those tor 1953 by bathythermograph. of a winter colder than usual may be expected to persist through the summer in the western side of the Gulf in the underlying strata of water. Such, at least, was the case in 1923, when the mean temperature of the water in Boston Harbor for the coldest month was 30.2°, following which the temperature at 80 to 155 meters in the par- tially enclosed sink off Cape Ann was 37.4° to 37.6° as late in the season as August 9, which contrasts with 40.1° to 42.8° there between the 9th and 31st of that month in 1913, 1914, 1915, and 1916 (4 stations). Table 11. — Water temperatures off Cape Ann in August 1926 and in August 1953 [Ar«a A, flg. 16. Temperature in °F.] Aug. 11, 1926 Aug. 19-20, 1953 Mean Depth (meters) Tempera- ture range Mean departure In 1953 0 . 64.4 48.7 42.4 40.6 40.5 64. 6-68. 0 61. 5-58. 0 47. 2-49. 0 44. 0-45. 5 44.0 65.9 54.4 47.9 44.5 44.0 -1-1. 5 20 40 _ -1-5 7 -t-5 5 100 -1-3.9 180 -1-3.5 From the available information it appears likely that the contrast between August temperatures in 1953 and in the earlier summers of record, for the water colunm as a whole, was in general about as we have outlined around the western and northern slopes of the Gulf. In the southwestern part of the basin, for example, near the 100-meter line (area B, fig. 16; table 12), mean temperatures at 0, 40, 100, and 150 meters were 2.2° to 4.2° higher at 3 stations on September 6, 1953, than on August 23, 1914. The summer of 1953 was relatively warm, as is illustrated by the temperatures of the water, surface to bottom, along the eastern part of the coast of Maine and in the vicinity of Mount Desert Island (area C, fig. 16; table 13). Here the temperature was 1.5° to 3° higher on August 16, 1953, from the surface to near the bottom (37-40 Table 12. — Water temperatures in southwestern part of basin in August 191/, and in September 1953 [Area B, flg. 16. Temperature In °F.] Depth (meters) Aug. 23, 1914 Sept. 6 1953 Range i Mean 0 67.2 43.8 39.6 41.7 69. 0-69. 9 46. 5-49. 5 43. 5-44. 0 43.7 69.8 40 47.8 100 4a8 150 . . 43.7 ' 3 stations. 316 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE meters) than on August 21, 1912, or on August 13, 1913, and between 3° and 8° higher than on August 5, 1923, a summer following a very cold winter. Temperature conditions were similar in the lower part of the Bay of Fundy, where readings of 12.5° C. (54.5° F.) at 25 meters and 12.3° C. (54.2° F.) at 75 meters for September 1953 contrast with September averages of 10.7° C. (51.2° F.) and 9.7° C. (49.4° F.) at these same depths for the period 1921-42.3 Table 13. — Water temperatures near Mount Desert Island in August 1953 and in earlier years [Area C, flg. 16. Temperature in °F.] Depth (meters) Aug. 21, 1912 Aug. 13, 1913 Aug. 5, 1923 Aug. 16, 1953 0 55 50.7 49.8 55 51.0 48.8 53.1 46.7 45 56 20 - 37-40...- 56.5 52.6 In the central and eastern parts of the Gulf, however, it was only in the underlying water layers that August-September temperatures were consistently higher in 1953 than in the earher years of record, and even so, they varied consider- ably according to locality. Thus the surface in the central part of the basin (area D, fig. 16; table 14) averaged about 1° colder (3 stations) on September 6, 1953, than on August 23,' 1914; whereas the underlying strata, to the greatest depth sampled, were 2.3° to 5.6° warmer in 1953 than in 1914. The situation was similar to the northward in the neighborhood of Mount Desert Rock (area E, fig. 16; table 15), where the readings at 40-150 meters were between 1° and 4.8° warmer on the average on August 16, 1953, than in any previous August of record, but where a Table 14.^ — Temperatures in central part of basin in August 1914 and in September 1953 lArea D, flg. 16. Temperature in "F.J Depth (meters) Aug. 23, 1914 Sept. 6, 1953 Mean depar- Tempera- ture range ' Mean, 1953 ture in 1963 0 66 5 46.1 39.1 41.2 43.2 64.6-66.6 51 -53 44.5-46 44. 2-47. 0 44. 5^6. 5 65.6 51.7 45.0 45.5 45.6 1 0 40 +5.6 +6.9 +4.3 +2.3 100 150 176 > 3 stations. ' Information supplied by letter by Dr. A. H. Lelm, Director, Atlantic Biological Station, St. Andrews, N. B., (or Prince station 6, midway between Campobello Island and The Wolves. recorded range of about 2.5° from year to year in the surface temperature was perhaps no wider than might be expected from one day to the next, from one stage of the tide to another, or within a short distance, in this region of strong tidal currents. Table 15. — Water temperatures off Mount Desert Rock in August 1953 and in earlier years ;Area E, flg. 16. Temperature In ' F.] Depth (meters) Aug. 13, 1913 1 Aug. 13, 1914" Aug. 6, 1923' Aug. 20, 1926' Mean 1913-26 Aug. 16, 1963 Departure in 1953 0 20 40 100 150 175 55.0 49.7 47.5 45.6 44.3 43.6 55.9 50.5 47 2 44.9 42.9 44.6 55.0 49.6 45.5 40.0 41.2 57.5 63.9 48.6 42 1 42.9 55.9 50.9 47.2 43.2 42.8 44.1 66 52 50.5 48.0 46.6 46.5 +0.1 to -1.5 +2.5 to -1.9 +1.9 to +5 +2.4 to +8 +2.2 to +5.3 ' Readings scaled from vertical graph for this station. ' Readings scaled in part from vertical graph for this station. In the eastern side of the basin, near the Nova Scotian slope (area F, fig. 16; table 16), the read- ings for 1953 were 1.4° to 4.0° higher than the combined means for 1912, 1913, 1914, 1915, and 1926, for the water column as a whole. But the surface was slightly the warmest in 1914, the 25-40 meter stratum of nearly the same temper- ature in 1953 as in 1913, and the bottom water at 220 meters of nearly the same temperature in 1953 as in 1912. It is doubtful, also, whether the differences between the readings for August 18, 1953, and those for July 23, 1914, and August 9, 1926, are indicative of any general periodic trend for the southeastern part of the basin, near the source of the indrafts of slope water (area G, fig. 16; table 17). While the 100- to 150-meter stratum was warmer in 1953 than in 1926 and the 20- to 40-meter stratum was 3.6° to 4.1° colder in that summer than in 1926, the bottom stratum was some 2° to 3° warmer in 1914 than in either 1926 or 1953. Table 16. — Water temperatures in eastern side of basin in August 1953 and in earlier years [Area F, fig. 16. Temperature n-F.l Depth (meters) Aug. 14, 1912 Aug. 12, 1913 Aug. 13, 1914 Aug. 20, 1928 Mean 1912-26 Aug. 17, 1953 Mean depar- ture in 1953 0 59 47.7 45.3 45.3 1 45.3 45.3 60.6 51.1 42.0 42.0 '42.6 42 6 63.5 43.5 41.5 42.9 1 42.6 42.6 61.9 43.9 40.1 41.2 42.4 61.2 46 6 42.2 42.9 43.2 43 8 62.6 50.0 46.2 46.2 45.2 45.2 + 1.4 25-40- 90-100 - +3.4 +4.0 145-150 - +2.3 200. 220 +2.0 ' Scaled from vertical graph for this station. CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 317 Table 17. — Water temperatures in southeastern part of basin in 195S and in earlier years (Area 0, flg. 16. Temperature In "F.) Depth (metere) July 23, 1914 Aug. 19, 1926 Aug. 18, 1953 o._ 59.5 '53 50.0 49.2 48.8 47.1 46.3 61.9 62.1 55.6 42.3 1 43.1 43.9 ■43.7 61.7 20 58 0 40 52.0 100 . _ 45 0 150 46 5 200. 44.0 250 >43 8 < Scaled from vertical graph (or this station. J Deepest reading was at 241 meters. Any continued alteration in temperatures for the summer-autumn period on Georges Bank is of special interest, because of the productivity of the Bank as a fishing ground. Unfortunately, the records are not clear in this respect, because the readings for 1914 and 1926 were taken so much earlier in the season than were those for 1953 that somewhat lower values at the surface and higher temperatures below the surface were to be ex- pected in 1953 than in the earlier summers, quite apart from any year-to-year fluctuation. This seasonal progression is illustrated by the tempera- tures taken on the eastern end of the Bank (area H, fig. 16; table 18) July 23, 1914, August 13, 1926, and September 13, 1953. But the mean temperature for the water column as a whole — nearly 8° lower in mid-August of 1926 than in mid-September of 1953 — at least suggests that the summer of 1953 not only was warmer on this part of the Bank, but that it was as much warmer as were the western and northern parts of the Gulf. Table 18. — Water temperatures on eastern part of Georges Bank in September 195S and in earlier years [Area H, fig. 16. Temperature in "F.) Depth (meters) July 23, 1914 Aug. 13, 1926 Sept. 13, 1953 0 52 '51 '51.4 '51.4 64 51.7 46.0 •44.7 43.3 20.. 60 40 60 54 5 70 Approximate mean 51.6 49.3 57 0 ' Scaled from graph for this station: readings at 30 meters and at 55 meters. ^Scaled from graph for this station; readings at 40 meters and at 70 meters. On the western part of Georges Bank, move- ments of the waters are so complex that wide differences in temperature within short distances, even on the same day, are not unusual. Readings at a pair of stations about 10 miles apart on the northwestern end of the Bank (area J, fig. 16; table 19) September 5, 1953, and at a second pair about 17 miles apart, on the southwestern end (area K), September 3^, afford a striking example of this regional irregularity'. Under such cir- cumstances the danger is obvious of mistaking regional or short-term variations for 3'ear-to-year differences, when temperatures for diflFerent sum- mers are compared. The present case is further complicated by the seasonal rise in temperature to be expected during the 5- to 6-week period between the dates of observation for 1914 (July 20), for 1916 (July 23), and for 1953 (September 3-5). Perhaps the most that can be said is that mean temperatures, which were some 6° higher in the upper 20 meters and some 4° higher at 40 to 50 meters in early September of 1953 than they were in late July of 1914 and of 1916 (table 20) on the western part of the Bank (areas J and K combined, fig. 16), are at least compatible with the somewhat higher temperatures that prevailed in the inner parts of the Gulf in 1953, as appears more clearly from the serial observations pre- viously discussed (p. 306). Table 19.^ — Regional variation in water temperatures on northwestern and southwestern parts of Georges Bank [Areas J and K, fig. 16. Temperature in °F.) Depth (meters) Northwestern Georges Bank Southwestern Georges Bank Station A Station B Station A station B 0 64.2 61.7 61.1 61.1 61.1 60.0 52.0 46.7 45.8 45.0 65.8 56.0 51.5 51.4 51.2 63.3 20 60.2 40 60.2 50 70 60.2 '60.2 61.7 49.5 54.7 60.7 ' Deepest reading was at 67 meters. Table 20.^ — Water temperatures on western part of Georges Bank in 1953 and in earlier years [Areas J and K combined; flg. 16. Temperature in °r.] July 20, 1914 and July 23, 1916 Sept. 3-4. 1953 Mean 1914 and 1916 Depth (metere) Num- ber of sta- tions Maxi- mum temper- ature Mini- mum temper- ature Num- ber of sta- tions Maxi- mum temper- ature Mini- mum temper- ature Mean 1953 0-20 40-50 4 4 65.5 56.3 52.0 60.1 5 5 65.8 62.5 52 45.9 54.8 52 61.1 56.3 In the eastern channel between Georges and Browns Banks (area L, fig. 16; table 21), the readings were about 2° lower at the surface for September 12, 1953, than for July 24, 1914, but 318 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE they were consistently higher, thence downward, with a maximum difference of 5.3° between the 2 series of readings at 120 meters, which was the deepest level at which the temperature was recorded in 1953. In the underlying strata this contrast is of the order to be expected from the difference between the dates when the observa- tions were taken in the 2 years. Furthermore, it may be no greater than can be explained on this basis, especially in view of the difference in the positions, for the 1953 station was about 21' miles farther out along the channel than the 1914 station, i. e., that much nearer the reservoir from which deep water enters the channel from the continental slope. Neither is the lower surface temperature for 1953 a safe index to the year-to- year fluctuation, for the surface had no doubt cooled 2° to 3° by the date of observation, whereas in 1914 the temperature was recorded when the surface was nearing its warmest for the year. Unfortunately, we have no salinity information for the 1953 station, which is the only reliable index to the amount of water from the continental slope that is entering the Gulf on any given occasion, or that had recently done so. Table 21. — Waler temperatures in the eastern channel between Georges Bank and Browns Bank in 1914 o,nd in 1953 [Area L, flg. 16. Temperature in °F.] Depth (meters) July 24, 1914 Sept, 12, 1953 Mean depar- ture in 1953 0 59.2 52.0 48.7 47.9 1 47.2 57.0 53.5 52 2 61.2 52.5 -2.2 20 + 1 5 40 +3.5 100 +3.3 120 +S. 3 ' Scaled from vertical graph for station. The Albatross III established only one station on Browns Bank inside the 50-fathom contour (area L, fig. 12) during her August-September 1953 cruises and recorded a bottom temperature of 50.9° F. on September 12 at 90 meters. This is slightly higher than the bottom temperatures reported by McLellan (1954, p. 410, fig. 4; p. 412, fig. 6) of about 42.8° to about 48° August 16-27, 1950, and of about 42.8° to slightly above 48° August 14-September 5, 1952. A bottom reading by the Grampus, July 24, 1914, of 47.3° is too close to McLellan 's to suggest that any long-term alteration of significance had taken place in the summer temperature of the bottom water on Browns Bank during the intervening 36 years. The comparison between temperature condi- tions in the Gulf of Maine in August and early September of 1953 and in the earlier summers of record may be summed up as follows: The entire water column down to 150 meters was warmer on the western side of the Gulf and around the northern periphery including the lower Bay of Fundy in 1953 than in the period 1912-26. This upward shift in temperature in 1953 as compared with 1922 amounted to 5° to 6°, at least, in Cape Cod Bay; to some 2° to 4.5° from surface to 150 meters off Cape Ann; to 1.5° to 3° from surface to 40 to 60 meters near Mount Desert Island; and to about 2° to 2.5° in the lower part of the Bay of Fundy at depths of 25 meters and greater. The upward shift in temper- atures may have been as great in the eastern part of Georges Bank as in the western side of the Gulf, though differences in the dates when these observations were taken prevent definite com- parison. Available data are at least compatible, with somewhat higher summer temperatures in 1953 than in previous summers on the western part of Georges Bank (p. 317). In the central and northeastern parts of the basin of the Gulf, at depths greater than 40 meters, the temperature was about as much warmer in 1953, compared to the period 1912-26, as it was farther to the north and west. We have no information to indicate any definite shift in temperature, either upward or downward, in the upper 10 meters or so, if allowance is made for differences in the location of the stations where the readings were taken and in the season and the stage of the tide when the observations were made in the different years. Thus, the surface was warmest in 1914 both in the central part of the basin and in the eastern side near the Nova Scotian slope, but warmest in 1926 in the north- eastern part near Mount Desert Rock. Readings taken August 1953 in the extreme southeastern part of the basin near the rising slope of Georges Bank were intermediate between those taken July 1914 and August 1926, at all depths from the surface down to 260 meters. Evidently, the effects on water temperatures of differences from year to year in air temperatures are masked in this region by the control exerted by warming by water from the continental slope, on the one CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 319 hand, and by chilling by the cold water that spreads westward past Cape Sable in the spring, on the other. Unfortunately, the August data for 1953 (fig. 17) fail to show whether the transition from sur- face temperatures as high as 57° to 65° F. off Massachusetts Bay to as low as 54° to 57° along the eastern part of the coast of Maine was as abrupt as the transition was from higher to lower temperatures in the area that the Grampus 44' 43' 42' 41" ■ 40° ■ PORTLAND^ CAPE SABLEi ■ 44' ■ 43* CAPE. ANN I ^_-~ 42" 41 40' 70' 69' 68' 67' 66 Figure 17. — Surface temperatures recorded, August 12-19, 1953. (Temperatures in degrees Fahrenheit.) 320 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE crossed in the offing of Portland, Me., in 1913, and a few miles farther eastward in 1 923 (Bigelow 1927, p. 589, fig. 47). A question of more general interest, from the ecological standpoint, is whether the higher summer temperatures of recent sum- mers, as compared with those of the earlier years of record, have appreciably weakened the barrier to the eastward spread of warm-water animals that was formerly set by the low surface tempera- tures that extended southward across the shelf from the elbow of Cape Cod and from the region of Nantucket Shoals. (For a general discussion of this temperature barrier, see Parr 1933, pp. 26-34 and 87.) There is no apparent reason, so far as summer- early autumn temperatures are concerned, why any species able to maintain itself in the upper 20 meters or so of water along the coast westward from Cape Cod during the first decade of the present century should not have been able to do so in the southwestern part of the basin of the Gulf and in Cape Cod Bay in 1953, for maxi- mum surface temperatures were about as high, there that year as they had averaged at Woods Hole, in Vineyard Sound, or in Buzzards Bay, during the period 1902-07 (table 22). Indeed, it had been known long before the upward shift took place that certain shallow, partially enclosed basins on the coasts of New Hampshire and of Maine — where the renewal of water from outside is slow — are warm oases, so to speak, where more-or-less permanent populations of warm-water animals exist, which do not range regularly north of Cape Cod along the open coast. The oysters that support a local fishery in Great Bay, N. H., and the hard clams (Venus) of Casco Bay and Table 22. — Surface water temperatures, in southwestern part of Gulf of Maine and westward from Cape Cod, in 195S and in 1902-07 ITemperatures in ° F.l Date Num- ber of sta- tions Surface temperature Locality Maxi- mum Mini- mum Mean 1953: August July 19 67.2 71.5 70 74.5 68.6 71.5 70 68 63 61.9 62.7 ' 70.7 Soutliwestern basin Oulf of Maine 1902-6: Woods Hole!- 1907: Vineyard Sound '. Buzzards Bay ' Sept. 5-6 August Aug. 16-29 -. Aug. 19-29 8 40 27 69 69 7 64.7 67.9 1 For details, see text above. 2 From Sumner, Osbum, and Cole, 191.1a; pp. 39 and 47. other Maine localities fall in this category, while the oysters of the southern shallows of the Gulf of St. Lawrence are a more striking example often cited. The upward shift in summer temperatures has involved the coastal belt along southern New England as well as the waters north and east from Cape Cod, the difference between the southwestern part of the Gulf and westward from Cape Cod being of about the same magnitude as formerly. This situation is illustrated by mean temperatures for the 2 warmest months combined for 1950 to 1953, of 65° to 67.2° F. at Boston contrasted with 69.7° to 71.7° at Woods Hole; also, by the distribution of the surface iso- therms to the eastward and to the westward of Cape Cod for early September 1953 (fig. 18). The contrast, too, between higher temperatures to the west and lower to the east, was not only about as great outside the islands (4° to 10° at the surface, 4° to 6° at 15 meters), in September 1953 as in the earlier summers of record, but the transition from the one to the other in the general offing of Nantucket Island and of Nantucket Shoals was almost as abrupt (compare fig. 18 with Bigelow 1933, fig. 35). No temperatures were taken in 1953 on Nantucket Shoals, but it is a matter of common knowledge that the surface there may be as much as 7° to 8° colder in summer than it is over the smoother bottom to the south- ward. This difference is the result of the active upwellings caused by the strong tidal currents that run over the shoals and around them. (For further details, see also Bigelow 1927, p. 595, and Parr 1933, p. 31.) Continuity between the warm water of Nan- tucket Sound and surface temperatures nearly as high in the southwestern part of the Gulf north of Georges Bank is similarly interrupted by areas of low temperature, caused by strong tidal currents running over the shoals that front more than three-fourths of the total breadth of the eastern exit from the Sound. In late August of 1925, for example, surface temperatures were 3° to 8° lower on Round Shoal (53° to 59° F.) than on Stellwagen Bank at the mouth of Massa- chusetts Bay (61°) or on Jeffreys Ledge off Cape Ann (62°). "(See Bigelow 1927, p. 1012, table 18.) No data were obtained for the eastern end of the Sound in 1953, but the regional temperature relationship was doubtless of the same order CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 321 43^ 42' 41- 40 - '^eRS 68 + 71' 70* 69* 68' -43' 42' 41 40' Figure 18. — Surface temperatures in the southwestern part of the Gulf of Maine and southward from the Martha's Vineyard-Nantucket region, September 1-13, 1953. Dots mark the locations where temperatures were recorded. (Temperatures in degrees Fahrenheit.) 322 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE that summer, for the tidal currents run as strong over the shoals and around them today as they did of old. Also, while the transition in tempera- ture from west to east off Nantucket is not as abrupt at depths greater than 10 to 15 meters as at the surface, the temperature averaged 10° to 12° lower in the southwestern part of the basin of the Gulf in early September 1953, even at 30 meters, than abreast of Martha's Vineyard (table 23) . Table 23. — Temperatures at SO meters and at 45 to 50 meters off southern New England and in the southwestern part of the basin of the Gulf of Maine, September 1-13, 1953 [Temperatures In °F.) At 30 meters At 45-60 meters Locality Num- ber of sta- tions Tem- pera- ture range Mean Num- ber of sta- tions Temper- ature range Mean South of Martha's Vine- yard South of Nantucket Southwestern basin of Gulf 13 14 12 53°-66° 52°-«3° 41°-56° 61.4° 58.1° 49.1° 13 9 8 51°-56° 51°-56° 45°-47.3° 57.5° 55.3° 46.0° The Nantucket boundary thus seems still to be a definite one for animals living within 10 to 15 meters of the surface and requiring summer tem- peratures higher than about 64°, while the low temperatures of the southwestern part of the Gulf similarly hinder expansion toward the northeast of animals that may be dependent on maximum yearly temperatures upwards of say 55° to 60° F. and that are restricted at the same time to depths of 25 to 30 meters or more. But there is no apparent reason why any animal, demersal or pelagic, that ranges indifferently between 10 and 50 meters (as many do), and that thrives in summer-autmnn temperatures down to say 55° to 60°, should not now range freely between the offing of southern New England and the south- western part of the Gulf, though confined to a narrower depth-zone in the Gulf. The possibilities of fauna] interchange between the head of Buzzards Bay and Cape Cod Bay via the Cape Cod Canal must also be borne in mind, though we have no definite evidence as to the role the canal may actually play in this regard. Autumnal progression In the western side of the Gulf, the surface was at its warmest for the year at about the same time in 1953 (middle to late August) as in earlier summers of record. No readings were taken in the Gulf in 1953 later than the middle of September; consequently, comparison of maximum temperatures for that year with those for previous summers cannot be extended to the underlying water, where maximum temperatures are reached progressively later in the season at increasing depths. The dispersal of heat downward, by the increasingly active vertical mixings that are characteristic of early autumn, seems at first to have followed the same schedule in 1953 as in earlier years, however; for while the surface off Cape Ann (area A, fig. 16; table 24) cooled, on the average, about 1.4° be- tween mid-August and September 9-10, the underlying waters down to 100 meters had mean- time warmed by some 2.5° to 3.3°. Hence, it may be assumed that when the maximum tem- peratures of the year were reached at successive depths, they were at least 2° to 3° higher in the western part of the Gulf in 1953 than was usual during the period 1912-26. Reduced to concrete terms, the vertical equalization of temperature characteristic of November, which took place at about 48° at the mouth of Massachusetts Bay in 1912 (Bigelow 1927, p. 980, table 4, station 10047), took place at about 50° to 51° there in the autumn of 1953. No information is available, yet, as to the autumnal progression of temperature for the eastern side of the Gulf in 1953. Table 24. — Temperatures off Cape Ann in August and September, 1963 (Temperatures In ° f.) August 12-15 September 9-10 Mean Depth (meters) Num- ber of sta- tions Tempera- ture range Mean Num- ber of sta- tions Tempera- ture range Mean depar- ture In Sep- tember 0 - 15 45 100 160 \ 64. 6°-68° 52. 0°-58. 0° 47°-49° 43. 6°-45° 44.6° 65.9° 56.6° 48.0° 44.4° 44.6° 5 5 6 3 1 63°-66. 1° 55°-65. 5° 47°-51° 45. 5°-50. 6° 44.6° 64.5° 59.3° 50.5° 46.9° 44.5° -1.4° -1-3.3° -1-2.5° -t-2.6° 0 Winter 1953-54 Data for the winter of 1953-54 are confined to bathythermograph readings taken February 16- 22, by the Asterias of the Woods Hole Oceano- graphic Institution, at a grid of 22 stations cover- ing the Massachusetts Bay region. It is probable that the temperature had risen fractionally above the winter minimum by that time, for the mean CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 323 was 0.7° higher for February 1954 in Boston Har- bor (36.6° F.) than for January (35.9°).' But the following comparison (table 25) between the temperatures taken February 16-18, 1954, and those taken botli 10 days earlier and 10 days later in the season in 1925 shows that winter chill- ing was not as severe in the Bay to the southward of Boston in 1954 as in 1925. The overall differ- ence between the 2 years for the southern part of the Bay as a whole was between 2° and 5°. More striking, the temperature seems not to have fallen as low as 33° anywhere in the open Bay at any depth in 1954, whereas it fell there (locally) in 1925, almost to the freezing point of salt water as it has in sundry other winters in the past (p. 303). Also, readings at 2 stations in the northern side of the Bay averaged 2° to 3° higher, surface to bottom, on February 18, 1954, than on February 13, 1913, or on March 4, 1921 — which were among the warmer of the winters that fell within the period of our earlier surveys (table 26). The mini- mum, also, seems certainly to have been no lower in the Bay in 1954 than in 1953, and may have been slightly higher, for while readings taken in the Bay averaged about 1° lower on February 18, 1954, than on March 31, 1953 (table 27), a some- what greater difference than this, but of the same order, might have been expected between the two sets of readings, on seasonal grounds. Table 25.^ — Maximum and minimum water temperatures in Massachusetts Bay Region, southward from the offing of Boston Harbor and seaward to a line from Cape Ann to the tip of Cape Cod, in February, 1925 and 193 4 [Temperatures in "F.] Table 27. — Water temperatures in Massachuietts Bay, in 1953 and 1954 (Temperatures In T.] Date Number of stations Surface temperature Bottom temperature Minimum Maxi- mum Mini- mum Maxi- mum Mini- mum ture at any depth 1925: Feb. 6-7... Feb. 24-28... 1954; Feb. 16-18 15 7 18 36.7 36.1 39.2 30.7 34.0 33.9 36.7 36.2 40.1 30.9 34.3 34.8 29.2 34.0 33.9 Table 26. — Surface and bottom temperatures in northern side of Massachusetts Bay in 1954 and in earlier years (Temperatures in °F.] Date Number of stations Surface temperature Bottom temperature Feb. 13, 1913... 1 1 2 37.1 36.0 38.8-39 37.6 36.6 39. 2-40. 8 Mar. 4, 1921 Feb. 18, 1954.. Date Number of stations Tempera- ture range surface to bottom Mean Mar. 31, 1953 2 4 41-41.3 39-40.7 41 2 Feb. 18, 1954 ' Information supplied by the U. S. Coast and Oeodetic Survey. This close correspondence between the winter temperatures of Massachusetts Bay in March 1953 and in February 1954, added to the consistently high level reached by the mean water temperature for the coldest month in Boston Harbor as far back as 1949 (year-to-year range, 35.9°-37.8° F.), shows conclusively that we are dealing here with a situation that has already persisted through 5 winters, not merely with 2 warm winters in succes- sion, such as have been known to interrupt the succession of colder winters in the past (p. 303). It is of special interest that through the winter of 1954, as during the preceding summer, the western side of the Gulf should have continued some 2° to 3° warmer than in earlier years, for the mean tem- perature in Boston Harbor was lower for that January (35.9°) than it had been for any month since February 1951. This with current reports of more ice than for many winters past in the partially enclosed situations along the coast of Maine, following severe weather in January, raises the question whether the upward shift in winter temperatures may have passed its climax in our Gulf. CHANGES IN ABUNDANCE AND DISTRI- BUTION OF MARINE ANIMALS Although valuable statistics on the landings of many marine species have been collected since 1887, for the most part these records are not con- tinuous prior to about 1930. It is not possible, therefore, to show long-term trends and fluctua- tions, except for a few species. Changes in fish- ing methods, efficiency, and effort, as well as changes in market conditions, tend to obscure the relations which may exist, so that one must fre- quently make general comparisons, based on minor fluctuations tliat may be related to environ- mental conditions when the major upward or downward trends may be due to other factors. 324 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE FLUCTUATIONS IN MACKEREL Landings of mackerel on the east coast of the United States from 1804 to 1930 have been esti- mated by Sette and Needier (1934). Their esti- mates for the period 1804-81 are based on records of inspection of barrels of mackerel in Massachu- setts (Goode et al., 1884). Landings for the period 1878 to 1904 were available from the re- ports of the Boston Fish Bureau. Since 1893, landings at the principal New England ports have been published by the United States Bureau of Fisheries and its successor, the Fish and Wildhfe Service. Except for the period 1804 to about 1820 when the fishery was beginning, the fluctua- tions in catch appear to reflect fluctuations in the abundance of mackerel. In figure 19, landings of mackerel are plotted against annual deviations in air temperature at New Haven that occurred 3 years earlier than the landings. Both curves are smoothed by 5-year moving averages. There is a marked tendency for rising temperatures to be associated with good catches 3 years later, the reverse with falling temperatures. From 1820 to 1890, four major peaks occurred in mackerel landings, each associated with a period of higher temperature. The degree of asso- ciation of landings with temperature may be de- termined by correlating the unsmoothed data, that is, correlating the landings in an individual year with the annual temperature deviation oc- curring 3 years earher. When this is done, the highly significant correlation coefficient of 0.554 is obtained for the 71-year period. Since air tem- peratures are not perfectly correlated with water temperatures and since it is the actual water tem- perature which may be expected to influence the mackerel, possibly a much higher correlation would be obtained if water temperatures were available. Following the high point reached in mackerel landings during the 1880's the catch dechned rap- idly and has never since reached its former high levels. From 1820 to 1890 the average annual catch was about 89 million pounds; between 1891 and 1950 it was about 33 million pounds. We see nothing in the tempera.ture record to account for this change in abundance, although the former 1807 1827 YEAR- (TEMPERATURES) J8.47 1867 1887 1907 1927 1947 1800 1820 1840 I860 1880 1900 YEAR (LANDINGS) 1920 19 40 Figure 19.— Landings of mackerel, 1804-1950, in New England (solid line) compared with annual deviations in air temperatures at New Haven, Conn., occurring 3 years earlier than the landings. (The curves are smoothed by 5-year moving averages.) CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 325 fluctuations in mackerel landings indicate tliat this decline was almost certainly due to natural causes rather than to fishing. After 1890 and to about 1925, there is a weak tendency for the mackerel landings to vary with temperature. After 1925 the landings are quite out of phase with temperature fluctuations, al- though both are at a much higher level than at the beginning of the period. The correlation between landings and temperature deviations for the period 1890-1954 is 0.207, which is about at the 10- percent level of significance (0.211). The spawning of mackerel, on botli tiie eastern and western sides of the Atlantic, appears to take place from south to north as temperatures reach a fairly well-defined critical level duriiig the vernal warming (Allen 1897-99; Orton 1920; Sette 1943). The length of the incubation period is greatly affected by the temperature of the waters in the surface layer. Sette reported that eggs hatch in 2 days at 70° F. and in 8K daj's at 50° F., and that the rate of development of the larvae also depends to some extent on the temperature. He estimated the survival rate of the 19.32 j-ear class during its planktonic existence to be in the order of magni- tude of only 1 to 10 fisli per million of newly spawned eggs. The period of planktonic life is shortened in warmer waters, and there is experi- mental evidence (Worley 1933) that the rate of mortality is less at higher temperatures. There- fore, relatively minor differences in the tempera- ture of the waters in which mackerel develop may produce wide fluctuations in the strength of the year classes. Furthermore, the spawning of mackerel in the western North Atlantic is largely confined to a coastal belt 10 to 30 miles in widtli (Sette 1943), where fluctuations in water tempera- tures may be expected to show greater correlation with land air temperatures than would water temperatures farther offshore. The influence of sea temperature on the move- ments of mackerel has been documented by Sette (1950). With regard to the appearance of mack- erel in the spring, and the northward advance, Sette considered that temperature has a limiting rather than a causal influence, water colder than 7° to 8° r. (44.6°-46.4° F.) acting as a barrier (op. cit., pp. 292-294). With a general warming of coastal waters, one might expect earlier arrival and later departure dates. Unfortunately, data on times of first arrival and corresponding tem})er- ature data are too fragmentary to determine this directly, but possible indication that this has occurred is shown in the monthly landing statistics for Gloucester. Between 1901 and 1935, nolandings of mackerel in April are shown in the Gloucester landings. Between 1935 and 1950, April landings of mackerel appear in 5 of tlie 15 years. Between 1901 and 1930, December landings of mackerel at Gloucester appear in 9 of the 25 years for which statistics are reported. From 1932 to 1950, December landings appear in 14 of tlie 19 years reported. Prior to 1939, no landings of mackerel are shown for the montli of January. Between 1939 and 1950, January landings are reported for 7 of the 12 j'ears. We have no satisfactory hypothesis to explain the great fluctuations in mackerel landings over the period of record; however, it is rather clear that the behavior of the mackerel is govern etl to a considerable degree by temperature. Data cover- ing fairly long periods of time show that the mack- erel has responded to fluctuations and trends in temperature, but in a maimer sufficiently complex to suggest the operation of other, perhaps indirect, factors. FLUCTUATIONS IN LOBSTERS Between 1940 and 1945, annual lobster landings in Maine increased from 7.6 million to 19.1 million pounds — a 250-percent increase and nearly three times the average landings made during the 1920's. Since 1945, the Maine catch lias averaged about 19 million pounds, a level not exceeded since the 1880's. The increase in lobster landings was observed not only along the coast of Maine but also in Massachusetts, in the Bay of Fundy, and in western Nova Scotia over about the same period. Such a widespread increase in lobster landings suggests an environmental change making possible the survival of greater immbers of lobsters to catchable sizes. Tagging experiments in Maine and length-frequency data collected by the Fish and Wildlife Service over the period 1939 to 1947 indicate that mortality rates remained very con- stant over the period of increase so that the in- creased catch was caused by an increase in abun- dance rather than by an increase in the amount of fishing. The only change in the environment known to have occurred was a general warming of coastal and offshore water. We shall, therefore. 326 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE attempt to relate changes in lobster landings to changes in temperature. A complete recoi'd of annual landings of lobsters since 1904 is available for Rhode Island in the an- nual reports of the Rhode Island Commissioners of Inland Fisheries (1904-35), and from federal statistics for recent years. These data, smoothed by 5-year moving averages, are presented in figure 20 together with January-July differences in water temperatures at Boothbay Harbor, 1906-49. The Boothbay Harbor data, previously shown in figure 8, page 299, have been given an additional smoothing by a moving average of 3. Tempera- tures at Boothbay Harbor were used for this com- parison because they were the only ones available which cover the period of these lobster landings. It has already been pointed out that trends in air and surface water temperatures are quite similar over an extensive range along the Atlantic coast. Temperature and landing data indicate a direct relation between water temperature and the availability of lobsters, a relation recently pointed out by Martin * for lobster catches in Canada. It is to be noted, however, that the curve of water temperatures in figure 20 represents a tendency toward warming of water conditions along the Atlantic coast in recent years, especially during the winter months, and that the catch of lobsters in Rhode Island has declined in the face of this warming. The relationship discovered by Martin is the reverse : Canadian lobster catches are greater in warmer years. Lobster landings for the Middle Atlantic States, for Massachusetts, and for Maine are presented for the years since 1918 in figures 21, 22, and 23. The downward trend in landings for the Middle Atlantic States is similar to that occurring in Rhode Island, while the trends in Massachusetts and Maine have been upward over the same period. * Martin, W. R., Long-term prediction based on climatic changes. Manu- script presented at a meeting of the Committee on Biological Investigations, Annual Meeting of the Fishery Research Board of Canada, January 1953. 9 pp. 2.8 - CO 2.4 O ^ 2.0 O Q. "^ 1.6 CO .2 — 0.8 0.4 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I - 6 - 4 I I I ' I ' I ' I I I ' I I I I ■ I I I I I I I I ' I I I I I I I I I I I I I I I O < LlI -4 -6 1904 1910 1915 1920 1925 1930 YEAR 1935 1940 1945 1950 Figure 20. — Landings of lobsters in Rhode Island, 1904 to 1949, as 5-year moving averages (solid line), and January- July differences in surface water temperatures at Boothbay Harbor, Maine, as deviations from the mean difference (dotted line). CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 327 16 1.4 () Z3 o Q. U. O 40 C/) Z o _J 30 20 MASSACHUSETTS LANDINGS MAINE PRICES 0 — 1937 1940 (/) UJ O Z) o Q. OC LlI Q. FiGiRE 28. — Landings of whiting in Massachusetts (1937-50) and the landings and price per pound of whiting in Maine (1937-52). not possible to decide, on the basis of the availabh^ data, whether this is true. It is probably safer to assume that some whiting have always wintered in the deeper parts of tlie Gulf, which waters were not fished until after 19,'57 wlien redfish boats began operations. Tliese boats do not ordinarily save any of the whiting they catch, so that whiting in the catches might liave gone tmiioticed for a number of years. Cursory examination of the weigh-out sheets of vessels landing at the port of Gloucester in Jan- uary, February, and March, 1952, indicates that small whiting boats can make consistently good catches during these months in subareas XXII E, which lies off the coast of Massachusetts, and XXII D, which lies off the coast of western Maine. The weigh-out sheets also show that whiting appears in small amounts, but regularly, as incidental <'atch for boats fishing in subareas XXII F and G. (See Rounsefell 1948, for a description of statistical areas.) A partial list of whiting landings at Gloucester 332 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE C/) Q 2 Z) O Q. U. O CO O 1937 1949 943 YEAR Figure 29. — The catch of whiting by pound nets in Massachusetts, 1937 to 1949. 1952 is shown in table 28 for the montlis of January and February 1952. This list includes trips in which whiting predominated in the catch, and the take of many of the trips consisted almost entirely of whiting. This is not to say, however, that whiting were not caught and landed by other vessels fishing in the same area. This fact is important, for the fishing for species other than whiting has been pursued in these areas for many years. The maximum depth of water in subarea XXII D is about 120 fathoms and in subarea XXII E about 150 fathoms. The larger part of both subareas is less than 80 fathoms deep. It is unlikely that these boats fished in the deeper waters, not only because no redfish appear in the catches but also because the deeper water, which lies more distant from port, probably could not be profitably fished in a 1-day trip. To determine more exactly the depths in which whiting are caught din-ing the winter months, interview slieets for the port of Provincetown were examined for January 1953. These sheets are prepared by a port interviewer and show, for each trip, the approximate position and the range of depths in which the vessel fished, as well as the weigh-out. Of 41 trips in which whiting were landed by small otter trawlers, 25 percent of the fishing was in depths of 30 fatlioms or less, 50 percent was in depths of between 30 and 55 fathoms, and 25 percent was in depths of between 55 and 70 fathoms. All of the trips were made to subareas XXII E and XXII G. Examination of monthlv landing records in Maine shows that whiting may be landed in any month of the year, although the landings during January, February, and March are small com- pared to the summer lantlings. These whiting probably do not come from the deeper waters of the Gulf of Maine, for these depths are fished only by the redfish boats, and the fisliermen rarely ti'ouble to save whiting when they encounter them. Recent data on the abundance of whiting in relation to other species on Georges Bank are CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 333 Table 28. — Landings of whiting at Gloucester, Mass., in January and February 1952 Subarca and date Type of gear ' Length of trip (days) Catch (pounds) Subarea XXII D: Jan. 6 Medium otter trawler do do 4.050 10 14 2.600 4, .WO 15 do 3.600 23 . do 3,550 Feb. 3 do 9.600 9 do 12 450 11 Small otter trawler 16,400 17 - do 7, 67.') 17 Medium otter trawler 3,875 27 ....do 2.285 27 do- - 7,600 Subarea XXII K: Jan. 3 -..- 900 4 do 9 375 4 5 9. 000 10 Medium otter trawler HOO 10 do... . 1.400 10 -- 1,740 14 do 380 15 - 6. 7.ifl 15_ do 4, 130 15 Small otter trawler 3. 000 16 300 21 do 525 28 Small otter trawler- -. 7.4,50 Feb. 2 8 do do 7.3.50 2,500 18 3.500 28 do 4,0.50 ' Small otter trawler, 5 to .50 gross tons; medium otter trawler. 50 to 150 gross tons. available from the otter-trawl investigations con- ducted there during the summers of 1948 to 1951 by the U. S. Fish and Wildlife Service. These investigations showed whiting to be the most abundant fish in tlie catches, outnumbering had- dock by 164 to 100, as contrasted with about 1 whiting to 8 haddock in 191.3 (p. 330). The pre- dominance of whiting in 1948 to 1951 was not due merely to its unusual abundance in depths or areas where haddock are not found, for in sub- area XXII J at depths between 30 and 60 fathoms, a favored fisiiing area of the commercial Jiaddock fleet, haddock outnumbered whiting by only about 1 .5 to 1 . Even allowing some latitude for tiie incom- pleteness of earlier observations, the evidence indi- cates major changes in the distribution and habits of whiting since Bigelow and Welsh (1925) de- scribed the distribution of the species. If one advances the hypothesis that there has been a general warming of coastal waters, which is indi- cated, the pattern of observed changes may be explained, at least in part. Accoi-ding to this hypotliesis, the coastal waters south of Cape Cod have become too warm for whiting to Ijc present in abundance. Since the warming may l)e expect- ed to be more pronounced close inshore where the pound nets are located, the decline will be first noted here, as in New York and New Jersey where IIul fish has practically disappeared from pound-net catches and in Massachusetts where the catch per pound net has declined to about one-third its former magnitude. The presence of wliiting in tlie Gulf of Maine during winter months is probably due to a general melioration of winter conditions as indicated by January temperatures at Boothbav Harbor, Me., and at St. Andrews, N. B. It might be objected that, as whiting tolerate a wide range of temperatures, the available data are not sufficient to explain the observed changes in distribution. We are not certain that this is a vaUd objection. The possibility that whiting may be able to tolerate the complete range of tempera- tures in the areas under consideration does not necessarily indicate that there are not optimum conditions which they prefer and seek. Even given optimum temperature conditions, whiting require food and the presence or absence of food will, of course, affect their distribution. Aside from the consideration that any fish may become adjusted to certain optimum environ- mental conditions because these conditions are related to its food supply, a temperature factor controlling distribution may be removed only one step from a direct relationship, according to this hypothesis. FLUCTUATIONS IN MENHADEN Occasional small menhaden catches have been reported for Massachusetts, but not farther north- ward, since this fish reappeared in abundance for the single year 1922 after an almost complete absence from about 1900. Its presence nortli of Cape Cod was noted in Maine waters in 1945, when small numbers were caught and used for lobster bait; in 1948,24,000 pounds were taken, and in 1949, 5,000,000 pounds. In 1951, 7,000,000 pounds were landed at (ilouccster, and in 1952, landings at the same port amounted to 26,000,000 pounds (Maine Coast Fisherman, 1952). The actual abundance of menlnulen north of Cape Cod since 1945 is not accurately reflected by catch figures. The catch in Maine in 1949, for example, was largely due to the presence of men- haden seiners from the southern fieet (Maine Coast P'isherman, 1949). Local vessels have not been equipped for menhaden seining and New 334 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE England conservation laws have made the oper- ation of out-of-state vessels difficult within the 3-mile limit where menhaden schools are most abundant (Bunker 1951). It is impossible to state to what extent the warming of inshore waters, as indicated by records of surface temperatures at Woods Hole, Boothbay Harbor, and St. Andrews, has affected the re- appearance of menhaden north of Cape Cod. According to Bigelow and Schroeder (1953), 50° F. is the coldest water temperature this fish will tolerate. Menhaden were reported abundant in the Eastport area prior to about 1840-45 (Goode 1879, p. 78), but they were not present in that area during the warm period of the 1870's when the menhaden fishery in the Boothbay Harbor region was at its height. In 1876, records of surface water temperatures at Eastport (Goode 1879, p. 291) show average July and August tem- peratures of 47.2° and 50.2°, respectively, which are sufficiently low to indicate unfavorable con- ditions for menhaden. In the same year, surface temperatures at Portland, Me., about 30 miles west of Boothbay Harbor, show July and August temperatures of 66.7° and 63.9°, respectively, so that the presence there of menhaden would not be surprising, as far as temperature conditions are concerned. It must be pointed out, on the other hand, that surface water temperature records at Boothbay Harbor over the period 1906-49 show that temperatures in that area have probably been sufficiently high for the presence of menhaden every year (see appendix table 1, p. 344). The presence or absence of a migratory species such as menhaden north of Cape Cod during the summer months is not necessarily determined by the inshore conditions where the fish is commonly caught, but may be determined b}" otlier conditions along its migratory route. Unfavorable conditions along the way may act as a barrier (see Role of Temperature in Faunistic Changes, p. 338). FLUCTUATIONS IN YELLOWTAIL FLOUNDER During the 1940's, a productive ycllowtail flounder (Limanda ferruginea) fishery developed on the southern New England fishing grounds, from Nantucket Shoals westward to Long Island (Royce, Buller, and Premetz, MS.«). In 1942, production of yellowtail amounted to more than 60 million pounds, but the catch fell away rapidly to about half that amount by 1944, and by 1949 to about 10 million pounds. It is natural under such circumstances to ascribe the decline to over- fishing, but Royce and his cowoi-kers found none of the usual symptoms of overfishing in the ex- tensive biological data collected during the rise and fall of this fishery. Paralleling the decline in catches of yellowtail on the southern New England banks was the increase in catches on the south- western part of Georges Bank from about 72,000 pounds in 1942 to more than 10 million pounds in 1949. Royce et al. point out that there have been at least three significant changes in the faunal com- position of the area used by the yellowtail flounder fishery during the 1940's. The Nantucket Shoals area produced many liaddock during the 1920's, the catch amounting to 13,000,000 pounds in 1928. These workers also point out that the haddock were caught in the same location and at exactly the same depth range whicli subsequently pro- duced yellowtail. They explain why yellowtail could not have been abundant during the period of the haddock fishery and also show that haddock were scarce on these same grounds during the period of the yellowtail fishery. Also significant is the fact that during the period of the haddock fishery yellowtail eggs and larvae were abundant to the south and west off the coast of New Jersey in areas where no yellowtail were found in numbers during the 1940's, at which time the demand for yellowtail was at its height. With the decline in abundance of yellowtail, the fishermen turned to "trash" species. The produc- tion of this fishery from the southern New England grounds, and largely from the same depths and locations where yellowtail were formerly caught, amounted to about 70,000,000 pounds in 1950 — approximating the peak landings of yellowtail at the height of that fishery. Royce, et al.° state that — Such changes in the habitat of a few species of fish must be evidence of more fundamental changes in the environ- ment and the entire complex of animal populations. In seeking an explanation we note that the known geographi- cal range of both haddock and yellowtail extends only a little south of the southern New England grounds but much farther to the north * * *. We notice, too, a re- treat of the haddock from the grounds west of Nantucket Shoals northeasterly to Georges Bank in the early 1930's » Royce, William F., Raymond J. Buller, and Ernest D. Premetz, 1966. Conservation of the yellowtail flounder Limanda ferruginea. V . S. Depart- ment of the Interior, Fish and Wildlile Service, Washington, D. C. (Un- puhlished MS.) CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 335 and a subsequent retreat of the yellowtail from off the N'ew Jersey coast in the 1920's to off southern \ew Eng- hmd in the early 1940's and then to Georges Bank about 1949 * * *. Perhaps these retreats have occurred because of the warming of the area, which has been reported by Conover (1951). NORTHWARD EXTENSION OF SOUTHERN SPECIES A great many reports of the northward exten- sion of the ranges of soutliern species have appeared in scientific htcrature since the 1930's. This may be attributed in part to an increased number of interested observers, but on tliis basis we would expect a nearly equal number of records of the southward extension of northern forms. We have noted only a few such southern extensions of range and only two in recent years: halibut landed March 13, 1946, near Reedville, Va. (Walford 1946), and four specimens of the Arctic cod (Boreogadus saicla) taken from a trap net in late January and February 1951, in Mirimichi Bay, N. B. (McKenzie 1953). In many localities, especially southward from Cape Cod, reports of southern forms became more frecjuent in the 1930's, a decade of peak air temperatures in the present warm cycle in the Northern Hemisphere (Willett 1950). At Sandy Hook Bay, for example, where observations of fish fauna were made for many years, three new rec- ords (although not new northern records) were reported by Breder and Nigrelli (1934) for 1932: the half beak {Hyporhamphus roberti), the gray snapper {Lutianus griseus), and the cutlassfish (Trichiurus lepturus). These workers comment on the rather unusual influx of southern species into that area during the summers of 1933 and 1934. They note that the cow-nosed ray (Bhinop- tera bonasun) was more common than ever before observed, the round pompano {Trachinotus falca- tus) was unusually abundant, and the gizzard shad {Dorosoma cepedianum) was first recorded in Sandy Hook Bay in October 1933. During the same decade, \«'e note the occurrence of the striped mummichog or killifish (Fundulus majalis) (p. 340) and the short big-eye (Pseudo- priacanthus nltus) in Massachusetts Bay, the lat- ter not having been reported there since 1859; •and of the butterfish (Poronotus triacanthus) in the (iulf of the St. Lawrence (Hoar 1937, Needier 1938) ; while a new northern record was establislied for the sting ray {Dasyatis say) (Smith and Griffin, 1939). Merriman (1939), commenting on tiie oc- currence of southern species in Connecticut waters in 1937, stated that "It seems probable that ab- normally high temperatures in the summer of 1937 were responsible for the unusual variety of southern species so far north." The unusual number of records of the white shark (Carcharodon carcharias) in the Gulf of Maine (Schroeder, 1938, 1939; Bigelow and Schroeder, 1953) is associated with the warming during this period. Schroeder (1939) describes this as an "uimsual occurrence" since "scarcely a dozen specimens" had been reported during the past century. It would be interesting to know if there have been any changes in the southern ranges of marine animals in New England and Middle Atlantic coast waters, especiall3' of the more northern forms. Unfortunately, there seems to be a paucity of observations bearing on this. It is perhaps of significance that the capelin (Mallotus rillosus) has not been reported from the Gulf of Maine since 1919, the spotted wolffish (Anarhichas minor) since 1927, the shanny {Leptoclinus maculatus) and the Arctic shanny {Stichae^l■'i punctatus) since 1930, while the staghorn sculpin {Gymnocanthus tricus- pis) and the spiny lumpfish {Eumicrotremus spi- nosus) have not been recorded since the last quarter of the 19th century. Reports of these species, which have been found in the Gulf at one time or another together with a number of others occurring in Nova Scotian waters, might well be expected in recent years, if such reports depended merely on the number and assiduity of observers. The trend, which began in the early 1930's, of an increased frequency of reports of southern species in more northerly waters has continued up to the present time. The following records, though by no means a complete listing, are some of the more interesting occurrences of recent years. Butterfly ray {Gymnura altavela). Consid- ered very rare on the Atlantic coast, a single specimen was recorded at Point Judith, R. I., in 1949 (Arnold 1951). Atlantic roi'Nd herring {Etrumeus sadina). Appear sporadically along the coast soutli of Cape Cod and are likely to come in schools, as happened at Long Island, N. Y., in 1890 (Bean 1903, p. 191) and Woods Hole in 1905 and 1908 (Sumner, Osburn, and Cole, 1913b, p. 741). Records in the Gulf of Maine prior to 1930 appar(>ntly are based on observations of one specimen or only a few 336 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE specimens. In 1937, round herring appeared in numbers at Campobello Island, N. B., at the mouth of Passamaquoddy Bay (Bigelow and Schroeder, 1953, p. 88). During August and September 1952, it was common in the Gulf of Maine along the coast as far east as Digdequash, N. B.; Scattergood (1952a and 1953) estimates that at least 210,000 pounds were landed. The fish was again present along the Maine coast dur- ing the summer of 1953. Atlantic lizard fish (Synodus foetens). Re- ported as relatively abundant in southern New England in the summer of 1949 (Arnold 1951). This fish is rare north of South Carolina (Breder 1948). Striped mummichog, or killifish {Fundulus majalis). A resident population has become established in Great Bay, N. H. (Jackson 1953). This fish was not recorded in Great Bay until April 1950, although records of the fishes of this region have been kept since 1908 (Jackson 1922). (See p. 340.) Striped mullet {Mugil cephalus). Jackson (1953) reports 14 taken at the mouth of Oyster River in Great Bay, N. H., between October 4 and 18, 1951. Bigelow and Schroeder (1953, p. 306) state that mullets are common as far north as New York, less so to Woods Hole, but so rarely do they stray past Cape Cod that there are only a half dozen records of them in the Gulf of Maine — each based on an odd fish. Sea horse (Hippocampus hudsonius). Two specimens were reported from the Maine coast in October 1953 (J. B. Glude, personal communica- tion). The specimens were collected by G. W. Coffin. The first specimen was obtained from a lobster fisherman at Kennebunk, Me., who found it in a lobster pot; the second was found clinging to a lobster-pot buoy rope at Pumpkin Ledges out- side Boothbay Harbor on October 18, 1953. The specimens were identified by Dr. Roland Wigley and are on file at the Woods Hole Laboratory of the Fish and Wildlife vService. Bigelow and Schroeder (1953, p. 316) state that— The sea horse is not common much beyond New York- Only a few are found each year about Woods Hole, chiefly in July, August, and September, and they so rarely stray past the elbow of Cape Cod that we have found only one definite (Provincetown) and one dubious (Massachusetts Bay) record of its capture in the inner parts of the Gulf of Maine, dead or alive, and one record for Georges Bank. Frigate mackerel {Auxis thazard). Was re- ported in great numbers in the vicinity of Point Judith, R. I., in August 1949 (Arnold 1951). Landings are reported for Massachusetts and Rhode Island for the years 1945 to 1950 and for the Middle Atlantic States for 1948 to 1950. Sumner, Osburn and Cole (1913b, p. 749) state that the fish is apparently rare in the vicinity of Woods Hole; Bigelow and Schroeder (1953) do not list it in the Gulf of Maine fauna. The capture of one specimen in a fish trap at Barnstable in Cape Cod Bay (F. J. Mather III, personal com- munication) in 1954 apparently is the first record of this fish north of Cape Cod. Striped bonito {Eufhynnus pelamis). One specimen was obtained from a fish trap at Barn- stable in Cape Cod Bay in 1954 (F. J. Mather III, personal communication). The only previous record in the Gulf of Maine is for Provincetown in 1880 (Bigelow and Schroeder, 1953, p. 336). Little TUNA {Euthynnus alletteratus) . Between 200 and 300 captured in a trap at Barnstable in the autumn of 1948 (Bigelow and Schroeder, 1953, p. 337) and 28 were taken from a trap in Cape Cod Bay in September 1949 (Schuck 1951a). These are first records for this fish in the Gulf of Maine. Common bonito {Sa>-da sarda). Scattergood (1948) reports a 310-mm. specimen collected from a herring weir on the southeast side of Campobello Island at the entrance to Passamaquoddy Bay, thus establishing a new record for the Gulf of Maine. Tuna {Thunnus thynnus). Although the tuna is a regular visitor to the Gulf of Maine, the young fish seem to have entered the southwestern part of the Gulf in much greater numbers during the past few summers than previously, although the local stock of large adults has not shown a corre- sponding increase. King mackerel {Scomberomonis regalis). This southern fish was once described as "not very common on our Atlantic coast" (Jordan and Ever- mann, 1896). Although sporadic in the statistics of landings for the Middle Atlantic States since 1877, it has appeared in the landings with increas- ing frequency in recent years. It is reported in Middle Atlantic States landings for the years 1931, 1933, 1937, 1946, and 1948 to 1954. The species is not reported from any of the New England CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 337 States for any of the years between 1919 and 1931 in which annual surveys of fish landings were made. It is reported for Massachusetts in 1931 and 1946 and for Connecticut in 1935 and 1946. Common dolphin {CorypJiaenahippurun). Was reported from Cape Cod Bay in August 1949 (Schuck 1951b) and in July 1951 (Bigelow and Schroedcr, 1953). The first record for the Maine coast was reported by Scattergood (1953) when a specimen was captured at Cape Elizabeth, Me., in 1952. There are no records for the Gulf of Maine before 1930. KiNGFisH {Menticirrhus saxitalis) . Scattergood (1948) reports a 47-cm. specimen removed from the mouth of a hair seal (Phoca vitulina) taken in a fish trap at West Point on the east side of Casco Bay, on August 2, 1941. FiLEFisH {Alutera ventralis). One specimen taken October 30, 1948, in 46 fathoms off Martha's Vineyard. Previously reported from the Tortugas region by Longley (Arnold 1949). Sea b.\ss {Centropristes striatus). A specimen taken in a lobster trap in the fall of 1950 at Corea, Me., and identified at the University of Maine (Maine Coast Fisherman, 1950b). Wood borer {Xylophaga dorsalis). Previously unreported on the coast of Maine, this wood borer caused extensive damage to lobster traps in the Mount Desert area in the winters of 1949 and 1950. It has since extended its range east and west (Dow 1950; Maine Coast Fisherman, 1950a, 1951). Sea hare {Tethys protea). A specimen was captured in a lobster trap off Woods Hole, Mass., in October 1953 (Hahn 1953). Two more were found on the beach of Nonamesset Island on December 15, 1953 (newspaper report), and J. Rankin (personal communication) found more than 20 in the area during the same period. Tube crabs {Pinniia and Polyonyi). In 1911, Pinnixa chaetopterana was the common com- mensal living in the tubes of Cha^topterits per- gamentaceus (Sumner, Osburn, and Cole, 1913a). At the present time this crab is rare in the Woods Hole region, having been replaced by the southern form Polyonyx macrocheles (J. Rankin, personal communication). Green crab {Carciniden maenaa). Extension of the range of the green crab north and east from Cape Cod since 1874 to Passamaquoddy Bay in 1952 has been described by Scattergood (1952b). Tlie documentary evidence gathered by Scatter- good establishes the slow progression of the species from Provincetowii in the 1870's to Casco Bay bj- 1905. Proctor's (1933) survey of the marine fauna of the Mount Desert region in 1933 does not list the green crab. Rathbun (1930) gives the range as New Jersey to Thomaston, Maine; she (1929) does not fist the green crab among the Canadian Atlantic fauna. As Scattergood (1952b, p. 6) points out, there were ample opportunities for the transportation and transplantation of this crab northerly by lobster smacks as early as Civil War times and by sardine carriers since 1900: The mere transportation of the crabs to other areas evidentlj' did not assure their establishing populations there. Conditions for the survival and successful repro- duction have to be present or new and permanent crab populations will not develop. Evidently such conditions were not always present in many Maine areas, for if the environment had been favorable, green crabs would have been established along the entire Maine coast before the early 1900's. If we knew what environmental changes have been necessary for the recent increased abundance and the greater dissemination of the green crab, we would probably understand why the crabs were not more com- mon in Maine waters at an earlier time. Since little is known about the life history and biology of the green crab, it is not possible to say what changes in the environment have made possible the greatly increased abundance of this crab northward to Passamaquoddj' Ba\\ Wallace and Glude (1952) present a number of observations indicating that exceptionally severe winters with heav3' ice along the shore kill the green crab. If this is true, then the higher temperatures in recent winters seem sufficient to account for the north- ward extension of the range of the green crab. Examination of surface water temperatures for tlie months of January and February at Booth- bay Harbor (appendix table 1) shows, for example, that the average temperature for the period 1906- 30 was 32.3° F. ; for the period 1931-49, the January-February temperatures averaged 34.2°; for the period 1945-49, the average was 35.1° — a most remarkable increase when one considers the effect on ice conditions along the intertidal zone. Prior to 1930, monthly averages below 32°, in- dicative of shore ice conditions, occurred with considerable frequency in the months of January, February, and March, so that it may well have been impossible for green crabs to establish per- 338 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE manent populations. Since 1930, months with temperatures below 32° have been rather infre- quent, and the persistent spread of the green crab northward would seem to indicate that sufficient numbers have survived the more severe winters to assure a more or less permanent population in the northern part of its range. ROLE OF TEMPERATURE IN FAUNISTIC CHANGES We doubt whether any marine biologist today would dispute that temperature of the water is the factor chiefly responsible for setting geographical boundaries to the ranges of marine animals along the seaboard of eastern North America. Conse- quently, any alteration in the temperature or any continuing trend, cither upward or downward, would be of great importance both ecologically, and from the commercial fisheries standpoint, if the alteration in temperature is wide enough and if it persists long enough to affect successful reproduction of the species or the well-being of the individual. Air temperature records, and the more limited sea temperature records, show that the warming trend since 1900 has not been steady, but ratlier has been interrupted by periods of cooling (fig. 1). In general, the amplitude of these fluctuations has shown a marked decrease since about 1880 — the result largely of rises in the minima. The pattern of increase within years also shows that the winter minima have risen while the summer maxima have increased relatively more slowly. The amplitude of fluctuations in winter tempera- tures has, however, been quite wide (fig. 4). These characteristics of the warming trend are important in considering the kind of faunistic changes which might be expected to result. Among the possible changes, we may consider the follow- ing categories: 1 . Establishment of resident populations north of Cape Cod of species that formerly were summer migrants. 2. Northern extension of ranges of resident populations already established north of Cape Cod. 3. Northern extension of ranges of summer migrants. 4. Changes in seasonal movements and distri- bution of permanent residents. 5. Changes in dates of arrival and departure of summer migrants. 6. Geographical changes in the abundance of permanent residents. In boreal latitudes, where the air climate is rigorous in winter but warm in summer, causing a wide seasonal difference in temperature of the water (as happens in the Gulf of Maine), the crit- ical questions are, How cold is the water during the coldest month of the year? and How warm is it during the warmest month? Perhaps it is not amiss to remind the reader that the mean temper- ature for the 3'ear as a whole plays only a minor role in this regard in the Gulf, or in similar situa- tions in general, except perhaps in the case of spe- cies that live at depths so great that they pass their entire lives within a very narrow tempera- ture range. We recognize, then, that changes in categories 1 and 2 depend primarily on the ability of the animal in question to withstand winter conditions, since temperatures suitable for spawning, though necessary, are not alone sufficient for the estab- lishment of resident populations. In category 3, the primary factor is undoubtedly the degree of summer warming, and may depend on conditions during only 1 or 2 months in the year, without reference to any overall trend in temperature. Categories 4, 5, and 6 depend to a much greater extent on the complex of conditions in all seasons of the year and on secular trends in temperature. It has long been common knowledge that no shoal-water animal could maintain a permanent population in estuarine situations around the shores of the Gulf, or even close in along the open coast, unless it were either hardy to temperatures nearly as low as the freezing point of salt water or able to escape the severity of winter by moving a few fathoms deeper along the sloping bottom. As an illustration, the mean for the coldest month was below freezing point of fresh water in 9 winters between 1922 and 1942 at Boston, Mass., and in 3 winters between 1934 and 1943 at Eastport, Maine. But there has not been a winter since 1948 when the local fauna was exposed to a mean tem- perature lower than 36° F. for as long as a month either in Boston Harbor or at Eastport — localities whicli taken together are representative of the western-northern periphery of the Gulf. Since the temperature of the water at the coldest CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 339 season is progressively higher from the siiore sea- ward all around tiie Gulf, not only at any given level but downward over the bottom as well, it is safe to say that in no winter since 1948 has any animal living as deep as 20 to 40 meters (whether on tlie bottom or in the mid-depths) anywhere in the Gulf been in danger of chilling if resistant to temperatures as low as 34° to 35° F., while any animal resistant to 38° to 40° has been safe at 100 meters anywhere in the open basin of the Gulf. A majority of the members of the fauna, whether bottom-dwelling or free-swimming, living along the middle Atlantic coast and shelf at these depths are no doubt as hardy as this or they would not maintain permanent populations there. Consequently, there is nothing apparent in win- ter temperatures such as may be assumed to have prevailed in the Gulf during the last 5 years or so to hinder the free interchange in shoal water of animal populations of the Gulf of Maine and of the coastal belt along southerly New England, Xew York, and perhaps Xew Jersey. This is in line with Parr's findings (1933, p. 87), based on a great number of temperatures taken at lighthouses and lightships, 1928 to 1930, that in winter the shallow belt along the mid-Atlantic coast was "equally in open continuity with the waters north of Cape Cod." This despite the fact that the mean temperature of the coldest month of the year at Boston was 3° to 4° lower during the period covered by Parr's observations (32.9°-33.6° F.) than during the past 5 years, 1948-52 (35.9°- 37.6°). Notwithstanding this continuity of conditions and the increase in winter temperatures, the quali- tative composition of the fish fauna is not uniform from north to south around Cape Cod in winter. We find, for example, that of some 30 species re- ported in the commercial landings in the New York-New Jersey area in January and February, 1952-54, only 8 are found north of Cape Cod in winter and they are permanent residents. Of the 7 leading species in these landings, only 1, the whiting {MeiiuccivM hilinearis) is reported in com- mercial lancHngs north of Cape Cod in winter. On the other hand, 4 of these 7 species are common north of the Cape in summer and the other 3 have been reported north of the Cape at one time or another. Tliere is, then, little evidence that the increase in winter temperatures has altered the general distributional pattern of the fisli fauna in the Gulf, with the possible exception of the whiting which has appeared in the commercial catches in winter only in recent j^ears (p. 329). On the other hand, we do not know of any ani- mal native to the Gulf, even among the cold- season spa\v^lers, that requires a temperature lower than 34° to 36° or so, either for its successful re- production or for any other stage in its life his- tory. Winter temperatures higher now than formerly can thus be classed as an improvement, from the faunistic standpoint, except perhaps for accidental strays from more-northerly seas that might reach the Gulf more often during colder periods, and might survive longer there, for instance, the Greenland shark and the capelin. Since it is not possible to deteimine whether some whiting have always wintered in the deeper parts of the Gulf, we cannot say that any marine animal formerly not found north of the Cape in winter has become a year-round resident under category 1 (p. 338). Sea temperatures and the commercial-landing data (p. 329), however, indi- cate that the whiting falls within category 4, since it appears safe to assume that this fish has always been a permanent resident of the Gulf. Evidence, too varied for analysis here, makes it likely that the spread of some animal populations from the south and west past Cape Cod toward the north and east into the Gulf of Maine is bounded more effectively by the distribution of temperature during the warmer part of the year than by regional differences in the degree to which the coastal waters chill in winter. Since the great majority of the adult animals of our middle At- lantic seaboard can survive in water that is as cool as the western side of the Gulf in summer, if not as cool as the nortlieastern side, the critical need here is for water warm enough for normal numbers to reproduce successfully. The oyster, surviving in a temperature close to the freezing point for a considerable period but requiring a temperature of 68° F. or higher for spawning, illustrates this category; the quahaug, or hard clam {Venus mercenaria), spawns successfully only in a temperature of about 68° to 70° but winters successfully where the water chills to 34° to 35°. Despite tlic liiglicr temperatures that have pre- vailed of late, tlie fauna of the Gulf — shoal water as well as deep — is composed today of much tlie same assemblage of species of fisli and of inverte- 340 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE brates as the fauna was half a century ago, indeed, as it was in the 1600's when Captain John Smith (1616, pp. 188, 196) and William Wood (1634, pp. 35-40) reported the abundance along the coasts of New England of cod, "frost fish" (tom cod), haddock, hake, cusk, striped bass, "pearch" (cunners), halibut, mackerel, smelts, herring, shad, eels, skates, lobsters, crabs, clams, "muscles" (mussels), "periwinckles," and oysters — the last, of course, in enclosed bays.' Corresponding to this conservatism of the Gulf of Maine fauna, in general, no conspicuous fish or invertebrate that is common along the coast west and south of Cape Cod, but which had not main- tained a regular population within the Gulf pre- vious to the recent upswing in temperature, is shown to have established itself there as a per- manent resident since the upswing. The weak- fish {Cynoscion), the scup {Stenotomus) , the toad- fish (Opsanus), and the blue crab {Callinectes) are among those species that would be expected to have so established themselves in the Gulf. The northward extension of the range of the striped mummichog (Fundulus majalis) to Great Bay, N. H., must be considered as falling in category 2 (p. 338) rather than 1, for it had probably established itself north of Cape Cod by 1939 (Bigelow and Schroeder, 1953), possibly bj' way of the Cape Cod canal (Schroeder 1937). Here again one must be cautious in attributing this extension to increases in winter temperatures subsequent to 1939 for it may merely represent the surmounting of a physical rather than a thermal barrier. The green crab (Carcinides maenas) — whose recent extension northward and explosion, so to speak, in population is rather fully documented (p. 337) — seems to have been resident earlier north of Cape Cod, at least locally. The tautog (Tautoga), whose status from year to year is followed with interest by local anglers, has not been more plentiful north of Cape Cod of late than in previous summers, perhaps even less so. Tiie occurrence of the various tropic strays (p. 335) may depend on short-term, and perhaps super- ficial, conditions that maj' have occurred fre- quently during the summer in the past, the present increase being distinguished chiefly by high tem- peratures sustained over periods of several years. ' The identities of the New England fishes listed by Smith as "Pinacks" and "Mullet" are not clear. In consequence, we might now expect more fre- quent records and greater numbers, but few unique occurrences. SUMMARY 1. A long-term upward trend in air tempera- tures in New England is evident from the record. The increase has been greatest for the winter months. 2. Upward trends in winter sea temperatures are shown for St. Andrews, N. B., Boothbay Harbor, Maine, and Woods Hole, Mass. The correlation of January water temperatures at Boothbay Harbor with January air temperatures at New Haven, Conn., and Eastport, Maine, indicates a long-term upward trend in surface temperatures corresponding to that for winter air temperatures. 3. Hydrographic data for the Gulf of Maine in 1953 and 1954 indicate an increase of from 1° to 5° F. throughout the water column since the period 1912-26 for most parts of the Gulf. 4. Northward shifts in the abundance and distribution of some important commercial species are indicated by a study of landing statistics and other data. These species include the mack- erel, lobster, menhaden, whiting, and yellowtail flounder. 5. Numerous southern species of fishes and other marine forms have extended their recorded ranges northward since 1930. At least two of these, the striped mummichog and the green crab, have established resident populations north of their earlier recorded ranges. But the recent upswing in temperature has not been accom- panied by any obvious general alteration in the composition of the fish or invertebrate fauna of the Gulf of Maine region. REFERENCES CITED Allen, E. J. I897-I899. Report on the present state of knowledge with regard to the habits and migrations of the maclcerel (Scomber scomber). Jour. Marine Bio- logical Association of the United Kingdom, vol. V (n. s.), pp. 1-40. Arnold, Edgar L., Jr. 1949. Unusual occurrence of fish on New England banks. Copeia, No. 4, pp. 299-.300. 1951. Northward dispersal of warm-water marine fishes in southern New England during the sum- mer of 1949. Copeia, No. 1, pp. 87-88. CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 341 Bean, Tarleton H. 1903. Catalogue of the fishes of Xew York. New York State Museum, Bull. 60, Zoology 9. 784 pp. BiGELOw, Henry B. 1927. Physical oceanography of the Gulf of Maine. Bull. U. S. Bureau of Fisheries, vol. 40 (1924), Part 2, pp. 511-1027. 1928. Exploration of the waters of the Gulf of Maine. Geographical Review, vol. 18, No. 2, pp. 232-260. 1933. Studies of the waters of the continental shelf, Cape Cod to Chesapeake Bay. I. The cycle of temperature. Massachusetts Institute of Tech- nology and Woods Hole Oceanographic Institu- tion, Papers in Physical Oceanography and Mete- orology, vol. 2, No. 4. 135 pp. BiGELOw, Henry B., and William C. Schroeder. 1953. Fishes of the Gulf of Maine. U. S. Depart- ment of the Interior, Fish and Wildlife Service, Fishery Bull. 74, vol. 53. 577 pp. (First revi- sion.) BiGELOw, Henry B., and William W. Welsh. 1925. Fishes of the Gulf of Maine. Bull. U. S. Bureau of Fisheries, vol. 40 (1924), Part 1. 567 pp. Breder, Charles M., Jr. 1948. Field book of marine fishes of the Atlantic coast from Labrador to Texas. 332 pp. G. P. Putnam's Sons. Xew York and London. Breder, Charles M., Jr., and R. F. Nigrelli. 1934. Fish notes for 1933 and 1934 from Sandy Hook Bay and other New York localities. Copeia, No. 4, pp. 193-195. Bunker, John. 1951. Pogie fishermen snarled with cease and desist orders. Maine Coast Fisherman, vol. 6, No. 2, p. 3. September. Clayton, H. Helm. 1927. World weather records. Smithsonian Miscel- laneous Collections, vol. 79. 1199 pp. 1934. World weather records, 1921-1930. Smith- sonian Miscellaneous Collections, vol. 90. 616 pp. Clayton, H. Helm, and Frances L. Clayton. 1947. World weather records, 1931-1940. Smith- sonian Miscellaneous Collections, vol. 105. 646 pp. CoNOVER, John H. 1951. Are New England winters getting milder? Weatherwise, vol. 4, No. 1, pp. 5-9. P'ebruary Dow, Bob. 1950. Trap destruction by borers increases in some areas. Maine Coast Fishermen, vol. 5, No. 3, p. 4. September. Fridriksson, Arni. 1949. Boreo-tended changes in the marine vertebrate fauna of Iceland during the last 25 years. Conseil Permanent International pour I'Exploration de la Mer, Rapports et Procfes-Verbeaux, vol. 125 (IB2), pp. 30-32. GooDE, G. Brown. 1879. The natural and economical history of the American menhaden. U. S. Commission of Fish and Fisheries, Report of the Commissioner for 1877, pp. 1-529. 30 plates. GooDE, G. Brown, Joseph W. Collins, R. E. Earll, and A. Howard Clark. 1884. Materials for a history of the mackerel fishery. U. S. Commission of Fish and Fisheries, Report of the Commissioner of Fisheries for 1881, pp. 91-531. Hachey, H. B., and H. J. McLellan. 1948. Trends and cycles in surface temperatures of the Canadian Atlantic. Jour. Fisheries Research Board of Canada, vol. 7, No. 6, pp. 355-362. Hahn, Jan. 1953. Strange sea hare found on Cape Cod. Maine Coast Fisherman, vol. 8, No. 5, p. 1. December. Hermann, Frede. 1953. Influence of temperature on strength of cod year-classes. Conseil Permanent International pour I'Exploration de la Mer, Annales Biologiques, vol. IX (1952), pp. 31-32. Copenhague. Hoar, William S. 1937. The occurrence of Porot^otus triacanthus in the Gulf of St. Lawrence. Copeia, No. 4, p. 238. International Commission for the Northwest At- lantic Fisheries. 1954. Statistical Bulletin, vol. 2 for the year 1952. Hahfax, N. S. Jackson, C. F. 1922. Ecological features of Great Bay, New Hamp- shire, and a preliminary checklist of its fish. Ecology, vol. 3, No. 1, pp. 48-54. 1953. Northward occurrence of southern fish (Fun- dulus, Afiigil, Pomatomus) in coastal waters of New Hampshire. Copeia, No. 3, p. 192. Jensen, Ad. S., and Paul M. Hansen. 1931. Investigations of the Greenland cod {Gadua callarias L.). With an introduction on the history of the Greenland cod fisheries. Conseil Inter- national pour I'Exploration de la Mer, Rapports et Procfes-Verbeaux, vol. LXXII (I). 41 pp. Jordan, David S., and B. W. Evermann. 1896-1900. The fishes of North and Middle America. U. S. National Museum, Bull. No. 47, Parts 1-4. 3313 pp. KiNCER, J. B. 1933. Is our climate changing? A study of long- time temperature trends. Monthly Weather Review, vol. 61, No. 9. pp. 251-259. September. Lauzier, Louis. 1952. Recent water temperatures along the Canadian Atlantic coast. Fisheries Research Board of Canada, Progress Reports, Atlantic Coast Stations, No. 53, pp. 5-7. April. 1954. Recent surface water temperatures along the Canadian Atlantic Coast. Cold years-warm years. Fisheries Research Board of Canada, Progress Reports, Atlantic Coast Stations. No. 58, pp. 6-10. April. 342 FISHERY BULLETm OF THE FISH AND WILDLIFE SERVICE LooMis, Elias, and H. A. Newton. 1866. On the mean temperature, and on the fluctu- ations of temperature, at New Haven, Conn., Lat. 41° 18' N., Long. 72° 55' W. of Greenwich. Trans. Connecticut Acad. Arts and Sciences, vol. 1, Part 1, Art. V, pp. 194-246. McKenzie, R. a. 1953. Arctic or polar cod, Boreogadus saida, in Miramichi Bay, New Brunswick. Copeia, No. 4, pp. 238-239. November. McLellan, H. J. 1954. Bottom temperatures on the Scotian shelf. Jour. Fisheries Research Board of Canada, vol. 11, No. 4, pp. 404-418. Maine Coast Fisherman. 1949. 21 southern boats are now fishing Maine waters. Vol. 4, No. 1, p. 3. July. 1950a. Sleuths hunt method to kill lobster trap borers. Vol. 5, No. 5, p. 16. November. 1950b. Vol. 5, No. 6, p. 14. December. 1951. Vol. 5, No. 7, p. 1. January. 1952. Vol. 7, No. 5, p. 1. December. Mavor, James W. 1923. The circulation of water in the Bay of Fundy. Part 2. The distribution of temperature, sa- linity and density in 1919 and the movements of water which they indicate in the Bay of Fundy. Biological Board of Canada, Contributions to Canadian Biology (n. s.) vol. 1, No. 18, pp. 353-375, plates 5-10. Merriman, Daniel. 1939. Notes on some marine fishes from Connect- icut, with comments on the scales of Elops saurus. Copeia, No. 2, pp. 113-114. July. Mitchell, J. Murray, Jr. 1953. On the causes of instrumentally observed secular temperature changes. Jour. Meteoro- logy, vol. 10, No. 4, pp. 244-261. August. Moore, Henry F. 1898. Observations on the herring and herring fisheries of the northeast coast, with special reference to the vicinity of Passamaquoddy Bay. U. S. Commission of Fish and Fisheries, Rept. of the Commissioner for the year ending June 30, 1896, pp. 387-442. Needler, A. W. H. 1938. Poronotus triacanihus in Malpeque Bay, Prince Edward Island. Copeia, No. 2, p. 96. Orton, J. H. 1920. Sea-temperature, breeding and distribution in marine animals. Jour. Marine Biological Asso- ciation of the United Kingdom, vol. XII (1919- 1922), N. S., pp. 339-366. Parr, A. E. 1933. A geographic-ecological analysis of the sea- sonal changes in temperature conditions in shallow water along the Atlantic coast of the United States. Bull. Bingham Oceanographic Collec- tion, vol. 4, Art. 3. 90 pp. Proctor, William. 1933. Biological survey of the Mount Desert region. Part 5. Marine fauna. 402 pp. Wis- tar Institute of Anatomy and Biology. Phila- delphia. Rathbun, Mary J. 1929. Arthropoda, Decapoda. Biological Board of Canada, Canadian Atlantic Fauna. 38 pp. 1930. The cancroid crabs of America of the families Euryalidae, Portunidae, Atelecyclidae, Cancridae and Xanthidae. U. S. National Museum, Bull. 152. 609 pp. Rhode Island Commissioners of Inland Fisheries. 1904-1935. Annual reports made to the General Assembly. ROUNSEFELL, GeORGE A. 1948. Development of fishery statistics in the North Atlantic. U. S. Department of the Interior, Fish and Wildlife Service, Special Scientific Rept. No. 47. 27 pp. February. Scattergood, Leslie W. 1948. Notes on some Gulf of Maine fishes. Copeia, No. 2, pp. 142-144. June. 1952a. Round herring appear on coast from West Point to Digdequash. Maine Coast Fisherman, vol. 7, No. 3, p. 8. October. 1952b. The distribution of the green crab, Carcinides maenas (L.) in the northwestern Atlantic. Maine Dept. of Sea and Shore Fisheries, Fisheries Circular No. 8, pp. 2-10. Augusta, Me. October. 1953. Notes on Gulf of Maine fishes in 1952. Copeia, No. 3, p. 194-195. August. Scattergood, Leslie W., Parker S. Trefethen, and Gareth W. Coffin. 1951. Notes on Gulf of Maine fishes in 1949. Copeia, No. 4, pp. 297-298. SCHROEDER, WiLLIAM C. 1931. An account of the fishes dredged by the Alba- tross II along the continental slope south of New England in February and March, 1929. Copeia, No. 2, pp. 41-46. 1937. Records of Pseudopriacanlhus alius (Gill) and Fundulus majalis (Walbaum) from the Gulf of Maine. Copeia, No. 4, p. 238. 1938. Records of Carcharodon carcharias (Linnaeus) and Pseudopriacanlhus alius (Gill) from the Gulf of Maine, summer of 1937. Copeia, No. 1, p. 46. 1939. Additional Gulf of Maine records of the white shark Carcharodon carcharias (Linnaeus) from the Gulf of Maine in 1937. Copeia, No. 1, p. 48. March. ScHUCK, Howard A. 1951a. Northern record for the little tuna, Eulhynnus allelleratus. Copeia, No. 1, p. 98. 1951b. New Gulf of Maine record for occurrence of dolphin, Coryphaena hippurus, and data on small specimens. Copeia, No. 2, p. 171. CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 343 Sette, Oscar E, 1943. Biology of the Atlantic mackerel (Scomber scombrus) of North America. Part 1 : Early life history, including growth, drift, and mortality of the egg and larval populations. U. S. Depart- ment of the Interior, Fish and Wildlife Service, Fishery Bull. 38, vol. 50 (1950), pp. 149-237. 1950. Biology of the Atlantic mackerel [Scomber scombrus) of North America. Part II. Migra- tions and habits. U. S. Department of the Interior, Fish and Wildlife Service, Fishery Bull. 49, vol. 51 (1952), pp. 249-358. Sette, Oscar E., and A. W. H. Needler. 1934. Statistics of the mackerel fishery off the east coast of North America 1804 to 1930. IT. S. Department of Commerce, Bureau of Fisheries, Investigational Rept. No. 19. 48 pp. Smed, Jens. 1949. The increase in the sea temperature in northern waters during recent years. Conseil Permanent International 1' Exploration de la Mer, Rapports et Proc^s-Verbaux, vol. 125 (IA2), pp. 21-25. 1953a. Monthly anomahes of the surface tempera- ture in areas of the northern North Atlantic in 1952. Conseil Permanent International I'Explo- ration de la Mer, Annales Biologiques, vol. IX (1952), pp. 16-19. 1953b. Variation of the surface temperature in the Northern North Atlantic during 1876-1952. Conseil Permanent pour 1' Exploration de la Mer, Annales Biologiques, vol. IX (1952), pp. 19-21. Smith, Captaine John. 1616. The generall historie of Virginia, New England and the Summer Isles. 2 vols. London. From London edition of 1629. Republished by Frank- lin Press, Richmond, Va., W. W. Gray, printer, 1819, 1 vol., 282 pp. Smith, Hugh M., and Robert A. Griffin. 1939. A sting ray (Dasyatis say), new to England. Copeia, No. 4, pp. 236-237. December. Sumner, Francis B., Raymond C. Osburn, and Leon J. Cole. 1913a. A biological survey of the waters of Woods Hole and vicinity. Bull. U. S. Bureau of Fish- eries, vol. 31 (1911), Part I, pp. 1-544. 1913b. A biological survey of Woods Hole and vicin- ity. Bull. U. S. Bureau of Fisheries, vol. 31 (1911), Part II, pp. 547-860. TAning, a. Vedel. 1949. On changes in the marine fauna of the North- western Atlantic area, with special reference to Greenland. Conseil Permanent International pour I'Exploration de la Mer, Rapports et Proies-Verbeaux, vol. 125 (I Bl), pp. 26-29. United States Coast and Geodetic Survey. 1918. United States Coast Pilot. Atlantic Coast. Section A, Serial No. 99. 305 pp. 1951. Surface water temperatures at tide stations, Atlantic Coast, North and South America. Special Publication 278. 57 pp. United States Weather Bureau. 1941-1953. Climatological data. New England. Annual Summaries, U. S. Department of Com- merce, Washington, D. C. Walford, Lionel A. 1946. New southern record for Atlantic halibut. Copeia, No. 2, pp. 100-101. Wallace, Dana, and John Glude. 1952. Plague of green crabs. Maine Coast Fisher- man, vol. 6, No. 7, p. 1. February. Willett, H. C. 1950. Temperature trends of the past century. Centenary Proceedings Royal Meteorological Soc, pp. 195-206. London. Wood, William. 1634. New Englands prospect by T. Cotes. 83 pp. London. Reprinted for E. M. Boynton, 1898. 103 pp. Boston. WoRLEY, Leonard G. 1933. Development of the egg of the mackerel at different constant temperatures. Jour. General Physiology, vol. 16 No. 5, pp. 841-857. May. 344 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE APPENDIX Monthly mean water temperatures listed in table 1 are based on the daily records of surface temperatures taken at 8 a. m., 12 noon, and 4 p. m. from 1905 to 1949 by the Branch of Fish Culture, of the U. S. Fish and Wildlife Service, at Boothbay Harbor, Maine. Similar records (table 2) were kept by the Branch of Fish Culture at Woods Hole for the period 1881 to 1942. The data for the period 1944 to 1952 were provided by the Woods Hole Oceanographic Institution. Appendix Table 1. — Monthly mean water temperatures, Boothbay Harbor, Maine, 1949 to 1905 (Temperature in °F.] 1949. 1948- 1947. 1946. 1945. 1944- 1943. 1942. 1941. 1940. 1939. 1938. 1937. 1936. 1935. 1934. 1933. 1932. 1931. 1930. 1929. 1928- 1927. 1926. 1925. 1924. 1923. 1922. 1921. 1920. 1919. 1918. 1917. 1916- 1915- 1914. 1913. 1912. 1911. 1910- 1909. 1908. 1907- 1906- 1906. Average, 1948-1906. January 38.4 34.6 36.6 33.6 36.5 34.1 32.8 34.6 36.2 31.4 34.1 34.4 36.0 33.7 32.0 31.8 38.0 36.2 38.0 32.3 34.6 36.0 33.4 32.7 29.6 34.3 30.1 31.8 36.4 29.9 34.5 28.1 33.0 33.0 31.4 34.1 36.6 31.7 34.9 34.6 33.7 36.2 32.2 34.9 33.9 February 36.0 32.4 35.6 33.7 33.6 34.0 32.9 32.0 33.2 29.1 32.0 31.7 36.0 31.3 32.0 30.0 38.0 36.0 36.5 31.2 33.1 33.0 32.6 29.8 31.5 30.9 29.8 31.0 33.2 29.4 34.5 29.5 29.9 31.3 30.7 29.9 33.4 30.4 31.5 33.5 32.3 32.4 29.0 31.4 32.3 March 37.0 34.8 36.5 37.5 35.8 34.0 33.6 33.9 35.5 30.0 32.0 32.3 37.5 37.4 34.8 31.9 38.2 35.9 38.7 33.1 33.7 34,9 34.6 32.2 35.5 34.6 30.6 34.9 36.7 35.6 36.8 33.0 31.9 30.8 33.1 31.9 35.7 33.2 34.3 37.2 36.5 34.8 32.2 31.3 32.7 34.4 April 42.9 40.5 41.1 40.4 43.6 38.6 37.0 39.8 41.7 36.2 34.6 37.5 41.3 40.7 41.1 41.1 41.6 40.2 43.6 39.6 37.7 39.1 38.6 36.1 40.6 39.2 38.6 41.6 43.1 39.7 41.6 38.8 36.6 38.3 39.9 37.1 40.7 38.9 40.0 45.3 40.7 40.1 37.4 37.7 38.0 39.' May 50.9 47.5 46.7 46.5 47.6 48.2 46.1 49.8 48.3 47.9 44.4 43.8 48.7 46.0 47.0 50.2 50.3 45.7 48.1 47.2 46.1 46.6 47.1 45.0 48.0 46.9 49.0 49,4 50.7 49.7 48.6 49.0 40,6 46.0 46.3 46.0 46.3 47.1 49.1 60.0 47.8 60.0 42.9 44.9 44.2 47.3 June 67.3 53.6 54.0 53.6 53.5 52.9 53.0 66.3 53.2 51.9 49.8 56.0 56.6 53.1 56.5 56,6 57.1 64.1 59.6 57.7 68.8 55.0 66.7 54.0 55.5 54.6 55.4 58.1 57.1 55.5 68.3 64.0 51.3 62.9 54.4 63.0 54.0 64.1 65.8 67.3 58.1 56.3 51.5 61.7 52.7 July 64.9 63.5 60.5 63.2 60.3 59.8 60,8 56,5 60.1 60.6 59.5 57.1 59.6 61.5 60.6 62.'4 61.6 61.6 58.7 64.2 62.6 57.9 63.1 60.9 61.2 61.4 61.6 69.6 63.9 63.4 61.6 62.8 62.3 69.8 59.4 60.1 68.4 61.7 61.6 64.8 64.0 61.2 62.4 57.7 58.5 59.4 60.9 August Septem- ber 61.2 63.4 60.4 61.7 60.5 59.9 60.0 57.0 58 9 60.0 60.2 64.2 61.6 62.3 60.0 62,1 62,6 60,4 60,9 58,7 64.6 60.4 61.4 61.2 61.1 58.5 62.4 60.7 64.1 60.6 62,0 62.6 59,4 58,8 59,8 61,3 62,0 61.2 63.1 62.6 62.9 69.7 62.1 60.1 61.2 66.9 60.5 58.8 58.3 58.7 54.7 57.3 52,1 56.3 64.7 64.3 67.9 57.3 67.4 59.1 65.8 56.9 58.0 57.2 66,5 57,4 66,8 66,3 56,1 56,9 57.8 59.0 59.3 67.0 67.3 56.6 53.9 64.6 57.2 56.4 66.9 67.3 66.4 59.4 69.0 68.9 57.3 66.7 54.3 57.0 October 48.0 66.9 52.5 51.0 51.0 61.7 62.8 50.4 49.2 48.0 49.8 51.1 48.0 50.9 46.9 48.3 50.5 60.1 63.7 49,0 51.7 51.1 50.3 46.8 50.4 50,0 49,4 51,4 63,6 50,4 49,8 48,0 47,3 47,7 50.4 53,4 50,7 50,1 50,9 63,4 63,1 48.4 49.0 47.9 50.4 Novem- ber 47.6 48.6 47.7 45.7 46.8 46.7 46.0 44.8 45.0 40.9 43.8 46.0 39.3 46.2 41.3 42.8 46.5 45.2 45.5 42.6 43.8 44.6 43,5 41,5 43,8 44,3 41,6 42,7 43,8 44,1 44,3 40,3 41,3 41,8 42,1 47.5 44.8 44.0 46.7 46.3 44.1 43.1 42.3 41.4 44.4 Decem- ber 42.5 40.0 41.3 37.1 39.2 40.1 37.4 39.5 39.7 33.7 37.3 41.7 36.0 37.9 36.2 36.6 38.9 38.3 37.8 32.7 38.0 38,7 36,6 36,8 36,8 40,4 33,4 37,2 39.8 36.8 37.5 31.9 36.4 36.3 34.6 41.8 38.9 39.6 36.4 38.5 37.2 39,0 35,5 36.8 37.5 Yearly mean 46.7 48.5 47.2 47.0 46.5 45.3 46.6 46.0 44,6 43,4 45,1 48,1 45,4 46,6 45,6 47,4 46,8 48,4 46,6 46,0 46,9 46,2 44,8 45,4 45,8 46,3 46,4 47,6 46.6 47.2 45.4 43.3 44.1 44.7 43.6 47.4 45.9 46.8 48,2 47,3 47.4 44.2 44.6 CLIMATE AND THE DISTRIBUTION OF MARINE ANIMALS 345 Appendix Table 2. — Monlhh/ mean water temperatures. Il'oorfs Hole, Mass.. 19-52 to 1881 (Temperature in °F.) Year January February March April May June July August Septem- ber Octotwr Novem- ber Decem- ber Yearly mean 36.1 37.1 39.6 37.7 30.0 35.8 32.9 32.4 35.0 34.9 36.2 36.7 30.5 34.0 31.9 30.6 36.4 38.4 34.0 45.2 46.7 42.3 54.0 56.0 50.1 55.4 52.2 57.9 53.0 54.5 59.2 65.0 65.0 62.0 65. 5 60.6 58.8 61.9 63.1 63.1 1951 ---- 70.6 69.7 71.6 69.1 71.7 68.4 70.7 70.0 71.7 71.3 72.5 71.3 72.4 68.6 70.6 71.8 70.2 66.2 67.9 68.0 69 6 66.3 68.6 68.2 61.4 59. S 61.4 58.3 60.9 61.1 59.1 59.2 50.4 52.9 52.0 50.4 49.4 53.6 50.5 48.0 41.3 42.9 41.9 43.8 38.4 42.9 37.8 38.1 53.6 1950 - 52.2 1949 . . . 1948 _ 35.2 36.2 39.0 38.0 44.6 43.5 45.3 48.0 42.1 51.2 1947 52.4 1946 -- 52.1 1945 52.0 1944 1942 33.3 33.6 29.7 33.3 35.8 39.9 31.2 30.8 32.7 39.3 41.1 1941 -- 30.6 30 0 31.4 36.2 37.4 29.5 28.6 29.8 35.7 35.9 32.3 33.4 34.1 37.7 37.7 37.6 33.6 32.5 35.9 41.9 40 3 40.fi 45.0 43.5 45.2 39.9 42.9 43.2 43.6 52.5 51.1 48.9 51.9 52.9 55.1 49.9 52.8 53.8 53.1 61. i 59.1 59.8 60.4 62.5 63.0 59.6 62.4 62.4 62.1 68.1 65.7 66.1 66.9 68.2 68.2 67.4 69.9 67.5 69.6 69 4 67.9 70.6 70.1 72.2 70 7 68.3 69.9 70.2 72.2 67.0 66.4 65.2 64.4 66.5 67.2 63.5 67.8 67.7 67.5 60.3 57.0 57.4 57.2 56.6 60.2 56.3 58.3 60.8 60.9 51.2 47.9 44.8 50.8 49.2 49.4 50.1 47.9 47.3 50.5 41.8 39.2 36.8 39.6 44.6 39.7 36.5 39.0 38.0 41.0 50.8 1940 - 49.0 1939 48.2 1938 - - 51.3 1937 52.6 1936 51.4 1935 47.0 1934 50.5 51.8 1932 1914 '- 33.9 39.2 31.6 34.7 31.5 31.1 34.1 30.1 31.3 32.0 32.0 38.1 34.0 33.7 37.8 42.0 45.5 42.3 41.3 46.9 51.8 52.9 51.8 52.9 55.3 62.4 62.0 65.0 62.5 62.6 66.4 69.6 69.4 71.0 70 4 68.7 71.7 69.5 71.1 71.0 67.2 66.8 66.9 66.3 66 9 60.5 60.2 60.1 57.8 60.4 49.1 51.2 51.5 49.1 48.0 39.1 42.7 41.5 41.5 36.7 50.4 52.8 1912 - 51.1 51.1 1910 51.6 1908 35.7 30.4 35.9 41.3 52.4 61.9 71.0 69.3 62.7 57.2 46.6 40.2 50.4 1906 33.0 30.7 29.3 32.7 31.5 33.2 34.6 33.4 36.1 32.0 32.9 34.0 35.0 29.6 36.8 33.1 40.2 36.6 31.1 32.6 32.2 34.6 32.4 34.4 29.0 29.0 32.0 29.1 29.6 31.2 30.8 34.0 31.6 31.7 29.6 32.7 29.8 31.9 34.8 36.4 32.2 30.4 33.2 30.9 29.8 34.5 31.2 33.1 30.9 35.0 32.6 33.1 39.1 36.2 33.4 33.8 34.5 39.0 35.4 32.4 "37.6 30.7 33.3 35.7 36.4 34.7 33.1 34.7 32.2 31.8 35.3 32.3 37.5 42.2 42.4 40.9 45 7 45.7 41.2 40.9 41.6 43.7 43.0 40.7 42.0 42.7 40.9 42.5 44.3 42.7 41.0 40.0 42.6 40.3 42.8 41.4 43.3 41.0 53.2 52.0 53.3 54.5 54.0 51.2 52.7 51.9 49.7 49 7 53.3 53.4 53.6 51.0 51.1 52.3 53.3 55.7 50.2 52.6 55.7 50.1 55.1 52.5 49.8 53.8 62.8 61.5 62.2 61.0 62.0 61.2 61.7 63.2 60.5 60.0 63.4 63.4 62.0 61.2 62.1 61.1 62.0 65.4 62.7 62.0 62.5 61.3 68.9 70.3 69.4 67.6 66.3 69.6 68.6 69.6 68.2 69.6 65.7 '76.4 69.6 69.9 66.9 69.4 68.7 66.6 70 6 68.7 71.4 70.9 69.9 65.8 68.0 69.1 69.9 70 9 69.4 70.8 70.3 70 9 69.9 69.4 73.3 71.8 68.6 71.1 70.7 68.9 72.0 69 1 71.8 68.1 66.3 67.0 66.6 67.3 68.7 68.8 66.0 68.2 67.6 66.7 67.6 67.2 67.4 67.1 69.1 68.0 67.2 64.3 66.5 68.1 65.9 60.0 59 4 57.7 57.3 60.7 62.2 60.1 57.8 59.6 59.4 58.2 59.5 59.6 60.4 58.8 57.6 59.0 57.0 54.3 58.9 60.3 55.6 48.4 48.1 46.5 47.8 52.0 47.8 50.7 49.4 49.9 49.9 53.1 49.5 48.1 52.4 50.6 47.4 48.0 49.8 48.6 48.2 51.0 51.4 36.2 39.8 34.4 36.5 38.8 36.8 39.8 41.8 37.1 41.9 41.2 39.7 39.2 40.8 37.3 42.0 36.7 44.4 39.1 40.3 38.0 39 7 51.1 1905 50.2 1904 49.1 .50.7 1902 51.1 50.4 1900 51.1 50.8 1898 51.4 50.9 1896 50.8 1894 --- 51.5 50.6 1892 51.1 1891 -- -- 51.1 1890 --- 52.1 1888 49.2 51.0 1886 51.0 1885 50.6 1884 64 9 61.6 71.3 69.9 65.1 55.9 47.7 48.0 50.5 37.1 34.6 59.0 34.10 32.15 35.07 42.83 52.88 62.15 69.01 70 33 67.01 58.97 45.92 39.69 o NEW GENUS AND TWO NEW SPECIES OF THARYBIDAE (COPEPODA CALANOIDA) FROM THE GULF OF MEXICO WITH REMARKS ON THE STATUS OF THE FAMILY By Abraham Fteminger, Fishery Research Biologist Two heretofore undescribed species representing a new genus of an obscure family of marine cala- noids, the Tliarybidae, have been found during studies on the copepod fauna of the Gulf of Mexico. The plankton samples that contained these forms are part of the extensive collections made by the U. S. Fish and Wildlife Service M/V Alaska in the Gulf of Mexico between 1951-1953. Apparently, this is the first record of tharybids from the western North Atlantic region. REMARKS ON THE THARYBIDAE Sars (1902) originally conceived of the family name Tharybidae to place Tharybis macrophthalma Sars, a northeastern Atlantic species of the mono- typic genus Tharybis Sars. A second genus, Psewlotharybis, was added by Scott (1909a) for two species from waters off the British Isles. Tanaka (1937) proposed a third genus (monotypic) for a species from Suruga Bay, Japan, which Brodski (1950) recently synonymized under Un- dinella Sars. Brodski also removed Undinella, together with its three species, from the Scolecith- ricidae and placed it within the tharybid complex. It is apparent from the available data tliat Brodski's tharybid revision was justified. The position of the Tharybidae in relation to tlie remaining families of the Calanoida has not been established satisfactorily. Although the family is usually included in Sars' section, the Isokeran- dria, it is now widely recognized that this section constitutes a heterogeneous complex unacceptable to a natural system. At present, it seems best to place the tharybids under Gurney's super- family, Paracalanina, in a subgroup following the Phacnnidac, the Scolecithricidae, and the Diaixi- NOTE.— Approved for publication November 9, 1956. Fishery Bulletin 116. dae. Although there is as yet little tangible evidence to support this move, it is noteworthy that these families are similar in the following basic characteristics: (1) the occurrence of senso- riform filaments and falcate spines on the second maxillae, (2) organization of the fifth legs, (3) segmentation of the swimming legs, and (4) form of the rostrum. Moreover, in the above order these families present an apparently natural and progressive "line of development, to the extent of secondary specialization of the appendages (e. g., atrophy of the male cephalic appendages, com- plexity of the sensoriform filaments, development of the fifth legs, spinulation and flattening of the swimming legs). Admittedly, these similarities and their serial relations have been somevvhat overgeneralized, but they strongly suggest a closer relation between these families than that generally held at present. The need for a formal description of the tharybid complex is urgent but the lack of sufficient details concerning most of the species prevents its presen- tation at this time. For example, attempts at precise separation of the tharybids and scole- cithricids meet with difficulty. The usual struc- tures of significance (such as body and appendage segmentation, as well as arrangement and setal ornamentation of appendages) are for the most part similar, or at least tend to intergrade between the two families. In this respect Scolecithricella ctenopus (Giesbrecht) and the new genus described present major obstacles to such a separation. Thus far a survey of the literature and study of a limited quantity of reference material iias revealed that only the swimming legs appear to exhibit consistent difi'erences. As to the thary- 347 348 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE bids, in addition to the lack of conspicuous spinu- lation on the anterior and posterior faces of the rami, the endopodal segments are not broadened, nor is the lateral margin lamelliform and produced distad in a prominent spiniform process. Otiier distinguishing characteristics may be forthcoming from consideration of the mouth parts, especially the mandibular dentition, and the number of setae on the first and second maxillae. It should be stressed that thorough revision of the Scoleci- thricidae, suggested previously by Vervoort (1951), and restudy of extant tharybid material is neces- sary to resolve the present confusion. Within the Tliarybidae further revision is needed to complete the foundation laid down by Brodski. It appears that PsevAotharyhis Scott must be removed from the family. According to Scott (1909a) the generotype {P. zetlandicus Scott) is distinguished from Tharybis by the lack of sensoriform filaments on the second maxillae and by the truncate nature of the mandibular gnathal lobe. As mentioned, the sensoriform filaments are characteristic of all other known species of Phaennidae, Scolecithricidae, Diaixidae, and Tharybidae. Although their function re- mains to be established, it seems obvious from preliminary examinations (unpublished data) that these filaments are significant, if not vital, sensory receptors. Moreover, many other workers have already suggested their probable sensory function. It does not seem reasonable, therefore, to assume that these filaments are secondarily lost in Pseudotharybis , nor is there any evidence that Scott's genus represents primitive tharybid stock. There are also fundamental diflFerence^ betiveen the tharybids and known species of Pseudotharybis regarding the mandibular gnathal lobe. In con- trast to that described by Scott for Pstudoiharybis, the tharybid gnathal lobe is typically elongate and the teeth, excepting the two most ventral, tend to be spiniform or styliform. Although Scott's (1909a, b) descriptions of both Pseudotharybis and his two species included there- in, zetlandicus and dubia, are inadequate for deter- mination of their actual phyletic relations, the eventual removal of these taxa from the tharybid complex appears a virtual certainty. For exam- ple, there are several interesting points of similarity between Pseudotharybis and Drepanopsis Wolf- enden. Pending restudy of Scott's material how- ever it is proposed that the genus and its species be treated as incertae sedis. Genus Parundinella, new genus Description. — Tharybids resembling those of the genus Undinella, according to Sars (1900:51) and Brodski (1950:275), although considerably smaller in size. Cephalon fused with thoracic segment I, thoracic segment IV fused with segment V; terminal portions of fusion seg- ment IV-V produced posteriad in lobiform or spiniform processes; genital segment symmetrical, swollen laterad at midportion, ventral portion at most only weakly produced. Abdomen with 4 segments in female, 5 in male; anal segment in both sexes reduced in length, considerably shorter than preceding segment. Rostrum with simple bifurcated basal portion; each bifurcation arti- culating with a slender filament. First antennae with segments 8-9, 24-25 fused in female; seg- ments 8-10, 20-21, 24-25 fused in male. Mandi- bular gnathal lobe elongate; only 3-4 monocuspi- date spiniform teeth present, confined to dorsalmost portion of lobe. First maxillae with inner lobe 1 moderate' in size, not broad or truncate. Second maxillae with lobe 5 bearing two sensori- form filaments, one normal seta, one enlarged falcate seta; terminal segments with 5-6 sensori- form filaments; filaments occasionally with apical flagellum. Exopodite of legs 1-4 three-segmented ; endopodite of leg 1 one-segmented, leg 2 two- segmented, legs 3-4 three-segmented. External exopodal spines of legs 2-4 ornamented with small denticles along upper and lower margins. Female fifth legs minute, two-segmented; terminal seg- ment with one or more distal acuminate processes and one robust medial spine bearing ventral spinules. Male fifth legs biramous, asymmetrical; left exopodite three-segmented, left endopodite two-segmented; right exopodite two-segmented, right endopodite in part fused with second basal segment. Remaining structures as in family. P. spinodenticula, new species, is herewith des- ignated the generotype of the new genus. THARYBIDAK FROM THE GULF OF MEXICO 349 SYNOPTIC KEY TO THE GENERA OF THARYBIDAE SrRincnts 24 and 25 of first antennae fused: mandibular gnathal lobe with 3-4 nionocuspidate spiniform teeth; female fifth legs with 2 segments; male right fifth leg with conspicuous endopodite; external exopodal spines of legs 2-4 bearing marginal denticles Parundinella Segments 24 and 25 of first antennae separate; mandibular gnathal lobe with more than 4 monocuspidate spiniform teeth; female fifth legs with 3 segments; male right fifth leg with endopodite vestigial or lacking; external exopodal spines of legs 2-4 not ornamented or bearing fine hairs 2 Thoracic segment I and cephalon partially separated; segments IV andV fused or separated; rostrum with bifid basal portion; first maxillae with inner lobe 1 not of greater surface area than remaining lobes and rami, endopodite extending to or beyond distal border of inner lobe 1, exopodite with 2-3 setae; maxillipeds with segment 2 of uni- form thickness; female fifth legs with segments more or less equal in length; terminal spine of legs 2-4 with fine shallow serrations; lobes of second maxillae crowded within distal half of appendage Undinella Thoracic segment I fused with cephalon, segments IV and V fused; rostrum with rounded knob-like basal portion bearing 2 long filaments; first maxillae with inner lobe 1 large, extending beyond remainer of appendage, exopodite with 5 setae; maxillipeds with segment 2 swollen, greatest thickness at midlength; female fifth legs with terminal segment longer than preceding 2; terminal spine of legs 2-4 with coarse deep serrations; lobes of second maxillae extend over more than half of appendage Tharybis It is noteworthy that several of the characters used in separating the 3 genera tend to intergrade as follows: (1) ratio of surface area of inner lobe 1 to remaining lobes and rami of first maxillae (largest in Tharybis; less in Undinella; smallest in Parundinella); (2) segmentation of female fifth legs (3 segments of about equal length in Undinella; basal and mid-segments considerably reduced in Thanjbix; loss of 1 segment, presumably the second, and reduction of basal segment in Parundinella); (3) segments 24 and 25 of the first antennae (separ- ate in Tharybis and Undinella; fused in Parundi- nella spinodenticula, new species, but partially separated in P. manicula, new species). Parundinella spinodenticula, new species Plate 1, Fir.s. 1-16; pi. 2, Firs. 1-11. Localities, Malerial.~GuU of Mexico: lat. 22° 20' X., long. 87° 31' W., (Alaska, cruise 4, station 11, 13 January 1952, 1 meter depth of plankton tow; 2 adult females, 1 intersexed individual); lat, 23° 18' N., long. 97° 38' W., (Alaska, cruise 2, station 11, 7 June 1951, 1 m, depth of plankton tow; 1 adult male). Measurements. — Specimens measured in dorsal view. Length of cephalothorax along midsagittal plane from anteriormost limit of forehead to post- erior margin of intersegmental membrane between thoracic fusion segment IV-V and genital seg- ment; length of abdomen from anterior margin of genital segment to posteriormost limit of right furcal ramus. Measurements made at lOO.x mag- nification with aid of ocular micrometer; speci- mens immersed in aqueous solution of 50 percent glycerine. Slender glass rods used to sii])port cephalothorax and abdomen in horizontal position during measurements of each specimen. Measurements given include total length (TL) and cephalothorax-abdomen length ratio (CAR). 1. Adult Fem.\le: (holotype) TL 0.84 mm., CAR 3.1:1; (paratypes) TL 0.83 mm.; TL 0.84 mm. 2. Adult M.\le: TL 0.74 mm.; CAR 2.7:1. Description. — A minute tharybid characterized at low magnifications by a robust ovate cephalo- tliorax with first antennae extending to thoracic fusion segment IV-V and a short slender abdomen about half as long as the fourth pair of swimming legs. Adult Fem.ale: Cephalothorax robust, greatest depth between maxillipeds and first legs; forehead strongly vaulted (pi. 1, figs. 3, 5; pi. 2, fig. 11). Rostrum with distal half of basal portion bifid, each process bearing a short rostral filament (pi. 1 , fig. 8). Terminal portions of thoracic fusion segment IV-V in lateral view produced in a tri- angular lappet with rounded apex, extending pos- teriad beyond midlength of genital segment (pi. 1, figs. 3, 5). Genital segment in dorsal view with moderate lateral swellings confined to anterior three-fourths of segment; segment narrowed abruptly just posterior to seminal receptacles (pi. 1, figs. 5, 16). Four abdominal segments and furcal rami with proportional lengths taken along midsagittal plane of 37, 19, 19, 8, 17 (=100). Furcal rami appro.x- imately twice as long as broad, each bearing 6 setae; 4 most terminal setae elongated, almost equal in length; lateral and medial setae shorter than ramus, unequal in length (pi. 1, fig. 5). Cephalic appendages differing in several details from corresponding ap|)endages of Tharybis. First antennae (pi. 1, fig. 4) with 23 visible seg- 350 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Plate 1. — Parundinella spinodenlicula, new species. Female: 1, Maxilliped; 2, Fifth legs; 3, Trunk, lateral view; 4, First antenna; 5, Trunk, dorsal view; 6, Mandibular palp; 7, Lobes of second maxilla; 8, Rostrum, anterior view; 9, Mandibular gnathal lobe; 10, Second antenna: endopodite with inner lobe of segment 2 damaged, lacking one or more setae; 11, First leg; 12, Fourth leg, abnormal: 13, Fourth leg; 14, Third leg; 15, Second leg; 16, Genital segment, fifth leg, terminal thoracic segment, ventral view. (All figures drawn with aid of camera lucida; all except figure 12 of holotype.) THARYBIDAE FROM THE GULF OF MEXICO 351 monts, about as long as cpphalothorax. Mandib- ular gnathal lobe (pi. 2, fig. 2) witii moro than half the cutting area comprised of molariform teeth, only dorsalmost 3 teeth setiforni. First maxillae (pi. 2, fig. 8) with endopodite two- segmented, not fused with second basal segment; inner lobe 1 with 12 robust spines; lobe less than half total area of appendage. Second maxillae (pi. 1, fig. 7; pi. 2, fig. 9) with lobe 5 bearing two sensoriform filaments, one seta, and one robust falcate spine; terminal segments with total of approximately 5 sensoriform filaments (exact number could not be determined), each somewhat elongated, exceedingly thin-walled, and flexible; filaments apparently unspecialized at apex. Max- illipeds (pi. 1, fig. 1) with first segment bearing three sensoriform filaments and 2 setae on prox- imal lobe. Swimming legs 1^ as figured (pi. 1, figs. 11, 13-15); leg 4 approximately twice as long as abdomen (pi. 1, fig. 3). Leg 1 with each segment of exopodite bearing an external spine; spines unequal in length; endopodite with rounded lateral shoulder bearing numerous spiimles. Ex- ternal exopodal spines of legs 2-4 with sagittiform denticles ornamenting upper and lower margins. Fifth legs minute, symmetrical, uniramous, two- segmented (pi. 1, fig. 2); basal fusion segment shared by left and right rami; each unisegmental ramus with total of 3 processes, an apical cone bearing spinules along lateral and medial margins, a robust medial spine bearing a ventral row of spinules, and a small lateral spur just proximal to apical cone. Adult male: Thorax, cephalic appendages except for first antennae, and swimming legs similar to those in female. First antennae with 21 visible segments (pi. 2, fig. 1). .Abdomen with 5 segments and furcal rami; second segment long- est, anal segment shortest. Fifth legs biramous; first basal segments par- tially fused, second basal segments separate, un- equal in length, left side longer. Right exopodite with first segment elongated, curving mediad; second segment reduced, bearing 1 apical seta. Right endopodite with first segment fused to basal segment, second segment reduced, weakly scler- otized. Left exopodite three-segmented, proximal segment longest; short distal segment bearing terminal and proximal spines, medial face hirsute. Left endopodite two-segmented, shorter distal seg- ment scoop-like. Types (cf. Localities, Material). — All deposited in United States National Museum. Female holotype, No. 99189, selected from material of Alaska, crui.se 4, station 11. Paratypes, Nos. 99190-99191. Remarks. — The new species can be readily identified by the following combination of charac- ters: the denticulated external spines of the swim- ming legs; the terminal lappets of thoracic fusion segment IV-V, which, in lateral view, are some- what triangular and cover most of the genital segment; the conical apex and medial spine of the female fifth legs; the unusual endopodite of the male right fifth leg, consisting of an elongated process fused to the second basal segment and bearing a reduced apical segment. Parundinella manicula, new species Plate 2. fig. 12; Pl. 3. Fins. 1-13 Localities, Material. — Gulf of Me.xico: Lat. 23° 18' X., long. 97° 38' W., (Alaska, cruise 2, .station 11, 7 June 1951, 1 m. depth of plankton tow; two adult females). Measurements. — Techniques as in preceding species. Adclt Female: two specimens, each with similar dimensions; TL 0.81 mm., CAR 3.0:1. Diagnosis. — In general resembles P. spino- denticula, differing chiefly in details of rostrum, thoracic fusion segment IV-V, and swimming legs. Terminal portion of each side of thoracic fusion segment in lateral view produced in short spini- form process barely overlapping genital segment (pl. 3, fig. 5). Rostrum with filaments longer than basal portion (pl. 2, fig. 12; pl. 3, fig. 7). Second maxillae with terminal segment bearing 6 sensori- form filaments; lobe 5 with 2 such filaments; at least 4 maxillary filaments with flagelliform apex (pl. 3, fig. 4). Leg 1 with shoulder of endopodite conical, bearing encircling row of spinules (pl. 3, fig. 12). Legs 2-4 with external spines of e.xopodite bearing fine hair-like denticles; endopodites of legs 2-3 with moderate number of spinules on posterior face (pl. 3, figs. 9-11). Fifth legs with two segments; basal segment reduced; terminal portion of distal segment with 3 outstretched spiniform processes fringed with hairs; medial margin of distal segment with robust spine bearing ventral row of spinules; face of distal segment ornamented with unevenly distributed spinules (pl. 3, fig. 13). Types (cf. Localities). — The two adult females, from which P. manicula is described, were dis- 352 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 2 Plate 2. — Parundinella spinodenlicula, new species. 1, Male first antenna; 2, Female mandibular gnathal lobe; 3, Male thoracic Segments IV-V, lateral view; 4, Male forehead, lateral view; 5, Male, dorsal view; 6, Male fifth legs; 7, Fifth legs of intersexed specimen; may be similar to those of penultimate male copepodite but genital segment in this specimen similar to that of mature female; 8, Female first maxilla, setae of rami, inner lobes 2-3, omitted; 9, Female second maxilla, terminal setae omitted; 10, Female abdomen, terminal thoracic segment, lateral view; II, Female forehead, lateral view. 12, Parundinella manicula, new species; Female forehead, lateral view. (All figures drawn with aid of camera lucida; figures l-II of paratypes, figure 12 of syntype.) THARYBIDAE FROM THE GULF OF MEXICO 353 Plate 3. — Parundinella manicula, new species. Female: 1, Mandibular gnathal lobe; 2, First antenna; 3, Abdomen, dorsal view; 4, Second maxilla; 5, Trunk, lateral view; 6, Genital senment, fifth legs, terminal thoracic segment, ventral view; 7, Rostrum, anterior view; 8, First maxilla; 9, Fourth leg; 10, Third leg; 11, Second leg; 12, First leg; 13, Fifth leg ramal segment. (All figures drawn with aid of camera liicida from syntyjies.) 354 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE sected and the parts mixed during study before their status as new species became clear. They are lierewith designated as syntypes since it is not possible to separate the dissected parts of one specimen from the other. Syntypes deposited in United States National Museum, No. 99192. Remarks. — The new species is readily separated from P. spinodenticula by the terminal spiniform process of thoracic fusion segment IV-V, the three spiniform processes of the fifth legs, and the slender, almost hairlike, denticles ornamenting the external spines of legs 2-4. LITERATURE CITED Brod.ski, K. a. 1950. Calanoida of polar and far eastern seas of the U.S.S.R. Tabl. anal. Fauna U.R.S.S., 35: 1-442, 306 text figs. (In Russian.) Sars, G. O. 1900. Crustacea. Sci. Res. Norwegian North Polar Exped., 1893-1896. I: 1-141, 36 pis. 1902. An account of the Crustacea of Norway, 4. Copepoda Calanoida. Bergen, 171 pp., 108 pis. Scott, T. 1909a. On some new and rare Entomostraca from the Scottish Seas. Ann. Mag. Nat. Hist., ser. 8, 3: 122-130, pis. 2-3. 1909b. On some new and rare Crustacea from Scottish waters. Ann. Mag. Nat. Hist., ser. 8, 4: 31-36, pis. 2-3. Tanaka, O. 1937. Copepods from the deep water of Suruga Bay. Japanese Jour. Zool, 7: 251-271, 19 figs., pis. 17-19. Vervookt, W. 1951. Plankton copepods from the Atlantic sector of the Antarctic. Vehr. Kon. Nederlandse Akad. v. Wetenschappen, Afd. Nat., Tweede Sectie, 47: 1-156, 82 figs. U. S. GOVERNMENT PRINTING OFFICE 1957 O — 414302 NEW CALANOID COPEPODS OF THE FAMILIES AETIDEIDAE, EUCHAETIDAE, AND STEPHIDAE FROM THE GULF OF MEXICO By Abraham Fleminger, Fishery Research Biologist The numerous and widespread zooplankton col- lections made by the United States Fish and AVild- life Service vessel Alaska in the Gulf of Mexico between the years 1951-53 offer the first opportu- nity for a comprehensive account of the epiplank- tonic copepods inhabiting the region. Previous studies on this fauna have been confined to limited coastal areas off Florida, Mississippi, and Ijouisi- ana (Schmitt 195-1). The present investigation of the Alaska collections has already revealed a ricli and varied fauna occurring in the surface waters of the Gulf. Approximately 100 calanoid species have been identified thus far, including nine previously undescribed populations. Three of these new forms, belonging to the families Aeti- deidae, Euchaetidae, and Stephidae, respectively, are described in this report. Family AETIDEIDAE BRADYIDIVS Giesbrecht Bradyidius arnoldi, new species Plate 1, figs. 1-13 LocuUiics, J/aferin?.— Gulf of Mexico: lat. 22°20' N., loiiR. 87°31' W. (Alaska, cruise 4, station 11, 13 January 19.52, surface plankton tow, four males) ; lat. 29°06' N., long. !>:rOO' W. [Alaska, crui.se 8, .station 2. 12 February 1953, 6 in. depth of plankton tow, one female) ; lat. 25°30' N., lonjr 97°06' W. (/l?. male rostrum ; 6. mandibular snathal lobe ; ( a ) stage V male : 1 1) ) adult female : 7. female second let;; H. female first leg: !). female fourth leg. endoixxlite : 10, female tliird leg. endopodite; 11, female genital segment, ventral view ; \2. male fifth legs ; 1.3. female .second maxilla. All figures drawn with aid of camera lucida : figures 3, 4 of holotype, remaining figure.s of paratypes. 358 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 16.7° C. to a high of 23.5° C. The new species occurred only in surface or near-surface night samples taken over the relatively shallow depths of the continental shelf. However, it seems reason- able to suspect that arnoldi is more typically a bathypelagic species, as are its congenitors. Con- sidering the nature of the tows containing aimoldi. vertical migration could readily account for the temporary presence of the species in surface wa- ters. Regarding subsurface conditions at the time and place of these tows, temperatures were found to be virtually identical at all depths. Therefore, from the available data arnoldi occurs at temperatures within the subtropical range whereas its congenitors appear to have been found only in temperatures confined to the cold tem- perate-boreal range, irrespective of latitude. Remarks. — The female of the new species is not readily separated from its congenitors, arnmtns, simiUs, and pacifcit.s. It appears to be charac- terized by (1) flaring of the posterior margins of the thoracic segments, (2) the more dorsal origin and the dorsal tilt of the spiniform processes ex- tending from thoracic fusion segment IV- V, and (8) details of the first pair of legs, including a reduced external spine on Eci, a thick spine on Res, and a broad truncate shoulder with recurved spinules on Ri. Other structures may ultimately prove more valuable for identification, such as the mandibular gnathal lobe and the length relation of the fourth leg and the abdomen. The male is easily distinguished from the other species. The most outstanding character is the relatively great length of the fifth legs, which are approximately twice the length of the abdomen. Furthermore, in the right fifth leg the length of the first segment of the exopodite is considerably greater than that of the endopodite. In ai'noldi the ratio of these segments is approximately 3.6 : 1, whereas in pae/fieu,^ and sii)uli.>i this ratio is about 2.0: 1, as calculated from Brodski (1950: text fig. 65), Vanhoffen (1907: pi. 22, fig. 30), and Sars (1902: pi. 21). In armatus the fifth legs lack endopodites. Despite the limited degree of morphological characterization and the lack of data on variabil- ity, it is proposed that this Gulf of Mexico form be given specific status because (1) the most out- standing character occurs in a sexual structure (i. e., fifth legs of the male) and because (2) it appears to be geographically as well as ecologi- cally separated (i. e., temperature range) from other forms of Bradyidiu.H. This species is named in honor of Edgar L. Ar- nold (I^. S. Fish and Wildlife Service), Fishery Researcli Biologist, in charge of plankton collect- ing operations aboard the Alaska during its sur- vey of the Gulf of Mexico. Family EUCHAETIDAE EUCHAETA Philippi Euchaeta paraconcinna, new species Plate 2, figs. 1-16 Localities, Material. — Gulf of Mexico : material obtained from 16 localities within lats. 22°21'-30''00' N., longs. 82°19'-96°00' W. at depths of 1-10 m. Plankton tows made during period from .Januar.v-.June, in years 19.")l-i53. ( .ila.'ika : cruise 1, station ."JO; cruise 4, stations 2S, .36; cruise 8, stations 5, 6, 30, 33, 34, 36 ; cruise 10, stations 17, 18, 25, 30; cruise 11, stations 16, 23, 24.) Material abundant, in excess of 100 individuals, mostl.v females. Onslow Bay, N. t\, M/V .itlaiitin: station 178 (May 1!».')3, oblique tow 30-1.5 m., one female) ; station 189 (.Tune lit.'JS, oblique tow 20-0 m., one female). Measurements. — All measurements from dorsal view at 32 X magnification; cephalothorax meas- ured along midsagittal plane from apex of frontal organ to posterior margin of intersegmental fold between thoracic fusion segment IV-V and genital segment ; length of abdomen from anterior margin of genital segment to articulation between fourth innermost seta and right furcal ramus. Other- wise, methods are as presented under preceding species. Measurements given include total length (TL) and cephalothorax-abdomen length ratio (CAR). Measurements grouped since variation similar in TL and CAR of specimens from various localities. 1. Adult female: 20 specimens selected at random, TL range 2.79-3.23 nun., mean with standard error 3.04 ±.0.34 mm., standard deviation 0.146 mm., CAR range 2.3-2.6 : 1, mean 2.4: 1. 2. Adult male: live specimens selected at random, TL 2.63 mm., CAR 2.7 : 1 : TL 2.49 mm., CAR 2.8 : 1 ; TL. 2.56 mm., CAR 2.7 : 1 ; TL 2.64 mm.. CAR 2.8 : 1 ; TL 2.63 mm., CAR 2.8: 1. CALANOID COPEPODS FROM GULF OF MEXICO 359 Diagnosis.— A western Atlantic-Gulf of Mexico population closely resembling an Indo-Pacific congenitor, E. cancinna Dana, as described by Giesbrecht (1892) and Scott (1909). Adult female: Differs from concinna chiefly in details of thoracic fusion segment IV-V and genital segment. Posterior terminal margins of thoracic fusion segment IV-V in lateral view symmetrical, rounded, neither side produced posteriad (figs. •2,4). Right side of genital segment with unique acuminate process and smaller, more postero- ventrah digitiform process, both extending laterad (figs. 4, 5, 8). Genital boss prominent, asymmetrical, extending posteroventrad from right side as conspicuous lobiform process (figs. 4. 7, 8) ; in dorsal view process barely visible (fig. 5). Posterior ridge of genital orifice with short lateroventral processes (fig. 7). Anteroventral portion of genital segment with small rounded swelling anterior and left of genital boss (figs. 2, 7,8). Adult male : Differs from concinnn with respect to minute details of left fifth leg. Iveft fifth leg with Re. slightly expanded at dis- tal portion; distal portion bearing short thick haii-s on lateral as well as medial margins, about five serrations on anterior portion of apex, and small cteniform row of hairs on remainder of apex (figs. 9, 10, 13). Base of terminal segment with two processes, somewhat as in concinna (vide Giesbrecht 1892: pi. 16, fig. 19), but proximal process smaller and ensiform, larger distal proc- ess with expanded basal portion and digitiform apex (figs. 9, 10, 13). Types.— X\\ deposited in United States National Museum. Female holotype, Xo. 99197, selected from material of Alaska, cruise 8, station 5; lat. 28°.54' N., long. 90°11' W.: 13 February 1953; 3 m. depth of plankton tow. Paratypes: Xos. 99198-99202. Further description. — Relation of rostrum and frontal organ similar in male and female, forming angle of approximately 90° ; distal portion of ros- trum curving moderately posteriad (fig. 1) . CAR as well as length relations of both abdominal seg- ments and six terminal segments of first antennae similar to those of concinna. calculated from Gies- brechfs (1892) description and figures. Cephalic appendages as in K. marina (Prestan- drea), except for first maxillae. First external lobe of first maxillae witli four elongated setae, one short seta. Endopodite of second maxillae with one spinulated seta as in Sewell's (1947) pro- visional SiJecies Group II. First legs (figs. 15, 16) apparently as in con- cinna. Female second leg with proximal two ex- ternal spines of Rej equal in size, larger than distal spine (fig. 11). I^gs 3-4 with external spines of Re3 of equal length. Left fifth leg of male with rudimentary endopodite. Egg cluster in spawning females with five-eight eggs extending posteroventrad from genital boss in single layer, two eggs wide. Individuals of both sexes often with stalked ciliates fixed to pos- terior thoracic segments and abdomen as in marina but smaller in size. Remarks. — In the new species, the female is readily separated from cancinna by the following : (1) the acuminate process extending laterad from the right side of the genital segment, (2) the form of the genital boss, especially the right lobiform process, which in doi-sal view barely protrudes lieyond the lateral limit of the genital segment, and (3) the rounded terminal portions of thoracic fusion segment IV-V. The new species also resembles E. consimilis Farran (Indo-Pacific region) and E. murrayi Sewell (Arabian Sea), the latter two known only from the female sex. It differs from can.fi milis in the same manner that distinguishes it from con- cinna. In contrast, murrayi could be confusing since its genital segment beare a weak protuber- ance at the midpoint of the right side and the ter- minal portions of thoracic fusion segment IV-V are rounded (Sewell 1947: text fig. 26A-D). However, in SewelFs species the protuberance is not acuminate, the genital boss is on the posterior half of the segment, and the right lobe adjacent to tlie genital orifice extends posteriad (ibid.). The genital segment, in particular the acuminate process extending laterad from its right side, ap- pears to distinguish paraconcinna from all other species of the genus. In the male, the new species is distinquished from cancinna only with difficulty. Apparently, the two are best separated by the distal portion of Re, of the left fifth leg. According to both Giesbrecht (1892: pi. 16, fig. 19) and Scott (1909: pi. 19, fig. 27), this segment in coTi^/ma terminates 360 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE SCALE , 0 5MM 'FIG 5 0 5MM Plate 2. — Eiicliaeta paraconcinna, new species. 1, female cephalon, lateral view: 2, female genital segment, lateral view ; 3, female mandibular gnathal lobe ; 4, female abdomen, lateral view ; .5, female, dorsal view ; 6, male abdomen, fifth legs, lateral view ; 7, female genital segment, ventro-lateral view ; 8, female genital segment, ventral view : 9, male left fifth leg. terminal segments, lateral view ; 10, male left fifth leg, terminal segments, medial view ; 11, female second leg, exopodite ; 12, male second leg ; 13, male left fifth leg, terminal segments, anterior view ; 14, male fifth legs, posterior view ; 1.5, female first leg ; 16, male first leg. All figures drawn with aid of camera lucida ; figure 4 of holot.vpe, remaining figures of paratypes. CALANOID COPEPODS FROM GULF OF MEXICO 361 in one or two rows of lon<; spinules, and tlie medial niarfrin l)ears sliort latenii spiniiles. In all of the available paraconcinna males the apex of this segment bears a row of closely spaced fine hairs and the anterodistal corner is ornamented with five- six subequal serrations; the lateral margin is fur- nished with short hairs. The male of E. plana Mori is also somewhat similar to the new species. From Mori's (1937) description, plana is distinguished from paracon- cinna by the absence of Ri and presence of large spinules lx)rdering Re-j on the left fifth leg, and by the acute angle formed by the rostrum and the frontal organ. I wish to thank Dr. Philip St. John (Brandeis University) for furnishing me with the records and specimens of the new species obtained by the Atlantic from the Cape Hatteras region. Family STEPHIDAE STEPHOS Scott Stephos deichmannae, new species Plate 3, figs. 1-15 Localities, Mutcriul.—GuU of Mexico: lat. 23°18' N., long. 97°37' W. {Alaska, cruise 2, station 11, 7 June 1951, 1 ni. (leptli of plankton tow) ; lat. 29°06' N., long. 9.3°0O' W. (Alaska, crui.se S, station 2, 12 February ID.'ni, 6 m. depth of ijlankton tow ) . Material consisting of 22 adult females, 4 adult males, numerous immature copepodites ; juveniles taken onl.v at cruise 2, station 11. Mea.si/femenfx. — Specimens measured from right lateral view at 100 X, magnification; length of cephalothorax from imaginary line between anteriormost limit of forehead to posterodorsal margin of intersegmental membrane between thoracic fusion segment IV-V and genital seg- ment ; length of abdomen from anterodorsal mar- gin of genital segment to posteriormost limit of right furcal ramus. Otherwise, methods are as presented under the first species described above. 1. Adult fem.\le: 22 sijecimens, TL range 0.62-0.73 mm., mean with standard error <).68±.0O6 uim., standard deviation 0.030 mm., CAR range 2.9-3..5 :1, mean 3.2:1. 2. Adult male: Four specimens: TL 0.61 mm., CAR 2.7 :1 ; TL 0.66 mm., CAR 2.7 :1 ; TL 0.66 mm., CAR 2.4 :1 ; TL 0.62 mm., CAR 2.8 ;1. Diagnosis. — A relatively small species somewhat resembling S. scotti Sars, S. fuU&ni Scott and Scott, and S. gyrans (Giesbrecht). Adult female : Differs from above species in details of thoracic fusion segment IV-V, genital segment, and fifth legs. Terminal portions of thoracic fusion segment IV-V asymmetrical, in dorsal view left side ex- tending more posteriad; in lateral view right side somewhat truncate and lacking spiniform process (fig. 3), left side extending posteroventrad in lobiform process (figs. 4, 7). Genital segment in dorsal view with asymmetri- cal lateral swellings, swelling of left side larger, right side bearing one lateral spine (figs. 1, 7) ; occasionally two lateral spines present. Posterior margin of genital orifice with elongated spine, differing from that in gyrans and fixJtoni; spine relatively straight, midventral, and extending posteriad to approximate midlength of following segment (figs. 1, 3). Fiftli legs similar in form to those of gyrans, difi'ering in spinal ornamentation; second segment with distal border bearing horizontal row of spinules; terminal segment with horizontal row of styliform spinules at approximate midlength, row of scalelike spinides on distal lialf of lateral margin, and fine hairs on medial margin (figs. 5,9). Tiip HAWAII 4 HUGH M SMITH CRUISE 15 a HUGH M SMITH CRUISE 16 15" o HUGH M SMITH CRUISE 18 15° p 140° 140° 140° 140° 1 1 4/ OJ/ o^ '»» 4J/ ft^ 0 36 5° O" 5° l„ aj^ &30 ti i,, »« li^' 43 0 34 o 5 7 10 11 5° 0° 5° t 53 MS Y' Y « ij7 »« is 1 ' 1 is? t3B 4^ 4/£J .„ 1.1,. 1 0 33 0 32 1 O J/ 1 1 o Of o? 8 33 3 )«, r^' "^^ r" »" Vso \<0 \" \„ i« 1.4, 1^' !.« 1 1 1 O 26 0 24 1 O/J 1 on 1 A« AV* 6/5 i/5 i 1 1 1 OI9 02I 1 \( 0° I55» 150* 145° 140° 135° 130° 125° 120° tl5° Fir; 1 RE 2. — Plankton-station positions of Hugh M. Smilh cruises 15 (May-June 1952), 16 (August 1952), and 18 (October- Xovember 1952). duration to a deptli of about 200 meters were eni- ploycd on most cruises. On the Hugh At. Smith cruise 14, multiple-net horizontal hauls were made at 7 stations. Methods of hauling and calcula- tion of sampling depth and amount of water strained have been explained in previous reports (King and Demond, 1953; King and Hida, 1954). TREATMENT OF SAMPLES IN THE LABORATORY First the few organisms with longest dimension greater than 2 cm. were removed from each sam- ple, identified as precisely as possible, and their displacement volume determined. Then the vol- imie of the remainder and bulk of the sample, i. e., those organisms with longest dimension less than 2 cm., was determined. In measuring the dis- placement volume, the plankton was poured in a draining sock of 56XXX grit gauze, to filter ofi" the preserving liquid. When the sample stopped dripping, it was transferred to a graduated cylin- der of appropriate size (usually 50 or 100 ml. ca- pacity). By means of a burette, a known volume of water was added to the drained plankton. The difference between the volume of the plankton plus the added liquid and the volume of liquid alone was recorded as the displacement or wet volume of that portion of the sample. Following the procedure at our laboratory, the volume of all organisms less than 2 cm. in length plus the volume of organisms 2 to 5 cm. in length that might be considered of significant nutritional value ^ were combined to give a single volume ' We consider annelids, chaetognaths, crustaceans, c^ptialopods, and flsb to he of significant nutritional value, and siphonophores, medusae, cteno- I)hores, heteropods, and tunicates as non-nutritious. Bigelow and Sears (ia39) and also Clarke (194(1) considered the crustaceans, chaetognaths. and nioUusks as being of high nutritive value. It Is our judgment, that the heteropod mollusks of the family I'terotracheldae, which are of common occurrence in the plankton of the troi)icftI and subtropical Pacific, do not belong with this group becau.se of t heir watery structure and should be classed with the noti-nutritlous forms. 368 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 30 25' 20 I5' 10' 5 162" 55' 50' *b' 40' J5' JC 5' 023 5' 6' 6"! 02Z 55' 50- S 000 02I 019 °^°. PALMYRA 1 0/5 OI6 0/7 55 50' OS 45' OS «■ »■ O/O 40' 15' ^5 30' JO' 25' 20' tS' ,0' 5 162° 55' 50' as' 40' 35' 3 Figure 3. — Plankton-station positions of Hugh M. Smith cruise 19 (January 1953). ISO" 155° 150" 145' 135" 130" IZS" 120° 115° no° Of' \ ^ 0* v^ p"^ \ 06 X \^ [^ OS ^ \ _\ oto X^ ) \ 0 ^HU < 13 i> < IS ore .„ 0 'e ! oil «?" °r' OlS CHRISTMAS loss o« 3 S3 oir o&> 059 1 0!9 O!- Oil OSS 3JJ MARQUESAS IS o)9 OJ'- Figure 4. — Plankton-station positions of Charles H. Gilbert cruise 15 (February- April 1954). ZOOPLANKTON OF CENTRAL PACIFIC 369 moasureinent for each sample. This figure was divided by the estimated amount of water passing through the net to obtain the vohime of zoo- plankton, as food, per unit of water strained. The contents of 6 samples obtained on cruise 15 of the Hugh M.Smith were counted for the purpose of comparing the catches obtained with 30XXX and .56XXX grit gauze nets. The counting method was essentially the same as that employed by King and Demond (1953). The zooplankton volumes have been examined by simple statistical analysis where it was apparent that a test of significance would aid in interpreting the results. Group comparisons, correlation, re- gression, and analysis of variance have been used, following Snedecor (1946). .Since it was not possible to design the sampling program to isolate sources of variation determined from a priori knowledge, in our analysis of variance we have been limited to a single criterion of classification with subsampling, i. e., a "completely randomized" design (Snedecor, 1946: 240-241). \Miile the method is conveniently adaptable to unequal subsampling, it is less sensitive and less efficient than one based on a more advanced experimental design. Inferences from the analysis are modified occasionally by consideration of the 0.95 fiducial intervals of means based on their individual variances. Although the distribution of the zooplankton volumes is slightly skewed to the left and the means correlated to some degree with the standard deviations, in tests of significance we have used untransformed data. Initially, various lots of data were transformed to logaritlims and em- ployed in statistical tests. The results and con- clusions in each were the same as those reached through an analysis of the imtransformed data. Snedecor (1946: 42, 252) states that little bias is introduced into the analysis of variance and the "t" test, by moderately skewed populations. We assume, therefore, that tlie moderate abnormality in the zooplankton population has little effect on tlie inferences made in this report. EFFECTS OF MESH SIZE ON ZOOPLANKTON CATCH Early in our zooplankton studies we adopted tlie 1 -meter, 30XXX grit gauze net (average aper- ture width 0.65 mm.) as being the best suited for our purposes. Nets of this mesh size retain the 418106 0—57 2 tuna eggs (of about 0.80 mm.) and tuna lar\'ae, the capture of which was one object of our sam- pling,^ but allow almost all phytoplankton to pass through the net; consequently, a relatively "clean" sample of zooplankton is obtained. Some preliminary hauls indicated that, at least on this occasion, nets of 56XXX (aperture width 0.31 mm.) and 72XXX grit gauze (aper- ture width 0.21 mm.) retained some of the larger phytoplankton as well as micro-zooplankton, thus making analysis and sorting of the sample more difficult. A comparison of the catch of Clarke-Bumpus samplers (with 5-inch mouth opening), equipped with 56XXX nets, with the catch of 1 -meter, 30XXX nets indicated that neither the sample volumes nor their variance differed appreciably between the two types of gear (Hida and King, 1955). The greater retention by the finer mesh of small Copepoda, Foraminifera, Appendicularia, and invertebrate eggs was at least partially com- pensated for in the large net of coarser mesh by the less successful avoidance of the net by the larger organisms. To obtain a more precise comparison of the catching abilities of 1 -meter nets of 30XXX and 56XXX grit gauze, 6 special hauls were made on tlie Hugh M. Smith cruise 15. A pair of con- secutive hauls, the first with a 30XXX net and the second with a 56XXX net, were completed at 3 stations on 140° W. longitude: station 45, at 7° S. in the South Equatorial Current, a "poor zone" in respect to zooplankton; station 52 at 1° N., in the "rich zone" of the equatorial diver- gence; and station 60 at 9° N. in the Equatorial Countercurrent, which in the eastern Pacific is also a "rich zone" (p. 377) . All were oblique hauls to an estimated 200 meters' depth and all were taken at night. As to volume, the catch of the 56XXX nets was about 1% to V/i times that of the coarser meshed 30XXX nets (table 9, appendix B). Un- fortunately, one of the samples contained an estimated 30 percent by volume of very small salps which were not separated from the bulk of the sample and therefore complicated the volume comparison. In respect to the number of organ- isms, the finer meshed nets retained 3 to 5 times ■^ Results ure reviewed In unpublished manuscript of W. M. Matsumoto entitlol. DesrriptioD o( Larvae ot Four Species of Tuna and Their Distri- bution in Central Pacific Waters. 370 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 1. — Average number of zooptanklers per unit of wafer strained and percentage composition of the catch obtained loith SOX X X grit gauze {apertures 0.65 mm.) and 56XXX {apertures O.Sl mm.) grit gauze at three stations of Hugh M. Smith cruise IB in June 1952 Organisms Foraminifera ._- Radiolaria Coelcnterata _, Chaetognatha.- Annelida Copepoda Ostracoda Euphausiacea. Amphipoda. — Shrimp Crustacean larvae.— MoUusca Tunicata Fish. Eggs Miscellaneous Total (or sample Station 45 (7°00' S.) Average number per 100 m.' 30XXX 56XXX 214 0 51 489 0 1,072 51 274 17 0 17 43 17 51 77 17 1.257 35 70 1,188 70 9,745 244 349 35 0 175 140 978 0 0 35 2, 390 14, 321 Percent composition 30XXX 56XXX 9.0 0 2.2 20.4 0 44.8 2.2 11.5 0.7 0 0.7 1.8 0.7 2.2 3.2 0,7 100.1 0.2 0.6 8.3 0.5 68.0 1.7 2.4 0.2 0 1.2 1.0 fi.8 0 0 0.2 99.8 Station 62 {VW N.) .\verage number per 100 m.' 30XXX 66XXX 166 104 139 521 0 2,811 17 174 17 17 0 35 330 52 156 0 4,529 1,665 0 166 1,332 33 17, 145 233 333 0 0 67 266 466 67 932 33 22,738 Percent composition 30XXX 56XXX 3.4 2.3 3.1 11.6 0 62.1 0.4 3.8 0.4 0.4 0 0.8 7.3 1.1 3.4 0 100.0 7.3 0 0.7 5.9 0.1 75,4 1.0 1.6 0 0 0.3 1.2 2.0 0.3 4. 1 0. 1 Station 60 (9°00' N.) Average number per 100 ra.' 30XXX 56XXX 16 0 317 269 32 1,473 0 48 0 16 16 111 1,061 0 0 190 3,649 299 0 220 439 140 6,924 439 220 0 0 60 299 778 40 120 120 10, 098 Percent composition 30XXX 66XXX 0.6 0 8.9 7.6 0.9 41.5 0 1.3 0 0.5 0.5 3.1 29.9 0 0 5.4 100.: 3.0 0 2.2 4.3 1.4 68.6 4.3 2.2 0 0 0.6 3.0 7.7 0.4 1.2 1.2 100.1 as many plankters as the coarser meshed nets. Table 1 gives for each sample the average number per unit (100 m?) of water strained and the per- centage composition for the major constituents. It appears that the greatest difference in the catch of the 2 nets is in the larger numbers of foraminifers and copepods retained by the 56XXX net. The results show a marked difference between the 2 nets in average size (volume) of individual organisms in the catch (table 2) ; plankters in the catch of the 30XXX net were about 3 times as large as those taken by the 56XXX net, primarily because of the difference in catch of small copepods such as the microcalanoids and cyclopoids. As the result of an increased catch of the larger zoo- plankton forms (coelenterates, salps) and fewer of the smaller forms (foraminifers, chaetognaths) , both nets yielded larger organisms, on the average, at the northernmost station (station 60). Table 2. — Average size {i. e., volume of catch divided by the number of organisms) of zooplankters captured in 30XXX and 56XXX grit gauze nets at three stations of Hugh M. Smith cruise 15 in June 1952 Item Station 46 (7°00' S.) Station 52 (1°00' N.) Station 60 (9°00' N.) 30XXX 56XXX 30XXX 56XXX 30 XXX 56XXX Average number of organisms per 100 m.> 2,390 3.22 13.5 14, 321 5.07 3.5 4.629 6.98 15.4 22, 738 12,21 .■5,4 3,549 8,04 22,7 10, 098 9,09 9,n Volume of catch, cc. per 100 m.3.. Average size of or- ganism, cc.xIO-*. It is obvious that these 2 nets of different mesh exercised a strong size-selection in sampling the zooplankton community. The question as to which net-size yielded the most reliable measure of abundance of zooplankton as potential fish food cannot be decided from the few data presented here. It is generally known that no one net or other sampling device will quantitatively sample the entire zooplankton community, and therefore the investigator must choose the method and gear that in his opinion will contribute the most toward his particular objective. Our objectives, to obtain a representative sample of the larger zooplankton forms and to retain all tuna eggs and larvae with a minimum of mesh-clogging, were realized, we be- lieve, with the use of 1-meter nets of 30XXX grit gauze. VARIATION IN CATCH WITH SAMPLING DEPTH On 7 stations of Hugh M. Smith cruise 14 in February 1952, horizontal hauls were made simul- taneously at 3 levels with open 1-meter nets. The hauls were of 1-hour duration; the nets were low- ered and raised at the start and end of the haul as rapidly as possible to minimize contamination in tlie intermediate and deep samples. All 7 sta- tions were off Canton Island in the South Equa- torial Current between 2°41' S. and 2°45' S. lati- tude at about 172° W. longitude, and were oc- cupied consecutively between 1315 and 0338 hours of February 9-10, 1952. Although the primary ZOOPLANKTON OF CKNTRAL PACIFIC 371 purpose of the sampling was to investigate the abundance and vertical distribution of tuna eggs and larvae in this area, some information was ob- tained on the variation in zooplankton volumes witli deptli and with time over a 14-hour i)eriod. According to the results shown in table 8B (appendix B) and figure 5, at each of the 7 stations (SI to S7) the largest volume of zooplankton was taken in the surface net. At station S5, hyperid amphipods were apparently swarming at the sur- face and resulted in an unusually large catch. O 40 O o 1 1 III 1 1 ' lo 1 1 - - o surface' « INTERMEDIATE • DEEP - a t o o o • o t t - - • • t • a 1 - i 1 1 1 1 1 1 1 1 1 1 1400 1600 1800 2000 2200 0000 0200 HOUR (ZT) FiouRE 5. — Variation in zooplankton volumes with hour of hauling, as obtained with horizontal hauls at the 3 depths: surface, intermediate (10.5-120 meters), and deep (210-240 meters): Hugh M. Smith crviise 14 February 9-10, 1952. The other samples were of mixed composition, typical of tiiis area. At four of the 7 stations the intermediate net, fishing just above the thermocline at a depth of 105 to 120 meters, caught more than the deep net fishing below the thermocline at 210 to 240 meters. These results arc generally similar to those obtained with Clarke-Bumpus samplers (employing 56XXX nets) on a series of 150 stations extending from 12° N. to 7° S. latitude along 150° N. longitude (Hida and King, 1955). The surface samples of the latter series averaged 60.7 cc./l,000m.^ the intermediate samples (from within the thermo- cline) averaged 29.2, and the deep samples (at 200-300 meters) averaged 16.6. Although this sampling period of about 14 hours is not adequate to demonstrate the diurnal cycle, there is some evidence of an "evening rise" between 1400 and 2000 hours, followed by a drop in catch at the intermediate and deep levels and then what is possibly the start of a "morning rise" at these levels. The parallel variation (r= 0.837, P<0.05) in volume of catch at the intermediate and deep levels is of interest and suggests that the zooplankton at these depths was behaving differently from that at the surface in response to varying illumination. ADJUSTMENT FOR DIURNAL VARIATION The hour of hauling provides an important source of knowledge of the variation in quantita- tive measurements of zooplankton abundance. Presumably, the difference between day and night hauls is due either to an augmentation in the upper strata of water by upward migration of the plankton at night or to a reduction in catch in the daytime owing to the greater ability of the plank- ton to dodge the net when there is light, or to a combination of the two. In some areas of the tropical Pacific the day-night difference is suffi- ciently great, if no correction is applied, to obscure the geographical and seasonal differences which are of primary interest in this study. Significant differences in zooplankton volume, associated with latitude, were observed among the night samples and not among the day samples on cruises 5 and 8 of the Hugh M. Smith in the central equatorial Pacific (King and Demond, 1953). In Hawaiian waters the volumes of night hauls have averaged about l}i times the volumes of da^' hauls (King and Hida, 1954). In the present instance during the 6 cruises in the equa- torial region on which sampling was conducted around the clock, niglit hauls yielded volumes about Uptimes the volumeof thedayhauls(table3), while the twilight hauls were intermediate in average volume.^ Some of the variation among cruises, as shown in table 3, may be due to differ- ences in season, longitude, and range of latitude sampled. Variations in the night/day ratio asso- ciated with the current system will be discussed later. ' Knr purposes of this comparison we deslRiiated the twiU^ht hours as 0430 to 073(1 and Ifi30 to 1930. which iHTioti.'i include suiiri.se and sunset and the heginningnnd end of twilight asspecided by the .\merican Nautical .\linanaci the day period was thereby limited to 0730-1630 and the night to 1930-04,30 hours. 372 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 3. — Differences in the average volumes of day, night, and twilight hauls and in the night/day ratios for six cruises of the Hugh M. Smith in the equatorial Pacific Cruise period Num- ber of sam- ples Zooplankton— mean volume, cc./lOOO m.J Night/ Cruise No. Night hauls Day hauls Twi- light hauls Total day ratio 2... i:::::::-:::: 11 14 Jan.-Feb. 1950... June-Aug. 1950.. Jan.-Mar. 1951... Sept.-Oct. 1951... Jan.-Mar. 1952... M ay-June 1952... ■24 51 87 '23 47 60 45.5 40.3 30.7 41.7 30.7 50.6 24.0 27.9 18.2 32.1 22.8 31.6 26.6 37.1 23.9 33 2 25.3 36.9 133.6 34.7 23 9 '36.0 24.6 39.9 1.90 1.44 1.69 1.30 15 . 1.60 39.9 26.1 30.5 32.1 ' Sections A and C only (King and Demond, 1953, table 1). ' Northbound section only. On tlie majority of cruises sampling was con- ducted around the clock so that there were about equal numbers of night and day stations. Under this S5'stem there rarely were more than two day stations or two night stations occupied consecu- tively. On certain cruises, however, such as cruise 18 of the Hugh M. Smith, and cruise 15 of the Charles H. Gilbert, hauls were made at about the same hour throughout the cruise; e. g., on cruise 18 all hauls were made near midnight, on cruise 15 between 1900 and 2000. The resulting data are most useful for within-cruise comparisons, but some modification is necessary if they are to be compared or combined with the results of the other cruises. An adjustment to remove the effect of diurnal change in zooplankton catch was described by King and Hida (1954). The method is based on the similaritj^ of diurnal variation in zooplankton abundance to the curve of the sine function when midnight is equated to the angle whose sine is + 1.0. The zooplankton volumes are increased or lowered dependent on the hour of hauling and adjusted to 0600 or 1800 hours, when the sine = 0. Since illumination is the major factor controlling the diurnal migration of plankton (Kikuchi, 1930; Gushing, 1951), solar time is used in the calcula- tions. The method as originally designed was applied to zooplankton volumes from the Hawaiian Islands area, where the geographical variation was slight and the night/day ratio rather uniform from cruise to cruise. On the long sections crossing the Equator we found considerable variation in the night/day ratio associated with latitude and the current system (p. 380) . and the geographical varia- tion is much greater than in the Hawaiian area. Although these factors lessen the accuracy and effectiveness of the method, it still provides a reasonably good correction for day-night differ- ences as judged by the significance of the "t" values and the night/day ratios for the adjusted volumes (table 4), and has therefore been applied to the equatorial data. Table 4. — Regression coefficients (b), "t" values and probability values for the sine transformation method of adjustment for 5 cruises of the Hugh M. Smith in the equatorial Pacific. I A comparison of the night/day ratios for the zooplankton volumes before and after adjustment indicates the general validity of the method). Number of samples b t P Night/day ratios Cruise No. Before adjust- ment After adjust- ment 5 51 87 23 47 60 0.0941 .1534 .0842 .1340 .1186 2.077 5.1M6 1.251 3.472 3.228 <0.05 <0.001 >0.05 <0.01 <0.01 1.44 1.69 1.30 1.34 1.60 8 11 0.96 0 95 14 15 1 03 Throughout this report we have employed the adjusted volumes in examining the variation in zooplankton abundance with respect to special features of the current system, with longitude, and with season. The data from cruises 5, 8, 11, 14, and 15 of the Hugh M. Smith were ad- justed by individual cruise. A pooled regression coefficient (b = 0.1 248) calculated from the com- bined data of these 5 cruises that covered large areas of the equatorial Pacific during which the stations were visited consecutively regardless of the time of day or night, was used in adjusting the volumes of cruises 2, 7, 9, 16, 18, and 19 of the Hugh M. Smith and of cruise 1 5 of the Charles H. Gilbert. On the latter cruises, sampling was not conducted around the clock, or there were too few data to be adjusted by individual cruises. Unadjusted volumes for the Hugh M. Smith cruises 2, 5, 7, and 8 have previously been pub- lished (King and Demond, 1953). The adjusted volumes for these cruises are provided herewith in table 14 (appendix B). DESCRIPTION OF THE ENVIRONMENT The general pattern of the Pacific equatorial current system has been described by Sverdrup and others (1942:708-712). In brief, the major surface currents of this region are the North and South Equatorial Currents flowing toward the west, and the eastward flowing Equatorial Coun- ZOOPLANKTON OF CENTRAL PACIFIC 373 tercurrent sandwiched in between. Although the boundaries of the Countercurrent may vary meridionally witli longitude and season, its south- ern and northern boundaries ordinarily occur near 5° X. and 10° X. latitude in the mid-Pacific. The South Equatorial Current is therefore on both sides of the Equator while the Xorth Equa- torial Current is confined entirely to the Xorthern Hemisphere. As previously stated, the Equator is the site of upwelling resulting from divergence of the surface waters. It is also the location of the newly discovered subsurface Equatorial Un- dercurrent flowing to the eastward (Cromwell and others, 1954). The region between the Equator and the southern boundary of the Coun- tercurrent is a zone of convergence. Under certain conditions, as described by Cromwell (1953) and Cromwell and Reid (1956), a sharply defined convergence or "front" * may be formed in the South Equatorial Current between the Equator and the southern boundary of the Countercurrent. The motion of these currents is either directly or indirectly the result of wind stress on the surface of the ocean, and it is logical that varia- tions in these currents are a reflection of variations in the prevailing winds or "trades." The Climatic Charts of the Oceans (U. S. Weather Bureau, 1938), based on averages of 50 years of observations, provide a general picture of the velocity and direction of prevailing winds in the equatorial Pacific. Average wind condi- tions for the months of March and August, which represent the extremes of the seasonal variation, are shown in figure 6. In the region of our zooplankton studies (110° W. to 180° long.), the charts show longitudinal and latitudinal as well as seasonal variations in tlie tradewinds. In an east-west direction along the Ecjuator there is a general decrease in intensity from Beaufort force 3 and 4 east of 160° W. longitude to force 1 and 2 west of that merid- ian. Between 100° W. and 140° W. the south- east trades are dominant O60 percent constant) along the Equator in all months of the year. Between 140° W. and 160° W. they are dominant from May to January; between 160° W. and 180° they are only of importance from Julj' to s Defined hy Cromwell (\9^) as "a pronounced oceanic convergence," and hy Cromwell and Reid (IQ.Vi^ as "a narrow band along the sea surface across which the density change,s abruptly" and "the surface temperature gradient Is often of the order of degrees per l/10() mile." October. At other months of the year the re- sultant wind at the Equator is from the east between 140° W. and 160 W., and from the north- east between 160° W. and 180°. Xorth of the Equator in the region of the Countercurrent, the period of strongest winds is from December to May when the northeast trades prevail. At other months of the year the winds are light and variable; in the eastern part of the region, from 120° W. to 140° W., the southeast trades exert a slight influence. Longi- tudinally the northeast trades reach their highest velocity between 140° W. and 170° W. longitude. According to the wind drift model of Cromwell (1953), convergence and sinking of the surface waters will occur to the north of the Equator in the South Equatorial Current during a south or southeast wind, and conversely to the south of the Equator under the influence of a north or northeast wind. A pronounced convergence or front has been encountered south of the Equator on only one of the many POFI hydrographic and fishing surveys. This is not surprising in view of the slight influence of north or northeast winds at the Equator in the eastern and central Pacific. Evidence of convergence north of the Equator has been observed, though, on several occasions. When the generally westward current near the Equator has a northward component, as during southeast trade winds, we anticipate that the zone of greatest zooplankton abundance will be north of the Equator, due both to the physical displacement of the organisms and the time lag in their development, with the peak of abundance occurring somewhere in the zone of convergence between the region of upwelling and the southern boundary of the Countercurrent. With a pre- vailing northeast wind the zooplankton maximum should theoretically occur to the south of the Equator and, with an east wind, more nearly on the Equator or with a double peak.* In summary, then, as a result of the direction and relative high velocity of the trade winds, we expect to find larger concentrations of zooplankton « Murphy and Shomura (IQ.Wb) have shown that the latitudinal variation in the zone of best-yellowfln-catch also coincides with differences in the pre- vailing winds. Fishing sections along 120° W. and 13n^ W. longitude, asso- ciated with southeast winds. Indicated the [>eak abundance to be north of the Equator; the catch along l.'iS^ W. and lfi9* W., associated with variable winds, showed the peak abundance to be nearly centered on the Equator, while a section along 180°, associated with northeast winds. Indicated the peak of yellowfin abundance to be displaced to the south. 374 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 140° 140° 150° 160° 170° 180° 170° 160° 150° 140° 130° 120° 110° 100° 90° 80° Figure 6. — Resultant direction and force of surface winds in the central equatorial Pacific during March (A), a month of light and variable winds on the Equator but strong northeast trades in the region of the Countercurrent, and during August (B), a month of strong southeast trades on the Equator and light winds along the Counter-current. [From Atlas of Climatic Charts of the Oceans, U. S. Weather Bureau, 1938. Arrows show resultant wind direction computed for each 5-degree unit area. Shadings indicate gradations of resultant velocities scaled in Beaufort units of wind force.] east of 160° W. tlian to the west of that longitude and also more to the north of the Equator than to the south. A narrow convergent zone which theoretically should concentrate the zooplankton is most likely to occur east of 160° W., and par- ticularly east of 140° W., because of the promi- nence of the southeast trades in that region. ZOOPLANKTON AND THE CURRENT SYSTEM Within the range of latitude sampled, there are certain natural subdivisions of the environment which may be established on the basis of the current structure. These may be defined as follows: (1) the North Equatorial Current from ZOOPLANKTON OF CENTRAL PACIFIC 375 2° 0° Z° S — LATITUDE — N Fku'rk 7. — Vertical tcmporature section (adapted from C'roniwell and Austin, 1054, figure 28) based on bathythermograph observations along 172° W. longitude, made February 27-March 12, li)51, on Hugh M. Smith cruise 8, showing the boundaries of the 6 areas u.sed in this study in relating variations in zooplankton abundance to particular features of the equatorial current systems. tlio iiortlipni limit of our sampling to the northern boiiiularv of the Countercurrent, a region of relatively shallow thermocliiie; (2) the Counter- eurrent with its boundaries being determined at the time of each crossing from vertical temperature sections, a region with shallow thermodine to the north, deepening to the south; (3) a zone of convergence in the South Equatorial Current extending (according to our definition) from the soutiiern boundary of the Countercurrent to 1^° X. latitude, a region of deep thermodine; (4) a zone of divergence and upwelling in the South Equatorial Current along the Equator from U$° X. to 1)2° S. latitude, evidenced by a doming of the isotherms, a reduction in surface temperatiu-e, and an increase in surface inorganic phosphate; (5) the South Equatorial Current from 1K° S. to 5° S. latitude, a region of deep thermodine; and (6) the South E{|uatorial Current from 5° S. latitude to the southern limit of our sampling (about 14° S.), a region of shoaling thermodine to the south. Figure 7 shows the boimdaries of these six areas superimposed on a vertical temperature section based on bathythermograph observations along 172° W. longitude. When the zooplankton volumes, adjusted for the day-niglit variation but disregarding differences related to longitude and season, are combined according to these natural divisions of tlie current system, we obtain the distril)ution siiown in figure' 8, with the greatest concentration of zooplankton occurring at the Equator (1K° X". to 1K° S.) in the region of divergence. Average volumes for the areas just north of the Ecjuator, i. e., the convergent zone and the Countercurrent were considerably higher than those for the cor- responding areas south of the Equator. The X'orth Equatorial Current and the South Equa- torial Current at the southern extent of our sampling were equally poor in zooplankton. From an analvsis of variance we conclude that the 2 o < 20 SEC (331 ISECIDIV|C0NV| CO ' "I — r (661 I (59) I (87) NEC (65) 14° 12° 10° 8° I I _Li_ go go iQO igo 140 150 igo S — LATITUDE — N FiGiRE 8. — Variations with the current system in zooplankton volumes (adjusted) for longitudes 120° W. to 180°, with the limits of the 0.!».5 fiducial interval shown for each mean. (The number of samples for each area is indicated in parentheses]. 376 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE lO o O 60 O o o_ z" o t- z < _J Q. O O M 40 - 20 1 1 1 1 1 1 "RONT 1 1 1 1 1 1 1 1 172" W. LONGITUDE 1 ■1 • - • • • • - 1 1 1 1 1 1 • 1 • • • • 1 1 1 1 1 1 1 • 1 o o o o o z < _l Q. O O M 60 40 - 20 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 FRONT io/-\o tAf 1 i-ML 1 /^ 1 -I- 1 1 i-i r- 1 I ItW ¥». UWIXUI 1 UUC • - • • • • - • • • 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 - S 8° Qo 2° 4° 6° 8° 10° 12° 14° 16° 18° 20° 22° 24° N LATITUDE Figure 9. — Relation of an oceanographic front to the distribution of zooplankton (adjusted volumes) as demonstrated by 3 series of stations along 172°, 158°, and 120° \V. longitude of Hugh M. Smith cruises 2, 5, and 18, respectively. ZOOPLANKTON OF CENTRAL PACITIC 377 differences among subdivisions of the current system are highly significant (F = 4.08 P<0.01). The degree of overlap in the 0.95 fiducial intervals of the means is shown in figure 8. The asymmetrical distribution of the zooplank- ton in respect to the Equator results, we believe, from the prevalence of the southeast trades during most of the year. The occurrence of the zooplank- ton peak at the site of the divergence in an area of newly upwelled water, rather than in "older" water to the north or south of the Equator, is somewhat surprising and may be evidence that, on the average, the northward and southward components in the westerly surface current at the Equator are slight compared with the rate of development of zooplankton. The distribution of zooplankton around well- marked fronts suggests a causal relation. Three well defined fronts have been observed on POFI cruises in the convergent or transition zone to the north of the Equator. On all three occasions strong southeast winds were experienced between the Equator and the region of the front. The latitudinal variation in zooplankton abundance as related to these fronts is illustrated in figure 9 for the three series of stations along 120° W., 158° W., and 172° W. longitude. On each of the three meridians the zooplankton abundance peaks south of the front and drops off sharply to the north. VARIATION WITH LONGITUDE To examine the east-west variation in zooplank- ton abundance in respect to divisions of the cur- rent system as previously defined, the adjusted volumes were first combined by 10-degree inter- vals of longitude disregarding season. Because of the shortage of data for some subdivisions, longitudes 170°\V. and 180°; were then combined as were 150° W. and 160° W.; 120° W. was grouped with 130° W. and 140° W. The lati- tudinal zooplankton distributions in the two western regions, 150° \V.-160° W. and 170° W.- 180°, are essentially alike (fig. 10) with peak abundance occurring at the equatorial divergence, and with the convergent zone next in importance. In the eastern region (120° W.-140° W.), we find the highest average volume in the Countercurrent with the area of divergence second in rank. Only in the Countercurrent are there significant differences among longitudes, as indicated by the 418106 O — 57 3 o 120° W- I40°W Q 150' W- 160' W A 170" W- -ISO" 60 50 - o o 10 - -T 1 I Ml Mill 1 1 . SEC 'sec]div]conV| 1 1 1 1 II'! CC j NEC 1 - 1 1 1 - - i! 1 1 - 111 H 1 ' ' i II ti .1 ~ 1 1 1 1 1 1 1 1 1 1 1 II 1 °^ 1 1 1 1 1 ^ 14° 12° 10° 8° 6° 4° 2° 0° 2° 4° 6° S — LATITUDE-N 8° 10° 12° 14° 16° 18° ' Only 1 sample. Figure 10. — Longitudinal and latitudinal variations in the distribution of zooplankton volumes (adjusted) with the data segregated into three longitudinal groups and in accordance with natural features of the current system. The limits of the 0.95 fiducial interval are indicated for each mean. lack of overlap in the 0.95 fiducial intervals of the means. And it is only in the eastern Pacific that production in the Countercurrent equals that of the divergent zone. WhUe these apparent rela- tions may change with further sampling and more complete seasonal coverage, we believe the results are logical in view of longitudinal variations in thermocline depth and winds. As previously mentioned, toward the northern boundary of the Countercurrent in the eastern and central Pacific, there is a doming in the isotherms (figs. 7 and 17) and the thermocline is relatively shallow; consequently high-phosphate water is within the photosynthetic zone and within the reach of wind-induced turbulence. To the westward the thermocline deepens (Sverdrup and others, 1942: 708), reducing the likelihood of such enrichment. Figure 11 shows the relation of the average zooplankton volumes for the range of latitude 8° N. to 11° N., and the depth of the 70° isotherm for four meridians (140° W., 150° W., 160° W., and 170° W. long.). The chosen range of latitude (8° N.-ll° N.) includes the doming in the isotherms at the northern boundary of the Countercurrent and represents the zotie of most shallow thermocline in the tropical Pacific. The results indicate a highly significant 378 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE inverse correlation (r =—0.688, P 0.05). Despite the statistical evidence that the differences among longitudes are not significant (with the ex- ception of that between 140° W. and 150° W. in the Countercurrent, as indicated by lack of over- lap of the 0.95 fiducial intervals of the means), the general picture of decreasing zooplankton abundance from 140° W. to 180° parallels certain changes in the environment. Along the Equator, with decrease in wind velocity from east to west, we may expect a corresponding decrease in up- welling and enrichment of the surface waters; in the region of the Countercurrent, the possibility of enrichment through wind-induced turbulence decreases from east to west with the deepening in thermocline. DIFFERENCES AMONG SEASONS AND YEARS It was pointed out by King and Demond (1953) that the zooplankton volumes taken in January and February in the equatorial Pacific averaged significantly less than those obtained in June and ZOOPLANKTON OF CENTRAL PACIFIC 379 July. With further sampHng the results showed a rather uniform level of abundance for the 9- month period April through December (King, 1954) with a reduction from January to March. Figure 13 shows the results of our sampling to date for the Countercurrent and for the equatorial region of the South Equatorial Current with the volumes combined, irrespective of longitude, into four quarterly periods of 3 months each. In the Countercurrent the highest average volume was obtained for the second quarter, April, May, and June, which occurs during the period when north- east trade winds are predominant at those lati- tudes. Along the Equator the last six months of the year, the period of strong southeast trade winds (Crowe, 1952) averaged higher than for the first two quarters. From an analysis of variance, however, we conclude that the differences among seasons are not statistically significant (F=1.87, P>0.05), but again differences between subdi- visions of the current system are highly significant (F = 8.38, P<0.01). If we segregate the data geographically ac- cording to divisions of the current system and seasonally into two 6-month periods, i. e., (1) January to June, which includes roughly the time of lightest winds along the Equator in the central Pacific, and (2) July to December, the period of 0 COUNTER CURRENT 60 D SOUTH EQUATORIAL CURRENT I (60) (74) (24) (55) 1 (17) (63) 1 (16) (20) 50 _ m s o H^ - - ^ 1 o r i o 1 1 ^-30 - - o Z { < 20 J 0. o o ■^ 10 - - 0 1 1 1 JAN-FEB-MAR APR-MAY-JUN JUL-AUG-SEP OCT- NOV- DEC Figure IS. — Seasonal variations in zooplankton volumes (adjusted) for the Countercurrent with boundaries at about 5° N. and 10° N. latitude, and for the South Equatorial Current from about 5° N. to 5° S. latitude; longitudes 120° \V. to 180° combined; the limits of the 0.95 fiducial interval are shown for each mean. [The number of samples for each season and each subdivision of the current system is indicated in parentheses.] • — — o JANUARY - JUNE ■-o JULY - DECEMBER O O 30 O ^ "1 — 1 — r NEC -LU I I I \ \ L TT "1 — I — I — r I \\ _1_L '-1-0-; 14. 12= 10= 8° 6° 4° 2° 0° 2° 4° 6° 8° 10° 12° 14° 16° 18° S — LATITUDE— N Figure 14. — Variation with the current system in (A) zooplankton volumes (adjusted) and (B) yellowfin long- line catch for the two 6-month periods, January-June, a period with northeast or light and variable winds, and July-December, a period of prevailing southeast trade winds (in the central equatorial Pacific). strong southeast trades, we find an interesting difference (fig. 14A). In both groups, the peak abundance in zooplankton occurred at the Equa- tor, but during the latter half of the year under the influence of the southeast trades the abun- dance continued high into the convergent zone. When the data from this zone are examined by means of the "t" test we find, however, that the mean for January-June is not significantly differ- ent (P>0.05) from the mean for July-December. Our data indicate that along the Equator there was considerable difference in zooplankton abundance among years. Figure 15 presents aver- age zooplankton volumes for the Countercurrent and the equatorial region of the South Equatorial Current which were visited repeatedly from 1950 to 1954. From an analysis of variance we may conclude that differences between the two sub- divisions of the current systems are highly signifi- cant (F=27.60, P<0.01), differences among years are also highly significant (F=7.33, P<0.01), but 380 FISHERY BULLETIN OP THE FISH AND WILDLIFE SERVICE O I50°W LONGITUDE □ I60°W LONGITUDE 4 I70°W LONGITUDE S 80 - O o o 60 20 - - (A) cour ITER CUf ?RENT - D o - D °. u - 100 = 80 (B) SOUTH EQUATORIAL CURRENT o o D 0.05). It is obvious that the differ- ences among years are derived principally from variations within the South Equatorial Current. The general agreement among longitudes is in line with results from the previous tests. Along the Equator the volumes for longitudes 160° W. and 170° W. averaged considerably higher in 1950 than in subsequent years. On longitude 150° W., August-September 1951 provid- ed much higher volumes than January and August 1952. In both the Countercurrent and the South Equatorial Current there is some indication of a rise in 1954. Possibly related changes are evidenced in other environmental factors. From a study of the rather sparse rainfall records available for the central equatorial Pacific,' Austin concludes that • In unpublished manuscript entitled, Review of Central Equatorial Pacific Oceanography, 1950-52. in the year 1950 the precipitation at Fanning Island (located at about 4° N. latitude, 159° W. longitude) was unusually low and Infers that southeast winds predominated throughout the year." On the other hand, judging by the climato- logical summaries, the years 1951, 1952, and 1953 may be considered as normal years in respect to rainfall and also, by inference, in respect to winds, i. e., with northeast and variable winds during the first 6 months and east to southeast winds during the latter half of the year. Therefore the year of highest apparent productivity in the zone of interest coincided with the year in which the southeast trades appear to have been unusually vigorous, thus perhaps causing the upwelling mechanism to operate more energetically. DIURNAL VARIATION AND THE CURRENT SYSTEM Many physical and biotic conditions influence the vertical movement of planktonic animals (Kikuchi, 1930; Cushing, 1951). The diurnal variation which we have observed in the zooplank- ton catch from 200-meter oblique hauls probably results from a combination of factors which in- clude: (1) vertical migration of the organisms in response to changes in illumination, and (2) their increased ability to dodge the net during daylight hours. In Hawaiian waters and in the central equatorial Pacific, night hauls yield catches about IK times the volume of day hauls (table 1). When the average volumes of night, day, and twilight hauls are segregated with respect to subdivisions of the current system, as in figure 16, we find a marked variation in the night/day ratio from north to south. In the North Equa- torial Current, Countercurrent, and convergent zone the ratios range from 1.31 to 1.43, while in the divergent zone and the South Equatorial Current to the southward the ratios are much higher, ranging from 1.76 to 1.94. This trend appears consistently in the individual cruises. The North Equatorial Current, an area of relatively shallow thermocline within the latitudes considered (fig. 7), has a very low night/day ratio; the convergent zone, with a deep thermocline. also has a low ratio, while the South Equatorial Current south of 5° S. latitude, which is an area « A study of the records had shown that a period of "doldrums" or nortl east winds bring heavy rains to the northern Line Islands. ZOOPLAXKTON OF CENTRAL PACIFIC 381 2 8 40 O o 24 9 16 (A) 1 — I — 1 — r 1 — I — I — r !/ II I II NT I I lo — o NIGHT SAMPLES I I J0--—0DAY SAMPLES I /°\ I 1° o TWILIGHT SAMPLES >•■;<> -I— - o- ,^ J \ UJ LL ■>, I I — I:- Mil \ L NEC 2° 0° 2° 4° 6° -LATITUDE— N FiGiRE 16. — Variations with the current system in (A) average volumes of night, day, and twilight hauls and in (B) the ratio of night to day zooplankton volumes. of moderate to deep thermocline, has a high ratio. We must conclude, therefore, that neither thermo- chne depth in itself nor the nig;ht/day ratio appears to he related to the general level of zooplankton abundance. Both high and low ratios are found in areas of poor zooplankton catch. We must leave this problem for the present without an explanation. SHORT-TERM VARIATIONS Two cruises of the Hiiijh M. Smith (cruise 11 and 15) crossing the ecjuatorial currents on 150° and 140° W. longitude provide information on temporal changes in zooplankton volume and distribution as related to changes in the physical environment. On cruise 11 in August-October 1951, the north- bound leg (stations 28-50) was worked immedi- ately after the southbound leg (stations 1-28). During the time interval (approximately 6 days) between crossings of tlic Erjuator, the wind (SE.) decreased from ai)out 20 knots to about 12 knots. As indicated by tlie change in positions of the 80° F. isoth(>i-m (fig. 17), the zone of mixing nt the Equator, i. e., the zone of cool, newly upwelled water, shifted to the south and narrowed in width during the 6-day interval. On the first leg the zooplankton maximum occurred at 1° X. latitude; on the second leg it occurred at 0° with a second peak of almost equal abundance at 2° S. latitude. These changes would seem to be evidence that during this 6-day period there was a shift in zooplankton distribution correlated with changes in zonal How. In the region of the Countercurrent during cruise 11, there was little change in winds within the interval (about 32 days) between sections, but there was a marked increase in rate of flow, as indicated by the broadening of the Counter- current and steepening of the thermocline. These changes in the current were accompanied by a significant change in the zooplankton distribution (fig. 17). At the time of the first crossing there was little variation among stations within the Countercurrent; at the second crossing, following an increase in the current velocity from 45 to 80 cm. /sec, there was a marked gradient in zooplank- ton concentration with the larger volumes being taken in the area of shallow thermocline at the northern boundary of the Countercurrent. Additional information on time changes in the environment and the distribution of zooplankton along a particular meridian was obtained during May 195.3 on Hugh M. Smith cruise 15 when 4 consecutive hydrographic and plankton sections were completed along 140° W. longitude with sampling from 9° X. to 7° S. latitude. The time Fici-RE 17. — Soutli and northbound sections of Hugh .U. Smith crui.se 11 in August -October Hlol, showing associ- ated changes in zooplankton distribution (adjusted volumes) and temperature along 150° W. longitude. ITemperature sections adapted from .Austin 1954a. ] 382 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE interval between the first and fourth crossings of the Equator was 16 days and from the start of sec- tion 1 to the end of section 4 was 23 days. Austin (1954b) summarizes the hydrographic changes during this period as follows : 1. The slope of the isotherms associated with the Countercurrent is greater in the fourth than in the first leg, suggesting an increased easterly flow of the Countercurrent. This was substantiated in the calculated velocities, 60 cm. /sec. on the first leg and 120 cm. /sec. on the fourth leg. 2. The 80° isotherm-surface intercepts for the fourth section have moved to the north and south of those for the first section. 3. The 70° isotherm shows considerably more doming at the Equator in the fourth section. 4. Between the first and fourth sections there is a generally southerly shift in selected isohalines. There is a similar change in the slope of sigma-t isopleths in the region of the Countercurrent. Selected sigma-t surfaces show a general displace- ment to the south when comparing the first and fourth sections. 5. The most apparent change in the phosphate sections is the deepening of the 0.8 and 2.0 ng at/L isopleths in the region of the Countercurrent which is associated with the suggested change in flow, and the change in configuration of the 0.8 ^gat/L isopleth near to and south of the Equator. Observations of wind speed and velocity along the four section lines, as diagramed by Austin 400 0 1 — I — I — \ I I \ rp I r SECTION 4 I A 80- / \ 0° r 2° 3° LATITUDE FiGiRE 18. — Variation in (adjusted) zooplaiikton voluiiK's and in the configuration of the 70° and 80° F. i.sothernis on section.s I and 4 of Hugh M. Smith cruise 15, along 140° W.. longitude in May-June 1952. (Temperature sections adapted from .\ustin 1954b.] (1954b), show this was a period of moderate and variable winds. Since we did not have observa- tions simultaneously to the north and south of the section lines, the changes with time are complex and difficult to summarize. In the region of the Countercurrent there appears to have been a re- duction in the northeast trade winds and an exten- sion to the northward of the moderate southeast trades. South of the Equator there was first a slackening in tlie winds followed by an increase, with the strongest winds of the cruise being recorded on the southern ends of the third and fourth sections. When the adjusted zooplankton volumes (table 9, appendix B) from the four series of stations along 140° W. are subjected to an analysis of variance with two-way classification, we find there are no significant (P>0.05) differences among the four sections but highly significant (P<0.01) differences among stations (latitudes). The latter significance results from the wide difference be- tween the high volumes obtained in the Counter- current and at the Equator and the low volumes from about 3° S. to 7° S. latitude. When we examine differences in zooplankton distribution between tlie first and fourth legs in relation to changes in the temperature structure at the Equator (fig. 18), we find that the increased distance between the 80° isotherm-surface inter- cepts (an indication of an increase in width of the mixing zone) was accompanied by a broadening of the zooplankton "rich zone." On the first section there was a single peak of abinidance directly on the Equator; on the fourth section there were two peaks, at about 1° S. and 1° N. latitude, with a trough at the Equator. In the Countercurrent the zooplankton catch was high in volume on all four sections. The suggested change in rate of flow in the Countercurrent was not reflected in any noticeable change in zooplank- ton abundance or distribution. It is difficult to explain or to draw conclusions from these events. In one instance (cruise 11) a change in rate of flow of the Cotuitercurrent was accompanied by a change in zooplankton distribu- tion; in the second instance (cruise 15) changes ni the CountorciuTent were not evidenced by any noticeable change hi the volume of zooplankton. On both cruises an increase in breadtii of tlie zone of divergence or mixing at tlie Etiuator was fol- lowed by a corresponding broadening in the plank- ZOOPLANKTON OF CENTRAL PACIFIC 383 ton rich zone. It does not seem likely that these rather quick responses of zooplanktdn to varia- tions in tlie pliysical environment are the result of immediate changes in l)iological productivity reflected in growth of the population, but are simply a shifting and perhaps dispersal or con- centration of the population associated with changes in the water mass. PHOSPHATE, ZOOPLANKTON, AND TUNA The primary objective of our zooplankton studies has been to obtain an estimate of the basic fish food present in different areas of the sea with the hope that this information would increase our understanding of variations in the abundance and distribution of the tunas. Where other factors, temperature for example, are not of a limiting nature, fast-swimming oceanic fishes such as the tunas will occur, we believe, in proportion to the amount of substance available for their nutriment. This does not mean that we expect to find a high positive correlation at all times and places between the volume of food and the abundance of tunas. In fact, it is probable that an inverse relation may exist locally after a period of intensive feeding. In general, however, when broad areas of the sea are being compared, we believe that high abun- dance of fish is most likely to occur in areas of high concentration of zooplankton and other forage organisms. The distribution of yellowfin tuna, A^eothunnux macropterufi (Temminck and Schlegel), summa- rized in figure 19, is derived from 12 cruises in the central equatorial Pacific during the years 1950- 53.'° The highest average catch (5.3 yellowfin per 100 hooks) was obtained in the convergent zone, with the second highest catch in the region of the divergence. Although the peaks in abundance do not exactly coincide, it is obvious that there is more than a casual relation between zooplankton and yellowfin. The best catches of bigcye, Para- thuiinu.'f ■^ibi (Temminck and Schlegel), were made in the North Equatorial Current and Counter- current (fig. 19). This species appears to respond in a different maimer than the yellowfin to the better foraging conditions in the convergent and divergent zones. A comparative study of the food of the two species failed to show differences in the The tuna ratch records employed In this report have resulted from explor- atory longline fishing conducted by POFI vessels and are analyzed in other POFI reports (Murphy and Shomura ig.Wa, ig.Wb, 1955; Shomura and Murphy 1955: Ivcrsen and Yoshida. 1956. o —• ZOOPLANKTON VOLUME o o SURFACE INORGANIC PHOSPHATE „ „ YELLOWFIN CATCH ° ' BIGEYE CATCH O O O O O 1 ! 1 1 SEC 1 1 1 i| i II 1 1 1 1 : < 1 1 ! iSEC IDIVICONV. CC ' NEC -.6 o -.6 3. o"^ 1 l/M ,1 ^ 1 / lSo^ 8'-^- < a o .-1--°' 1 1 ° 1 \ 1 Uj" - a. a 1 , 1 1 °--. 1 2;o5- -.2 1 1 1 t. 1 1 K ^ - 1 1 1 1 1 "--, 1 1 ll 1 h III 1 1 1 i'" I 1 60 « O 5.0? 2.0 14° 12° 10° 8° go 4= 2° 0° 2° 4° 6° 8° 10° 12° 14° 16° 18° S — LATITUDE — N FioORE 19. — Variations with the current system in yellow- fin and bigeye catch on longline gear, zooplankton volumes (adjusted) and surface inorganic phosphate, for the range of longitude 120° \V. to 180°. The tuna catch data are derived from cruises 7, 11, and 18 of the Hugh M. Smith, cruises 11, 12, 13, U. 15, 16, and 18 of the John R. Manning, cruise 1 of the Charles H. Gilbert, and cruise 1 of the Cavalieri. The phosphate data are from cruises 2, ,5, 8, 11, 14, 15, 16, IS, and 19 of the Hugh M. Smith. diet which might explain this marked difference in distribution (King and Ikehara, 1956). Measurements of inorganic phosphate performed on POFI hydrographic cruises during the years 1950-53 show that the zone of divergence and the South Equatorial Current immediately south of the Equator contained the highest concentrations of this basic chemical nutrient while the North Equatorial Current contained the lowest (fig. 19). This variation may result from unequal utilization of phosphate and/or the unequal mixing of high and low phosphate water to the north and south of the Equator as the result of the asynunetrical effects of the southeast winds. As evidenced Viy the zooplankton and yellowfin catch, the greatest organic productivity occurred on, or to the north of the Equator. The difference in degree of north- ward displacement for the two eutropliic levels, zooplankton and tuna, may to some extent be indications of the lag periods in their development and may also be related to the slow northward drift in the surface cm-rents under the inlluence of east and southeast winds. When long series of stations extending in a north-south direction are examined, we usually find a highly significant positive correlation be- 384 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 5. — Correlations of adjusted zooplankton volumes (cc./lOOO m?) as the A'l variate, with A'2 variate the surface inorganic phosphate or yellowfin longline catch from same locality Xj variate Motor vessel and cruise No. Range o( latitude Degrees of Correlation freedom coeffi- cient (r) 22 0.771 41 0.678 50 0.365 20 0.631 25 0.381 34 0.277 24 0.286 58 -0. 294 16 0.101 22 -0. 177 Inorganic phosphate, fig at/L. Inorganic phosphate, Mg at/L. Inorganic phosphate, /xg at/L. Inorganic phosphate, Mg at/L. Yellowfln catch pi-r 100 hooks Inorganic phosphate, Mg at/L. Yellowfin catch per 100 hooks Inorganic phosphate, ttg at/L. Inorganic phosphate, Mg at/L. Yellowfin catch per 100 hooks Hugh M. Smith— 2... Hugh M, Smith— 5... Hugh M. Smith— 8-.. Hugh M. Smith— 11.. Hugh M. Smith— 11.. Hugh M, Smith— 14.. John R. Manning — 11 Hugh M. Smith— 15.. Hugh M. Smith— 18.. Hugh M. Smith— 18.- 24° N.-S" S. 27° N.-5° S. 21° N.-14° S 19° N.-4° S. 15° N.-5° S. 9° N.-8°S.. 8° N.-8° S... 9° N.-7° S.., 9° N.-9° S... 9° N.-9° S.. <0.01 <0.01 <0.01 <0.01 0.05 >0.05 >0.06 <0.05 >0.05 >0.C8 • In this instance zooplankton volumes obtained by Hugh M. Smith cruise 14 were correlated with longline catches of John R. Manning cruise 11, the two cruises occurring during the same period of time. tween surface inorganic phosphate and zooplank- ton volumes (table 5). With a short series of sta- tions the correlation may be nonsignificant as that for the Hugh M. Smith cruises 14 and 18, or even be significantly negative as for cruise 15. The latter is perhaps an example of an inverse relation resulting from high utilization. The cor- relation of zooplankton volume and yellowfin catch was significant (P = 0.05) for Hugh M. Smith cruise 11, but non-significant for cruise 11 of the John R. Manning and cruise 18 of the Hugh M. Smith. Within the equatorial "rich zone," from the southern boundary of the Count ercurrent at about 5° N. latitude to 5° S. latitude, zooplankton and yellowfin showed a gradient of increasing abund- ance between 180° and 150° W. (fig. 20). The » — = ZOOPLANKTON VOLUME o -o SURFACE INORGANIC PHOSPHATE „ „ YELLOWFIN CATCH O 40 O o O 30 1 1 1 1 1 1 1 - -.8 o^ -<>- i^— -—-'-^-'^ - o X 0. \ - -.2 1 1 1 1 "-0 1 1 1 - o o 2 S o 160 160 140 WEST LONGITUDE 130 120 FifitiRE 20. — Longitudinal variations in yellowfin longline catch, zooplankton volumes (adjusted) and surface inorganic phosphate for the South Equatorial Current from the southern boundary of the Countercurrent, at about 5° N. latitude to 5° S. latitude, with the data segregated by 10-degree intervals of longitude. yellowfin catch continued high at 140° W. and then dropped off sharply to the east, while zoo- plankton volume varied somewhat irregularly to the east but remained moderately high. The variation in surface inorganic phosphate was roughly just the reverse (fig. 20), with high concentrations on the eastern and westernmost longitudes and low values in between. We have no empirical explanation at present for this distribution of phosphate. It may possibly result from differences in rate of utilization as the most productive areas appear to be the mid-longitudes. In the equatorial region of the central Pacific, July, August, and September was the period of best yellowfin catch (fig. 21). It was also the period of highest zooplankton abundance, al- though the quarter October, November, and December was essentially of equal rank. Phos- ZOOPLANKTON VOLUME SURFACE INORGANIC PHOSPHATE YELLOWFIN CATCH ' 1 1 1 A "S. ^^'^ "\ ■ "■"■ — o _ O o _ =s ''^ ^' - -.6 :i. <>*r -- -. ,•— \ - UJ \- . X If) * o X Q. -.2 1 1 1 1 - 2 S q JAN-Fee-MAR APR-MAY-JUN JUL-AUG'SEP OCT-NOV-OEC Figure 21. — Seasonal variations in yellowfin longline catch, zooplankton volumes (adjusted) and surface inorganic phosphate for the South Equatorial Current from the southern boundary of the Countercurrent at about 5° N. latitude to 5° S. latitude, with the data segregated into quarterly periods of 3 months each. ZOOPLANKTON OF CENTRAL PACIFIC 385 phate again showed an inverse correlation, par- ticularly with the yellowfin catch. Figure 14B demonstrates tlie difference in yellowfin catch for the two 6-month periods: (1) January-June, a period of generally light, variable or northeast winds, and (2) July-December, a period of strong southeast trades. With the change in winds during the latter half of the year there was apparently a shift to the northward in the area of best catch. The zooplankton exhibited a general increase during this period, especially in the convergent zone. SUMMARY AND CONCLUSIONS 1. This is the second report of the Pacific Oceanic Fishery Investigations on variations in zooplankton abundance in the central Pacific; it presents the results of 270 quantitative hauls made on eight cruises during the years 1951 to 1954. Data from earlier cruises, included in a previous report (King and Demond, 1953), were also utilized in this study. 2. The majority of the collections were ob- tained with 1-meter nets of 30XXX grit gauze (aperture widths 0.65 mm.). For comparison, a few haids were made with 56XXX nets (aperture widths 0.31 mm.). Oblique hauls to 200 meters' depth were employed at most stations. The results from a short series of horizontal hauls are included. 3. The displacement volumes of all samples were measured in the laboratory. For each sample there was calculated the volume of the more nutritious zooplankton per unit of water strained. Counts were made on six samples to examine the composition of the catch from nets of different mesh size. 4. The catch of 56XXX grit gauze nets (aper- ture widths 0.31 mm.) was about IK to 1^4 times greater in volume than that of the catch of the 30XXX nets (aperture widths 0.65 mm.). The number of plankters retained by the finer-meshed net was 3 to 5 times that retained by the coarser- meshed net. At three stations, two rich and one poor, the catches for the two nets were generally proportional. 5. Horizontal hauls made simultaneously at three levels showed that the greatest bulk of zooplankton was near the surface even in the daytime, rather than at depths just above or l)elow the thermocline. 6. The night hauls yielded volumes averaging 1 .57 times the volumes of day hauls; twilight hauls were intermediate in volume. To reduce these differences associated with hour of hauling, a method of adjustment was employed based upon the similarity between the diurnal variation in zooplankton abundance in the upper 200 meters and the curve of the sine function, with midnight equated to the angle whose sine is +1.0. 7. When the adjusted zooplankton volumes were combined according to natural subdivisions of the equatorial current system, disregarding differences associated with longitude and season, we found the greatest concentration of zooplankton occur- ring at the Equator in the region of upwelling and divergence. Average volumes for the convergent zone and the Countercurrent were greater than for the South Equatorial Current south of the Equa- tor. This asATnmetrical distribution of zooplank- ton in respect to the Equator may result from the prevalence of southeast trade winds in this part of the Pacific. 8. As determined from exploratory longline fish- ing conducted by POFI, in the central equatorial Pacific the greatest abundance of yellowfin tuna occurred in the convergent zone just to the north of the area of highest zooplankton abundance, and although the peaks did not exactly coincide, there was a high degree of co-variation in yellowfin and zooplankton in respect to the current system. 9. Oceangraphic fronts occurring in the transi- tion zone between the Equator and the southern boundary of the Countercurrent appeared to de- marcate areas of high zooplankton abundance on the south from areas of poor to moderate abund- ance on the north. 10. The Countercurrent in the east-central Pa- cific produced unusually high zooplankton vol- umes. As this is an area of shallow thermocline, with high-phosphate water within the photosyn- thetic zone and within reach of wind-induced tur- bulence, conditions are more favorable for plank- ton production than fartlier to the westward where the thermocline deepens. 11. Within the equatorial region there was a west-east gradient of increasing zooplankton abun- dance from 180° to 150° W. longitude which was correlated positively with average wind velocity and inversely with thermocline depth, and was closely paralleled by a gradient of increasing yel- lowfin catch. East of 140° W. the vellowfin catch 386 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE dropped sharply, but zooplankton volumes re- mained liigh. 12. Largest zooplankton volumes occurred in the quarter, July, August, and September, with October, November, and December essentially equal in rank, and the lowest in January, February, and March. Best yellowfin catches were obtained in July, August, and September. 13. Zooplankton volumes averaged considerably higher in the year 1950 than in 1951, 1952, and 1953. There was some indication of a rising trend in 1954. 14. The ratio of the volumes of night hauls to day haids ranged from 1.31 to 1.43 in divisions of the current system north of the Equator, and from 1.76 to 1.94 at the Equator and in the South Equatorial Current to the southward. The night/ day ratio did not appear to be related to thermo- cline depth or to general level of plankton abun- dance. 15. With an increase in breadth of the mixing zone associated with the divergence at the Equa- tor, there was a corresponding broadening in the zooplankton rich zone. On one cruise an increase in rate of flow of the Countercurrent was ac- companied by a marked cliange in the distribution of zooplankton within the cmrent. These obser- vations indicate that the zooplankton was quick to respond to physical changes in the environment by dispersal or concentration of the population following changes in the water mass. 16. On long series of stations extending from the phosphate-poor North Equatorial Current to south of the Equator, we found highly significant ])ositive correlations between zooplankton volume and surface inorganic phosphate. On series covering a sliort range of latitude the correlation was insignificant or even negative. Although the highest concentration of phosphate occurred in the divergent zone at the Equator, agreeing in this respect with zooplankton, longi- tudinally and seasonally there was some evidence of an inverse relationship with zooplankton and yellowfin; this may result from differences in rate of utilization. 17. Zooplankton distribution was rather uni- form tiu'oughout the island waters of Palmyra, an atoll lying in the Countercurrent at about 162° W. longitude. Sampling along four station lines extending from a few hundred yards from the outer reef to about 14 miles offshore revealed no significant change in zooplankton abimdance with distance from land (appendix A). LITERATURE CITED Austin, T. S. 1954a. Mid-Paeific oceanography III. Transequa- torial waters, August-October 1951. U. S. Fish and Wildlife Service, Spec. Sci. Rept., Fisheries No. 131, 50 p. 1954b. Mid-Pacific oceanography, Part V, Trans- equatorial waters, May-June 1952, August 1952. U. S. Fish and Wildlife Service. Spec. Sci. Rept.: Fisheries No. 136, 86 p. Bir.ELow, H. B., and M.^ry Sears. 1939. Studies of the waters of the Continental Shelf, Cape Cod to Chesapealce Bay, III. A volumetric study of the zooplankton. Mus. Comp. Zool., Mem. 54(4):183-378. Cl.\rke, G. L. 1940. Comparative richness of zooplanl^ton in coastal and offshore areas of the Atlantic. Biol. Bull. 78(2) :226-255. Cromwell, Townsend. 1951. Mid-Pacific oceanography, January through March 1950. U. S. Fish and Wildlife Service, Spec. Sci. Rept.: Fisheries No. 54, 76 p. 1953. Circulation in a meridional plane in the central equatorial Pacific. Jour. Mar. Res. 12(2):196-213. 1954. Mid-Pacific oceanography II. Transequa- torial waters, June-August 1950, January-March 1951. U. S. Fish and Wildlife Service, Spec. Sci. Rept.: Fisheries No. 131, 180 p. Cromwell, Townsend, R. B. Montcomery, and E. D. Stroup. 1954. p;quatorial Undercurrent in Pacific Ocean revealed by new methods. Science 1 19(3097) :648- 649. Cromwell, Town.send, and J. L. Reid, Jr. 1956. Astudy of oceanic fronts. Tellus 8(1) :94-101. Crowe, P. R. 1952. The seasonal variation in the strength of the trades. Institute of British Geographers, Trans- actions and Papers (1950), No. 16:25-47. CUSHING, D. H. 1951. The vertical migration of pianktonic Crus- tacea. Biol. Rev. 26(2): 158-192. GR.'iH.^M, H. W. 1941. Plankton production in relation to character of water in the open Pacific. Jour. Mar. Res. 4(3):139-197. HiD.\, T. S., and J. E. King. 1955. Vertical distribution of zooplankton in central equatorial Pacific, July-August 1952, U. S. Fish and Wildlife Service, Spec. Sci. Rept.: Fisheries No. 144, 22 p. Iversen, E. S., and H. O. Vo,shid.\. 1956. Longline fishing for tuna in the central equatorial Pacific, 1954. U. S. Fish and Wildlife Service, Spec. Sci. Rept.: Fisheries No. 184, 33 p. ZOOPLANKTON OF CENTRAL PACIFIC 387 Jerlov, \. ("i. 1053. Studies of the equatorial currents in the Paoific. Tellus 5(3):308-314. KiKucHi, Kexzo. 1030. Diurnal migrations of plankton Crustacea. Quart. Rev. Biol. 5(2): 189-206. Kisc. J. K. 1054. \'ariations in zooplankton abundance in the central equatorial Pacific, 1950-1052. Fifth Meet- ing, Indo-Pacific Fi.sheries Council, Sympo.xiuin on Marine and Fresh-water Plankton in the Indo- Pacific, p. 10-17. King, J. E., and .Io.\n Demoxd. 1953. Zooplankton abundance in the central Pacific. U. S. Fish and Wildlife Service, Fish. Hull. 54(82): 111-144. King, J. E., and T. S. Hid.\. 1954. Variations in zooplankton abundance in Hawaiian waters, 1950-52. V. S. Fish and Wild- life Service, Spec. Sei. Rept.: Fisheries Xo. 118, 66 p. King, J. E., and I. I. Ikehar.4. 1956. A comparative study of the food of bigeye and yellowfin tuna in the central Pacific. U. S. Fish and Wildlife Service, Fish. Bull. 57(108): 61-85. MooBE, H. B., H. OwRE, E. C. .Ioxes, and T. Dow. 1953. Plankton of the Florida Current. III. The control of the vertical distribution of zooplankton in the daytime by light and temperature. Bull. Mar. Sci. of the Gulf and Caribbean 3(2):83-95. Murphy, G. I., and R. S. Shomur.\. 1953a. Longline fishing for deep-swimming tunas in the central Pacific, 1950-51. U. S. Fish and Wild- life Service, Spec. Sci. Rept.: Fisheries Xo. 98, 47 p. 1953b. Longline fishing for deep-swimming tunas in the central Pacific, January-June 1952. U. S. Fish and Wildlife Service, Spec. Sci. Rept.: Fish- eries Xo. 108, 32 p. Mtrphy, G. I. and H. S. Shomura — Continued 1955. Longline fishing for deep-swimming tunas in the central Pacific, August-Xovember 1952. V. S. Fish and Wildlife Service, Spec. Sci. Rept.: Fish- eries Xo. 137, 42 p. Sette, O. E. 1955. Consideration of mid-ocean fish production as related to oceanic circulatory systems. Jour. Mar. Res. 14(4):398-414. Sette, O. E., and Staff of Pacific Oceanic Fishery Investigations. 1954. Progress in Pacific Oceanic Fishery Investiga- tions, 1950-53. U. S. Fish and Wildlife Service, Spec. Sci. Rept.: Fisheries Xo. 116, 75 p. Shomura, R. S., and (i. I. .Murphy. 1955. Longline fishing for deep-swimming tunas in the central Pacific, 1953. L. S. Fish and Wildlife Service, Spec. Sci. Rept.: Fisheries X'o. 157, 70 p. Snedecor, G. W. 1946. Statistical methods. 4th ed., xvi, 485 p. Iowa State College Press, Ames, Iowa. Stroup. E. D. 1954. Mid-Pacific oceanography. Part I\'. Trans- equatorial waters, January-March 1952. U. S. Fish and Wildlife Service, Spec. Sci. Kept.: Fish- eries Xo. 135, 52 p. SvERDRUP, H. v., M. W. JoHXsox, and R. H. Flemixo. 1942. The oceans, their physics, chemistry, and gen- eral biology. .\, 1087 p. Prentice-Hall Inc., Xew York, X. Y. I'. S. Weather Bureau. 1938. Atlas of climatic charts of the oceans, vi, 65 p., U. S. Govt. Printing Office, Washington, D. C. Wentworth, C. K. 1931. Geology of the Pacific equatorial islands. Bernice P. Bishop Mus. Occas. Papers 9(15): 1-25. APPENDIX A ZOOPLANKTON DISTRIBUTION ABOUT AN OCEANIC ISLAND Palmyra Island lies 352 nautical miles north of the Equator and about a thousand miles south of Honolulu. In relation to other islands of the Line Islands group, Palmyra is located about 33 miles southeast of Kingman Reef and 120 miles northwest of Washington Island. The island is an atoll consisting of 40 to 50 small islets arranged in a rectangle about 4 miles long and IK miles wide. The islets rest on a shallow reef platform 6 miles long and 2 miles wide with the long axis of the platform extending in an east-west direction. Outside the 10-fathom line the submarine slope is steep, ranging from 2,500 to 3,000 feet to the mile and descending to the general depth of about 15,000 feet (Wentworth 1931). Occupying latitudes 5°52' N. to 5°54' N. at approximately 162° W. longitude. Palmyra lies close to the soutliern boundary of the Counter- current and ordinarily is batlied by it throughout the year. The surface current was flowing to the east at the time of our observations, as was to be expected, since, in the region of the Line Islands, the southern boundary of the Countercurrent has alwaj's occurred south of 5)2° N. latitude on the numerous crossings of POFI vessels. The zooplankton abundance about the island was investigated in January 1953, on Hugh Af. Smith cruise 19, by running lines of stations out to the north, south, east, and west, starting as close to the reef as the vessel's safety permitted and extending out to a maximum of about 14 miles (fig. 3). A total of 20 hauls were made, all at night. With the exception of a single haul made on the shallow shelf west of the island which yielded a sample about twice the average volume of the 200-meter oblique tows, the results indicated ■' The salinity and temperature data collected on this cruise indicated that the waters about Palmyra were also rather uniform as to chemical and phys- ical conditioDs; e. g., the maximum variation in surface temperature was less than 0.5° C. 388 a rather uniform distribution of zooplankton throughout the island waters." From an analysis of variance we conclude that the difTerences between the four series of stations were not signifi- cant (F = 0.896, P>0.05). There was a slight indication of an inverse relation between the zooplankton catch and distance from land (fig. 22) ; a regression analysis showed, however, that this trend was not significantly different (b=— 0.431, P>0.1) from a random distribution. The variation about Palmyra was less than we found in two series of stations extending offshore from Oahu, Hawaii. Here the largest volumes occurred at one or two miles from shore and the difference between stations was significant (King and Hida, 1954). Although the sampling was entirely inadequate for any broad conclusions, it appears evident that at the time of our visit to Palmyra there was no definite gradient in zoo- plankton abundance along four station lines extending from a few hundred yards from the outer reef to about 14 miles ofi'shore. 3U " 40 1 1 1 1 1 1 1 1 1 1 1 1 S 0 • 0 0 0 30 • . -- u • • • • z 0 • • h- • • ^ 20 z ' •. • • — r < • • _) Q- 0 • S 10 1 1 1 1 1 0 2 4 6 8 10 12 14 16 DISTANCE FROM 10 FATHOM LINE-MILF.S FioiTRE 22. — Zooplankton volumes (adju-sted) in relation to di.stance oflshore from the 10-fathom line, Palmyra I.sland, January 1953, Hugh M. Smith crui.se 19. APPENDIX B Table 6. — Zooplankion volumes obtained on cruise 9 of the Hugh M. Smith, with collection data station No. Oblique tows, 200 m. depth; 1-meter nets, 30XXX grit pauze: I ■2 ... 9-. 10- 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Position Latitude 25°40' N 22°45' N 19°33' N 16°05' N WIS' N lO'lO' N 7°27' N. 5''35' N. 4°23' N.. 2°43' N. 1''28' N.. 0''27'S.. rai's.. 2°04'S.. 3°05'S.. 4°32'S.. 4''28'S.. 3°59'S.. 0°32'S.- 2°18' N.. 5°12' N.. S'OS' N.. 11''24' N. 14°40' N 17''40' N. 20''34' N Longitude 175°24' 173°16' 171°32' 169°06' 166°45' 164°49' 163°05' 161°44' 160''02' 158°15' :58°02' 160''20' 164°05' 168°16' 172°07' 172°48' 171°38' 17I''26' 170°33' 168°43' 166 46' les-ii' 163°37' 161''49' 159°57' 158°15' W.. W.. w.. w.. w.. w.. w.. w.. w.. w.. w.. w.. w.. w.. w.. w.. w.. w.. w.. w... w.. w.. w.. w . . w.. vv . Date May May May May May Mav May June June June June June June June June June June June June June June June June June June July 21. 1951 22, 1951 23. 1951 26, 1951 27. 1951 28. 1951 29, 1961 3, 1951 5, 1951 7, 1951 10, 1951 11, 1951 12, 1951 13, 1951 17, 1951 19, 1951 20, 1951 21, 1951 24, 1951 25, 1951 26, 1951 27, 1951 28, 1951 29, 1951 30, 1951 1, 1951 Time 1 1202-1235 1210-1237 1214-1238 1224-1250 1233-1301 (') 1246-1316 1545-1614 1358-1430 1203-1235 1206-1243 1200-1235 1144-1216 1224-1254 1408-1438 0804-0834 1207-1240 1209-1242 1213-1243 1222-1251 1232-1304 1233-1302 1240-1307 1250-1322 1253-1324 1200-1231 Water strained, m.' I Zooplankton, cc./lOOO m.i Sample volume 3304.0 1932. 5 1385. 4 1312.7 ' 1366. 4 1505. 0 1480.8 1517.2 1656.3 3026. fi 2531.3 2302.6 1930.0 1997.6 1647 5 1975.4 2336. 1 2103.3 1994. 2 2275. 1 1978, 8 1521.6 2371.7 2716. 2 2622.3 I ' Apparent solar time. ' Adjusted for day-night difference by the sine transformation method using a pooled regression coefficient (0.1248) . ' Based on an estimated meter reading. * No sample. 10.3 1.5.7 23,0 18,0 10,5 18.7 35.0 22.5 29.5 14.4 21 2 22.1 19.4 23.2 24 3 30. 1 25.1 17.6 15 7 17.8 25 2 14.2 8,2 11,3 17 5 Adjusted volume - 13.7 20.9 30.6 23.9 13.9 24.7 40.4 28.4 39.3 19.1 28.2 29.4 25 8 29.4 28.7 40.1 33.4 23.4 20.8 23.6 33.4 18.8 10.8 14.9 23.1 389 390 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 7. — Zooplanklon volumes obtained on cruise 11 of the Hugh M. Smith, with collection data Station No. Position Date Time ' Water strained, m.3 Zooplankton, cc./lOOOm.i Latitude Longitude Sample volume Adjusted volume ' Oblique tows, 200 m. depth; 1-meter nets, 30XXX grit gauze: 14°37' N 13°06' N 11°31'N _. 9°41' N... 8°58'N 7°59'N e-ss'N S'Sg'N 4'49' N 4°00'N 2° 55' N. ... 1°59' N 2°02' N 2°03'N 1°53'N 2-02' N 1°S8'N l°2rN I'M'N.. 2°02'N 2°03'N 2°00'N CSS'N.. 0'01'S r03'S 2°ors 3°34'S 4°57' S 160°12' W 149°68' W 150°02' W 15O°01' \V 160°08' \V 149°61' W 149°61' W 149' 55' W 150°02' W 150°02' \V 150°09' \V 150°04' W 151°39' W 153°05' W 165''18' \V 156°13'\V 1.57' 32' W 157°18' W 156°1,5'\V 154°43' W 153°04'\V 151'18' \V 149°54'\V 149°5fi' W 150°16' W 150°08' W 150°05' W 15O°04' W 150°00' W 149° 57' W 16O°02'\V 149°58' W 150°00' \V 150°01' \V 149°.53' W 150°03' W 150°00' W 149°56' \V 150°00' W 149° 57' W 149°57' W 149°58'\V 150°13'W 15O°02' W 150°03'\V 149' .56' W 150'03' W 150°0r\V 150°42'W 151°19' W Aug. 24, 1951 Aug. 25, 1951 Aug. 26, 1951 Aug. 27, 1951 Aug. 28. 1951 Aug. 29,1951 Aug. 30. 1951 Aug. 31. 1951 Sept. 1,1951 Sept. 2, 1951 Sept. 3,1951 Sept. 4, 1951 Sept, 5, 1951 Sept. 6.1951 .Sept. 12. 1951 .Sept. 13, 1951 Sept, 14, 1951 Sept. 15. 1951 Sept. 16. 1951 Sept. 17. 1951 Sept. 18. 1951 Sept. 19. 1951 Sept. 20. 1961 Sept. 21. 1951 Sept. 22, 1951 Sept. 23. 1951 Sept. 24. 1951 Sept. 25, 1951 Sept. 26. 1951 do do Sept. 27, 1951 do do Sept. 28, 1951 do Sept. 29, 1951 do do Sept. 30, 1951 do do Oct. 1. 1951 do Oct. 2, 1951 do do Oct. 3, 1951 do Oct. 4. 1951 0851-0922 0820-0853 0830-0902 0759-0828 0720-0754 0754-0832 0734-0805 0742-0810 0731-0804 0739-0812 0738-0807 0741-0811 0729-0800 0729-0758 0729-0759 0718-0749 0713-0746 0731-0803 0725-0753 0733-0803 0742-0813 0744-0812 0748-0818 0750-0820 07,52-0808 0743-0813 0748-0820 0914-0944 0000-0033 0906-0937 1728-1758 0219-0249 1309-1338 2341-0011 0849-0920 1724-1756 0201-0231 1048-1119 1929-2000 03,53-0423 1307-1338 2243-2312 0718-0749 1604-1634 0052-0123 0918-0948 1744-1815 0165-0226 1701-1733 0756-0826 1437. 7 1223. 7 1898, 0 1363.2 1852. 4 2838.5 1693. 4 1540,6 19.54. 6 699.9 1699. 8 1935. 4 1034. 0 1458.4 1773.6 1752. 6 1683.2 1827. 2 1368.8 1455. 1 1492.3 1294. 5 1311.3 1471.2 1133 4 1187.4 1546.0 1337. 3 1882.7 1635.0 1389. 6 1506.7 1124.0 1569.9 1544.3 1628. 2 1531.2 1072. 2 1660.0 1632. 2 1932. 1 1465. 9 1790.3 1733. 7 1471.2 1596.4 1782. 5 1596,4 1942. 7 1966. 3 5.9 29.2 15.8 26.3 15.2 26.2 25.7 28.8 32.8 45.1 33 0 36.0 28.2 35.5 46,9 33.8 26.6 26,7 34.3 34.1 55.8 52.0 63.1 66.7 61.3 48.8 30.1 41.1 53.4 44.3 65.0 71.9 56.7 47.2 36.9 38.2 51.8 38.9 18.3 28.7 26.2 51.0 24.0 29.5 29.8 10 8 15.7 23.2 13.7 12.3 6 8 2 33.0 3 17.9 4 29 4 5 16.5 6 29 2 7 27.9 8 31.7 9 35.6 10 49.7 11 36 3 12 39.6 13 . 30 6 14 38.5 15. . 50.9 16 . 36. 7 17 28.4 18 29.0 19 37.2 20 37. 1 21 61.5 22 57.3 23 69.6 24 62.5 56.5 26 53.7 33.2 28 47.6 29 4°00'S 3°00' S 2°00'S 1°00'S 0°02'N r02'N ..._ 2°orN .- 3°00'N 3°58' N 4°.59' N 5°56' N 6°51' N 7°5r N... 8°67' N 10°00' N 10°58' N I1'59' N 13°00'N U'Ol'N 15°02'N 17°00' N 19°00' N 44.1 30 51.3 66.1 32 62.1 33 -- 68.1 34 38.9 42.3 36 38.8 44.2 38 39. . . 46 9 16.8 40 .- 26.1 31.4 42 43... 44 45 46 . .. 42.4 26.1 32.1 24.7 12.6 47 15.7 48 -. 19.8 49 . 50 14.2 13.7 I Apparent solar time. - .Adjusted for day-night difference by the sine transformation method. ZOOPLANKTON OF CENTRAL PACIFIC 391 Table 8A. — Zooplankinn volumes obtained on cruise 14 of the Hugh M. Smith, and rollertion data Station No. A . Oblique hauls to 2(H) m. (li'plti: l-rnc'tcr nets, 30XXX Brit pauze: 1 - 2 3 4 S _ fi 9 in 11 12.... 13 ._ 14 15 16 17 18-- - .-.- 19 20 - 21 22 23 24 _ 25 26 27 28 29. 30 31.... 32... 33 34 35 36 37.... 38 39 40 _ 41 42 43 44 _. 45 46 _ 47 _ Position Latitude 7''57' N S'SS' N 4055' N 3''54' N 2''56' N 1°54'N 0°54' N 0=06' S. i°ia' S- 2''00' S. 3''00' S. 4°00' S. 4° 58' S. 5°53' S. 6''50' S. 7°56' S- 7°or S. e'os' s. 5"'04' S- 4°03' S. 2°56' S. r52' S. 0°59' S. 0°02' N I'OS' N 2°07' N 3°06' N 4''06' N 5'^06'N 6°06' N 7°04' N 8°03' N 8''58'N 4°59' S. 4°03' S. 3°04' S- 2°04' S. foa' S. 0°02' S. 1003' N 2°05' N 3°04' N, 4°05' N 4''57' N 5''56' N Longitude 1,54°57' 154°57' 154''55' 154''5r 154°5r 154"'59' 155°03' iss'ir 155° 14' 155°08' 165°03' 154°68' 155°07' 155°00' 155°03' I55''05' 179° 53' 179°49' 179°58' 179° 58' 179°58' 180°00' 180°06' 180°03' 180°or 179° 58' 179°57' 179°57' 179°55' 179°50' 179°44' 180°01' 179°58' 179°55' 168°59' 168°58' 168°59' 168°57' 168°54' 169°00' 169°00' 168°57' 169°00' 168°57' 168°5fl' 169°00' 168°.V1' Date Jan, 27, 1952 do Jan. 28, 1952 do .. Jan. 29.1952 do do Jan. 30, 1952 do Jan. 31.1952 do do Feb. 1, 1952 do Feb. 2. 1952 do Feb. 15.1952 do do Feb. 16,1952 do Feb. 17, 1952 do do Feb. 18, 1952 do Feb. 19.1952 do do Feb. 20,1952 do do - Feb. 21, 1952 do Mar. 1. 1952 Mar. 2, 1952 do ....do Mar. 3,1952 do Mar. 4,1952 ....do Mar. 5. 1952 do ....do Mar. 6,1952 ....do Time I 1125-1151 2059-2129 0740-0810 1635-1711 0135. 0205 1024-1054 1937-2008 0803-0836 1831-1916 0418-0457 1127-1159 2051-2122 0633-0709 1846-1918 0421-0458 1333-1402 0102-0137 1040-1110 1946-2016 0616-0550 1462-1522 0026-0067 1053-1122 1918-1961 0506-0539 1601-1637 0156-0227 1051-1124 1954-2025 0452-0525 1358-1428 2264-2326 074.'H)ai4 1707-1737 1706-1737 0166-0224 1131-1202 2211-2241 0805-0833 1800-1831 0354-0426 1544-1613 0116-0147 1025-1050 1910-1936 0416-0449 1742-1811 Water strained, ni.3 1014.4 1334.6 1128.6 1533.9 1325. 4 1033. S 11.60.5 1330.0 2064. 2 1647.3 1479.6 1511.2 1394.6 1246. 3 1334. 6 1316,0 1878. 2 1670. 5 1546. 1 1765. 6 1307. 1 1389. 7 1342. 1 1642.0 1214.6 1505.8 1067. 7 1738.6 1472. 9 1699.6 1461,3 1606, 2 1296,3 1329,9 1779,1 1506,7 1158,3 1028,6 994, 1 1228,4 1176,7 946,0 11.50,5 828,5 1091.4 1.583.8 1465.4 Zooplankton, cc./IOOO m.> Sample volume 23.5 43.2 10. 1 22.6 28.0 24.5 27. 6 8.4 35.1 42.7 22.7 26.3 16.5 25.0 27.8 13.3 9.6 16.0 17.3 19.9 23.0 31.4 15,0 34,1 31,7 23.6 35.4 10.1 35.4 23.8 16.9 20.9 20. 1 16. 1 11,9 26,9 10,7 19.5 24.6 41.6 61.5 28.8 43.2 22.5 12.3 31.5 21.6 Adjusted volume = 31.9 34.1 11.8 24.5 21.4 32.7 23.7 10.0 32.4 38.4 30.8 21.1 17.9 23.1 26.0 17.6 7.2 20.2 14.8 19.4 28.6 23.2 20.2 29.9 30.1 26.9 27.5 13.6 30.3 22.0 21.8 15.4 23.4 17,0 12,6 20,6 14,6 14,8 29.4 40.5 54.0 ,33.6 32.7 30.1 11.1 28.3 21.6 ' Apparent solar time. ' Adjusted for day-night difterence by the sine transformation method. Table 8B. — Zooplankton volumes obtained on cruise I4 of the Hugh M. Smith, with collection data Special station Horizontal hauls at various depths: SI . Sample Position Latitude 2°4I'S . , 2°41' S 2°41' S 2°41'S.... 2°41' S 2°4I' S 2°43' S.... 2°43' S .. 2°43' S 2° 44' S 2°44'S 2°44'S... 2°45' S . 2°45' S 2°45' S 2°41' S 2°4!' S.... 2°4I'S. . 2»4I'S., . 2°41'S.... 2°41' S Longitude Date Feb. 9, 1952 do do do do do do do do .do .do .do .do .do do Feb. 10, 19,52 .. .do .. .do ...do ...do ...do ... Time I 131.5-1443 1322-1438 1329-1429 1557-1724 1603-1719 1610-1711 1848-2021 1856-2012 1903-2004 2034-2205 2042-2169 20.50-2161 2221-2365 2232-2347 2242-2341 0029-0157 0036-0149 0043-0142 022.5-03.54 0230-0345 0238-0338 Estimated depth of haul, m. ?10 105 0 210 105 0 210 105 0 240 120 0 240 120 0 220 110 0 240 120 0 Water strained. 2694. 6 2594. 1 2591. 9 3378. 0 2508. 1 2504.2 1433. 7 1309.3 2317. 2 1835.5 2472. 1 2302.8 1378. 6 1843. 1 2332. 7 1448. 6 2175.2 2616. 5 1118. 1 1137.2 2464. 7 Sample volume cc./IOOO m. 3 15.4 23.9 32.2 22.6 30. 1 47.8 29.2 33.7 4,5.5 22.4 23.2 49.2 25.4 23. 6 : 1,57, 8 27,8 26.9 44. 1 41.8 39,3 44.2 .\pparenl solar time. ' Estimated .50 percent amphipods. 392 FISHERY BULLETIN OF THE FISH AND "WILDLIFE SERVICE Table 9. — Zooplankton volumes obtained on cruise 15 of the Hugh M. Smith, with collection data station No. Oblique tows, 200 m. depth; 1-meter nets, 30XXX grit gauze: 10- 11- 12- 13- 14- 15- 16- 17- 18- 19- 20- 21- 22- 23- 24- 25- 26- 27- 28- 29- 30- 31- 32- 33- 34- 35- 36- 37- 38- 39- 40- 41- 42- 43- 44- 45- 46- 47- 48- 49- 50- 51- 52- 53. 54. 56. 56- 57.. 58. 59. 60. Special hauls with 56XXX grit gauze nets: 45— - - 52. 60.- -- Position Latitude 8°59' N. 7°51'N. 7°00' N- 6°36' N- 6°00' N- 5°02' N. 4''05' N- 3''09' N . 1°67' N. 0°53' N- O'll'S.. 1°02' S- 2°00' S-. S^OO' S.. 4°00' S, 5°30' S., 7°00' S. S''26' S. 3052' S- 2°50' S. r46'S. 0°43' S- 0°00'--- 1°02' N. 2°04' N. 3°06' N . 4°09' N- 5°10' N. 6007' N- 7°04' N- 8°00' N. 7°04'N. 6°06' N- 5°00' N. 4°01' N. 3°00' N- 2°00' N- 0°58' N. 0°04'S- roe' S_ 2°08' S- s^iy s. 4°05' S- 5''32'S. 7°00' S. 5°29' S. 3°43' S- S-OO' S _ 2°00' S- 0°57' S- 0°01' S_ 1°00' N 2°06' N 3°11'N 4°00' N 5°00' N 5"'54' N 6°47' N 8°00' N 9°00' N 7°00' 1°00' 9°00' Longitude 39°56' W.- 39°46' W-- 40°00' W-- 39°44' W.. 40°00'W-- 39°58' W.. 39°58' W-- 39°53' W- 40°03' W-- 40°02' W-- 39°52' W.. 39°52' W-- 39°50' W - - 39"'52' W-_ 40°05' W-- 39°57' W-- 39°58' W - . 40°00' W-- 40°04' W-- 40°06' W-- 40°09' W-. 40°11' W.. 39°59' W-- 39°58' W-. 39° 57' W-. 40°0fl' W-. 40°06' W-. 40<'I2' W-. 40°11' W.. 40°10' W-. 40°00' W-. 40°04' W-. 39°58' W-. 39°61' W-, 39° 55' W-. 39°57' W.. 39°56' W_. 39°56' W . . 39°54' W . . 39°51' W- 39°50' W_. 39°49' W_. 39°49' W.. 39°54' W-. 40°00' W-. 40°05' W- 40°04' W- 4n°00' W- 39°53' W- 39°45' W _ 39°43' W . 39°51' W. 40°02' VV- 40° 13' W. 40°09' W_ 40°04' W- 40°00' W- 39°56' W_ 39°S2' W- 40°00' W- 40°00' W- 39°51' W- 4n°on' W - Date May 28, 1952 - -do May 29, 1952 -..-do -do May 30,1952 ---.do -do - May 31, 1952 --..do --.-do June 1, 1952 -...do . -do June 2, 1952 do June 3, 1952 . do June 4, 1952 ---.do -do June 5. 1952 --.-do -do June 6, 1952 -do...- June 7. 1952 - -do -do June 8, 1952 --.do .do --- - June 9, 1962 .do... June 10, 1952 ...do .do June 11, 1952 ---.do .do June 12, 1952 . .do -do ... . June 13, 1952 .do - June 14, 1952 June 15, 1952 .do ... June IB, 1962 ...do -do June 17, 1952 - .do .do -. June 18. 1952 -do -do June 19, 1952 - .do ...do June 13, 1952 June 1 7, 1952 iM'ie 19. 1952 Time ' 0714-0745 1822-1855 0258-0328 1018-1049 1710-1743 0151-0223 0946-1016 1821-1863 0326-0401 1219-1249 2121-2161 0468-0630 1240-1311 2009-2040 0610-0640 1821-1851 0661-0722 1928-1968 0659-0730 1454-1529 2225-2260 0644-0713 1315-1345 2240-2308 0805-0832 1701-1733 0214-0264 1116-1147 1848-1920 0334-0407 1043-1112 1869-1929 0541-0614 1605-1641 0107-0137 0927-0954 1757-1828 0243-0314 1131-1210 2123-2161 0606-0637 1313-1344 2133-2203 0918-0946 2166-2226 1641-1611 0726-0756 1444-1614 0019-0049 0937-1007 1621-1653 0116-0148 1051-1120 1924-1954 0259-0329 1023-1054 1748-1823 0351-0421 1419-1449 2244-2315 2231-2302 0156-0228 2323-2354 Water strained, m.s 1,421.8 1,617.5 1.731.8 1, 868. 8 2, 059. 1 1. 896. 4 2. 167. 6 1.510.8 2, 466. 2 1,702.0 2, 060. 9 1, 947. 7 1, ,172. 7 1. 523. 7 1,323.8 1.361.6 1, 716. 6 1.350.5 1, 793. 9 1. 849. 9 1, 326. 1 1,321.5 1, 250. 6 1, 377. 5 1,251.3 1,391.4 !, 948. 1 1, 438. 7 1, 569. 8 1. 758. 9 1. 396. 0 1.441.8 1, 978. 1 1, 856. 1 1, 493. 2 1. 874. 0 1.773.8 1.279.3 2. 463. 1 1.811.1 1, 727. 0 1, 706. 5 1, S79. 2 1, 379. 6 1, 399. 6 1, 142. 1 1,729.4 1.723.2 I, 856. 6 1, 577. 7 I, 703, 4 1, 383. 2 1, 240. 3 1, 447. 7 1. 296. 8 1, 699. 0 1, 903. 3 1.614.0 1. ,632. 0 1,516.6 1.374.2 1.441.8 1, 202. 7 Zooplankton, ec./lOOO m.3 Sample volume Adjusted volume 2 40.1 43.7 70.7 51.4 37.4 42.4 27.4 29.4 22.5 29.5 »88.0 ,63.8 17.1 30.0 20.2 16.0 16.5 14.1 23.1 25.3 38.0 33.0 23.3 44.7 23.2 39.0 48.8 44.8 58.7 I 88.6 57.5 70.4 46.8 22.2 27.1 48.3 22.4 25.1 21,5 5 106. 7 43.8 21.6 45.9 29.6 32.2 20.6 23.0 18.7 68.0 57,4 28.2 69.8 36.5 32 0 28.7 23.7 42,0 .59,1 41.8 «80. 4 50.7 122. 1 90.9 44.1 41.6 59.3 66,4 39.3 33 4 34,8 28.0 19.2 38.6 70.3 51.3 22.3 26.6 20,7 15,2 17.7 12 6 25.4 30.2 29.4 35.4 29.8 34.3 27.2 40.9 39.5 ,68.7 54.7 77.3 75.0 64.1 46.8 24.9 21.0 60.4 21.9 20.7 28.2 85.3 44.9 27.9 36.7 37.0 26.2 23.6 26.8 22,7 51.9 72,8 31.0 54.6 47.5 28.5 24.1 30.6 42.0 61.6 51. 5 61,7 ' Apparent solar time. ' Adjusted for day-night difTerence by the sbie transformation method. 3 Estimated 60 percent euphausiids. < Estimated 30 percent euphausiids. s Estimated ,60 percent amphipods. ' Estimated 30 percent salps. ZOOPLANKTON OF CENTRAL PACIFIC 393 Table 10. — Zoo-plankton volumes obtained on cruise 16 of the Hugh M. Smith, with collection data .■station Xo. Position Date Time 1 Water strained, m.> Zooplanljton, cc./lOOO m." Latitude Longitude Sample volume Adjusted volume ' Oblique tows. 200 m. depth; 1-meter nets, 30XXX grit gauze; 31 32 - . 3°12'S 3°12'S 3°a6'S .. 3"'04'S 2°67'S 0°06'N 0°22'N O-SQ'N i''4r X - 1'>58'N 2-15' N 2°29' N - 2°39'N 2''47'N 149°28' VV 149°33' W 149°40' W Aug. 5. 1952 Aug. 6. 1962 do 1236-1309 0040-0119 1233-1305 0103-0134 0030-0100 0042-0117 003,5-0108 0037-0109 0028-0101 0027-(K),59 0038-0110 0027-0057 0025-00,58 0028-0101 1356.7 1430.3 1241.7 1236.8 1683, 2 13.55. 6 1239.9 1702.6 13,55. 4 1669. 2 1513.9 1498. 6 1862. 6 2033.3 IS. 8 21. 1 20.7 33.0 26.6 38.9 31.6 37.5 37.3 42.2 J68.6 50.8 37.2 45.4 20.9 33 27 5 34 149°42' W 149°58' W Hg^SS' W 149°36' W 149°43' W 149°59' W 150<'15' W 150°35' W 150°47' W.... 151°17' W.... 151°38' W Aug. 7, 1952 Aug. 8. 1952 Aug. 10, 19,52 Aug. 11, 1962 Aug. 13, 1952 Aug. 15.1952 Aug. 16, 19.52 Aug. 17, 1952 Aug. 18, 1952 Aug. 19, 1962 Aug. 20, 1952 25 2 36 37 29 4 38 , 23 9 39 ... 28 4 41 28 1 42 31.8 44.4 38.3 43 - 44 45 46 34 2 ' ApiKirent solar time. • Adjusted for day-nipht difference by the sine transformation method using a pooled regression coefficient (b = 0.1248). ■• Estimated 3()-40 percent siphonophores. Table 11. — Zooplankton volumes obtained on cruise 18 of the Hugh M. Smith, wi th collection data station Xo. Position Date Time ' Water strained, m.s Zooplankton, ce./lOOO m.s Latitude Longitude Sample volume Adjusted volume 2 Oblique tows, 200 m. deptli; 1-meter nets, 30XXX grit gauze; 2 9''00' N 7°06'N 5°50'N 4''53'N 4°01'N 3°03' N -. 1°5.5' N.- 1°02'N 0°39'N 1°04'S 3°06'S... 4°50'S.. 6°37'S 8°40'S 4°12'S 2015' S 0°27'S 1°10'N 1°42'N 3°15'N 4''08'N 5''42' N 7°35'N g'OO'N 120°50' W 120'>00' W 120''16' W 119°69'W 120''06' W 120°05' W.... 120°14' W.-.. 120°17' W 120°14' W 120°07' W 120°10'W 120°2r W 120°24' W 120°35' W... 130°14'W 130°ir W 130°09' W 130°08' W 130°22' W 130°14'W 130°07' W 130°60' W 13ri4' W 131°46' W Oct. 19. 1952 Oct. 21. 1962 Oct. 22.1952 - - .do Oct. 24. 1952 Oct. 26.1952 Oct. 26, 1952 Oct. 27. 1952 Oct. 28,1952 . do --- Oct. 29. 1952 Oct. 31.1962 do ... Xov. 1. 1952 Xov. 5.1962 Xov. 6. 1962 Xov. 7. 1962 Nov. 9, 1952 .. .do .. Xov. 10. 1952 Xov. 11. 1952 Xov. 12. 1952 Nov. 13. 1952 Nov. 15, 1952 0024-0050 0215-0245 0231-0304 2341-0010 0016-0048 0020-0051 0120-0200 0126-0154 0130-0200 2326-2356 2325-2356 0021-0100 2321-23.52 2322-2355 2246-2316 2245-2319 2245-2319 0038-0125 2304-2338 2340-0009 2351-0020 2241-2316 2333-OOOfi 0032-0105 1629.2 1237. 5 < 2489. 5 1454.8 1632, 4 1643. 8 3048.7 1458.5 < 1465. 7 1617.6 1363. 8 766.0 1367. 3 884.5 874.1 1121.3 1307. 0 » 1812.0 1306.8 1398. 2 714.3 ' 1622.5 1180.4 596.7 3 9.7 48.3 54.1 54.0 39.3 36.3 32.9 61.8 69.3 55.9 42.7 48.0 37.7 38.3 43.5 54.0 34.4 21.9 41.7 56.4 64.5 55.8 30.2 67.1 4 38 7 5 . . . 43 3 6 7 8 9 29.6 27.3 26.4 47.6 53.5 42 0 10 11 13 15 32 1 17.. 19 28 3 21 28 8 . 24 32 9 26 40 8 28 30 16.6 31 4 31 32 42.3 33 48 4 34 42 3 36 37 43 0 ' Apparent solar time. 3 Adjusted for day-night difference hy the sine transformation method using a pooled regression coefficient (h = 0.1248). 3 Doubtful volume; most likely the net was not properly washed down. * Based on estimated meter readings. 394 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 12. — Zooplankton volumes obtained on cruise 19 of the Hugh M. Smith, with collection data 1 Apparent solar time. 3 Adjusted for day-night difference by tlip sine transformation method using a pooled regression coetficient (b = n.l248). 3 Based on an estimated flow-meter reading. * This station was located on a shoal with depth of water about 10 fathoms: the haul was made between the surface and about 5 fathoms. Table 13. — Zooplankton volumes obtained on cruise 15 of the Charles H. Gilbert, with collection data station No. Position Date Time' Water strained, m.3 Zooplankton. cc./lOOOm.J Latitude Longitude Sample volume Adjusted volume ' Oblique tows, 200 m. depth; 1-meter nets. 30XXX grit gauze: 2 - 31°50'N 29°38' N 27°4D' N 25°49'N 23°48'N 21°36' N - 19°53'N 8''54'N - 5°05' N 3°43'N.. - 2''08'N 0-32' N 1°37'S.. 3°S5'S 5°35'S. 7°29'S 8°42'S 8°58'S 8°53'S 6''25'S 5°04'S 3°2.5'S 2°17'S 0''42'S l-U'N 2''23'N 2009' N 3°16'N 5°00'N... 6°42' N 119°48' W lig-SO' \V 120°10' W 120°02'W 119°30' W 120°08' W 119°10' W IIO^IO' W iio-ie' w 110°34'W 110°03' W 110°55' W 111''28' W 112'>17' W 113°52' W 114<'49' W 115°39' W 121028' W 132°07' W 155''04' W 155°08' W 155°20' W.-._ 155°10'W 154°50' W 154°58'W 155»26'W 167°05' W 155°13' W 1,54°41' W 154°47' W Feb. 19.19.54 Feb. 20.19.54 Feb. 21. 1954 Feb. 22.1954 Feb. 23, 1954 Feb. 24, 19.54 Feb. 2.5, 1954 Mar. 3. 19.54 Mar. 5, 19,54 Mar, 6. 19.54 Mar. 7. 19.54 Mar. 8, 1954 Mar. 9. 1954 Mar. 10, 19,54 Mar. 11,19,54 Mar. 12. 1954 Mar. 13. 19.54 Mar. 15. 1954 Mar. 18, 19,54 Apr. 9. 1954 Apr. 10. 19.54 Apr. 11. 19.54 Apr. 12. 19.54 Apr. 13,1954 Apr. 14, 19.54 Apr. 15,19,54 Apr. 16.19.54 Apr. 20,19.54 Apr. 21. 19.54 Apr. 22.19.54 1908-1949 1900-1928 1852-1918 1856-1921 1857-1929 (') 1900-1925 1847-1921 1935-2006 1902-1924 1916-1946 1904-1940 (■) 1901-1938 1854-1923 1856-1927 1902-1932 1903-1937 18.57-1924 1926-1957 1910-1942 1916-1949 1917-1946 1917-1948 1916-1949 1909-1948 1904-1933 1917-1947 19!li-194S 1913-1944 2586. 6 1132.8 863.5 1053. 5 1086. 5 58 3 .54.9 J 17.8 33.0 16,9 52.8 49.8 fi 16.5 g . 30.6 10 15.3 14 1166,5 2112.2 8.53, 8 .564, 0 1261.2 2619, 1 13.2 s 12.5. 1 .53.6 82.4 64,7 6 172.5 12.0 16 116.2 19 - 46.5 21 74.7 23 57.3 26 ; 156.3 29 1096. 5 546.1 1725.5 1448. 8 1537. 6 1203. 2 1250.5 1398. 4 1428.5 1255.7 1292.2 2075, 8 2257. 4 1999. 3 1348. 5 1375.2 1682.0 47.4 57.7 21.6 23.7 23.6 14.8 17.7 19.6 16.5 20.9 54.6 33.6 33.4 34. 5 "83.4 52.1 31.2 42.9 31 .- . 53.6 33 19.5 35 21.5 21.4 39 13.4 55 15.7 57 .- 17.5 59 14.6 61 18.5 63 48.3 65 29.7 67 30.3 69 31.3 72 73.8 74 46.1 76 . . 28.3 ' Apparent solar time. 2 Adjusted forday-night difference by the sine transformation method using a p,)oled regression coefficient (b = 0.124fi). ■^ Small hole (^i") in bag of net at end of haul. * Sample not quantitative. a Principally euphausiids. 8 Plus about 12 qts. of salps discarded. ' No sample. * Estimated 60 percent salps. ZOOPLAXKTOX OF CENTRAL PACIFIC 395 Tabi.k 14. — Zooplankton volumes, rc./WOOm.', adjusted for hour of hauling, for Hugh M. Smith cruises S, 5, 7 and 8 \¥oT unadjusted volumes together with station position, date, lime ot hauling and description of method, refer to King and Demond (1953)] Station Xo. Cruise 2 Cruise 5 Cruise? Cruises Station Cruises 10.4 13.9 23.0 14.2 22.8 19.1 48.2 23.7 19.0 14.9 8,4 32.1 31 3 27.0 17.9 25.4 28.6 25.6 41.5 33.0 37.3 57.7 49.3 86.5 46.9 54.2 28.2 27.0 8.4 29.4 41.5 24.4 29.0 73.0 35.3 76.7 57.3 79.8 71.8 36.6 29.2 25.6 42.8 24.2 27.9 26.4 17.3 22.0 2n.fi 22.2 42.7 23.6 32.0 97.5 37.7 16.2 25.0 39.9 24.8 47.2 36.3 35.6 32.7 18.7 13.5 20.6 77.1 64.0 43.5 75.3 87.5 25.8 54.6 17.9 17.7 20.0 12.9 14.0 53 54 55 56 57 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 24.9 25.2 17.8 16.7 30.3 26.9 27.6 22.5 10.9 11.1 18.5 23.5 15.9 46.2 24.4 12.4 16.3 19.9 16.9 24.1 28.8 67.9 40.1 29.0 22.3 47.4 29.2 11.6 19.1 - 9.4 25.9 21 0 rj .. ... 9.9 8.3 13.7 25.5 31.1 40.8 18.6 36.1 ]4 _ ._. 24.4 37.8 29.2 16 23.5 41.2 24.1 18 17.5 19 _ 56.5 24.4 12.8 21 65.3 14.1 16.2 2"^ ... - 42.9 16.9 15.3 26.7 21.4 14.1 18.4 11.1 18.3 42.9 13.3 27.7 28.5 18.5 44.8 27.5 25.1 48.0 31.5 44.3 44.9 73.1 38.5 51.9 92.5 37.8 27.6 75.9 33.7 21.6 11.6 11.4 13.3 8.8 14.1 17.7 15.0 18.2 21.3 14.9 9.7 11.2 46.6 13.2 13.9 10.8 9.9 12.2 3.7 23.2 10.7 2.9 13.2 13.6 31.2 U. S. GOVERNMENT PRrNTING OFFICE : 1957 O - 418106 EARLY DEVELOPMENT, SPAWNING, GROWTH, AND OCCURRENCE OF THE SILVER MULLET (Mugil curema) ALONG THE SOUTH ATLANTIC COAST OF THE UNITED STATES By William W. Anderson, Fishery Research Biologist The biology, cliemistry, and physical oceanog- raphy of the waters adjacent to the coast of the United States from Cape Hatteras, N. C, to lower Florida arc little known. Collections by the Fish and Wildlife Service motor vessel Theodore N. Gill provided basic material and data for the study of these conditions as a result of a series of cruises made during 1953 and 1954. This paper presents the findings on spawning, early development, growth, habits, and occur- rence of one species. The data contribute to the tremendous task of determining the identity and biology of the numerous fish larvae, a knowledge of which is prerequisite to the understanding of fish populations in the area. Opportunity to add to the meager information available on spawning, early life history, growth, habits, and the occurrence in offshore waters of the silver mullet, Mugil curema Cuvier and Valen- ciennes, began off the coast of southern Florida. A school of spawning individuals was located (fig. 1) and several running ripe females and ripe males were taken in dip nets. The eggs were fertilized, and during the next several days a developmental series of eggs and larvae was obtained. This series was carried forward with specimens about 2.5 mm. to 5 mm. long, from material taken in routine plankton tows; dip-net collections pro- vided a series of specimens ranging in length from about 6 to 25 mm. Finally, our seine collections from the outer beaches and estuarine marsh areas in Georgia extended the series to juveniles up to about 120 mm. in length. The silver mullet is known on the Atlantic coast from Cape Cod to South America, and on the Pacific coast from about Chile to the Gulf of California. Along the Atlantic coast of the United States it is taken in commercially significant quan- tities only in Florida, where it constitutes about one-twentieth of the total mullet landings of about .30 million pounds annually. The principal Note.— Approved (or publication Februarys, 19i7. Fishery Bulletin 119. fishing gears are gill nets, trammel nets, beach seines, and stop nets (Idyll 1949). I greatly appreciate the assistance of George A. Rounsefell and Frederick H. Berry for critical review of the manuscript. My special thanks are extended to Jack W. Gehringer for review, also for numerous other aids during the study. METHODS Eggs were fertilized in several small culture jars and in a glass aquarium of several gallons' capac- ity. It was not necessary to strip either the females or males as eggs and milt were running during the handling of the fish. Sea water from the spawning location strained through No. 1 plankton silk was used as the culture medium. For some unknown reason the eggs in the large aquarium failed to develop, but hatching was accomplished in all of the culture jars. In an effort to carry some of the larvae through at least the end of the yolk-sac period, they were handled in several ways: some were retained without changing water in the original jars in which fertili- zation took place; some were placed in new water from wherever the ship was at the time; but the majority were removed to fresh jars containing a new supply of strained water from the spawning location. The latter procedure met with the best success. Most of the larvae were dead after 36 hours, but one larva survived for 45 hours. After fertilization a series of eggs was preserved every 2 hours for the first 12 hours and then every 4 hours until hatching at 40 hours. Larvae were preserved at hatching, and then 4, 10, 16, 23, 32, 42, and 45 hours after hatching. All eggs and larvae were preserved in 5 percent buffered formalin. Biological, oceanographic, and chemical methods used aboard the Theodore N. GUI, general work plan, cruise plans, and objectives are given by Anderson, Gehringer, and Cohen (1956). 397 398 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 82" 81° 80° 79° 78° 7 7° 76° 75° 74° 7 3° 3 5° NORTH CAROLINA v^ ^f .-^ ^'' . 34° SOUTH CAROLINA /^ }^ / / ( pjy / 33° 32° o* / xr / / W _7 ;■■■ ■ / / 31° - _T X S' ^ — I .^ 9 / 1 9 9 30° - \ \ ; 1 "^ \ X 0 IX X \ '. 1 A 20 FATHOMS \ -.It \ "'I 100 FATHOMS 29° \ X :0' X| X ^ \ . \ ^ AXIS OF GULF STREAM 28° \ • i ' X 27° ' ft ^=:;^^^, 1 /i' 1 ' i? 2»« 25° V-^'/i 'I Vi ^^ 1 \)i 1 1 Figure 1. — Capture of silver mullet larvae in plankton tows indicated by circles; dip-net collections are shown by x's. (See tables 1 and 2 for occurrence data.) General location of seining shown by blacked-in area on Georgia coast; spawning school of silver mullet indicated by black dot off lower Florida coast near 20-fathom line (Theodore N. Gill, station 4). SILVER MULLET 399 Soine collections to supplement the material and (lata collected with the MA' Theodore N. GUI in offshore waters have been made on a semi- monthly basis since 1953 at three localities in Georgia. One station is on the open ocean beach on St. Simons Island, a second is in the marshy estuarine area behind the barrier islands, and the third is up the Altamaha River at about tidewater limits. Oidy beach and marsh material are in- cluded in this paper. All descriptions of eggs, larvae, and juveniles are based on preserved material urdess otherwise stated. Measurements of eggs and the larvae up to about 25 mm. in length were made with a stereoscopic microscope and a micrometer eyepiece. The larger specimens were measured with calipers. Both standard lengths (in small specimens from the tip of the snout to tip of the urostyle) and total lengths (in small specimens from the tip of the snout to tip of the finfold or caudal fin) were determined. For discussions of body proportions only standard lengths were used. In other dis- cussions, larvae less than 7.0 mm. standard length are referred to in total lengths, and those 7.0 mm. and larger in standard lengths (figs. 3 through 16 follow this procedure). Original measurements were used to construct the curves portraying rates of growth of various body parts, and changes in certain body propor- tions. I have followed the general approach used by Ahlstrom and Ball (1954) in presenting the larval development in that sequences of fin formation, body proportions, and pigmentation are discussed. Egg development and yolk-sac larvae are pre- sented separately. DEVELOPMENT OF THE EGG The pelagic eggs of the silver mullet are splieri- cal in shape and contain single, large oil globules. In unfertilized eggs (fig. 2a) the yolk appears as an unsegmented opaque mass with little, if any, perivitelline space. The oil globule is pale yellow and located at the top of the yolk mass. The surface of the eggshell has a finely scratched or etched appearance. A series of 10 unfertilized eggs ranged in diameter from 0.77 to 0.86 mm., witli an average of 0.82 mm.; and tlic diameter of the oil globules ranged from 0.27 to 0.32 mm., with an average of 0.30 mm. Two hours after fertilization (fig. 2b) the eggs had reached the 32-blastomere stage and had developed a perivitelline space ranging from 0.04 to 0.12 mm. wide. A series of 10 eggs at this stage ranged in diameter from 0.86 to 0.92 mm., with an average of 0.90 mm.; and had oil globules with diameters ranging from 0.27 to 0.32 mm. and averaging 0.30 mm. The small increase in average diameter of the egg appears to result from absorption of water with an accompanying expansion of the eggshell and development of the perivitelline space. From this stage of develop- ment until hatching the diameters of the eggs and oil globules maintain about the same range of sizes and averages. The perivitelline space and yolk mass vary as the embryo develops and some of the yolk material is used up (fig. 2, c to i). Four hours after fertilization (fig. 2c) the blasto- disc was well formed and berrylike in appearance. The segmentation cavity was present 8 hours after fertilization (fig. 2d), and the embryonic shield was well advanced in 12 hours (fig. 2e). The embryo was well differentiated 16 hours after fertilization (fig. 2f). The optical vesicles are well defined, eight somites are visible, and the blastopore is closed. The tail has not begun to separate from the yolk, which at this =tage has a granular appearance. Irregular lines of pig- ment spots are present on the dorsal surface of the embryo, one on each side of the notochord, extending from just behind the head onto the tail section. At 24 hours after fertilization (fig. 2, g, h) pupils have developed in the large eyes, 24 myo- meres are discernible, and the tail has started to separate from the yolk mass which remains granular in appearance. Melanophores in the rows of pigment spots, on the dorsal surface of the embryo (one row each side of center line) are now more closely set and extend from just back of the eyes to that portion of the tail which is free from the yolk mass. A few scattered melanophores appear on the sides of the embryo. After 32 hours (fig. 2i) the embryo has a well- developed finfold and the tail free for about one- third the length of tiie body. In addition to the dorsal rows of pigment spots, melanophores are present on the ventral aspect of the embryo and are more numerous on the sides of the bodv. 400 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Figure 2. — Various stages of development of silver mullet eggs: a, unfertilized eggs; b, 2 hours after fertilization (32 blastomeres) ; c, 4 hours after fertilization (blastodisc well formed, cells small) ; d, 8 hours after fertilization (seg- mentation cavity forming); e, 12 hours after fertilization (early embryo); f, 16 hours after fertilization (embryo); g, 24 hours after fertilization (lateral view of embryo) ; h, 24 hours after fertilization (top view of embryo) ; i, 32 hours after fertilization (lateral view of embryo). SILVER MULLET 401 YOLK-SAC LARVAE Hatching: began 40 hours after fertilization, and no increase in number of larvae was evident after 42 hours. The larvae floated at the surface, and movement consisted lar^jely of occasional jerking actions. Only 3 newly hatched larvae were preserved, and these had a size range of 1.63 to 1.76 mm. total length, and an average of 1.69 mm. The larae hatch in an undeveloped state, lacking a mouth and fins, and without pigment in the eyes (fig. 3). The large oil globule is located, in a large part, in the posterior half of the yolk sac. Pigmentation consists of ventral rows of pigment spots, in addition to the dorsal rows, and a few scattered melanophores on the sides of the body and head. Four hours after hatching (fig. 4) the body had lengthened (11 specimens ranged in total length from 1.74 to 2.15 mm., an average of 1.89 mm.) and there was considerable shrinking of the yolk mass; the finfold was more developed and begin- ning to constrict in the caudal region. Pigmenta- tion remained essentially the same. At 16 hours (fig. 5) there was a further increase in length (7 specimens ranged in total length from 2.30 to 2.47 mm., an average of 2.36 mm.), the yolk mass was much reduced, the finfold had reached its ma.ximum development, and no basic change in pigmentation had occured. The larvae were dying rapidly after 32 hours, and it was apparent none would survive much longer. Consequently 6 of the more active ones were preserved at this stage (ranging in total length from 2.49 to 2.68 mm. and averaging 2.56 mm.). Other than this increase in length, a smaller yolk mass, and minute pectoral fin buds; tliese larvae were similar to those at 16 hours after hatching, and are not figured. All larvae were dead 45 hours after liatching, and the last to die appeared to have been infected by fungus. For this reason I have not considered specimens older than 32 hours. The develop- mental series is continued with material from [)lankton tows. Two specimens from the plankton material, both 2.56 mm. total length and from the same tow! indicate that the yolk-sac larva begins a rapid transition at about 2.5 mm. total length. In the least developed of the specimens (fig. 6), the larval hump has disappeared and small pectoral buds are evident. The oil globule has shrunk, but remains comparatively large, and the yolk mass has decreased. The mouth is not diflFerentiated, and the eyes have no pigment. Body pigmenta- tion remains essentially the same, irregular dorsal and ventral rows of pigment spots with a few scattered melanophores on body and head. The finfold is more constricted in the caudal region. The second 2.56-mm. specimen (fig. 7) has a distinct head, and the mouth has developed. The eyes are pigmented, and tlie pectoral fin has enlarged into a fleshy-based, fan-shaped fin FiGCRE 4.— Yolk-sac larva, 2.15 nir FiGCRE 5.— Yolk-sac larva, 2.47 mm. FiGi RE 6. — Yolk-sac larva, 2.5(1 mm. FiGCRE 3.— Xi'wly liatclicd larva, 1.7C> mm. 4-.>:n82 0— 57 2 FiGLRE 7.— Yolk-sac larva, 2.56 mm. 402 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE without rays. The oil globule is still visible, but the yolk mass has been largely absorbed. Pig- mentation is basically unchanged. Thickening of the head and of the body anterior of the anus is very noticeable. I believe the yolk-sac stage ends at about this period of development. DEVELOPMENT FROM LARVAL TO JUVENILE STAGES Development from about 3.5 to 110 mm. is included as larval, early juvenile, and juvenile. A developmental series is illustrated in figures 8 to 16. Each set of characters will be discussed separately and development briefly outlined. DEVELOPMENT OF FINS Caudal. — A fully developed caudal fin has 14 principal raj's (of which 12 are branched) and 14 Figure 8. — Larva, 3.7 mm Figure 9. — Larva, 4.0 mm. Figure 10. — Larva, 4.7 mm. Figure 11. — Larva, 5.3 mm. to 17 secondary rays. For details of caudal skele- ton of M. curema, see Hollister 1937. At about 3.7 mm. total length (fig. 8) there is no indication of development of the caudal fin other than a deeper constriction of the finfold in the caudal region. A thickening is evident on the ventral side of the urostyle at about 4.0 mm. total length, but no true rays have formed (fig. 9). At about 4.7 mm. total length (fig. 10), 14 rays have developed ventrally which will be the principal caudal rayo, the urostyle has tipped upward, and the fin shape has changed. The caudal has de- veloped a well-rounded form at about 5.3 mm. total length (fig. 11), the urostyle has reached its maximum flexing, and about 23 rays are visible. The fin is much broadened at about 7.0 mm. standard length (fig. 12), and 25 or 26 rays are present. A full complement of rays (14 principal and 15 secondary) is visible at about 14.5 mm. (fig. 13), although none of the principal rays are branched ; and tlie fin has begun to fork. Branch- ing of the 12 principal rays (2 principal rays do not branch) has occured by 25 mm. standard length, and forking seems complete by 110 mm. (figs. 14, 15, and 16). Dorsals. — When fully developed the first dorsal has 4 spines, and the second dorsal has 1 spine and 8 branched rays. The dorsal bases are first evident and developing raj's visible in both fins when the larvae are about 4.7 mm. total length (fig. 10). The 4 spines of the first dorsal and the 1 spine and 8 soft rays of the second dorsal are quite evident at about 5.3 mm. toial length (fig. 11). The last ray of the second dorsal is branched by 14.5 mm. standard length, 7 of the 8 soft rays are branched by 25 mm., and all are branched by 50 mm. Final fin shape is reached when the juveniles are between 50 and 110 mm. long (figs. 13, 14, 15, and 16). Anal. — The fully developed anal fin has 3 spines and 9 branched soft rays. Developing rays and the anal base are first evident in larvae about 4.7 mm. total length (fig. 10). The full complement of rays (12) is present in larvae about 5.3 mm. total length (fig. 11). The two spines are discernible and the last ray has branched by a larvae size of about 14.5 mm. (fig. 13). (Young silver mullet have 2 spines and 10 soft rays, rarely II, 9: the third spine develops from the first ray which starts as a seg- mented ray and fuses into a spine when the fish is SILVER MULLET 403 about 30 to 40 mm. long, and I consider the larval period to end at this time.) The last 9 rays are l)ranched in specimens about 25 mm. long (fig. 14). Final fin shape is reached in juveniles between 50 and 110 mm. long (figs. 15 and 16). Pectorals. — The pectoral fin buds are first evi- dent in yolk-sac larvae about 2.5 mm. total length, but the first rays do not develop until the larvae are about 5.3 mm. total length, when about 5 or 6 rays are present in the upper part of the fin. The number of rays increases to about 10 by a larval size of 7.0 mm. standard length, and the full complement (15 to 17) is reached by a larval size of 14.5 mm. Ray development and changes in fin shape at various larval and juvenile sizes are demonstrated in figures 6 to 16. Figure 12. — Larva, 7.0 mm. V. , ^..•.n.,v/•>.vV,r•..J,.»•A-:^/,;^-r-:^^^ ^■;.i r *g Figure 13. — Larva, 14.5 mm. Figure 14. — Larva, 25.5 mm. 404 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Ventrah. — The ventral fin buds are first evident at a larval size of about 4.0 mm. total length. Rays are visible by 7.0 mm. standard length, and 1 spine and 5 rays are present by a larval size of 14.5 mm. (figs. 9 to 13). BODY PROPORTIONS As occurs in many fish, the greatest changes in body proportions of the silver mullet take place between hatching and a length of 4 to 5 mm. The head length, eye diameter, and body depth (at pectoral) all increase at a rapid rate during the initial period; slow down when the larvae are between about 5 and 25 mm. long; and stabilize at a constant rate from 25 mm. to adult size. These relations are demonstrated in figures 17, 18, and 19. The slight upward shifts in the regression lines for head length and body depth (at pectoral), which occur in fish about 25 to 30 mm. long, take place at the time the young mullet leave the open ocean and move to beach and estuarine habitats. I believe that a greater abundance of food avail- able on these inshore nursery grounds results in an initial rapid growth reflected in these body parts. Young mullet, between 20 and 25 mm. long, from the open ocean appear thin and never seem to have full bellies, whereas young mullet taken from the estuarine areas usually have stomachs somewhat distended. The distances from snout to insertion of first dorsal fin, snout to insertion of second dorsal fin, and snout to insertion of anal fin increase at a remarkably uniform rate from a larval size of 3-4 mm. to adults. These relations are given in figures 20 and 21. Jacot (1920, p. 223) said that the developmeni, of M. curema was much as in M. cephalus but without a definite silvery stage and with a con- stant rate of development of the various parts and of the individual. However, Jacot's smallest specimens were 20 mm. long, and he gives no other measurements. J Figure 15. — Juvenile, 50.0 mm. Figure 16. — Juvenile, 110.0 mm. SILVER MT7X.LET 405 PIGMENTATION Pigmentation of the embryo and of the larvae through the yolk-sac stage has been described and illustrated. At about 3.7 mm. total length, the pigmentation continues to consist essentially of pigment spots along the dorsal and ventral aspects, with scattered melanophores on the head, sides of body, and throat (fig. 8). Development of pigmentation from this stage is a process of inten- sification and spreading onto the head and sides of the body, so that by a larval size of about 14.5 mm. the specimens are so densely covered by large and small pigment spots as to appear almost black (this is especially intense on the caudal peduncle). No pigment is present on any of the fins during this period. The development of ^ I 1 1 MM n 1 1 1 1 1 1 M 1 1 1 1 1 MIL 80 — ■■ 60 - ■/ — 40 - •^ - 20 10 J ' 8 : f — 5 5 6 - J - z «r^ X 4 1- - ^ •" o ^ J/' «■ z • 5- t^ liJ _1 Jf^ o 2 — ^/\ "■ < jj' 111 z * X 1 - — .8 _ * r^ ^ .6 ^ t _ .4 - - .2 " 1 1 1 1 1 1 1 1 1 f 1 1 1 1 1 1 1 1 1 1 1 1 1 11 1 2 4 6 8 10 20 40 60 80 100 200 400 600 STANDARD LENGTH IN MM. Figure 17. — Relation of head length to standard length.' ' Specimens below 25 mm. long are plankton and dip-net material from open ocean; ttiose 27 to 122 mm. are seine material from marsh and beach are-is and those 151 mm. and larger are the specimens from the spawning school. 406 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE pigmentation between larvae lengths of 3.7 and 14.5 mm. is illustrated in figures 8 to 13. By the time young silver mullet have reached a length of 25 mm., the pigmentation has decreased in intensity. The specimens have a peppered appearance, with lighter pigmented areas appear- ing on the lower part of the head and belly, and a scattering of pigment spots on the dorsal and caudal fins (fig. 14). Juveniles at a length of about 50 mm. have so few pigment spots on the belly that it is beginning to appear white, and pigment is present on the anal fin (fig. 15). Juveniles 110 mm. long are heavily pigmented on the dorsal surface of the body to about the mid- line, where the intensity of color decreases rapidly so that the lower third of the body from head to caudal fin is silvery or white; head pigmentation is largely on the dorsal surface, but patches of 1 1 I I ( M I 1 r I I MM 1 I I I I I I I 80 60 - 40 20 1 5 10 8 ui < uj 1.5 >- lij J I I I I I I J_ I t M IJ- J I I t I I ' See footnote 1, p. 405. 16 8 10 20 40 60 80 too STANDARD LENGTH IN MM. Figure 18. — Relation of eye diameter to standard length.^ 200 400 bOO SILVER MULLET 407 pigment are present under the eyes and on the opercles; and all of the fins have pigmentation (fig. 16). At this size both freshly caught and preserved specimens appear blue-black on the dorsal surface grading to silvery white bellies. Jacot (1920) observed that M. curema does not go through a definite silvery stage such as occurs in M. cephalus. (His smallest specimens were 20 mm.) SCALES, PREORBITAL, AND TEETH Scales. — A detailed account of scale develop- ment and characters in young of both M. curema and M. cephalus was given by Jacot (1920). Preorhital. — The serrated or toothed preorbital bone (common to all Mugilidae) becomes visible when larvae are between 7 and 14 mm. long. There is wide variation in the number and size of the serrations. _ 1 I 1 1 1 1 II 1 1 1 1 1 II II 1 I 1 1 1 1 1 '_ 80 ~ — 60 - / — 40 - y / - 20 - •« — 5 j/^ 5 J^ 2 ''^ - 10 — _ i 8 — >% ^ CE — — O 6 » _ (- ' Q. 4 - < - y. - X — • 3^ _ LU ^^ Q -y • >- Q ■I O 1 — •r m — /'• _ .3 ~" y^ ft- — .6 .» / ^_ .4 - /• \ - .2 1 1 1 1 1 1 1 1 1 1 1 1 1 Mill \ 1 1 1 1 II 1 200 ' See footnote 1, p. 406. 4 6 8 10 20 40 60 80 100 STANDARD LENGTH IN MM. Figure 19. — Relation of body depth at pectoral to standard length.* 400 600 408 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Teeth.— Schultz (1946, p. 389) states with re- gard to teeth in Mugil, "the teeth in the lower hp are setiform or ciliform, partly embedded or con- spicuous ; teeth in upper hp similar. The outer row of teeth in both Hps is usually more prominent, with simple tips, and if inner rows occur these are either bifid or trifid, at least on adults (apparently the teeth in certain species of this genus become bifid or even trifid in large adults)." He further stated, "In this genus I find that in small speci- mens of certain species the teeth have simple tips, but later the inner teeth have bifid tips and in the largest adults some possess trifid tips. The teeth of the outer row usually have simple tips, but in some large specimens these are bifid too." Teeth develop in young silver mullet at about 1 1 1 1 Mill 1 1 1 1 1 Mil 1 1 1 1 MIL 800 : — 600 — — s 400 _ I s. z - - CO z 200 — - u. :•■•' _l • * < OC. 100 - •• — O Q 80 - a • - d 60 ^ , • _ z ? * CM ~ • — Q 40 . .a" ^ ^ z < — •• H .•' • w — 20 — 1 _ u. * t" o y ^ m w 'f - z o 10 ^ ''* y _ l- V J^ ^ UJ 8 - • - Z 6 I • * • mm / • o 4 _ • • • _ h- • • • 3 ^ ^ O * z •• to 2 1 1 1 1 Mill 1 1 i 1 1 III! 1 1 1 Mill 10 20 40 60 60 100 STANDARD LENGTH IN MM. 200 400 600 FiGUKE 20. — Relation of the distances from snout to insertions of first (lower line) and second (upper line) dorsal fins to standard length.* * See footnote 1, p. 40S. SILVER MXJLLET 409 20 mm. standard length. In stained material cf this size the teeth of the upper lip are in a single series, have simple tips, and number about 12 to 15 on each half of the lip (about half protrude from the fleshy lip). There are no teeth in the lower jaw. At about 30 mm. the teeth in the single row in the upper lip are larger, number about 30 to 35 in each half of the lip, all protrude from the lip, and have simple tips that curve inward. The lower lip has a single row of simple-tipped teeth which are smaller than those of the upper lip. These point straight outward from the thin lip edge, number about 30 on each half of the lip, but barely protrude beyond the lip edge. In specimens about 50 mm. long the simple- tipped teeth in the single row in the upper lip number about 30 on each half of the lip, but now appear to have flattened tips which are bent _ 1 1 1 1 Mill 1 1 1 1 1 MM 1 1 1 I 1 1 1 '_ 800 — - 600 — - 400 - - z s 200 — - z ^ . — ..».• z .»• b. 100 - • - _l 80 _ -• ^ < • Z ■" • ~ < 60 ™ y — u. « ^ o z 40 — • - o 1- (T •• ill • V) 20 — • _ z fr o ./ t- t- 10 - /• _ o M * . o 8 Ub • • • . z _ ^ w 6 4 2 - • 1 I • • • •*• • 1 t 1 • 1 III 1 1 1 1 1 1 1 1 1 1 1 1 i 1 1 1 1 1 200 2 4 6 8 10 20 40 60 80 100 STANDARD LENGTH IN MM. Figure 21. — Relation of distance from snout to insertion of anal fin to standard length.' 400 600 ' See footnote 1, p. 405. 410 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE sharply inward (having a resemblance to a leaf rake). The simple-tipped teeth in the single row of the lower lip are close set, number about 50 in each half of the lip, but still barely protrude from the thin-edged lip. At about 100 mm. length there are two rows of teeth in the upper lip. The larger teeth in the outer row number about 50 to 55 on each side and have simple tips which are somewhat flattened and incurved. The teeth of the inner row are much smaller than those of the outer row, have simple tips, and number about 30 in each half of the lip. The simple-tipped teeth in the single row in the lower lip are very close set, number about 75 to 80 in each half of the lip, point forward from the thin-edged lip, but do not protrude much beyond the margin of the lip. GROWTH SPAWNING their first year reach a length around 200 mm., and may be maturing at that age. The second growth line on figure 22 relates to a group of young from later spawning which is present in the beach and marsh areas in August (this represents the last major recruitment). Starting at 20-mm. length and applying the 17 mm. per month growth rate, this group could be expected to reach a size of about 105 mm. in December. The third growth line relates to the latest spawned stragglers which are about 20 to 30 mm. long in October and reach a size of about 70 mm. by December (fig. 22). The effects of winter water temperatures on growth of silver mullet in Georgia cannot be followed since the young appear in April when the water temperatures have warmed and apparently leave this coast when waters cool in late fall and .,^, , ,, ,. .. 1 J earlv winter (table 5) . Although the samphng program was not planned specifically for this study, growth of mullet may be estimated for Georgia from material from the seine collections. Three growth lines are indicated Little has been published on the spawning habits in figure 22. The first is in reference to the earliest of the silver mullet. Based on presence of the spawned group which appears first on Georgia young (20 mm. and up) in estuarine waters of beaches late in April (earliest record is April 22). North Carolina, Jacot (1920, p. 226) suggested a Assuming that from 3 to 4 weeks were required for rather protracted spawning period of mid-April to larvae to reach a standard length of 17 to 24 mm., mid-August with a peak about May. I relate the April recruits to spawning in late March or early April. From April to October the growth ''^:,i^^Z&::T;^rll::^'i^^^ can be followed from the upper limits of the size s, and 6 ranges (after October larger juveniles appear to move out of the area). This growth, computed from time of hatching, apparently progresses at a rate of about 17 mm. per month, so that juvenile fish about 120 mm. long late in October would have been spawned about April 1 and are about 7 months old. Hildebrand and Schroeder (1928, p. 197) state that M. curema in southern waters has an average length of about 10 inches (250 mm.) and a maxi- mum length of about 14 inches (350 mm.). As the mature individuals taken from the spawning school off the Florida coast averaged 189 mm. for the males, 209 mm. for the females, and about 198 mm. for males and females combined, they apparently were relatively young fish. If we extend the growth line of about 17 mm. per month At GUI regular station 4 off the southern Florida beyond October and to a size of 200 mm., it seems coast near the 20-fathom line (fig. 1), between the quite reasonable that silver mullet at the end of hours of 2200 and 2315 on April 25, 1954, thou- Sta- tion No. Date Position Number of larvae No. N. Lat. w. Long. 2 to 5 mm. 6.1 to 7mra. 7.1 to 10 mm. 2 3 14 19 25 26 31 42 43 48 49 54 59 61 63 65 75 8 3 14 Apr. 23, 1953 Apr. 25, 1953 Apr. 26. 1953 Apr. 26. 19,53 Apr. 27, 19.53 Apr. 27, 1953 May 5, 1953 May 6, 19.53 Mav 6, 1953 May 6, 1953 May 7, 19.53 Mays, 1953 Mav 8, 1953 May 8, 1953 May 9. 1953 Mav 10, 1953 Julv 26, 19.53 Apr. 25, 1954 Apr. 27, 1954 27° 01' 29° 01' 29° 39' 30° 20' 30° 20' 31° 00' 31° 57' 32° 12' 32° 23' 32° 12' .33° 03' .33° 22' 32° 53' 33° 15' 33° 42' 34° 39' 28° 18' 27° 00' 29° 00' 80° 04' 80° 08' 80° 23' 80° 35' 80° 12' 79° .59' 79° 18' 79° 33' 78° 43' 78° 25' 78° 21' 77° 37' 77° 04' 76° 23' 76° 56' 75° 53' 79° 26' 80° 03' 80° 10' 2 1 5 4 15 2 5 8 1 3 7 2 2 2 2 2 2 2 "i" 1 2 2..- 2 2 3 6 6 3 1 11 2 .- 1 1 1 SILVER MULLET 411 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 APRIL — — MARCH FEB. — JAN. — y y y y y DEC. — y ~~ y y y NOV. ^^ /... ^^^^^"^^ ^ y y y OCT. — __- — ^y^ y SEPT. ^-t^^"^ ^*^ ^^^^■^^*^* AUG. — ^*— «?^^^^ - -p. — JULY _...-. .^ JUNE ■■— -py^ MAY — -py^ — APRIL — ^ y^ • ' 1' i 1 1 1 i 1 1 \ 1 1 1 1 1 1 1 1 1 i 1 10 20 40 60 80 100 120 140 160 180 200 STANDARD LENGTH IN MH. Figure 22. — Monthly size ranges of young silver mullet, M . curema, taken from two seine locations in Georgia (horizontal solid line is St. Simons Beach, and horizontal dotted line is Sapelo Marsh). Data of March 1953 to January 1956: the size ranges represent the smallest to the largest specimens taken during any given month in this period. When only a few isolated specimens are involved they are indicated by a single black dot (see table 3 for occurrence data). Three growth lines are indicated; the first (lower) relates to young, starting at about 20 mm. length, in April; the second (middle) relates to young, starting at about 20 mm. length, in August; and the third (upper) relates to young, starting at about 20 mm. length, in October. 412 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 2. — Numbers of larvae of the silver mullet, M. curenia, at various sizes taken by dip netting on Gill cruises 1 through 9 (none were caught on cruises 1, 5, and 9) Station No. Date Position Number ot larvae Cruise No. N. Lat. W. Long. 6 to 10 mm. 10.1 to 15 mm. 15.1 to 20 mm. 20.1 to 25 mm. Over 25 mm. 2 Spc. 5 Apr. 17.19.53 Apr. 19,1963 Apr. 23,1953 Apr. 24,1953 Apr. 26, 1953 Apr. 26, 19.'.3 Apr. 27,1953 Mav 4, 1963 May 5,1953 May 7,1953 May 8,1963 May 11,19.53 Julv 23,1953 July 25,1963 July 28.1963 July 29.1953 Aug. 11,1953 Aug. 12, 1953 Oct. 13,1953 Oct. 13,1953 Oct. 14,1953 Oct. 1.5,1963 Oct. 16,1953 Oct. 26,1953 Apr. 25, 19.54 Apr. 27,1954 Jun. 25.1964 Sept. 12, 1954 Sept. 13, 1964 30° 00' 26° 19' 27° 01' 29° 00' 29° 40' .30° 20' 31°00' 31° 21' 31° 38' 33° 03' 32° 53' 35° 13' 26° 21'. ... 27° 00' 30° 20' 30° 18' 34° 32' 34° 18' 27° 40' 28° 18' 29° 00' .30° 17' 31° 00' 32° 34' 27° 20' 29° 00'.... 29° 38' 28° 17' 29° 40' 77° 00' . 1 1 1 1 1 2 Std. Sta ... 76° 44' 2 .... 3 80° Oi' 1 2 2 13 20 to 21 80° 33' . 2 80° 57' ... 2 28 80° 36' 1 3 1 2 31 79° 59' 2 . . 35 80° 65'. ... 1 1 2 37 80° 14' 1 2 54 78° 21' 77° 04' 76° 32' 2 7 2 61 2 Cape Hatteras 9 5 3 Std. Sta 76° 46' 1 3 3 80° 03' 1 3 1 2 3 26 80° 36' 3 26 80° 12' 76° 49' 1 3 69 11 1 3 70 76° 32' 4 7 79° 18' . . 1 8 79° 26' 79° 48' 80° 11' 1 1 M 16 A 26 1 1 32 80° 23' 52 77° 48' 2 6 4 80° 03' 1 1 16 79° 26' 1 7 17 79° 36' 79° 28' 1 8 8 1 8 18 80° 00' 1 1 sands of mullet were around the vessel. On occasions paired fish were observed swimming side by side, but in general there was just a dense mass of fish milling and splashing at the surface. These fish were very difficult to catch with a dip net, but 12 specimens were captured (6 males and 6 females), which proved to be silver mullet, Mugil curema. Standard lengths of the males were 151, 176, 186, 189, 203, and 228 mm.; and of the females, 194, 195, 202, 210, 217, and 234 mm. All of the females were running eggs, and all the males were running milt; in fact the specimens coidd not be handled without eggs or milt stream- ing from them. This was a spawning school. Plankton tows were taken February 16 to March 7, 1953, during cruise 1; and April 22 to May 14, 1953, cruise 2; July 16 to August 12, 1953, cruise 3; October 7 to November 12, 1953, cruise 4; January 21 to February 23, 1954, cruise 5; also, April 15 to April 28, 1954, cruise 6, a partial cruise. Collections were examined and yielded silver mullet larvae as shown in table 1 (no larvae were taken on cruises 1, 4, and 5); specimens were largely early stage larvae under 5 mm. long. Locations of capture are Olustrated in figure 1. All silver mullet material taken by dip netting on GUI cruises 1 through 9 is shown in table 2, and locations of capture are illustrated on figure 1 . These specimens ranged from about 6 to 25 mm. in length. Table 3. — Occurrence of young silver mullet (M. curema) in seine collections of two areas in Georgia [Arranged to show dates and approximate abundance; size ranges illustrated in figure 22] Date Sapelo Marsh Date St. Simons Beach Mav 6 1963 Number 11 Apr. 22, 1963 Number 9. M'iV 9 1965 Many Many Many Many Many Many 7... Many Many Many 14 Apr. 25, 1955 1. Mav 18 1954 May 4, 1954 4. May 20, 1953. -. Mavis, 1954 Mav 24, 1956 3. Many. June 3 1954 Jiine 3, 1953 1. June 4, 1953 June 9, 1955 July 3, 1953 20. Tnnp 9 1955 1. J imp 17 1954 July 21, 1953 3. July 21, 1954.... 7. Aug. 6, 1956 4. Julv 3 1953 \ug. 11, 1953 2. Julv 6 1954 15 Aug. 19, 1955. 1. 22 Aug. 26, 1953 6. Tulv 21 1953 2 Sept. 9, 19M 1. Tulv 21 1954 23 Oct. 5, 1956 15. 12 Oct. 14, 1955 1. Aug 5, 1955 18 Oct. 29, 1954. Nov. 3, 1955 1. Aue fi 1954 1 4 41. Nov. 16, 1956 Many. \ue 19 1955 8 Dec. 2, 1956 2. Aue 2fi 1953 1 Dec 19, 1955 3. 3 3 3 Mnv Ifi 1955 2 From lower Florida to North Carolina, early- stage larvae (2 to 5 mm.) were captured near the 20-fathom line and offshore to about the axis of the Gulf Stream, with the location of capture most frequently between the 20- and 100-fathom SILVER MULLET 413 lines on the outer Continental Shelf; none were taken on the inner Continental Shelf (fig. 1). These facts, together with the location of a spawn- ing school of mullet near the 20-fathom curve off the lower Florida coast, are strong evidence that the spawning of the silver mullet occurs at sea over the outer Continental Shelf from Florida to North Carolina. That the spawning season for silver mullet ex- tends from late March or early April until Sep- tember with peak spawning during April, May, and June is evident from several types of data: (1) No larvae were found in plankton tows from February 16 to March 7, 1953 (cruise 1); early stage larvae were taken in plankton tows from April 22 to May 14, 1953 (cruise 2); and one larva was found in plankton tows from July 16 to August 12, 1953 (cruise 3); (2) young silver mullet 20 to 25 mm. long were dip netted as early as April 17 off the lower Florida coast, and a very few under 10 mm. in length were dip netted as late as October (table 2) ; (3) the earliest date young silver mullet were taken along Georgia beaches by seine was April 22 (these ranged in length from 17 to 24 mm.); and some specimens from 20 to 30 mm. long were taken in October, although scarce after August (fig. 22 and table 3). Assuming a minimum of 3 weeks and a maximum of 4 weeks for the larvae to reach a size of about 20 mm. after hatching, the extreme dates of spawning I have suggested are reasonable. Peak spawning of April to June is evident from num- bers of young dip netted and seined during periods that would yield young from spawning during these 3 months (tables 2 and 3, and fig. 22). Spawning of the silver mullet begins with the rise of water temperatures in early spring. In table 4 we see that during late February and early March the surface temperatures over the Con- tinental Shelf in general averaged under 20° C, but exceeded 20° C. during April and May. Perhaps rising temperatures stimulated spawning activity and are more important than absolute temperatures, since spawning continues into summer when water temperatures are much warmer. MOVEMENTS AND HABITAT OF LARVAL AND JUVENILE FORMS From data contained in tables 1, 2, and 3, in- formation presented in figures 1 and 22, and Table 4. — Average surface water temperatures (° C.) and salinities (°/oo) for the several sections of the area from Cape Hatleras, N. C, to the Florida Straits for Gill cruise 1 (Feb. 10 to Mar. 10, 1953), Gill cruise 2 (Apr. 16 to May 15, 1963), and Gill cruise 3 (July 15 to Aug. 16, 1953) [The southern section comprises that part of the area Ivlng between Jupiter Inlet and Jacksonville. Fla.; the central section from Jacksonville, Fla. to the North Carolina-South Carolina boundary; and the northern section from this point north to Cape Hattenis. In general, the inner shelt com- prises waters lying between the beach and about halfway from the shoreline to the lOO-fathom line (mostly waters less than in fathoms deep); the outer shelf, the remaining area out to the inofathom line; and olTshore, that part of the area from the lOO-fathom line to beyond the axis of the Qulf Stream] Season and localitv Northern Central Southern °C. °/oo °C. °/oo °C. °/,. Winter- 13.6 19.4 M.O 20.7 21.7 25.2 27.7 27.7 28.3 34.5 35. P 36.3 3t.4 35.4 36.2 35.5 3.18 35.9 14.1 17.7 22.8 21.6 22.4 25.8 28.4 28.2 29.0 34..') 36.0 36.2 34.0 35.6 36.2 35.1 35.7 36.9 19.4 20.7 24.4 21.7 23.5 28.5 27.5 27.9 29.0 Outer shelf 36 2 Oflshore 36 2 Spring- Inner .shelf Outer .shelf .36.2 36 3 Offshore 36 1 Summer- 35 9 Outer shelf 36 0 Offshore 35 9 discussions on development and spawning, several facts relating to movements and habitat of larval and juvenile mullet from hatching to a length of about 120 mm. are evident. The larvae spend the first several weeks of their lives in the open ocean as far out as the axis of the Gulf Stream. At a length of about 17 to 25 mm. these larvae move inshore (no silver mullet more than 25 mm. in standard length was taken by dip netting in offshore waters). In Georgia waters (about center of the South Atlantic coast of the United States) these young are taken first on the beaches late in April and then in the marshy estuarine areas in May. Recruitment of young from the ocean con- tinues in volume until August, after which re- cruitment is erratic and small. The young live and grow in the rich estuarine habitat until October or November at which time they have reached lengths up to 120 mm. At this time the young mullet apparently move from the marshy estuarine areas to the outer beaches. By December they appear to depart from the beach area, and perhaps the Georgia coast, as the species was not taken in the seine collections from January until late April when the young (17 to 25 mm.) appear. Trends in temperature and salinity of the marsh and beach habitat are shown in table 5. Jacot (1920, p. 223), in regard to young M. curema at Beaufort, N. C, states, "At Beaufort they have not been recorded earlier than May 414 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 5. — Average monthly temperatures (° C.) and saHnilies (°/(;o) /<"■ the two areas in Georgia where seine collections were taken [The averages were derived from data obtained at time of seining and cover the period March 1953 to January 1956 (sec table 3)] Sapelo Marsh St. Simons Beach " C. "/oo ■> C /oo 12.8 13.1 20.8 23.7 26.3 29.1 31.9 30.4 27.2 22 2 m.i 12.4 23.1 19.7 22.1 19.7 20.2 23.4 26.3 26.8 20.7 23.1 31.2 25.6 11.4 14.6 17.8 21.1 26.2 27.3 29.6 30.0 28.0 23.8 19.9 14.2 28.2 30.1 30.2 April 28.0 May 29.3 32.2 July - 33. S 32.8 31.5 October 28.9 31.1 31.1 25th, but there is reason to beheve that they could be found even as early as late April." He further indicates (fig. 3, p. 203), the smallest specimens normally procurable are 20 to 21 mm. long, and specimens of that size occur until September. In summary, Jacot (1920, p. 227) makes these addi- tional general observations: (1) the young are abundant in the bays and estuaries of our Atlantic coast and develop rapidly; (2) in the fall the young school and migrate south; and (3) after their first year, white mullet are seldom caught north of Florida. Thus we have the same sudden appearance of the young in the beach and estuarine waters in both North Carolina and Georgia during the spring at a length which is comparable (17 to 25 mm.) and which coincides with the length (25 mm.) above which the young were not taken in dip-net collections at sea. In both North Carolina and Georgia the population of juvenile silver mullet, resulting from the previous spring spawn- ing, seems to disappear from the estuarine and beach areas by late fall or early winter, so that these young are apparently scarce or absent entirely during the winter. It is quite likely that a similar situation exists in South Carolina waters. Jacot thought the juveniles were migrating south and intimated they were moving to Florida waters. I have no direct evidence to substantiate or refute this theory. The only silver mullet more than 25 ir.m. long which I have taken in offshore waters along our South Atlantic coast were the mature and spawning fish off the lower Florida coast near the 20-fathom line. It is more logical to assume a southward migration along the Florida coast where a habitat ascribed to mullet in general is available than to assume an offshore movement. My guess is that the exodus of juveniles is hastened by cooling of the waters during the fall, but whether or not temperature is the major controlling factor is not known. Perhaps the young would have migrated even- tually regardless of temperature, but we cannot disregard the correlation of their disappearance with rapidly falling water temperatures in our marsh and beach areas during late fall and early winter (table 5). There does not seem to be a population of adult silver mullet in our inshore waters of the Carolinas and Georgia commensurate with the numbers of young there during spring and summer. LITERATURE CITED Ahlstrom, Elbert H., and Orville P. Ball. 1954. Description of eggs and larvae of jack mackerel (Trachurus symmeiricus) and distribution and abundance of larvae in 1950 and 1951. U. S. Dept. Interior, Fish and Wildlife Service, Fish. Bull., No. 97, vol. 56, pp. 209-245. Anderson, William W., Jack W. Gehringer, and Edward Cohen. 1956. Physical oceanographie, biological, and chemi- cal data. South Atlantic Coast of the United States, M/V Theodore N. Gill Cruise 1. U. S. Dept. Interior, Fish and Wildlife Service, Spec. Set. Rept. Fisheries No. 178, 160 pp. Higgins, Elmer. 1928. Progress in biological inquiries, 1926. U. S. Dept. Commerce, Bureau of Fisheries, Doc. 1029, In Rept. Comm. Fish, for 1927, pp. 517-681. Idyll, Clarence P. 1949. Stop-netting on the west coast of Florida. State of Florida, Bd. of Cons. Tech. Ser. No. 3, 23 pp. Jacot, Arthur Paul. 1920. Age, growth and scale characters of the mullets, Mugil cephalus and Afugil curema. Trans. Am. Microscop. Soc, vol. XXXIX, No. 3, July 1920, pp. 199-229, pis. XX-XXVI, 7 figs. Hildebrand, Samuel F., and William C. Schroeder. 1928. Fishes of Chesapeake Bay. Bull. U. S. Bureau of Fisheries, vol. 43, Part 1, 388 pp. Hollister, Gloria. 1937. Caudal skeleton of Bermuda shallow water fishes. II. Order Percomorphi, suborder Per- cesoces: Atherinidae, Mugilidae, Sphyraenidae. Zoologica, vol. XXII, Part 3, No. 17, pp. 265-279. Schultz, Leonard P. ■ 1946. A revision of the genera of mullets, fishes of the family Mugilidae, with descriptions of three new genera. Proc. U. S. Nat. Museum, vol. 96, No. 3204, pp. 377-395. U. S. GOVERNMENT PRINTING OFFICE : 1957 O— 4231B2 LARVAL FORMS OF THE FRESH-WATER MULLET (AGONOSTOMUS MONTICOLA) FROM THE OPEN OCEAN OFF THE BAHAMAS AND SOUTH ATLANTIC COAST OF THE UNITED STATES By WILLIAM W. ANDERSON, Fishery Research Biologist Among the Mugilidae in the dip-net collections, made dnring cruise 5 (Jan. 20 to Feb. 20. 195-1) and cruise 9 (Xov. 3 to Dec. 12, 1954) of the Fish and Wildlife Service motorvessel Theodore N. Gill, were 10 larval specimens of fresh-water mul- let. AgonoKtomuK monticola (Bancroft). I have found no previous record of the capture of this species from the open ocean. .1. manticola is a fresh-water species of mullet. Its habitat has been given by Jordan and Ever- mann (1896, p. 819) as, "fresh waters of the West Indies and eastern Mexico, Vera Cruz, etc." Evermann and Mareh (1902, p. 115) record it as found in "fresh waters of the West Indies and eastern Mexico." Sjjecific mention is made of its abundance in the fresh- water streams of Porto Rico. Meek and Hildebrand (1916, p. 335 ) give its luibitat as "Me.xico; Ontral America; botli slopes of Panama and the West Indies'"; mentioned specifically are specimens from Cuba and Guatemala. Beebe and Tee- Van (1928, p. 92) state that it is j'l'esent in "fresh-water streams of the West Indies and the east coast of Mexico"; mention is made also of specimens from fresh- water streams in Haiti. Schultz (1949, ^>\>. Ill and 112) lists the occurrence of specimens in the fresh-water rivers of Venezuela. .Jor- dan, Evermann. and Clark (1930, p. 254) state, that tliis species inhabits "fresh waters of the West Indies and the Atlantic and Pacific streams of Mexico and Central America." Carr and Coin (1955, p. 84) give its luibitat as snuill. swift streams, and its Florida range as, "streams of the Atlantic drainage and Pinellas Comity on the AVest Coast." I liave located lu) reference to A. monticola occuning in streams in tiie Bahamas. In addition to presenting occurrence records of the young of A. monticoht in ocean waters oft' the Bahamas and the South Atlantic Coast of the United States, this report gives a description of the yoimg, and presents tlie body proportions of hirval and juvenile stages. It compares certain 42;il81 o— 57 anatomical characters with the young of Mugil curetna Cuvier and Valenciennes and with Mugil cephalus Linnaeus, and also discusses the pos- sibilitj' that ^4. monti.cola is a catadromous fish. I appreciate the aid given by Giles W. Mead, Jr., and Frederick H. Berry in reviewing the manuscript. I also wish to thank Jack W. Gehringer for review of the manuscript and for other help during the coui-se of the study. My thanks also go to officials of the United States National Museum and the Museum of the Uni- versity of Florida for loan of specimens. MATERIAL Specimens from tlie Theodore N. Gill collec- tions were captured at 6 locations oft' the Bahamas and the Florida and Georgia coasts: Off Elbow Cay, Bahamas, 2(;°20' X. and 76°44' W., 1 specimen (28.1 nun. standard length) on Jan. 23 and 1 specimen (28.6 mm.) on Jan. 24. Of!' Eleuthera Island, Bahamas, 25°30' N. and 76°30' W., 2 specimens (24.1 mm. and 31.3 mm.) on Jan. 27, and 1 specimen (29.4 mm.) on Jan. 28. Oti' lower Florida coast, 28°00' N. and 79°00' W., 1 specimen (25.5 mm.) on November 16. Off central Floiida coast, 29°00' X. and 80°32' W., 1 specimen (30.6 mm.) on Nov. 17. Off central Florida coast, 29°00' N. and 79°48' W., 2 speci- mens (26.4 and 27.0 mm.) on Nov. 18. Oft' tiie lower (ieorgia coast, 30°57' N. and 79° 14' AV.. 1 specimen (27.3 nun.) on Xov. 19. See figure 1. For comparison, I examined 2 specimens of .1. monticola (33.0 and 39.0 nnn. standard lengtli) U. S. Xational Museum Xo. 101483, taken in Kio Carolina, Porto Rico, P>b. 17, 1934; and fiom the University of Florida collections, 7 specimens (31.5, 47.0. 55.9, 59.0, 64.5, 75.5, and 119.0 nun.) taken approximately 2 to 4 miles upstream from the mouth of the Yateras River, Cuba, on Sept. 16, 1952; also from the University of Florida col- lections taken from fresh waters of Florida, 3 specimens (44.8, 48.0, and 56.0 mm.) from Cedar- 415 416 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 32° 3 1° ^ J* / 3 • - 30° -v\ \ ' 29' \ \ t 2^ "^ \ ; t \ i 1 2/ A '^ ; -^x — 26' 25' 8 2° 81° 80' 79' 78' 77' 76° Figure I. — X indicates the areas of eapture of the fresh-water mullet, Agonostdmiis moiiticoUi. Arrows on (lashed line show approximate axis of Gulf Stream. Dotted line represents the 20-fathom curve. LARVAL FORMS OF FRESH-WATER MULLET 417 Cow Creek i^ "I'le east of Sainsulix Junction on Florida Hijjhway 75, Volusia Vo., captured Feb. 21, 1948; 1 specimen (32.4 mm.) from Volusia Co.. taken in 1948; 1 si)ecimen (81.5 mm.) from Cedar-(^ow Creek, Volusia Co., taken in June 1949; 3 specimens ((Ki.l, ()8.0, and 75.5 mm.) from Cow Creek, Volusia Co., taken on May 23, 1949; 4 specimens (94.4, 96.5, 105.0, and 105.3 mm.) from an artificial fresh-water pond near St. Aujrustine Beach, St. Johns Co.. taken August 30, 1949; 2 specimens (62.0 and 63.8 mm.) from Spruce Creek, Volusia Co., taken April 29, 1950; and 1 specimen (26.7 mm.) from Spruce Creek, Volusia Co., taken Nov. 8, 1951. METHODS Metliods employed aboard the vessel Theodore \. GUI are given by Anderson, Gehringer, and Cohen (1956). Descriptions are based on preserved material, and measurements of larvae up to about 35 mm. standard lengtli were made with a stereoscopic microscope and a micrometer eyepiece. The larger specimens were measured with calipers. Original measurements were used in constructing tlie graphs to portray rates of growth of various body parts, and changes in body proportions. DESCRIPTION In general appearance tliese sea-stage young of A. nwnticola closely resemble the sea-stage young of both M. curema and M. cephalus, the most strik- ing difference being tliat ^4. monticola appears much more slender and with a longer caudal peduncle. All are heavily pigmented and appear blackisii. Figure 2 illustrates a 31.3-mm. speci- men. Observed under magnification, the presence of ctenoid scales quickly separates A. monticola from these two species of Mwgil. FINS Dorsal: — All e.xcept two of the 34 specimens examined Iiad a dorsal Hn formuhi of IV-I, 8. One of the Florida specimens (96.5 mm.) had 5 spines in the first dorsal, arranged in a peculiar manner so as to appear as two fins; the 1st, 2d, and 3d spines were connected by membranes, the 3d and 4tli spines were not connected by a mem- brane, but the 4th and 5th spines and the 5th spine and the body were connected by membranes (tliis aj)i)ears to have resulted from an abnormal division of the 3d spine, altliough all spines were of normal size). One specimen from Cuba (64.5 mm.) had only 6 soft rays instead of the usual 8. Anal: — All except one of the specimens ex- amined had an anal fin formula of II, 10 (one Florida specimen, 81.5 mm. had II, 9). In figure 3 the details of the anal fin of cleared and stained specimens of A. mo-ntieola, M. ewema^ and M. cephalus at comparable sizes (about 30 mm.) are illustrated. Both spines and the first soft ray in A. monticola are shorter and more slender than in .1/. curema and M. cephalus. A notable dif- ference at this size occurs in the last ray. In both M. curema and M. cephalu.s tliis ray has two main branches, each of which is in turn deeply forked, while in A. monticola only the anterior main branch is forked. A difference in the number of anal spines re- ported for ^4. montkola occurs in the literature. Jordan and Evermann (1896) on page 809 in their key to the genera of Mugilidae indicate 2 anal spines for Agonostomus Bennett; on page 818 under tlie generic description is the statement, "Anal spines usually 2, the first soft ray slender and often taken for a spine"; pages 819 and 820 describe 4 species of Agonosfomus, including A. monticola, each of which is given an anal fin formula of III, 9. Evermann and Marsh (1902) Figure 2. — A young Agonn.itomuit nioiitii-oln. 31.3 mm. long, capturetl in the open ocean. 418 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE A. MONTICOLA M. CUREMA M. CEPHALUS >-j*i!«Vf'iK*'S:'*/'-'' Figure 3. — Anal fins from a 29.4-mm. specimen of Affonostomiis monticola, a 30-mm. Miigil curema, and a 30-mm. Mtiffil ccphalus. (Camera lueida drawings from cleared and stained specimens.) on page 112 in tlieir key to the Mugilidae give 2 anal spines for Agonosfomu.s, on page 114 undei" the generic description "Anal spines usually 2, first soft ray slender and often taken for a spine" ; Evermann and Marsh give A. monticoJa an anal fin formula of III, 9. Meek and Hildebrand (1916) on page 333 under the generic description of Agonostomiis state, "anal spines 2, the first one minute, often hidden in the skin"; and on page 334, they show an anal fin formula of II, 10 for A. monticola. Hubbs (1944) on page 72, in his discussion of the anal fins of various fish, reports that Mugilidae, "Instead of having only one thin flexible anal spine all mugilids have 2 thick pun- gent anal spines, and the third ray with age trans- forms (in the Mugilinae but not in the Agonos- tominae) from a flexible, paired, articulated soft ray into a pungent, solid, unsegmented spine (Jacot, 1920:207-208)." Schultz (1949, p. 112) gives an anal fin count of III, 9 for specimens of A. monticola from Venezuela. Carr and Goin (1955, p. 83) give an anal fin count of III, 9 for specimens from Florida. The 34 specimens I have examined (ranging from about 24 to 119 mm. standard length) have only two anal spines. The third ray, even in the largest specimens, shows no evidence of trans- forming into a solid unsegmented spine ; the seg- ments in this slender ray are difficult to see unless examined under magnification. Pectoral: — Of the 34 specimens examined, 3 had 14 rays, 26 had 15 rays, and 5 had 16 rays. Placement, shape, and extension of this fin are illustrated in figure 2. Caudal: — The caudal skeleton of a cleared and stained specimen (29.4 mm. standard length) is illustrated in figure 4. There are 14 principal rays of which 12 branched ; and 19 secondary rays, 9 dorsal and 10 ventral. Of interest, at this size, is the peculiar branching of the 8 middle principal rays, which end in three tips (as do the last dorsal and anal fin rays). In M. curem,a and M. cepha- lus at comparable size, the 8 middle caudal rays and the last dorsal and anal rays end in 4 tips. PREMAXILLARY, MAXILLARY, AND PREORBITAL BONES Schultz (1946) in his revision of the genera of the family Mugilidae pointed out the importance of these bones in separation of the various genera. In figure 5 are illustrated these bones from cleared and stained specimens: a 29.4-mm. .4. monticola and 30-mm. M. curema and M. cephalus. The great similarity of these bones in the two members of Mttgil is striking, as are their differences from A. monticoln (which has a very wide, serrated posterior margin; tips of maxillary and pre- maxillary extending farther below the posterior margin of preorbital ; a deep hook about midway on the rear edge of premaxillary ; and entirely difl'erent shape of tips of maxillary and pre- maxillary). In some larger specimens of .4. monticola thei-e are serrations along the front edge of the preorbital in addition to those along the posterior edge (serrations on the posterior edge vary from about 9 to 13). LARVAL FORMS OF FRESH-WATER MULLET 419 I MM Figure 4. — Caudal skeleton of a 29.4-mm. Ji^&hos^owim* montirola. (Camera lucida drawing from a cleared and stained specimen. ) TEETH The larval sea-stage of .4. monticola have a band of strong, sharp-pointed, conical-shaped teeth in the upper and lower jaws, of which the outer ones are the larger (fig. 6A and 6B) ; bands of sharp- pointed, conical-shaped teeth on the vomer and palatines (tig. 6C) ; and strong, sharp-pointed conical teeth on the tongue arranged in several patches; one center patch near the tip of the tongue, followed by a single row of about 8 to 10 teeth along each side of the tongue, and a small isolated center patch of smaller teeth near the rear of the tongue (tig. fil)). There are no teeth in the lips of .4. ftionticola. In M. rureiiKi and M. rephalux there are no teeth on the vomer or palatines, and the setiform or ciliform teeth occur in the upper and lower lips. PIGMENTATION The sea-stage specimens of A. monticola are all rather heavily pigmented (some more than others) so that they appear black on tbe upper half of the body shading to a lighter, yellowish color on the ventral surface. Pigment spots are smaller and closer together on the upper part of the body and larger and less numerous on the lower half of the body. In the size range of sea- stage specimens (24.1 to 31.S mm.) the first and second dorsal fins and the caudal fin have pig- mentation, but the anal, pectorals, and ventrals do not. See figure 2. This is in sharp contrast to the coloi-ation of specimens examined from fresh-water streams. These tend to be brownish on tlie dorsal surface, becoming blotched brownish to below the mitlline of the body, and white on the belly and the lower sides of the body and head. There is a dark spot at the caudal base. MISCELLANEOUS STRUCTURES Gill rakerx. — The following counts were ob- tained on the lower limb of the first arch: 14 in a 25.5 nun. specimen; Hi in specimens 29.4 ami 31.5 nun.: 17 in a .'iil-mni. specimen; and 1!> in 420 nSHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE I MM. FiGUKE 5. — Premaxillary, maxillary, and jn-eorbital bones in a 29.4-nim. Agoiiostomiis monticola, a 30-mni. Miigil curcma, and a 30-mm. Mugil cephalus. (Camera lucida drawings from cleared and stained specimens.) LARVAL FORMS OF FRESH-WATER MULLET 421 specimens of 68 and 105 mm. Meek and Hilde- brand (191(!, p. 834) jrive cri)] rakers on tlie lower limb of tlie first an'b for -I. mont'icoJa as ranging from 17 to 20. It appears that the full comple- ment of gill rakers develops when the tish are l)etween ;}() and 40 mm. long. Tliis increase in number of gill rakei-s with increase in size also occurs in other related species. A 30-mm. M. I MM. I MM. I MM. FiiiiKK ti. — .\irai)genieiit ami form of teeth in a 2!l.4-iiiiii. MioniixtomHH monticola: A, left iialf of premaxilary ; B, left lialf of lower jaw ; (', vomer and palatine bones ; I), tongue. (Camera lucida drawings from a cleared and stained sjieeimen.) citrenia had 30 gill rakers on the lower limb of tlie first arch whereas adults have more than twice that number; a 30-nnii. M. cephaluK had about 25 gill rakers and a 60-mm. specimen had alx)ut 40 (lower limb first arch). Scales: — All specimens examined had ctenoid scales. Figure 7 shows arrangement of spines on a scale from a 28.fi-mm. specimen. There were 39 to 42 (most frequently 40 or 41) rows of scales in a series from the upper angle of the opercle to tlie middle of the caudal base ; and 11 or 12 rows of scales between the second dorsal and the anal fin. BODY PROPORTIONS Measurements from the 34 sj^ecimens examined (ranging from 24.1 to 119 mm., standard length) were nsed to estimate the relation between stand- ard length and eye diameter, head length, body deptli (at pectoral), and distances from snout to insertions of the first dorsal fin. second dorsal fin, and anal fin. These relationships are illustrated in figures 8 and 9. and the rate of development of body parts, within the range of sizes studied, is apparent. The upward shift in the regression line for body depth (at pectoral) occurs between sea- stage and fresh-water stream material. I have observed the same occurrence in .1/. cure ma be- I MM. KiciHK 7. — Scale from a 2,s.t>-mm Afloiioatoniiix monticola, showing arrangement of spines. (Scale taken above midline of Ixidy between second dorsal and anal tins.) 422 FISHERY BULLETIN OF THE FISH AND WILDLIFE SEHVICE 1 1 —^ 1 1 1 — r- 1 1 I 90 » A . 70 - - 60 " X X 30 " isrf & .^* ^ 40 * v-^^ ~ 30 - X O^ .o'^ - to 20 ■ ^^ / X X %^ iLl • • ;6'' A \- •• >^ UJ ^^ Z / ^^ i ■ \/ 10 ■• <5 — - ^^^ - 8 " V ••• A ^ - ..r X X ^ - 6 ^ • « ^ ^ .^ Xi* <(. 4 ™ ■ ^^ " 3 — • .•^^ — 1 1 1 1 1 1 1 1 I 1 20 30 40 60 80 100 STANDARD LENGTH IN MM. 150 Figure 8. — Relation of standard length to eye diameter, head length, and distance from snout to insertion of anal tin. (Dots represent sea-stage larvae from GiU collections, solid sijuares indicate I'uerto Rico speci- mens, triangles represent Cuba specimens, and X's the Florida specimens.) LARVAL FORMS OF FRESH-WATER MULLET 423 90|- 70 60 SO 40 30h I- 10 8 6 _ 4 _ 3 - 1 1 1 1 1 1 1 1 1 I - A ^ " ■" ~ - — ^ '^ x" C» ■ ^ . .» j^' ^ r ^" — x^*' .^ o<^ . ^ •• "• -^ •V ^ ■ * ^ — <^ — ^ ■ O^' ^ ^ X -\ - ••fr • *°^ - — • — — — — - - - 1 1 1 t 1 1 1 1 1 1 20 30 40 60 80 100 STANDARD LENGTH IN MM. ISO Figure 0. — Uelation of standard length to body depth (at pectoral), and distance from snout to insertions of first and second dorsal tins. ( Dots represent sea-stage larvae from dill collection.s. solid sipiares indicate Puerto Rico specimens, triangles represent Cuba specimens, and X's the Florida siiecimens.) 424 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE tween the sea-stage and inshore specimens, and attribute it to a greater abundance of food in the inshore areas, which is reflected in a pot bellied appearance (Anderson 1957). OCCURRENCE OF LARVAL AND JUVENILE FORMS The literature presents interesting facts regard- ing sizes of A. monticola obtained from fresh- water streams. Evermann and Mai'sh (1902, p. 115) say, ''It is very abundant in the fresh-water streams of Porto Rico, and is much used as food ; many were collected at Caguas in the Rio Loiza and Rio Caguitas and in Rio Bayamon at Bay- amon, of all stages of growth from 1.5 to 11 inches" (minimum size taken, 1.5 inches, is about 38 mm.). Meek and Hildebrand (1916, p. 335) had 118 specimens in their collection from Panama which ranged from 40 to 255 mm. ; taken from lowland streams, some from brackish water, and a few from ujiland streams. Beebe and Tee-Van (1928, p. 92) found the species in sea- ward flowing streams of the Cul-de-Sac Plain at Souire Mariani in Haiti ; they had 49 specimens ranging from 43 to 188 nun. Schultz (1949, pp. Ill and 112) lists collections from several rivers in Venezuela, and the smallest specimen listed was 22 mm. A series of specimens in the University of Florida collections from the Yateras River, Cuba, ranged from 31.5 to 119.0 mm. The University of Flor- ida's collection from streams in Florida contains si^ecimens ranging from 26.7 to 105.3 mm. From these records it appears that about 22 mm. is the smallest size at which this species has been taken from fresh-water rivers ; a strange oc- currence if the species spawned and developed in fresh water. The specimens taken by the Gill in ocean waters off the Bahamas, Florida, and lower Georgia coasts (ranging in size from 24.1 to 31.3 mm.) assume greater significance in view of this, and very strongly point to a sea spawning for ^4. monticola, with the young remaining at sea until they have reached a size of about 20 to 35 mm. This would make the species a truly catadromous fish. I have located only one reference in support of placing A. monticola in the list of catadromous fishes, and this reference refers only to the genus, and not to a particular species, although I suspect that A. monticola is involved. Myers (1949, p. 94) in his discussion of terms for migratory fishes states that "The mullet, Agoiwstomm, may per- haps be added to the small known list of truly catadromous fishes. Dr. C. L. Hubbs tells the writer that he has found the 'Querimana' stage of this genus 'in completely salt water in Acapulco Bay,' Mexico, but it is, of course, possible that Agonostomus may belong in the next division." (The next division referred to is "Amphidro- mous," defined by Myers as, "Diadromous fishes whose migration from fresh water to the sea, or vice versa, is not for the purpose of breeding, but occurs regulai-ly at some other definite stage of the life-cycle.") SPAWNING Placing the time of spawning of the larval speci- mens taken off the Bahamas, Florida, and lower Georgia coasts must be based on two things, one factual and one an assumption. Specimens 24.1 to 31.3 mm. were taken during late January 1954, and specimens 25.5 to 30.6 mm. were taken in mid-November 1954. Assuming a period of 4 to 6 weeks for the larvae to attain a length of 20 nun. after hatching, the (tHI specimens would have been spawned from September to December. I believe that the larval specimens taken by the Theodore N. Gill probably were spawned much farther south and the larvae were carried north- ward by the prevailing currents. This would partially explain our failure to capture smaller specimens. Also, I suspect that the occurrence of the species in Florida streams has resulted from "seeding" by specimens carried northward by the Gulf Stream from possible spawning off the Cuban coast. The largest specinuen in the University of Florida collection (from Florida) is 105 mm., and it is not known whether or not the species is maturing in waters of that State. The collection of specimens from St. Augustine Beach, Fla., is the most northward recorded occurrence in fresh waters on the Atlantic coast. LARVAL FORMS OF FRESH-WATER MULLET 425 LITERATURE CITED Anderson, William W. 10')~. Karly development, spawuiuK. Krowtli and (ic- curieiue of the silver mullet [MiKjil run ma) along the South Atlantic Coast of the United States. Fish and Wildlife Service, United States Department of the Interior. Fisli. Bull. 119, vol. r.7, pp. :i\)l-4U. A.NDERSON, WiLLi.\M W.. .I.\cK W. Gehringer, and Ed- ward Cohen. 19.56. Physical oi'eanopcrapliic, hiolojiical, and chemi- cal data, South Atlantic Coast of the United States, M/V Theodore N. Oill Cruise 1. Fish and Wildlife Service, United States Department of the Interior. Sp. Sci. Rept., Fisheries 178, pp. 1-160. Beebe, William, and .Iohn Tee- Van. 1928. The fishes of Port-au-Prince Bay, Haiti. Zoolofiica, vol. X, No. 1, pp. 1-270. Carb, Archie, and Coleman J. Goin. 19,i5. Guide to the reptiles, amijhibians, and fresh- water fishes of Florida. University of Florida Press, pp. 1-341. Evermann, Barton Warren, and Millard Caleb Marsh. 1902. The fishes of Porto Rico. Bull., U. S. Fish Comm., vol. XX, Part 1, 1900 (1!K>2), pp. 51-350. HuBBS, Carl L. 1944. Fin structure and relationships of the phal- lostethid fishes. Copeia, Xo. 2, June 30, 1944, pp. 69-79. .Jordan, David Starr, and Barton Warren Evermann. 1896. The fishes of North and Middle America. Bull., U. S. Nat. Museum, No. 47, Part I. pp. 1-1240. .Jordan, D. S., B. W. Evermann, and H. W. Clark. 19.30. Checklist of the fishes and fishlike vertebrates of North and Middle America. Report X'. S. Comm. Fish and Fisheries, 1928, Part 2, pp. 1-670. Meek, Seth E., and Samuel F. Hildebrand. 1916. The fishes of the fresh waters of Panama. Field Museum of Natural History, Publication 191, vol. X, No. 15, pp. 217-374. Myers, George S. 1949. Usage of anadromous, catadromous and allied terms for migratory fishes. Coi)eia, No. 2, June 30, 1949, pp. 89-96. oCHULTZ, Leonard P. 1946. A revision of the genera of mullets, fishes of the family Mugilidae, with descriptions of three new genera. Proc. U. S. Nat. Museum, vol. 96, No. 3204, pp. 377-395. ScHULTZ, Leonard P. 1949. A further contribution to the ichthyology of Venezuela. Proc. U. S. Nat. Museum, vol. 99, No. 3235, pp. 1-211. U. S. GOVERNMENT PRINTING OFFICE: 1957 FECUNDITY OF THE PACIFIC SARDINE (SARDINOPS CAERULEA) By John S. MacGregor, f'ishery Research Biologist Extensive studies of the biology of the Paeific sardine {Sarduwps caerulea), designed to lead to an understanding of the causes of fluctuations in the size and distribution of the population, have been carried on since 1949 under the sponsorship of the California Marine Research Committee. These studies, the California Cooperative Oceanic Fisheries Investigations, have been made by the California Academy of Sciences, the California Department of Fish and Game, Hopkins Marine Station of Stanford University, Scripps Institution of Oceanography of the University of California, and the South Pacific Fishery Investigations of the U. S. Fish and Wildlife Service. One of the requisites of these studies is a measure of the size of the sardine population. The number of individuals in a spawning population of fishes can be estimated if the following are known: (1) the total number of eggs produced per year by all females in the population, (2) the average number of eggs produced per year by each female in the population and (.3), the sex ratio in the population. This metliod has been one of those used for several years by the South Pacific Fisheries Investigations to estimate the size of the Pacific sardine popula- tion along the coast of California and Baja Cali- fornia. To determine the average number of eggs produced per year by each female it is necessary to determine the average number of eggs spawned per batch and the number of batches produced. The annual egg production per female used in the population estimates has been based on fecundity data for eight female sardines given by Clark (1934). The purpose of the present paper is to record more extensive data on the fecundity of the Pacific sardine (item number 2 above). I wish to express my appreciation to my col- leagues, E. H. Ahlstrom and T. M. Widrig for helpful suggestions, F. E. Felin for making the age determinations, A. M. Vrooman for assisting in counting and measuring the eggs and preparing the figures, and W. M. Morton for collecting the material, and to members of cooperating agencies for their assistance. Note.- AiiprovcU for puWioUion tVbriiiiry 18, 1957. Fishery Bulletin 121. MATERIAL AND EQUIPMENT The 13 samples of sardines examined were taken from the commercial fishery at San Pedro, Calif., in November and December, 1945, and January and February, 1946, and included 1,270 individ- uals. These fish, originally collected for a mor- phometric study, were fixed and preserved in formalin in gallon jars. Some of them were subse- quently transferred to alcohol. Ovaries and ovary samples for ovum counts and diameter measurements were weighed on a Sartorius Selecta balance. The ovum counts and diameter measurements were made on projected images of the ova using the scale reading device described by Mosher (1950). The ovary samples were placed on microscope slides for weighing and projection. Slides were prepared by placing two threads parallel to each other and to tlie long axis of the slide, along one surface of the slide, and securing them at the ends with cellophane tape. The threads were spaced a distance apart somewhat less than the diameter of the microscope field and served as guide lines. The use of threads rather than etched lines kept the ova from falling on the line and controlled the spread of the mounting medium. METHODS A pair of ovaries was drained for a few minutes on paper toweling and weighed to the nearest 0.001 gram. A small sample from one ovary was then placed on a microscope slide and weighed to the nearest 0.0001 gram. A drop of glycerin was placed over the sample; the ova were teased out and spread into three strips separated by the two guide-line threads. A second microscope slide was then placed over the first and the two slides were fastened at the ends with cellophane tape. The formalin-hardened ova were not distorted between the slides prepared in this manner so long as the pressure applied to the slides was not excessive. A second mount was prepared in a similar manner, using a sample from the other of the pair of ovaries. The use of two samples for each pair of ovaries 427 428 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 80 70 _ 60 _ 50 _ 40 _ O 30 UJ IT ^ 20 < UJ s 10 < fe 80 Q: 70 UJ CD i 60 z 50 40 30 20 10 0 STAGE A Number of ova not estimated Fig. l-a Fig l-b STAGE "C" Estimated number of ova 02- 20mm 640,000 .22-.3emm 43,600 10 14 18 .22 .26 .30 OVUM DIAMETER IN MILLIMETERS Figure l-a, b. — Sardine iiitra-ovariaii ovum diameter frequencies. FECUNDITY OF THE PACIFIC SARDINE 429 provided a clieck on the aeciiracy of tlie ovum counts and yielded a larger sample tlian could liave been easily examined on one slide. Eacli sample of ova was projected in the scale l)rojeclor at a magnification of 50X. Tiie pro- jected ovum diameters were measured to the nearest millimeter. "Wiicn doubled, this measure- ment gives the actual ovum diameter in hun- dredths of a millimeter by 0.02 mm. units. If the sample of ova contained no distinct (by size) group of developing ova, the diameters of the larger ova in the sample were measured. The largest ovum diameter was used as a measure of the ovarian development for these early stages when no modal or other central value could be obtained. Figure 1-a and b and figure 2-a and b show ovum diameter frequencies from ovaries in these stages of development. The dotted lines in figures 1 and 2 represent 95 percent confidence limits of the counts. If the ovary was advanced enough so that a distinct group of yolked ova had differentiated by size from the smaller yoli\ed ova, enough ova (generally 100 to 200) 0.22 mm. or larger (i. e., containing yolk) were measured to delimit clearly this most advanced group of developing ova. When the size range of this group was determined, all ova in that size range on the slide were counted. Figure 2-c, d, e, f, g, and h shows ovum diameter frequencies from ovaries in these stages of develop- ment.' The most advanced groups of ova shown in these figures consist of opaque ova except the group shown in figure 2-h which consists of trans- lucent ova. Xo perivitelline space has formed in ' The mullimodal distribution of diameters of developing intraovariun ova of the Pacific sardine was first demonstrated and figured by Clark (1934). r'V- 0 30 -%- O tlJ CD z 3 ^V^ 30 20 10 XO-,- ^^ Et -^ fl "tfl^ L^ ^fezb3^ EiL: Fig. 2-G STAGE "b" Estimated number of ova .02- 20mm 510,000 .22-. 36mm 20,000 Fij 2-b STAGE "C" Estimoted number of ova 02- 20mm 640,000 .22-.40mm 43,600 Fig 2-c STAGE "E" Estimaled number of ova 22- 40mm 52,000 42- 52mm 29,500 Fig 2-d STAGE "E" Estimated number of ova 22-44mm 69,100 46-. 60mm 35,300 Fig 2-e STAGE "F" Estimated number of ovo .22- 46mm 39,600 .48- 70mm 29,100 22 26 30 34 38 42 46 50 54 58 .62 66 70 OVUM DIAMETER IN MILLIMETERS Figure 2-a, b, c, d, e. — Sardiiu' iiitra-ovariaii ovum diamotor frequeiicie.-; (.-ice p. 4'iO). 430 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE the intraovarian ova, and hence the ovum dia- meters are also yolk diameters. In figure 2-i yolk-diameter distributions of 343 planktonic sar- dine ova in early developmental stages are shown for comparison with figure 2-h. Figure 2-j sliows the ovum diameter distribu- tion of degenerating ova from the ovaries of a sardine taken in May of 1951. None of the San Pedro 1945-46 sardines used in this study were found to be in this condition. Clark (1934) grouped developing ova in the largest mode into twelve stages of maturity. As I will have occasion to refer to some of these stages in comparing my observations with those of Clark, I am including a description of these stages: lMM.\TrRE Stage A; Frequencies with a mode between 0 and 0.2 mm. only. Maturing (Opaque Eggs) Stage B: Frequencies with the last mode between — 0.22 and 0.26 mm. Stage C: Frequencies with tlie last mode between — 0.26 and 0.3-1 mm. Stage D: Frequencies with the last mode between — 0.34 and 0.44 mm. Stage E: Frequencies with the last mode between — 0.44 and 0.54 mm. Stage F: Frequencies with the last mode between — 0.54 and 0.64 mm. Stage G: Frequencies with the last mode between — 0.64 and 0.74 mm. Stage H: Frequencies with the last mode between — 0.74 and 0.84 mm. f-N- o 10 0 20 10 0 20 10 0 70 60 50 40 30 20 10 0 20 10 0 ^ f-l--^' n^- wjti 11-- Fig. 2-f STAGE "G" Estimated number of ova .22-.62mm 86,600 .58-. 74mm 45,700 Fig. 2-g STAGE G Estimated number of ova .22-.50mm 53,000 .66-.7emm 31,000 Ett=i rftrFK-- STAGE I Estimoted number of ova .76-.96mm 20,400 Fig. 2-h ^ s Fig. 2-1 YOLK DIAMETERS eggs from plankton ^v- A STAGE Fig 2-1 LNt .4:?T-r ^l:>4^T:i:JbJfe^ .22 .26 .30 .34 .38 .42 .46 .50 .54 .58 62 .66 .70 74 78 .82 .86 .90 .94 .98 OVUM DIAMETER IN MILLIMETERS Figure 2-f, g, h, i, j. — Sardine ovum diameter frequencies. FECUNDITT OF THE PACIFIC SARDINE 431 .MATiRiNti (Transparent Eggs) Stage I: Frctiucncies with the lust iiiodf lictwocn 0.84 uiid 0.04 mm. Stago J: Frequencies with tlic lust mode above 0.94 mm. Mature (Transparent EIggs Free in- the OviDtCT) Stage K: Freqiienoies with the hi.st mode above 0.!)4 mm. and the ripe eggs segregated. Spent Stage L: Frequencies with ova hirger tlian 0.20, but these eggs degenerating. The numlxT of ova ii) the most advanced group was calcidated by multiplying the weight of the ovaries by the number of these ova in both samples and dividing by tlie total weight of both samples. It was also determined from each of the two samples separately as a check on the sampling variation. All (587) of the female sardines in the 13 samples were used for ovum diameter measurements. Ovum counts were made for all fisli (116) that contained a distinct group of developing ova. The weight of both samples from a pair of ovaries equalled approximately 2 percent of the weight of that pair of ovaries. The deviations of the 116 pairs of ovinn estinuites from their respective combined (>stimates was 1,700 eggs at one stand- ard deviation level. The right and left ovaries of each pair were at the same stage of maturity in every case, and neither ovary gave a consist- ently higher or lower count than the other. As pointed out by Clark (1934), there are no apparent differences in the relative immbers of ova in eacii size group in the different parts of the ovary. RELATION BETWEEN FECUNDITY AND LENGTH Figure 3 shows 116 fecundity observations obtained from the January and February, 1946, samples plotted against their respective fish lengths. The regression line, fitted by the method of least squares, of the form Y=a-\-bX (where Y=tlie number of ova in the most advanced mode in thousands, A''=the fish length in milli- meters and a and b are constants) is also shown. Clark's (1934) 8 observations are plotted for comparison. It is apparent from figure 1 that the fecundity of individual fish of the same length can vary considerably. This variation ^vr- _L _L _L _L _L _L _L J_ 170 180 190 200 210 220 230 240 STANDARD LENGTH IN MILLIMETERS 250 260 270 Figure 3. — Rehitioii between fecundity and standard length for IIG sardines (San Pedro, Jan. -Feb., 1946). Open circles are Chirk's (19:54) observations. 432 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE {Sy = 8.76 thousands of ova) is more than 5 times as large as the variation attributable to sampling and counting techniques. There has been some difiPerences of opinion among various authors as to whether length, length squared, or length cubed should give the best straight line correlation witli fecundity. Lehman (1953) correlated the fecundity of 22 shad (Alosa sajndissim.a) with their lengths. His equation (least squares) for the line of average relationship is r=— 462.691+40.090X in which F=the number of ova in thousands, and ^Y=the length in inches. Clark (1934:21), referring to eight fecundity ob- servations on the Pacific sardine, states: By the method of least squares from the formula N=FL^, where A^ indicates the number of eggs; L, the length; F, a constant; and x, the exponent expressing the relationship between the number of eggs and the length of the fish; x was found to have a value of 1.9868. This suggests that the numbers of ova produced by individual sardines increase as the square of the length. But because the calculations were based on very scanty data, these conclusions can be tentative only. Simpson (1951j expressed fecundity observa- tions for 256 plaice {Pleuronectes platessa) by the formula F=KU, where F=the number of ova; K, a constant; and U, the cube of the length of the fish. The line of best fit (least squares) for the 116 sardine fecundity observations is plotted in figure 4 for each of four different relationships. Table 1 gives the I'-intercepts (a), slopes (b), standard errors of estimate of Y (Sy) and coeffi- cients of correlation (r) of the four lines for the 116 fish and for each of the five samples which together constitute the 116 fish. The standard errors of estimate of Y for 10-millimeter length intervals are presented in table 2. Table 3 gives the number of ova in the most advanced group calculated by each of the four different formulas at each of 9 different 10-millimeter length inter- vals. It is apparent from these comparisons that convenience, rather than theoretical considera- tions should be the deciding factor in selecting one of the regression formulas to describe these data. The same is true for Lehman's (1953) fecundity ^v- 60 D O O UJ o Y=a+bX Y=a*bx2 Y=a+bX^ - _ Y=bX^ -L _L _L _L 180 190 200 210 220 230 STANDARD LENGTH IN MILLIMETERS Figure 4. — Comparison of fecundity-length regression lines. 240 , FECUXDITY OF THE PACIFIC SARDINE 433 data for shad. Using his formula }'=a + 6A'', S„ = 40.2 and a = -462.7. Using Y=a + bX\ Sy=39.'.i and a = 19.0. As a approaches zero the formula Y=bX^ would also give an S„ approxi- mating the above two. Theoretically the number of ova is dependent upon the volume of the ovary, a three-dimensional function, and tlierefore shoukl l)etter correlate with the cube of tiie length, length itself being, of course, only linear. In Simpson's (1951) paper on the fecundity of the plaice, the straight line re- gression of fecundity on length cubed is consider- ably better than that of fecundity on length squared or length. The significant differences between the regressions in this case are brought out by the greater relative range in sizes of fish used in calculating the regression lines. Measured as a percentage of the length of the shortest fish used, the range of Simpson's plaice is 162 percent, that for Lehman's shad is only 66 percent and for the 116 sardines 48 percent (56 percent including Clark's 8 fish). Table 1. — Comparison of some possible fecundity-length relationships Sample Y=a+bX Y=tt+bX' Y=a+bX I Y=bX3 SP-8(17fish) a (thousands)..- 6 . . -25.4 .239 5.33 .547 -112.9 .668 5.02 .786 -43. 1 .358 5.49 .478 -87.8 .574 4.09 .839 -104.3 .674 10.2 .730 -76.0 .514 8.76 .541 -1.4 .000594 5.32 .549 -44.9 .00164 5.02 .786 -4.3 .000825 5.50 .476 -27.7 .00137 4.16 .833 -32.1 .00158 10.3 .725 -22.9 .00124 8.74 .643 +6.6 .00000195 5.30 .553 -22.3 .00000534 6.03 .785 +8.6 .00000253 5.49 .478 -7.4 .00000430 4.29 .822 -8.0 .00000475 10.4 .719 -4.3 .00000386 8.76 .642 0.0 S, (thousands) 5.48 SP-9 (9 fish) . . a (thousands).- 6 - - 0.0 0()0009f>=i S, (thousands) - r 5.97 .677 0.0 .00000338 5.60 .445 .0 .00000352 4.44 .808 0.0 .00000386 10.6 .703 0.0 .00000341 8.79 .638 SP-11 (41 fish) a (thousands) b . S, (thousands) SP-12 (5 flsh) 0 (thousands) b . ■Sy (thousands) SP-13 (44 fish) a (thousands)- 6 Sy (thousands) r.- All samples (116 flsh) a (thousands) S, (thousands) r Table 2. — Standard error of estimate of Y for ten milli- meter length intervals using formula: 1^=0 + 6.^ |y=N'o. of ova In thousands; Jf=lcngth In mm.) Length (mm.) Fish (number) (thousands) 170-179 - I 10 27 21 36 15 3 2 1 180-189 4.68 190-199 6 60 200-209 6 80 210-219 8 91 220-229- - 7 33 230-239 240-249..- 250-259 - -- Total 116 8 76 Table 3. — Calculated number of ova in most advanced mode for sardines of different lengths Length (mm.) Number or ova (in thousands) Y=a+bX Y=a+bX' Y=a+bX' Y=bX' 170 11.4 16.6 21.7 26.8 31.9 37.1 42.2 47.4 52.5 12.9 17.3 21.9 26.7 31.8 37.1 42.7 48.5 54.6 14.7 18.2 22.2 26.6 31.4 36.8 42.7 49.1 56.0 16 8 180 19 9 190-,.- 23.4 200 27 3 210 31 6 220...- 36.3 230 41 6 240 250 63.3 Obtaining a sample of Pacific sardines suitable for fecundity determinations with a size-range comparable to that of plaice would be an impossi- bility. The size range of the sample of spawning sardines might be doubled by including large fish (270 to 290 mm.) from the northern part of the sardine's range and small spawning fish from the southern part of the range. (I have examined some 145-155-mm. sardines taken off the southern tip of Baja, California, that had apparently devel- oping gonads.) This procedure could introduce a considerable error. Simpson found that there were geographical differences in fecundity among stocks of plaice, those from the Baltic containing many more eggs than those of comparable sizes from the North Sea. In figure 5 the data for eacli of the 5 samples have been described b}' regression lines, using the 434 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 180 190 200 210 220 230 240 STANDARD LENGTH IN MILLIMETERS Figure 5. — Fecundity-length regression F=6A''. 250 formula Y=bX^. The use of this formula makes the curves much easier to compare, since the Y- intercept is taken as zero for all samples and the slopes may be compared directly. When the curve is fitted by the least squares method to the data, using the formula Y=a + bX^, both the a and b values are determined by the data. The sum of the plus deviations will equal the sum of the minus deviations, while the sums of the squares of the deviations will be at a minimum. When the formula Y=bX^ is used, the T-intercept is forced to be zero, and both of these conditions cannot be met if the distribution of data is at all skewed. A regression line about which the sura of the plus deviations equals the sum of the minus deviations can be obtained using a value obtained 2F from the formula b= 2(X') A line about which the sum of the squares of the deviations is at a minimum can be obtained using b= ZiX^Y) Al- ^{xy though there is little difference between the b val- ues derived by these two methods ^ from the data for the 116 sardines, the latter method has been used because the standard errors of estimate and ' Simpson apparently used the formulai) = z:f -^^ j -^iV which also gives a 6 value differing very little from the other two for the Ufi sardines. correlation coefficients based upon it should theo- retically be more nearly comparable with those obtained in conjunction with the least squares methods in which the J^-intercept is determined by the data. Figure 5 shows that samples SP-8 (Jan. 11) and SP-9 (Jan. 24) almost coincide and that samples SP-11 (Feb. 9),SP-12 (Feb. 21), and SP-13 (Feb. 27) show an apparent increase in fecundity as the season progresses. Because of the restricted range of lengths and the great variation in ovum count at any given length, no significance is attached to this apparent temporal fecundity increase. In figure 6 are plotted the regression lines for each of the 5 samples, fitted (least squares) by the formula Y=a+bX (which differs most from the formula Y=bX^). In table 4 each of the 10 possible com- binations of pairs of b's is compared by t test for the possibility of significantly different slopes for both formulas. There are no pairs of regression slopes that are significantly different when both formulas are taken into consideration. The ap- parent significant difi"ereiices in the slopes of the pairs, SP-8 and SP-9, and SP-8 and SP-13, when the formula Y=a+bX is used are primarily a re- sult of the comparatively great variation in Y values and the restricted range in X values. FECTJNDITi' OF THE PACIFIC SARDINE 435 Table 4. — Comparison of regression coefficien ts n Y^a+bX Y= 6X5 t P t P SP-8 22 54 18 } ^" } 46 } '" } 49 1 ^^ } *' 45 1.944 0.834 1.359 2.411 1.381 0.313 0.016 0.859 0.659 0.235 <.05 <.4 < 1 <.01 <. 1 <•? <.9 <.3 <.4 <.8 0.0085 0.652 0.433 0.803 0.400 0.321 0.404 0.072 0.363 0.100 SP-9 <.9 SP-8 SP-Il <.5 SP-8 SP-12 <.6 SP-8 SP-13 <.4 SP-9 SP-11 <.6 SP-9 SP-12 <.7 SP-9 SP-13 <.6 SP-11 ^ SP-12 SP-11 <.9 SP-13 <■ 7 SP-12 SP-13 <.9 RELATION BETWEEN FECUNDITY AND WEIGHT The fish used for this fecundity study had originally been preserved in formalin, but many of them were subsequently transferred to alcohol. The fish in alcohol were noticeably diiTerent in appearance from those in formalin. The sub- cutaneous fat deposits of the formalin preserved fish showed through the skin and scales as a white background wherever pigment did not conceal them. These fat deposits were not apparent in the alcohol preserved fish. The formalin pre- served fish also weighed considerably more than alcohol preserved fish of comparable lengths from the same sample. Only a few fish from sample SP-1 1 were in a jar containing alcohol, and of tliese only one was a female for which ovum counts were made. The weights of the other 40 females for which ovum counts were made are plotted in figure 7 as the independent variable, with numbers of ova as the dependent variable. Two different regression lines have been fitted to the data. As will be shown for this sample, the weight ' of a sardine is a much better indicator of its fecundity than is its length. It can also be shown that for practical purposes the F-intercept can be taken as zero in the fecundity-weight regression without any significant loss of accuracy (table 5). Table 5. — Comparison of two fecundity-weight regressiotis Item Y=a+bX Y=bX a -1-2.94 .2402 4.91 .624 0 0 b. 2628 S, .. 4 92 r 622 The use of the formula Y=bX enables one to determine directly the total weight of spawning ' The weight used is round weight. If the ovary weight is subtracted from the round weight, the regression is not changed significantly (a = 4-3.3, 6 = 0.25, S,=5.1, r=0.60). 180 190 200 210 220 230 STANDARD LENGTH IN MILLIMETERS Figure 6. — Fecundity-length regression Y—a + bX. 240 250 4:;tiUi ()_,i7_ 436 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE female sardines required to produce the estimated number of ova present on the spawning grounds for any one season (assuming that each sardine spawns only the most advanced mode of ovarian ova). If the weight-fecundity relation of sample SP-11 is taken as representative of that of the stock of spawning sardines, the weight of spawning females can be easily estimated on the basis of 262.8 ova per gram or 238.4 million ova per ton of spawning female sardines. COMPARISON OF FECUNDITY-LENGTH AND FECUNDITY-WEIGHT REGRES- SIONS It has been shown that for the size-range of sardines in these samples, there is very little differ- ence among the regression lines of fecundity on length, on length squared, or on length cubed. For purposes of comparison, the fecundity-length regression line was computed for the same 40 fish used for the fecundity-weight regression line. The partial regression coefficients and the multiple regression of fecundity on length and on weight * were also computed (table 6). It would appear from this comparison that the fecundity of these sardines is not only better correlated witli their weights, but that the correlation with length is merely a reflection of the very good correlation between length and weight. This comparison may be tested further by the use of condition factors (Clark 1928). In the present case the condition factor, K,^ equals the weight of the fish in grams divided by the cube of the length in millimeters and multiplied* by 10^ (to give a three-digit whole number). Condition ' Over this length-weight range the cubic relation of length to weight is obscured, and the length-weight regression will approximate a straight line. 5 The use of condition factors "corrected" by subtracting the ovary weight from the fish weight before computing ii gives results that are almost identical to those obtained from K computed from round weight. Theoretically it might also be argued that the fish which contained more ovarian ova or larger ova, or both, had produced them at the expense of other tissues and there- fore the inclusion of ovary weight in fish weight should have a corrective value rather than the opposite. 120 130 140 150 FISH WEIGHT IN GRAMS 160 Figure 7. — Relation between fecundity and fish weight for 40 sardines (San Pedro, Feb. 9, 1946). (Dashed line, K=.2628X; solid line, y=2.94-f .2402X). FECXJNDITi' OF THE PACIFIC SARDINE 437 factors in tlieorv (assuniiii UJ a - 5 ■10 1 r 1 \ \ r 15 Lv _L _L J_ J_ _L _L J_ no 115 120 140 145 150 125 130 135 CONDITION FACTOR Figure 9. — Deviations of fecundity-weight regression plotted against condition factor (Sample SP-11). tionship. From an examination of weiglit-lengtli data of commercial fishery sardine samples over a 12-year period, it is evident that tiie average K value of the sardine population varies greatly not only within a year but also from year to year. Included in the data recorded in the course of routine sampling of the commercial sardine fishery are the individual lengths of the 50 sardines that make up a sample and their total weight. An approximate K value can be obtained for any sample by dividing the mean-sardine-weight of that sample by the cube of the mean-sardine- lengtli of that sample. An approximate K ob- tained by this method for the 40 sardines used for the weight -fecundity regression in this paper is 128.87; the mean K value for these same fish is 128.55. Approximate K values obtained from the San Pedro commercial fishery samples for January of various years illustrate how much year to year variation can occur (table 7). Table 7. — Condition factors of sardines collected in January during 1941-42 to 1953-54 Month and year Average length in millimeters of sardines in 50-flsh sample— Season 170-179 180-189 190-199 200-209 210-219 220 and above All sizes No. samples Av. K No. samples Av. K No. samples Av. K No. samples Av. K No. samples Av. K No. samples Av. K No. samples Av. K 1941-42 Jan. 1942 Jan. 1943 2 1 117 112 12 6 16 116 118 121 39 30 30 117 118 122 6 17 12 117 118 124 1 1 3 110 120 124 60 54 63 117 1942-43 118 1943-J4 Jan. 1944.- 2 123 122 1944-45 1946-16 Jan. 1946 2 1 126 126 9 6 2 5 fi 1 8 124 129 132 127 124 124 129 8 6 4 1 4 28 2 127 128 132 132 127 127 125 1 2 128 128 20 16 10 11 15 30 19 5 3 126 1946-47 Jan. 1947 2 3 1 2 115 131 129 113 127 1947-48 Jan. 1948 Jan. 1949 1 123 131 1948-49 4 3 128 121 128 1949-60 Jan. 1950 123 1950-61 Jan. 1951 1 9 5 3 128 130 137 134 127 1951-52 Jan. 1962 129 1952-53 Jan. 1963 137 1953-54 Jan. 1954 134 FECUNDITY OF THE PACIFIC SARDINE 439 180 190 200 210 220 230 STANDARD LENGTH IN MILLIMETERS Figure 10. — Regression line y=0.51-lA' — 76.0 adjusted for various K values by formula deviation = . 372X^ — 47.9. In figure 10 are plotted the regression lines of fecundity-length for tlie 116 sardines adjusted for various K values on the basis of the computed deviation values owing to K (weight correction) found in sample SP-11. On the basis of this figure it can be seen that a 20-unit K value differ- ence between two groups of sardines of similar length distribution would result in a difference of 7.4 thousands of ova in tlie most advanced grouj) of developing ovarian ova for each fish. Tlie above is, of course, purely theoretical when applied to years other than 1946. Sample SP-11 is prob- ably representative of the population present in the San Pedro area in January and February of 1946 with respect to K. The mean K value of sample SP-11 (taken on Feb. 9, 1946) is 128.6. The approximate mean K value obtained from thirty-four Sfl-fish samj)los from the San Pedro commercial fisiiery in January and February of 1946 is 128.1. What actually happens in years of very high or very low K values of the Pacific sardine may be very different from what can be postulated to happen on the basis of fecundity values extrapolated to these high or low K values. The pre-spawning-season condition of a fish may determine not only how many ova will develop per batch of developing ova, but also how many batches will spawn, or if spawning, will take place at all in that season. RELATION BETWEEN FECUNDITY AND AGE The mean lengths of each age group arc com- pared with those obtained by Phillips (1948) for the 1945-46 commercial sardine fishing season at San Pedro in table 8. The sampling method upon which Phillips' data are based was designed to sample the commercial catch representatively throughout tlie fishing season. Although he used fewer fish, a much larger nimtber of samples are included in his data. 440 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Age-fecundity data are presented in figure 11 and table 9 for the 113 sardines for which both fecundity data and age data were available. These show that the correlation between fecundity and age is not so good as that between fecundity and length or weight for these samples. Simpson (1951) also found a poorer correlation between age and fecundity than between fecundity and length or weight for plaice. On the other hand, Lehman (1953) obtained a better correlation be- tween fecundity and age than between fecundity and length or weight for shad. ONE AT 101.8 ANNULI 1 2 3 4 5 6 YEAR CLASS 1944 1943 1942 1941 1940 1939 AGE 2 3 4 5 6 7 Figure 11. — Relation between fecundity and age. Table 8. — Mean lengths of 9-year classes of sardines taken in the 1945-46 commercial fishery at San Pedro Samples SP-1 through 13 Phillips Number of amiuli Year Class Number of fish ■ Mean length Number of flsh 1 Mean length 0 1945 1944 1943 1942 1941 1940 1939 1938 1937 256 299 369 215 74 28 10 4 4 Mm. 161. 1 184.0 199.4 210.7 213 4 227.1 244.0 237.5 245.5 Mm. 1 - 2 105 337 242 151 44 14 1 1 182 200 3 -. 206 4 6 6 213 218 224 7 248 g 254 1,259 895 Table 9. — Age-fecimdity regression data using formula Y = a + bX (Y = No. of ova in most advanced mode: X = No. of annuli ') Sample Number offish a 6 Si/ r SP-8..- SP-9 17 9 39 5 43 113 14.0 14.5 24.0 23.8 16.4 17.7 3.83 2.97 3.36 3.38 8.00 5.28 5.64 7.52 5.66 5.56 11.28 9.82 0.438 377 SP-11 SP-12 .448 .674 SP-I3 . 665 All - .525 1 Males and females. ' Number of annuli+l = age. NUMBER OF SPAWNINGS PER YEAR The maturing ovaries of a sardine contain yolked ova of two or more size groups. This point was illustrated in figure 2. Clark (1934) pointed out this fact and used it as a basis for concluding that individual sardines spawn more than once during each spawning season. As the number of batches of ova spawned per season is as critical a point in determinations of fecundity as the number of ova spawned per batch, and much more difficult to assess, I will review evi- dence for and against multiple spawnings. The data presented in this paper are based on sardines collected during the period November 1945 through February 1946. It is unlikely that any of the sardines had spawned during this period. These months, with the possible excep- tions of January and February (when nominal amounts of spawning have been found), are in advance of the sardine spawning season in the soutliern California area. As no samples were available after February in 1946, the changes that took place during and after the spawning season could not be studied. However, some of the data obtained from the January and February samples do bear on the problem of multiple spawning, and these will be summarized. It is assumed by workers on fecundity that only ova in the most advanced mode will be spawned at a given time or batch. Hence, the number of ova in this mode is taken to represent tlie number of ova spawned per batch. The ratio of number of smaller yolked ova to number of ova in the most advanced mode has been taken as an indication of the number of batches that could potentially be spawned in a season. The number of yolked ova found in a mature ovary just prior to the initial spawning of a season would represent the total number of ova spawned in that season if (1) all the yolked ova were subsequently matured and spawned, and (2) if no ova were subsequently added to the group of yolked ova. FECUNDITi' OF THE PACIFIC SARDINE 441 Although the 1946 data are too restricted in time months in 1946 (table 10). A discussion of Clark's to show any trend in ratios of ova in the largest data is given in the latter part of this section. mode to smaller yolked ova during the spawning season, thev are of value in showing the ratios Table XO.^Frequeticies of ratios of all smaller yolked ova , , . ' . . , . I , to the number of yolked ova in the most advanced group found in maturing ovaries during January and „^ 5„„ P^^^o ,„ January and February 1946 February. The data are summarized by stage of development in table 10. There are three aspects of this tabulation to which I would like to call attention: (1) There is no change in ratio between stages E, F, and G (as defined by Clark 1934:13). This indicates that no additional ova were being added to the group of smaller yolked ova while the more advanced group was developing from stage E to G. (2) There is marked variation in the ratio of large ova to small in different fish; this is equaUy true if the largest mode is in stage E, F, or G. The range in ratios is from 0.3 to 3.2. (3) The ratios are markedly lower than those reported for San Pedro in January-February of 1929 and 1930 by Clark (1934). The ratios given by Clark for 3 ports and several seasons are summarized in table 11. Clark com- pared the ratio of ova in stage G (the last abundant stage before the ova become translucent and are spawned) to those of all smaller yolked ova. The ratios reported by Clark (table 11) during January and February 1929 and 1930 at San Pedro are about twice as large as those I found for the same Table 11. — Ratios of all ova between 0.20 and 0.59 mm. (stages B to F) to all ova larger than 0.59 mm. for all females in stage G (from Clark, 1934, table U V- ■''') "nd number and percentage of adult female sardines in stage G {nwde ■(>4--73 mm.) and stages A {resting-mode between 0 and 0.20 mm.) and L (spent) (after Clark, 1934, table table 8, Monterey, p. SO; table 9, San Diego, p. 32) Ratio stage (after Clark 1934) of most advanced group of developing ova D E F a Total Median diameters .34-.43 .44-.53 .54-.63 .64-.73 .34-.73 2 3 1 1 4 1 1 5 1 1 6 7 1 2 1 4 2 4 3 4 2 1 1 4 3 5 5 5 4 3 3 2' 3 8 1 2 2 2 3 2 4 4 9 4 10 - -- 7 1.1 _ I 8 1 2 8 1 3 11 1.4 - 1 13 1 5 9 1 6 1 2 1 1 2 1 5 1 7 2 4 3 2 1 7 18 8 19 -- 4 20 4 3 22 3 4 2 4 1 1 2 5 1 1 26 1 1 2 7 1 1 2 8 29 2 1 3 30 3 1 i 1 2 1.3 38 1.5 45 1.5 28 1.5 113 Mean ratio- L5 San Pedro, p. 27; Monterey San Pedro San Diego Date Ratio stages B-F/G Stage G Stages A and L Ratio stages B-F/Q stage G stages A and L Ratio stages B-F/O Stage Q Stages A and L 1929 Number 0 3 1 2 2 0 0 0 0 2 5 16 14 2 0 2 19 20 2 Percent 0 10 33 7 10 0 0 0 0 4 14 47 44 7 0 6 60 69 6 Number 2 0 1 22 18 30 20 42 33 8 5 2 9 23 16 2 0 1 27 Percent 7 0 33 76 90 97 95 100 100 14 14 6 28 85 37 6 0 4 84 3.36 3.80 2.87 2.38 2.26 2.39 2.07 2.11 2.95 3.06 2,02 2.92 2.84 Number 9 18 10 5 6 2 5 2 0 9 17 20 2 16 2 12 16 2 2 Percent 17 33 50 9 17 25 14 4 0 18 33 37 9 48 5 27 34 6 6 Number 0 0 1 6 2 6 36 40 0 2 5 3 7 4 0 0 0 2 Percent 0 0 5 2 17 25 17 77 100 0 4 9 13 21 11 0 0 0 6 Number Percent Number Percent TT^i-ITI-ai 5.06 IV24-V23 3.96 4.16 2.17 2.88 V-24-Vr21 VI 22-VII-21 VII 22-VIII-20 VlII 21-1X18 IX19-X18 1930 1:15-11:13 3.88 3.72 2. 85 2.22 2.42 2.34 2.71 3.12 8 32 13 28 65 62 0 0 0 0 11:14-111:14 0 III:15-IV:13 0 IV:14-V12 V13-VI 11 1931 I-6-II'3 11:4-111-4 III 5-IV-2 IV:3-V-2 V:3-V:31 442 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE I interpret the differences in the ratios between 1929-30 and 1946 as real. I seriously question that they could be due to differences in techniques or interpretation of modes by Clark and myself. There have been three investigators working on the data I have presented in table 10, and the three of us have obtained comparable results. If the ratios are any indication of the average num- ber of batches spawned per female sardine, then the number of batches must vary considerably from year to year. One implication of this is that when basing population estimates on egg surveys and fecundity data, it may be necessary to have current determinations of both egg abundance and average fecundity for that season. I find that there is no significant difference among ratios of the 113 females considered in table 10 when grouped by size (standard length) rather than stage of maturity (table 12). Clark postulated that larger fish mature earlier than the smaller fish and spawn over a longer time interval. My study does not indicate a higher ratio (of smaller ova to ova in the most advanced mode) in larger female sardines. Hence, if larger fish spawn more batches of eggs in a season than smaller fish, these additional batches are not evident in maturing ovaries. Therefore if larger fish are to spawn more batches of ova than smaller fish, one of the following two conditions must be fulfilled: either additional ova are added to the mass of smaller yoked ova from the reservoir of non-yoked ovocysts, or, if this does not occur, then larger fish must spawn a larger proportion of the smaller yoked ova present in the initial maturation period than do smaller fish. If the first condition obtains, then there would be little significance to ratio changes during the season. If the second condition obtains, then only part of the smaller yoked ova initially present in the developing ovary would eventually mature and be spawned. Hence, smaller fish should have a larger number of batches of degenerating ova at the termination of their spawning than the larger fish. With regard to the 1946 samples, the average ratio of 1.5 to 1 shows that an average of only 2.5 batches could be spawned if all yolked ova were matured and spawned, and no additional yolked ova were added during the spawning season. If a portion of these degenerate, then the average number of batches spawned per female sardine must be between 1 and 2.5 batches. Andreu (1951) postulated that the peak spawn- ing season of the European sardine, as measured by the occurrence of planktonic eggs (Hickling Table 12. — Frequenries of ratios of number of all smaller yolked ova to the riiituher of yolked ova in the most advanced group (taken as unity) grouped by fish size (standard length) Fish length in millimeters Ratio 170- 179 180- 189 190- 199 200- 209 210- 219 220- 229 230- 239 240- 249 250- 259 To- tal 0.0 0.1 0.2 - 0.3 1 1 0.4 1 1 0.5 1 1 0.6 0.7 2 2 2 1 1 4 3 1 3 2 1 1 2 3 0.8 1 2 1 4 0,9 4 1.0 1 5 3 1 4 r 1 2 2 1 3 2 4 6 1 2 3 2 1 1 1 2 7 1.1 _ 1 8 1 2 1 1 8 1.3 . - 3 ..... 11 1.4 ..- 1 1 13 1 5 9 1.6 .. 5 1.7 1 1 7 1 8 2 8 1.9 1 4 2.0 2 1 4 2 1 1 1 1 3 2.2 2 4 23 2 4 1 1 2.5 . 1 1 26 1 1 2 7 1 1 28 29 1 1 1 3 3.0 3 1 3 2 1 1 1 1.1 Total. _- Mean ratio 1 1.4 9 1.3 26 1.4 20 1.4 36 1.6 15 1.6 3 2.1 2 1.1 113 1.5 1945) off Plymouth, was too restricted in time to allow the sardines to mature and spawn more than one modal group of eggs in a spawning season. This could also be inferred for the Pacific sardine from table 12, and the ovum growth rates postulated by Clark (1934:29): ". . . probably slightly more than two months are necessary for eggs to grow from stage C [last mode between 0.26 and 0.34 mm.] to stage G [last mode between 0.64 and 0.74 mm.]," and ". . . approximately three or four weeks will elapse before females in stage G reach maturity." Clark's figures show modal egg diameters for groups of translucent ova ranging from 0.80 to over 1.30 mm. According to Ahlstrom (1950:134- 135), "Fertilized eggs average about 1.70 mm. in diameter (range 1.35-2.05 mm.)," pre-cleavage eggs "taken during the four hour period before midnight are considerably smaller in diameter than are those with some embryonic development. Since the yolks are of similar size in both groups. FECUNDITl' OF THE PACIFIC SARDINE 443 the difference lies in the width of the perivitelline space, wliich is nearly wanting in pre-cleavage eggs taken during this period; such eggs averaged only 1.20 mm. in diameter (range 1.02-1.38 mm.)." Measurements of the yolk diameters of 343 sardine eggs (fig. 2i) from 8 plankton samples taken at various times and localities showed the mean yolk diameter to be .84 mm. (range .74-1.00). The mean egg diameter was 1.66 (range 1.40-1.90), indicating that this group of eggs is comparable with those constituting Ahl- strom's data (mean =1.70 mm.). One sardine of those used for this study contained translucent j-olked ovarian ova. The mean of 118 diameter measurements of these ova was .84 mm. (range .76-.96 mm.). As there was no perivitelline space, these measurements are also yolk diameter measurements. Clark's (1934) larger diameter measurements of ovarian ova apparently include a perivitelline space as well as the yolk and therefore are not comparable with the ovarian ova, .20 to .84 mm. (mean diameter), which include yolk diameter only. Referring to sardines containing translucent, yolked ovarian ova, Clark (1934:11) states, "In the 11 years of study only 39 have been found among the thousands of females which have passed under observation." This would indicate that this stage is very transi- tory. In table 13 I have summarized data on the length of the spawning season off southern Cali- fornia as determined from sardine egg surveys. Between 70.7 and 97.2 percent of the spawning in this area occurred within a period of two months, and between 88.2 and 99.6 percent occurred within a period of 3 months. The peak month (italicized in table 13) was different in each year, indicating the amount of variation that can occur in the time of spawning. Clark (1934:19) bases the case for multiple spawning in the sardine on four lines of evidence: "The multiplicity of modes in the ova diameter Table 13. — Percent of spawning occurring in each month of 1940 and 1941, 1950, and 1951 spawning seasons off southern California ' Year Feb. Mar. April May June July Aug. 1940 U.O 17.0 0.0 0.0 29.7 47.1 0.0 1.3 il.O 24.1 2.4 1.3 18.3 8.7 21.3 89.6 1941 3.1 76.9 25.9 1950 1951 0.4 2.0 0.0 0.0 ' Based on sardine egg survey data given by Sette and Ahlstrom (1948) and Ahlstrom (1954). frecjuency curves; the high degree of correlation between the growth of these successive groups of eggs; the occasional presence in the ovary of a few ripe unspawned eggs accompanied by a new ripening group; and the decrease, as the breeding season advances, in the numerical ratio between succeeding batches of eggs and the largest size group . . ." The presence of different size-groups of yolked ova in the developing ovaries of any species of fish has been accepted by various authors as a, if not the criterion of the existence of multiple spawning in that species. However, it is also known that in a number of species of fishes at least some of the yolked ova of intermediate size are not spawned, but instead degenerate and are resorbed. This does not necessarily mean that none of the intermediate-sized ova will be spawned. Clark (1925) concluded that multiple spawning occurred in the grunion {Leuresthes tenuis). The ovaries of these fish contained a group of inter- mediate-sized yolked ova from which a group of ova to be spawned developed and segregated by size at approximately two-week intervals. At the end of the spawning season Clark (1925 : 22) found that, "In all cases, eggs in the intermediate group were undergoing a process of degeneration and resorption." Hart and McHugh (1944) figure the distribution of ovum diameters for three species of Osmeridae found along the Pacific Coast of British Columbia. As these three species participate in inshore runs their spawning activities are much better known than those of species that spawn offshore. One of these species, the eulachon {Thaleichthys. pacijicus), migrates into rivers to spawn. The spawning migration lasts from mid-March to mid-May. Each female apparently spawns only one batch of ova; the ovaries of a ripe female contain only the group of matured ova to be spawned and verv small ova that contain no yolk. A similar condition is found in the Pacific herring {Ciupea pallasi). I had the opportunity to examine five ripe females of this species. I found only one mode of yolked ova, all larger than 1.00 mm. in diameter and the residual non- yolked ova, all smaller than 0.20 mm. This species is a demersal spawner, the eggs being deposited on eel grass and seaweed. The spawn- ing of the one group of matured ova may take 444 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE place over a period of several days (Fraser 1922). Another Osmerid studied by Hart and McHugh (1944) is the silver smelt (Hypomesus pretiosus). These fish spawn on beaches during high tides and in most months of the year. The ovaries contain yolked ova of intermediate sizes. Schaefer (1936) has also figured the ovarian ova diameter distributions of this species. The above three authors concluded that this species spawns several batches per season. The third Osmerid for which ovarian ova di- ameters are figured by Hart and McHugh (1944) is the capelin (Alallotus catervarius). Of this species the authors say, "Spawning takes place in various localities in the strait of Georgia during late September or the month of October . . . Spawn- ing takes place in the evening at high tide right at the water's edge . . . The size frequencies of the eggs [fig. 10] suggest that the mature capelin spawns more than one batch of eggs as there ap- pears to be one group of small eggs becoming differentiated from the general mass which com- prise the residual ovarian tissue after the ripe eggs are spawned out. It is not known at the present time whether any such second spawning occurs. It is possible that more or less nomad schools of capelin move around the southern part of the strait of Georgia spawning on suitable beaches as the eggs ripen. There is no racial evidence against such a belief but the only evidence in favor of such a supposition is the observations on the ovaries taken in conjunction with the lack of second spawnings occurring on the same beaches." Thus we find three possible types of spawning occurring among these shore spawning species : (1) only one group of yolked ovarian ova with only one known spawning (eulachon, herring); (2) more than one group of j'olked ovarian ova with multiple spawning (grunion, silver smelt); (3) more than one group of yolked ovarian ova with only one known spawning (capelin). The degree of correlation between the modal value of the most advanced group of developing ova and the modal value of the secondary group of developing ova was used by Clark (1934) as an indication of multiple spawning in the Pacific sardine. June (1953) also used this method for the yellowfin tima (Neofhunnus macropterun). By this method of correlating modal values, Clark (1934) obtained a coefficient of correlation of 0.70 for modal diameters below 0.80 mm. and 0.72 for modal diameters above 0.80 mm.; June (1953) obtained a coefficient of correlation of 0.855 for his tuna data. There may be some doubt concerning the inter- pretation of these correlations because of the mathematical restrictions imposed upon the plot- ting of the data. Although there are biological limits to the plotting of any biological data there should be no mathematical limitations; that is, it should not be mathematically impossible to plot any point. In the present case the convention of always plotting the larger diameter on one axis and the smaller on the other limits the plots to a triangular area rather than a rectangular area. It can be demonstrated that while a regression line based on a random plotting of points (positive numbers) in a rectangular area wOl have a Y- intercept (a) equal to the mean of Y, a slope ib) of zero and a coefficient of correlation (/•) of zero, a regression line based on a random plotting of points in a traingular area with a hj-potenuse of slope 1.00 passing through zero (i. e., coincidence of the two modes in the present case) will have a F-intercept of zero, a slope of .500 and a coeffi- cient of correlation of .500. Regarding the "decrease, as the breeding season advances in the numerical ratio between suc- ceeding batches of eggs and the largest size group," Clark (1934:19) states: "The change in the ratio from approximately 4-1 to 2-1 suggests that eacli fish may mature an average of three batches of eggs, although this number may be higher, for this study furnishes no data to deter- mine whether growth from the immature to the maturing class accompanies growth within the maturing sizes." Her ratio data are given in table 11 along with data on the numbers and percentages of fish in stage G (the last abundant stage before the eggs become translucent and are spawned) and in stages A (resting) and L (spent). The data for numbers and percentages are based on sardines larger than 199 mm. in length; the ratios appar- ently include a few smaller sardines in some months. High percentages of stage G fish were shown for Monterey between March 15 and May 13 in 1930 and between March 5 and May 2 in 1931 (the 1929 Monterey data are not used as they are based on only 8 stage G fish for the season) , fol- lowed in each case by high percentages of stages A FECUNDITi^ OF THE PACIFIC SARDINE 445 (resting) and L (spent) fish in the next hiiiar period. This would indicate a short peak spawning period in the Monterey region during tliese two years. Clark shows a decrease in ratios in 1930 at Mon- terey from 3.88 in the January-February lunar period (1:15-11:13) to 2.22 in the April- May lunar period (IV:14-V:12). Assuming that the values adequately represent the ratios during the two periods (the higher is based on 2 speci- mens), the decrease in ratios amounts to 1.66, hence the average number of batches spawned per female could approximate 2.66 batches. The situation at San Pedro is more difficult to interpret. In 1929 Clark shows a high ratio of 3.80 (based on 18 G stage fish sampled between 11:24-111:25), a low of 2.07 (based on 5 stage G fish sampled between VII:22-VIII:20) for a total decrease in ratios of 1.73; this approximates the decrease given above for Monterey in 1930. It should be noted that most of the decrease in ratios in 1929 at San Pedro also occurred during a two- montli period. The ratio of 3.80 had decreased to 2.38 by lunar period IV:24-V:23 for a drop in ratios of 1.4. During the four succeeding lunar months the further drop in ratios amounted to only 0.3. The 1930 data show no real change in ratios during the season. A drop in ratios from 3.06 to 2.02 occurred between two lunar periods (between 11:14-111:14 and III:15-IV:13), but the ratios again jumped to 2.92 and 2.84 in the two following lunar periods. At the end of the spawning period at both ports spawning females still contained several modes that had not been spawned. These yolked ova must have degenerated and been resorbed. In fact, it may be characteristic of fish wliich mature several modes of ova, that one or more of the modes will be resorbed eventually rather tlian spawned. LENGTH AT FIRST MATURITY To facilitate handling of the data, the samples have been grouped by lunar periods ^ in table 14. Clark (1934) considered as maturing all those fish having a group of ovarian eggs with a modal diameter of .22 mm. or larger (i. e., yolked ova). ' The Pacific sardine fishery in California Is carried on at night, when the fish schools can he located by inmlne.scence. There is generally no flshing (luring the full-moon period. Lunar periods are assigned inirnbers in con- nection with the sampling program of the California Cooperative Oceanic Fisheries Investigations. Using this critcriDii, the 250 fish taken in lunar periods 325 and 326 are compared in table 15 with Clark's data for the months of February, March, April, and May of the years 1929, 1930, and 1931. Table 14. — Samples grouped by lunar periods Lunar period Sample Date Number of females 322 323 . f SP-1 \ SP-2 SP-3 SP-4 SP-5 SP-6 \ SP-7 SP-8 SP-9 SP-10 SP-11 SP-12 SP-13 Nov. 10, 1945 Nov. 13, 1945 Nov. 26, 1945 Dec. 5, 1945 Dec. 10, 1945 Dec. 28.1945 Jan. 5, 1946 Jan. 11,1946 Jan. 24, 1946 Feb. 1, 1946 Feb. 9, 1946 Feb. 21, 1946 Feb. 27, 1946 »« 119 324 325 326 123 146 .04 Maximum ovum diameter, the measurement used to describe the stage of maturity for the 1946 material, is the diameter of the largest ova- rian ovum present in a sample of the ovary. An average value (mode, mean, or median) of the group of developing ova cannot be determined ac- curately for early developmental stages when the developing ova are not completely dift'erentiated from the mass of immature, non-yolked ova that are always present in the ovary. For developing groups of ova that have differentiated enough from these immature ova to form a distinct size group, the maximum ovum diameter is about .07 mm. greater than the median ovum diameter of that size group. This difference is probably less for developing groups of smaller ova than for groups of larger ova. Table 15. — Numbers of sardines and percentages containing developing ova |San Pedro, Calif.) Lunar period (1946) Lunar 1929-1930-1931 325 (Jan. 24, IH-rlod 326 Lunar periods Clark (1934) Standard Feb. 1 and 11) (Feb. 21 325 and 326 (Feb. through length and 27) May) meters) Total Percent Total Percent Total Percent Total Percent num- matur- num- matur- num- matur- num- matur- ber ing ber ing ber ing ber ing 120-129... . 0 0 0 1 0 130-139.--. 0 0 0 0 14M49.. 0 1 0 1 6 4 6 150-159 - 11 9 24 17 35 14 4 0 160-169..,. 34 29 23 43 47 35 21 24 170-179 - 8 63 2 100 10 50 21 38 180-189.. fi 100 10 100 16 100 36 58 190-199.. 13 92 20 100 33 97 62 89 2(X)-2fl9... 25 96 9 100 34 97 83 210-219.. 31 97 9 1 100 40 98 94 99 2'20-229 13 100 3 i 100 IB 100 85 100 230-239.. 4 100 0 4 100 79 99 240-249 . 1 100 2 100 3 100 75 97 ;>.'iO-2,59.... 0 1 100 1 100 67 100 2(iO-2fi9.... 0 0 0 54 98 270-279.... 0 0 0 36 100 280-289..-. 0 0 0 10 290-299..-. 0 0 0 2 446 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 16. — Comparison oj maximum ovum diameters ISO 160 170 180 190 200 210 220 230 240 250 260 STANDARD LENGTH IN MILLIMETERS Figure 12. — Sardine length frequency distribution for each of three maximum ova diameter ranges (ova to 0.20 mm. solid lines, 0.22 to 0.40 mm. dashed lines, 0.42 mm. and larger dotted lines). Figure 12 shows fish length-frequency distribu- tions for each of three maximum ovum size ranges and for each of the five lunar periods. The first category of ovum diameters, to 0.20 mm., covers the range of non-yolked ova. The maximum ovum diameter observed for any fish during lunar period 322, when the ovaries were in a "resting" stage, was 0.40 mm.; this size was used as a brealv between the two categories of yolked ova. This is also the maximum ovum diameter attained by any fish less than 175 mm. in length during lunar period 326, when all but two of the larger fish contained ova that exceeded this value. Table 16 compares maximum ovum diameters of fish more than lis mm. in length from the November 10 and 13 saniples with those under 175-mm. and those over 175 mm., taken in the February 21 and 27 samples. Hange of ovum maximum Lunar pf Nov. 10 Fish mc 175 mn riod 322 and 13 Lunar period 326, Feb. 21 and 27 diameters re than n. long Fish less than 175 mm. Fish more than 175 mm. .... to 0.20 mm... 0.22 ton.40mm... 0.42 and above Number 62 33 0 Percent 65 35 0 Number 33 15 0 Percent 69 31 0 Number 0 2 51 Percent 0 4 96 Total 95 100 48 100 53 100 15 ~I 1 1 1 1 1 1 1 1 1 1 1 1 M 1 1 1 1 1 1 1 1 1 1 1 NOV 10 TO NOV 13 _ 10 - - 5 : "\ _ n ^.i/N. \/^A ID NOV 26 TO DEC 10 _ 5 / yV\ _ 0 )..yn^:Ar:Ci^/\ ,-... 10 _ DEC 28 TO JAN. II 5 ^ - n /-^.,^--.^.^A•;.■ .-■.-...•■■■■-.. 10 _ JAN 24 TO FEB 9 _ 5 - / V^.^-^^/"^' '-../I. A A n ' ••■. 10 FEB 21 TO FEB 27_ 5 - fV y\ 0 1 / 1 1 >si/i Va^\ I .-1 -■■i,'-f-^i t,'^-Y-r-i 1 1 1 0 D8 .16 .24 32 40 .48 .56 .64 .72 .80 .88 .96 MAXIMUM OVUM DIAMETERS (IN MILLIMETERS) Figure 13. — Maximum ovum diameter frequency distri- bution for each of three sardine length ranges (148-174 mm. solid line, 175-189 mm. dashed line, 190-253 mm. dotted line). Figure 13 shows the maximum ovum diameter frequency distribution for each of three fish size ranges. The above data show a very definite growth of the ova in fish more than 175 mm. in length and negligible growth of the ova in fisli less than 175 mm. The state of development of the ovaries of the fish less than 175 mm. long in lunar period 326 (late February) is very similar to the FECtJNDITi' OF THE PACIFIC SARDINE 447 state of (levelopment of the ovaries of the fish above 175 nun. in k'lifitli in hniar period 322 (mid- Novemher). Clark (1934:22) states that, " * * * adolescent fish may start to mature eggs but these eggs may fail to reach a ripe state and eventually degenerate. Such a condition has been claimed for the hake by Hickling (1930), but we have not been able either to prove or disprove the possi- bility for tie sardine." As no samples are avail- able for tlie 1946 spawning season after February 27, tiie ultimate fate of the ovarian ova of these smaller sardines cannot be determined. If the presence of yolked ova in the ovary is the criterion of maturity, all fish more than 175 mm. in length and 31 percent of the fish under 175 mm. in lengtli from the samples of Februarj- 21 and 27 had attained maturity. If a maximum ovum diameter exceeding 0.40 millimeters (i. e., the maximum ovum diameter attained by any fish in the samples of November 10 and 13, when the ovaries are "resting") is the criterion of maturity, all fish 190 mm. or more in length, 82 percent of the fish 175 to 189 mm. in length and none of the fish 174 mm. or less in lengtii from the samples of February 21 and 27 had attained maturity. AGE AT FIRST MATURITY Numbers and percentages of sardines at each of three maturity stages for each age group and each lunar period are presented in table 17. If the presence of yolked ova in the ovary is the criterion of maturity, 28 percent of the one-year- old fish and 100 percent of the two-year-old and older fish were maturing in lunar period 326 (samples of February 21 and 27). But the pres- ence of yolked ova in the ovaries may not neces- sarih' mean that a fish will develop and eventually spawn those ova. Actually the one-year-old fish were less advanced in respect to ovarian develop- ment in lunar period 326 (samples of February 21 and 27) than the two-year-old and older fish were in lunar period 322 (samples of November 10 and 13). In any case the one-year-old fish taken by the commercial fishery cannot be considered repre- sentative of the one-3'ear-old fish in the population at large. One-year-old sardines are not often taken by the commercial fishery, and those tliat are taken are undoubtedly the larger specimens of that age group. Table 17. — Age at first maturity Age and year class Maximum ox'um diam- eter (in mil- limeters) Limar period 322 Lunar period 323 Lunar period 324 Lunar period 325 Lunar period 326 Number Percent Number Percent Number Percent Number Percent Number Percent (0 to .20 25 0 0 100 0 0 37 8 0 82 18 0 31 12 0 72 1-year-old (1945) <.22 to .40 -. 28 1.42 and over. 0 25 100 45 100 43 100 (Oto.20 ■^.22 to .40 1.42 and over. 2-years-old (1944) 17 7 0 71 29 0 23 15 0 61 39 0 4 14 3 19 67 14 0 7 3 0 70 30 0 5 24 0 17 83 24 100 38 100 21 100 10 100 29 100 (0to.20 •^.22 to .40 1.42 and over. 3-years-oId (1943) 22 15 0 59 41 0 12 13 1 46 50 4 13 35 14 21 56 23 12 23 3 33 64 0 0 17 0 0 100 Total ._ 37 100 26 100 62 100 36 100 17 100 roto.20 .22 to .40.... 1.42 and over 4-years-old (1942) 9 10 0 47 S3 0 3 10 1 21 71 7 3 15 17 9 43 49 1 5 26 3 16 81 0 0 6 0 0 100 19 100 14 99 35 101 32 100 6 100 f0to.20 ^22 to .40 (.42 and over. 5 to 9 years old (1934-41) 9 4 0 69 31 0 5 11 0 31 69 0 0 S 2 0 80 20 0 2 11 0 15 85 0 0 5 0 0 100 Total 13 100 16 100 10 100 13 100 5 100 448 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE SUMMARY 1 . Ovarian ovum counts and diameter measure- ments are presented for 13 samples of sardines con- taining 587 females taken by the San Pedro commercial fishery during the period November 10, 1945 to February 27, 1946. 2. Throughout the size range of fish studied there are no significant differences among the regression lines of fecunditj' expressed as the number of ova in the most advanced group of yolked ova on length, on length squared, or on length cubed. 3. The estimated number of ovarian ova present in the most advanced group of yolked ova in a sardine 200 millimeters in length is 26.8 thousands (using the formula Y=a-\-bX). The rate of increase or decrease in estimated numbers of ova is 5.14 thousands per increase or decrease of 10 millimeters of length. There is considerable variation in the number of ova produced by in- dividual fish at any given length. The standard error of estimate of y is 8.76 thousands of ova for the above formula. 4. The correlation between fecundity and fish weight is better than that between fecundity and fish length. The number of ova present in the most advanced group of yolked ovarian ova equals 263 ova per gram weight of sardine. 5. Both the fecundity-length and fecundity- weight correlations are better than the fecundity- age correlation. 6. No conclusions can be drawn from the material used in this study regarding the number of groups of ova that a sardine will spawn in one spawning season. The evidence presented in the literature by various authors is inconclusive and then- conclusions are contradictory. 7. All fish two or more years old and all fish more than 175 mm. in length from the February 21 and 27 samples contained yolked ovarian ova (0.22 mm. or larger in diameter) ; 28 percent of the one-year-old fish and 31 percent of those less than 175 mm. in length also contained yolked ovarian ova. None of the fish in these latter groups con- tained ova greater than 0.40 mm. in diameter, while 91 percent of the two-year-old and older fish and 96 percent of the fish greater than 175 mm. in length contained ova greater than 0.40 mm. in diameter. The stage of ovarian lievelopment of the one-year-old sardines less than 175 mm. in length on February 21 and 27 was much like that found in the older and larger fish taken about 3 months earlier (on November 10 and 13, 1945). LITERATURE CITED .\hlstrom, Elbert H. 1950. Influence of temperature on the rate of de- velopment of pilchard eggs in nature. U. S. Dept. Interior, Fish and Wildlife Service, Spec. Sci. Rept.: Fisheries No. 15, pp. 132-167. 1954. Distril^ution and abundance of egf^ and larval populations of the Pacific sardine. U. S. Dept. Interior, Fish and Wildlife Service, Fishery Bull. 93, vol. 50, pp. 83-140. .Andreu, Buenaventur.\. 1951. Consideraciones sobre el comportamiento del Ovario de sardina (Sardina pilchardns Wall?.) en relacion con el proceso de meduracion y de freza. Boletin del Instituto Espanol de Ocean- ografia. No. 41, 16 pp. Cl.\rk, Frances N. 1925. The life history of Leuresthes tenuis, an atherine fish with tide controlled i-pawning habits. California Division of Fish and Game, Fisli Bul- letin No. 10, 51 pp. 1928. The weight-length relationship of the Cali- fornia sardine (Sardina caerulea) at San Pedro. California Division of Fish and Game, Fish Bulletin No. 12, 58 pp. 1934. Maturity of the California sardine {Sardina caendea), determined by ova diameter measure- ments. California Division of Fish and Game, Fish Bull. No. 42. 49 pp. Phaser, C. McLean. 1922. The Pacific herring. Contriijutions to Cana- dian Biology, No. VI, pp. 103-111. Hart, J. L., and .1. L. McHiTnii. 1944. The smelts (Osmeridae) of British Columl)ia. Fisheries Researcli Board of Canada, Bull. 64, 27 pp. HlCKLIVG, C. F. 1930. The natural history of the hake. Part III: Seasonal changes in the condition of the hake. England. Min. Agri. Fish., Fish Invest., Ser. 2, vol. 12, No. 1, 78 pp. Ref. from Clark (1934). 1945. The seasonal cycle in the Cornish pilchard, Sardinia pilchardns Wall). Journal Mar. Biol. Assn., vol. XXVI, No. 2, pp. 115-138. June, Fred C. 1953. Spawning of yellowfin tuna in Hawaiian waters. U. S. Dept. Interior, Fish and Wild- life Service, Fishery Bull. 77, vol. 54, pp. 47-64. Lkh.man, Burton A. 1953. Fecundity of Hudson River shad. U. S. Dept. Interior, Fish and Wildlife Service, Res. Rept. No. 33, 8 pp. FECUNDITY OF THE PACIFIC SARDINE 449 MosHKR, Kenneth U. 1950. Description of a projection device for use in age determination from fish scales. U. S. Dept. Interior, Fish and Wildlife Service, Fish. Hull. .54, vol. 51, pp. 405-107. PiiiLLii's, Julius B. 1948. Growth of the sardine {Sardinops caerulra), 1941-1942 through 1946-1947. California Fish and (lame, Fish Bull. Xo. 71, 32 pp. Sette, ()sc.\r E., and Ki.bert H. .\hlstro.m. 1948. F'stiniations of abundance of the eggs of the Pacific pilchard (Snrdinops caerxdea) oft southern California during 1940 and 1941. Sears Founda- tion: Journal of Marino Research, vol. V'll, No. 3, pp. 511-542. SiMPSO.V, A. C. 1951. The fecundity of the plaice. Ministry of Agriculture and Fisheries, Fishery Invest. Xo. 5, vol. 17, Series II, 27 pp. U. S. GOVERNMENT PRINTING OFTICE 1956 FECUNDITY OF NORTH AMERICAN SALMONIDAE By George A. Rounsefell, Fishery Research Biologist This paper is the first in a projected series in whicli the autlior proposes to compile and evahiate the pubhshcd information on various phases of the hfc history and conservation of North Ameri- can salmonids. The available information is so widely scattered that merely bringing it together will facilitate the expanding research. Further- more, even a hasty perusal of the literature reveals large gaps in our knowledge. Once these gaps are clearly seen, there is a much better chance of their being filled. The primary purpose, however, is to discover througii comparison of the same life phases of the different species and genera, the relation between the fish and the ecological factors in their environ- ment. Since emphasis has been placed on material that would aid in developing principles, and as I am making the study as complete as possible without assistance, I am not including minor items of information. Original data are presented for Karluk River sockeye. Although not indigenous, the brown trout, Saimo Irutta, is included in this study as a thor- ougldy naturalized species. European and Asiatic literature is used sparingly, either to aid where knowledge of the North American stock is de- ficient, or to corroborate tlie North American findings. Fecundity is an especially interesting topic in the Salmonidae because the comparatively small number of large eggs suggests (as other re- searchers have proved, e. g., Rounsefell and Kelez, • 1938, Rounsefell, 1949 and in ms.') a demonstrable relation between the reproductive potential of the spawning stocks and the numbers of young sur- viving. Neave (1948) has also pointed out that the variation in egg number between species of Oncorhynchus is related to the varying vicissitudes of their life historv. ■ Factors causing decline in sockeye salmon of Karluk River, Alaska. V. S. Department of the Interior, Fish and Wildlife Service, Washinpton, D. C. Approved for publication. February 8, 1957. The relation between size of spawning stock and number of young produced is fundamental to studies of changes in population size. The survival from spawnings cannot always be de- termined at an early stage, but is more usually measured at some later stage of the life history. In this paper we are concerned with quantitative measurement of the reproductive potential of the spawning stock. Such measurements are usually gross estimates derived from one of the following bases : 1. Relative abundance of the adult population. This will usually be in pounds of fish caught by some standard amount of fishing effort (a stand- ardized unit of gear fishing a certain period of time). 2. Relative abundance of the eggs or larvae. This usually is a summation of the density of eggs (in the case of pelagic eggs) per cubic meter over the water area inhabited by the particular population under consideration. Estimates of abundance of species spawning in the littoral zone, e. g.. Pacific herring {Clupea pallasi), may be based on miles of shoreline utilized for spawning. 3. Actual numbers of mature adults. These numbers may be an actual count of the individuals or maj- be statistical estimates of population size. These measures of reproductive potential are each based on one or more of the following assumptions: 1. That the number of eggs spawned is in direct proportion to the number of mature adults and their mean weight (or length). For this to be true, the relation between size of fish and fecundity must be linear. Moreover, if the size composition of the adult population varies from year to year, then the theorem is true only if the regression of eggs on size passes through the origin, i. e., the regression formula must be of the form y = bx. 451 452 FISHERY BXJLLETIN OF THE FISH AND WILDLIFE SERVICE 2. That the annual sex ratio remains constant. 3. That the regression of number of eggs on size of fish does not vary between years. 4. That the size and/or age at maturity does not vary between years. 5. That the number of eggs is a function of fish size independent of age. 6. Tliat there is no annual variation in the proportion of the eggs retained by the females in spawning. The foregoing assumptions are usually not fully satisfied so that the variability of an ap- proximate measure of reproductive potential in critical experiments may be so large as to obscure the very factors, the effects of which the biologist is seeking to measure. This variability between numbers of mature adults and actual reproductive potential has long been recognized, and biologists have attempted to discount it by substituting an estimation of the total annual egg deposition for number of adults as being a better measure of reproductive potential. This paper is confined to an analysis of the factors causing variation in the relation be- tween number of eggs and number of mature adults. After making the necessary allowance for dif- ferences in size of fish, a wide range in fecundity still exists between races of the same species from different localities. For instance, McGregor (1922, 1923a) found that the king salmon of the Sacramento River have a far higher fecundity than those of the Klamath River. Thus, if y is number of eggs and x is length of the fish in centimeters, the formulae for the regressions of number of eggs on length are — Klamath River Log y'= .00682 .Y + 3.01 116 Sacramento River Log K=. 00319 A' + 3.56836 The Klamath River fish (65 specimens) ranged from 61 to 107 centimeters in length (average, 82.6), with a geometric mean of 3,754 eggs. The Sacramento River fish (50 specimens) ranged from 59 to 110 cm. (average, 92.4) and had a geometric mean of 7,298 eggs. At 85 cm., the calculated geometric means for the two popula- tions are 3,894 and 6,912 eggs, an increase of 78 percent in number of eggs for the Sacramento River fish when compared with king salmon from the Klamath River. The question arises as to the causes and the biological significance of such a great difference in fecundity between populations of the same species. It is recognized that harsher ecological situations impose lower survival rates on some races. Assuming that the number of eggs can be in- creased by selection (as seems to have been done for domesticated strains of trout), then the number of eggs may well differ genetically between various wild races of salmonids. In the case in point there is good reason to believe that the variation in egg number is not caused by variation in the marine environment since, as McGregor pointed out (1923b), Sacramento River and Klamath River king salmon occur together in the ocean troll catches. That the fecundity of fish of the same length may even differ widely between populations spawning in different portions of the same river system is shown by Aro and Broadhead (1950) for the sockeye salmon of the Skeena River. For 3 years, 1939, 1948, and 1949, the female sockeye of small Lakelse Lake (5.5 sq. mi.) averaged 58.9 cm. in length (58.1-59.6) with an average of 3,816 eggs (3,699-3,888); while for the 3 years of 1946, 1947, and 1949, the female sockeye of the large upriver Babine Lake (171.8 sq. mi.) averaged 58.5 cm. (57.1-60.1) in length with an average of only 3,181 eggs (3,056-3,389). In assessing the significance of differences in fecundity between fish of various localities, it becomes important to measure the variation within localities. Some of the important factors within localities to be considered are — 1. Size of \ the fish in relation to number of eggs- 2. Age at maturity. 3. Size of the eggs. 4. Seasonal trends in fecundity in the same locality. 5. Annual variation in fecundity. RELATION OF SIZE OF FISH TO NUMBER' OF EGGS Combining his own observations with those of Titcomb (1897), Ricker (1932) states that the relation between number of eggs and- length of fish is curvilinear for the eastern charr, or brook trout, Salvelinus fontinalis. The number of eggs varied from 80 in a 5.1 -inch charr to 5,630 in a 22-inch charr. However, Allen (1956) points out FECUNDITY OF NORTH AMERICAN SALMONIDAE 453 that Titcomb's data are of limited value since Titcomb stated that some of the charr had ap- parently dropped part of their eggs before being captured. Osgood Smith (1947) obtained a linear relation between the logarithm of tlie number of eggs and the body length of 29 eastern charr, but inasmuch as his specimens were from such diverse localities as California, Ontario, and Xortli Carolina, the results cannot be regarded as conclusive. The number of eggs of eastern charr from four localities is shown according to size of fish in table 1 and figure 1. These data show that the dif- ferences in egg number between localities are too great to permit combining localities in studying the egg number-fish size relation. When the curves for the separate localities are examined, it becomes apparent that the number of eggs in- creases appro.ximately as the weight of the fish, since the logarithm of egg number plotted against the logarithm of fish length approximates a straight line, as does the logarithm of fish weight against the logarithm of fish length. Table 1. — Fecundily of eastern charr, Salvelinus fontinalis 2 500 1000 CO o LlI li. o a: LlI m 500 250 100 -\ 1 1 — I — r -n/ew JERSEY (HATCHERYI 20 25 30 40 50 60 70 80 100 FORK LENGTH (CM.) Figure 1. — Relation of egg number to body length in Salvelinus fontinalis. Locality and fork lenplh W voniing (beaver pond) : 12.70 cm 14.44 cm l(i.82cm 19.65cm - New Jersey (hatchery stock): 29.69 cm ' 31.00 cm 32.1.1cm 35.21 cm 36.50 cm 38.00 cm - Michigan (streams): 10.40 cm » 12.T3cm 15.07 cm 17.41 cm , 19.74 cm _ 22.08 cm 24.42 cm - Quebec (Laiireiitides Park): 13.75 cm 16.25 cm 18.75 cm 21.25 cm 23.75 cm 28.25 cm --- 28.75 cm.. 31.45 cm 35.50 em Number .K v.'raee or fish number of eggs 2 148 5 191 5 275 2 376 5 916 4 1.028 30 1.114 22 1.249 4 1.611 8 1.867 38 104 91 169 59 268 24 395 15 525 8 643 4 753 4 131 14 177 10 200 11 280 14 362 13 505 3 732 6 970 2 1,469 Authority Allen (1956). Hayford and Em- body (1930). Cooper (1953). Vladykov and Leeendre (1940). 1 Converted from standard length by (actor 1.1. :Converted from total length by factor 0.92. The logarithm of the number of eggs shows a closer linear relation to length than does the actual number of eggs when specimens are avail- able over a wide range of length. However, over the rather narrow ranges of length at maturity found in Oncorhynchus, the difference is usually trifling and can be disregarded in computing. Extensive data on the relations between number of eggs and length and weight of the fish are given by Foerster and Pritchard (1941). Correlations between egg number and fish length and between egg number and fish weight are shown for Cultus Lake sockeye for each of 6 years (1932-35, 1937, and 1938) and for pijik salmon from McClinton Creek, Masset Inlet, for eich of 6 even-numbered years from 1930 to 1940, inclusive. In order to compare the values of the two series of correlations, we have combined the correlation coefficients for the various years by transforming the r values into z values (Fisher 19:10, p. 171). The value of /• for the combined samples is ob- tained from the weighted average value of z. The results are as follows: V'aiut of correlaticn of egg number with — FM length Fitit weight 0. 57 0. 56 0. 35 0. 40 Sockeye salmon (Cultus Lake).. Pink salmon (McClinton Creek). This means that in the sockeye about 31 to 32 percent of the variation in number of eggs is 454 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE associated with change in length of the fish; but in the pink salmon this association is much weaker, only about 12 to 16 percent. The two combined regression lines for number of eggs on length and on weight of the fish in figures 3 and 4 of Foerster and Pritchard (1941, pp. 58, 59) obviously have much steeper slopes than the regressions for the individual years, showing that these lines do not represent the regressions within years. Since the mean annual lengths of the fish varied in the same direction as the average number of eggs, these combined lines represent chiefly regression between years and are therefore of no utility in predicting egg number for various fish lengths within any individual year. As the relation between egg number and fish length within any year appears to be so weak in pink salmon, it is of interest to determine what factor is controlling egg number. One factor for which measurements are available is sea tem- perature at Ketchikan, Alaska, which is just across Dixon Entrance from Masset Inlet and slightly east of it. To determine the role of sea temperature we have made a covariance analysis using the pink salmon data from McClinton Creek, Masset Inlet, B. C, prepared by Foerster and Pritchard (1941), as follows: Year Mean length (AT,) Mean sea tem- perature in degrees Fahren- heit at Ketchi- kan, July to Sept. Mean number of eggs (Y) 1930 Cm. 51.1 51.6 52.7 53.0 .=3.0 54.0 56.7 57.0 55.3 56.4 54.8 54.9 1,535 1940 1,619 1938 1,698 1934 1936 1,804 1.899 1932 -- 1,758 Average 52.67 55.68 1,719 The results of the test are as follows: Number of eggs (Y) Correlations of Y with X's Standard regressions of Y on X's. Fish length | Sea tempera- ture (-ST.) I (ATj) 0. 7801 0. 0310 -0. 8692* -0.8314' i?=0.8593 (N. S.) Y=3.834 A'l - 115.616A'2-)-7,954 Standard error of 3's=0.21536 t for p,, ,=0.83138/0.21536=3.860 P of .05=3.182. The relation between the average number of eggs in McClinton Creek pink-salmon females and the summer sea temperature at Ketchikan, with fish length held constant, is shown in figure 2. The correlation, r,2, of .Yi with X^ is —0.9011 and is statistically significant. Obviously, both annual mean fish length and annual mean egg number are negatively correlated with sea tem- perature. The annual differences in mean egg count in pink salmon are a function of sea tem- perature, because it is the principal factor con- trolling average fish length. 1900 1500 54 55 56 57 58 MEAN SEA TEMPERATURE JULY-SEPT Figure 2. — Relation of annual mean egg number of McClinton Creek pink salmon, O. gorbuscha, (body length held constant) to mean July-September sea temperature at Ketchikan. The above analysis does not mean that the regression between egg number and fish length within years is invalid, but that the within- years regression for the combined samples can only be obtamed by reanalysis of the original data to eliminate the portion of the total regres- sion accounted for by regression between years. The problem of the relative effects of mean annual size and sea temperature on egg number for the sockeye salmon is complicated by the effect of varying age at maturity which will be discussed later. In addition to the between-years difference in egg number at any particular length, there is also the difference between rivers mentioned previously in the case of the king salmon populations of the Klamath and Sacramento Rivers. A better il- lustration of this is perhaps afforded by the data from Eguchi, Hikita, and Nishida (1954) on chum FECUNDITY OF NORTH AMERICAN SALMONIDAE 455 salmon, Oncorhynchu-s keta, in Japanese waters. They point out that chum salmon from Hokkaido rivers have a larfj:er number of eggs tlian chum salmon from rivers in South Kui'ile; however, analysis of their data shows that in both areas there is a significant difference between individual rivers. The analysis, based on data from 7 rivers in Hokkaido and 109 specimens of chum salmon, is as follows, using data for the left ovary only to simplify the tabulations: D. F. Sums of squares and products Errors of estimate Source of variation Sx' Sxil Sk> Sums of squares D. F. Mean square 108 2250. 0300 64603.0 6, 992, 007 5,137,112 107 Between rivers 6 102 666.0752 1583. 9548 28678. 8 35924. 2 2, 369, 067 4, 622, 940 Within rivers 3, 808. 177 101 37, 704. 7 For test of sigrnlficance of adjusted means _ . 1, 328, 935 6 221 489 2 F=221,489.2/37,704.7=5.87. For Pof .01, F=2.99 Similarly for South Kurile rivers, 7^=3.50 with /' for P of .01 of 3.17. For the combined data (243 specimens) which come from 13 rivers, F= 13.74 with F for P of .01 of 2.27. This shows that there is a tendency for each river to have its own regression line for egg number on fish length. The fact of differences between the regressions for chum salmon from different rivers results in three regression lines (three center curves of figure 3). The total regression (dotted line) >- < > o o UJ CD 1500 400 300 1200 I 100 1000 900 800 T T MEANS • - HOKKAIDO X - SOUTH KURILE 53 54 55 56 57 58 59 60 62 63 FORK LENGTH (CM.) FioiRK 3. — Hchitioii of egg iiumbir (left ovary) of Japanese chum salmon, O. kela, to body length to illustrate total between-rivers, and within-rivers regressions. 456 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE includes both the within- and the between-rivers regressions. The within- and the between-rivers regressions each has a useful connotation. If one wishes to estimate from the lengths of the fish, the number of eggs contained in a sample of chum salmon from a particular river, then from the between-rivers regression one obtains an estimate of the average number of eggs per female (left ovary) in accordance with the average length of the entire sample. If, however, one wishes to determine the difference in egg number (left ovary) between fish of different lengths within the same sample, then the slope of the regression would follow the within-rivers slope. RELATION OF AGE AT MATURITY TO NUMBER OF EGGS The best material available on the effect of age at maturity on egg count is in unpublished data for the sockeye salmon of Karluk River, Alaska, as follows : Num- Ages Method of Year hernf avail- Lengths enumeration of Dates sampled speci- able taken eggs ' mens 1926 2 40 No... Yes Number in 5 Sept. 15, 1926. gm. 1938 65 Yes.. Yes Weigllt of a Aug. 1-fi. 1938. counted sample. 220 Yes.. Yes. Actual count June9-July6, 1939; no dates for indi- viduals. 1940 155 Yes.. All 60 cm.... Volume— 200 eggs. June 2-Sept. 13, 1940. 1941 114 Yes.. All 60 cm.... Volume— 200 eggs. June 9-Sept. 8, 1941; no sampling July3-Aug. 11. 1943.-.. 182 No... Daily aver- age. No information. June 29-Aug. 18, 1943. * Left and right ovaries estimated separately. 2 Summary published in Gilbert and Rich (1927). Some measure of the reliability of these data is contained in figure 4, which sliows for 1938 the average weight in grams for 1 egg of the right ovary plotted against the weight of 1 egg of the left ovary for 41 Karluk River sockeye of ages Sg, 63, and 64. The samples were taken from salmo'i captured at the mouth of the river so that there is great variation in the stage of maturity of the ova, but the figure shows that the eggs in the two ovaries are maturing at the same rate. Since the data from the two ovaries form two independent estimates from the same fish, their close agreement gives confidence in the consistency of this method of calculating the number of eggs. to .02 ,03 .04 .05 .06 .07 .08 .09 JO ,11 ,12 V/EIGHT IN GRAMS OF ONE EGG- LEFT OVARY Figure 4. — Paired observations of egg weights in right and left ovaries of sockeye salmon, 0. nerka, of ages Bi, 63, and 64, of Karluk River in 1938. It is interesting to note that although the eggs in the left and right ovaries maintain the same rate of egg maturation the total number of eggs in the two differ noticeably. Figure 5 shows that for low total number of eggs the right ovary contains as many eggs as the left or more; however, as the total number of eggs rises the proportion in the left ovary becomes increasingly greater than in the right. Kendall (1921, pp. 195, 197) says. As the ova approach maturity, the left ovary is nearly or quite always the longer, and it extends, tapering, to the posterior end of the abdominal cavity. 600 800 1000 1200 1400 1600 I BOO 2000 2200 2400 2600 NUMBER OF EGGS IN LEFT OVARY Figure 5. — Relation between the egg number in the right and the left ovaries of the sockeye, 0. nerka, of Karluk River in 1939. FECUNDITY OF NORTH AMERICAN SALMONIDAE 457 These backward extensions of the ovaries are formed by the maturing and enlarging ova filling the previously crowded interlamina spaces at the posterior end of the ovary, thus stretching it longitudinally. This increasing disproportion between the left and right ovaries in fish with larger numbers of eggs is logical since in a fish with few eggs the posterior portion of the body cavity would be relatively empty. Fish of the same size with more eggs would have to utilize this space and the left ovary, which is usually longer than the right, would thus be proportionately larger. However, for the chum salmon in Japanese waters, the data of Eguchi et al. (1954) show no significant differ- ences in egg number between the two ovaries. For 243 chum salmon the averages are 1,134 in the left ovary and 1 ,146 in the right ovary. Sock- eye salmon from little Bare Lake in the Red River system of Kodiak Island contain more eggs in the right than in the left ovaries (personal commu- nication from Philip R. Nelson). It is interesting to speculate whether this is a genetic difference or induced by the great environmental difference between Bare and Karluk Lakes. Probably the best explanation of this dispro- portion in the size of the two ovaries is given by Brown and Kamp (1942, p. 196). In discussing the brown trout, Salmo trutta, they say — In the brown trout, the posterior portion of the intestine usually bends strongly to the right, thus crowding the right ovary at its caudal end. The length of the ovary is in- versely proportional to the degree of crow'ding. However, the left ovary is not always the longer. One fish was observed to have a longer right ovary and it was Interest- ing to note that this specimen had an intestine which bent to the left instead )f the right. In one or two fish the ovaries were of approximately equal length, with the jntestine bending neither to the right nor the left. They found in 8 brown trout averaging 36 cm. in standard length that the right ovary was 133 mm. long and weighed 32.4 grams, while the left ovary averaged 169.5 mm. and weighed 42.6 grams. In discussing the effect of age at maturity on number of eggs in Oncorhynchus there are two questions: (1) Is the number of eggs determined by length of residence in fresh water or length of residence in the sea? (2) Does the number of eggs for any given length of fish increase or de- crease with age? These questions cannot be answered by the pink salmon data because they leave fresh water immediately after emerging from the gravel, and because they invariably mature in their second year. The following tabulation has been made for the Karluk River sockeye salmon, showing the average number of eggs in relation to the period of residence of the salmon in fresh water and in the ocean. Summers in ocean Fresh-water age ' Ocean age and year sampled ' In third year In fourth year Summers in fresh water Number of flsh Average number of eggs Summers in fresh water Number of fish Average number of eggs 2-year ocean age: 1938 1936-37 1937-38 1938-39 1939-40 1934-35 1935-36 1936-37 1937-38 24 36 80 45 3,430 3,055 3.421 3,708 1933-35 1934-36 1935-37 1936-38 10 7 25 36 2,972 2,674 J940 - -- 3.549 1941 .. . ,__ 3,668 3-ycar ocean age: 1938 - 1935-37 1936-38 1937-39 1938-40 1933-34 1934-35 1935-36 1936-37 9 58 23 20 2,S3I 2,973 2,926 3,011 1932-34 1933-35 1934-36 1935-37 >2 13 1 11 3,610 1939 2,866 1940 ._ 3.459 1941 ,. 3,160 ' Since smolts enter the sea from early to late spring and reenter the rivers as adults from spring to early fall the ocean age gives number of ocean summers, but 2 years at sea may vary from about 23 to 27 months. The growing seasons are of paramount importance to this discussion. * Fresh-water age' is from the lime the eggs are deposited (from late June to November) until smolts enter the sea (from May to July), so that a fresh-water age of 3 can vary from about 29 to 3ii months in fresh water, but the sununers spent in the lake after hatching are the periods important to this discussion. 3 Not corrected for length of fish. The data for 1940 and 1941 are for 60-cm. fish, so to make the data for 1938 and 1939 comparable to data for the other years it was necessary to obtain the number of eggs for 60-cin. fish from the regressions of eggs on length of fish. These regressions were computed separately for the left and the right ovaries and the counts for each, calculated from these regressions, were then combined. In order to discount environmental effects the 4273!>:5 O — 57- 458 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE averages were compared according to spent in each environment as follows: seasons Ocean age Summers in lake 2-year 3-year Difference Number offish Average number of eggs Number of fish Average number of eggs 1934-35 -. . 24 36 80 10 25 3,430 3,055 3,421 2,927 3,549 58 23 20 13 11 2,973 2.926 3.011 2,866 3,160 457 1935-36 --. --- 129 1936-37 410 1933-34-35 106 1935-36-37 389 3,285 2,987 ' 298±74. 7 Fresh-water age Summers in ocean 3-y ear 4-year Difference Number offish Average number of eggs Number offish Average number of eggs 1936-37 .- 1937-38 24 36 80 45 58 20 3,430 3,055 3.421 3.708 2.973 3,011 10 7 25 36 13 11 2,972 2.674 3.549 3,668 2,866 3,160 458 381 1938-39 - - . , - -128 1939-40. _.. 1936-37- 38 40 107 1938-39-40 -140 Average 3,266 3,148 118±103. 7 I (-3.99, P for 0.05=3.75, 4 d. f. The significant difference of 298 eggs between fish spending 2 summers at sea and those spending 3 summers, but with similar lake histories, is fairly clear evidence that younger ocean-age fish have higher fecundity than older ocean-age sockeye of the same size. The rather consistent difference between 2- and 3-ocean-age fish would also indicate that the ocean environment is relatively stable as the two groups were not at sea during identical yeai's. If we now turn the analysis around and note the egg counts for 3- and 4-fresh-water-age fish with identical ocean histories that spend 2 and 3 summers in fresh water, the difference in fecundity is neither consistent nor significant. This could be interpreted to mean that fecundity does not differ between fish of 3 and 4 fresh-water ages, but since there are obvious differences between year classes, owing probably to lake conditions, such a conclusion is not fully warranted by these data. What is required are data over a sufficiently long period to discount these fresh-water en- vironmental effects. RELATION OF EGG SIZE TO EGG NUMBER Surprisingly few records have been published on actual size of ova of Salmonidae, investigators being content to speak of size in a purely compara- tive sense. For instance, Belding et al. (1932, p. 214) say — In general the size of the egg depends upon the size of the parent salmon, the larger specimens producing the larger egg. Also, the size of the egg varies with the sal- mon of the different rivers. The material used in this study permits its division into two classes, large and small eggs. There is no relationship between the size of the egg and the length of the incubation period. Gilbert (1915, p. 57) also used only a compara- tive measure of size. He says in speaking of British Columbia sockeye, A similar difference, but even more pronounced, is found among certain lots of eggs collected by Mr. Stone in Smith Inlet, those from Quey Creek being markedly smaller than those from the Gelulch and Delelah Rivers. It re- quired 74 Quey Creek eggs to fill a tube which would hold only 38 from the Gelulch and the Delelah. Perhaps the chief reason for this lack of data on size of ova is that the salmon taken by the commercial fishery are in various stages of egg maturation. Thus, at Karluk River many of the sockeye taken in the fishery may not spawn for at least another month. This is reflected in the weight of sockeye eggs at Karluk ranging from .03 to .095 grams (fig. 4). The same late maturation is found in the Atlantic salmon. Speaking of S. salar in Norway, Dahl and S0mme (1944, p. 39) say- In the grilse, which have spent more than a year in the sea, the GW/TW [ratio of gonad weight to total weight] is still practically in the same undeveloped stage in the early part of the season. A gradual development in the relative size of the sexual organs asserts itself as the fish- ing season advances, but the main growth towards matur- ity takes place after the fish have entered the rivers. They agree with Belding et al. that the indi- vidual egg size is partially dependent on fish size, saying (op. cit., p. 22), "It is a well known fact that in large salmon the ovaria as well as the single ova arc larger than in salmon of small size." Egg size is regarded bj' Svardson (1949, p. 120) as resulting from natural selection. He states — Summing up it can be said that the evidence now at hand shows that competition among fry gives the larger FECUNDITY OF XORTH AMERICAN SALMONIDAE 459 fry better survival chance.s. A selection pressure ii' favour of large eggs therefore certainly exists and this selection must work until the eggs are so few that no note- worthy competition for food exists among the fry. While we must agree that larger fry generally have better survival rates, the reason given by Svardson — intraspecifie competition — may some- times liave little bearing on the matter; un- doubtedly, there are other important factors. For instance, Robertson (1922) has pointed out that the race of small-sized sockeye salmon that spawns in Harrison Rapids, a tributary of the Fraser River, produces larger eggs than the other races of sockeye in the Fraser. This may be related to the fact that this is one of the few races of sockej^e in which the young go to sea as fry, since large, vigorous fry would be required to survive in sufficient numbers to maintain the population. It should be noted, moreover, that among the Pacific salmons (table 2) the smallest eggs are found in tlie sockeye which normally spend the longest time in fresh water. Size can be attained only by the sacrifice of number. In each ecolog- ical situation there is some point at which, on the average, the forces favoring size are exactly balanced by those favoring number. This point must vary between river systems, tending to produce genetic variation between populations for egg size and number. Table 2. — Size and weight oj eggs and fry of certain North American Salmonidae [Asterisk (•) indicates diameter calculated from volumetric measure by Von Bayer conversion table] Species and area Eggs Sac fry Fry after yoik is absorbed Authority Diam- eter Weight Length Weight Length Weight Oncorhynchits: tshauytscha: Wusliinpton Mm. Gm. Mm. Gm. 0.520 Mm. Gm. 0.509 Chapman (1938). Rich (1920). Columbia R 35-40 •6.3 •7.8 •7.9 .\rt'a not recorded. Bower (1910). Stone (1897). Rich (1920) Do - 35-40 30-JO ktta: Area not recorded _ 0. 24-. 62 Kobayashi (1953). Watanabe (1956). Chapman (1938). Beal (1955). Shapovalov and Taft (1954). Skud (1955). Pritchard (1944). Bower (1910). Chapman (1938). Dahl (1917). Vladykov (1954). Belding and Hyde (1932). •7.4 0.232 kisulch: Oreen R., Wash .. 0.284 •8.4 •7.2 Scott Creek, Calif.. Qorliuscha: 971-1800 32-38 per pint caO. 3 McClinton Cr., B. C - nerka: •6.3 Baker R., Wash 0.192 Salmo: satar: Rivers in Norway. . 5.3-7.0 6.8 0. 1-. 2 0.200 Oaspe. P. Q 17.4 17.7 18.1 18.8 16.0 0.180 0.141 0.133 0.146 0.110 Morell R., P. E. I 27.5 26.7 27.9 0.057 0.124 0.144 Xt'w Brunswick _. - Do Pollilt R., X. B 5.4 •6.1 •6.6 5.1 •5.5 0.090 Vladykov (19.M). Brice (1898). .\rea not recorded .. gairdneri: Do Scott Cr., Calif ._ _ Shapovalov and Taft (1954). Curtis (1935). Brice (1898). Brown and Kamp (1942). Irving (1955). Lord (19.30). Vladykov (1954). Do g. aiina-fjonita: Cottonwood L., Calif 15.0 trulla: .\rca not recorded •4.2 4.94 4. 3-5. 1 •4.5 5.8 5.7 4.9-5.4 •5.2 •6.5-5.6 •5.4 •5.8 5.0 4.7 4.4 4.0 •4.6 •4.2 Madison R.. Mont c. /eiriji: Yellowstone R Crj^«/jromfr; namaycush: Baldwins Mills, P. Q 0.125 0.125 16.2 19.1 0.080 0.075 Twin Mountain, N. H Eschmeyer (1955). Royce (1951). Do Adirondack Lakes, N. Y Xorthville. .Mich Bower (1910). Brice (1898). Vladykov (1933). Vladykov (1954). Do. Area not recorded. ... Satrelinus: a(;)in!/s; Area not recorded. aiireolus: .New Uampton, N. H n.050 0.040 0.040 17.1 15.0 13.9 0.055 0.033 0.040 fontinalis: L. Jacqucs-Cartier, P. Q .New Hampton, N. H Do Bower (1910). Brice (1898). 460 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE The data given in table 2 are from several sources. Undoubtedly exhaustive search of the literature would reveal more data on the subject; however, these suffice to give a general picture. Because data on ova size are missing or very scanty for several species we have included length and weight of sac fry and free-swimming fry. Egg size, in general, is correlated with the average size of the species. Thus Oncorhynchus tshawytscha, the largest species, has the largest This is, however, only a generalization. ova. It may be noted that the fry of the small pink salmon, 0. gorbuscha, are larger than those of the Atlantic salmon, Salmo salar. That is, the tend- ency toward large eggs and fry appears to be a characteristic of the genus Oncorhynchus. Be- cause of the scarcity of data from actual measure- ment of diameters of mature eggs, it is felt that the scattered material brought together in table 2 cannot be wholly relied on to give a true picture of egg size. However, corroborative evidence can be obtained by an indirect method. In figure 6 (data from appendix table 4), the average number of eggs is plotted against the 2 3 4 5 6 AVERAGE WEIGHT (KGJ Figure 6. — Relation, by species, of average egg number to average weight. (Data from appendix table 4.) average weight of the females for each species. (See also appendix tables 1, 2, and 3 for detailed information on egg numbers, by species and locality.) Obviously the data fall into two general groups: fluvial anadromous Oncorhynchus, which show a low number of eggs for their weight, and lacustrine anadromous sockeye and members of the other 3 genera, which show a large number of eggs for their weight. That this difference in egg number for comparable weights is not due to a difference in the shape of the fish is indicated by the very close correspondence between the length-weight relation for all of the genera (fig. 7, data from appendix table 5). 3.0 o o 1.4 1.6 1.8 LOG OF FORK LENGTH (CM.) 2.0 Figure 7. — Relation, by species, of the logarithm of mean weight to the logarithm of mean fork length. (Data from appendix table 5.) It must therefore be concluded that the lower egg number in the fluvial anadromous species of Oncorhynchus can be due only to one of two causes: either the eggs form a smaller percentage of the total weight of the fish or the eggs are considerably larger. Despite the paucity of available information, the true cause of the lower egg number in these species can be confi- FECUNDITY OF NORTH AMERICAN SALMONIDAE 461 dently ascribed to egg size if we consider the data on weight of fry in conjunction with that of egg diameter (see appendix table 6). Thus, the sac fry of 0. tshaioytucha weiglied 2.9 times the upper limit given for Salmo salar. There is general agreement that, within the genus OncorhynchuK, the largest eggs are found in lahawytscha and the smallest in nerka. O'Malley (1920) gives the following number of eggs of each species required to fill a hatchery basket: Species: Thousands of eggs 0. tshawytscha 20-30 0. kisutch 30-35 C. keta 33-38 O. gorbuscha 40-50 0. nerka 50-60 Bean (1893, p. 30) says of the pink salmon, 0. o X CO CO CD UJ U_ O tr LU m O - 0. NERKA A - 0. KISUTCH Values read from published regression curve. 3 3 to 8 specimens per year. * Partial duplication of fish in previous total or totals. 5 Marked Cultus Lake O. nerka caught outside of Fraser River. " Standard length of 103 females converted to fork length by factor 1.1. "> Standard length of 5 females converted to fork length by factor 1.1. Appendix Table 2. — Number of eggs at maturity in North American Salmonidae of the genus Salmo Average number of eggs Sampling Number of eggs per kilo- gram Species and area Number in sample Year Age Average fork length Average weight offish Authority salar: Miramichi R., Canada Gulf of St. Lawrence 7,678 9,409 13,883 12,313 2,400 3,900 5,600 7,600 326 765 1,102 1,577 1,914 2,930 1,113 1,100 1,750 750 900 •3,660 • 2, .594 «2,508 1,238 1,383 1,164 1,279 1,061 1,208 163 340 15 16 537 2 sea 2 sea 3 sea (') Cm. ' 71.4 '74.8 ■86.7 ■83.6 40.0 50.0 60.0 70.0 20.0 30.0 37.5 •31.9 MO. 8 •51.8 Kg. 4.26 5.07 8.68 7.57 1,802 1,856 1,594 1,627 Belding (1940). Do Do -- Do Do Do gairdntTi: Scott Cr., Calif.s 1932-33 Shapovalov and Taft (1954). Do Do - Do Do - -- Do. g. aqua-bonila: Cottonwood Lakes, Calif 450 Curtis (1935). Do. Do Do Do clarki: 10 10 10 104 55 1953 1953 1953 0.573 1.180 2.394 2,752 1,622 1,224 Irving (1955). Do Do . Do Do c. lewisi: Yellowstone R -- Lord (1930) c. henshawi: Blue L., Callf.3.-- 1941 1941 1941 1941 1940 1941 1942 1936 1936 1936 1936 1952 (») (') (2) 35.0 40.0 35.0 40.0 0.33 0.50 0.30 0.44 3,333 3,500 2,600 2,045 Calhoun (1944). Do Do- Do.s 38 Do Do Do Heenan L., Calif 214 310 320 1 18 14 4 14 78 Smith, 0 (1947). Do - Do Do - - Do iTutta: Madison R., Mont 2 3 4 5 <37.1 <37.6 MO. 7 Ml. 4 0.51 0.55 0.66 0.68 2,427 2,515 1,764 1,881 Brown and Kamp (1942). Do Do Do - Do Do- Convict L., Calif Nielson (1953) Michigan streams 34.0 Cooper (1953). ' These values from weight-length regression of fig. fi. ' Previously spawned. 3 Readings from published regression curves. * Standard length converted to fork length by factor I.l. 5 First spawning. • Eggs stripped by hatchery. FECUNDITY OF NORTH AMERICAN SALMONIDAE 467 Appendix Table 3. — Number of eggs al maturity in North American Salmonidae of the genera Cristivomer and Salvelinus Average number of eggs Sampling Number of eggs per kilo- gram Species and area Number in sample Year Age Average fork length Average weight offish Authority Cristivomer: namayeush: 7,943 3,383 4,253 4,995 8,667 8,881 11.603 13,836 11, 789 25 9 15 13 17 8 6 2 2 !72 12 23 '21 6 211 51 J 42 28 53 20 239 2 5 5 2 77 29 1927 1951-53 1951-53 1951-53 1951-53 1951-53 1951-53 1951-53 1951-53 Cm. 72.7 160.7 '65.4 170.4 1 75.3 178.5 '83.2 ■87.9 193.9 Kg. .1.00 2 81 3.36 4.26 5.26 6.31 7.48 8.75 1,589 1,204 1,266 1, 173 1,648 1,408 1,551 1,581 Dymond (1928). Eschmeyer (1955). 1)0 Do. ))o Do. Do Do. Do Do. Do Do. Do - --.. Do. Do Do. Do 1,424 1,011 Do. H siscowet' Lake Superior 4,387 3,589 3,645 2,726 909 1,313 1,443 4,927 1,183 1,414 254 148 191 275 376 399 «410 «640 •950 •1,400 1950-54 1950-51 1950-51 1951 1945^8 1949 1949 1950 159.2 56.0 4.34 Do. f^atrelinua: alpinus: 13-22 13-22 7-9 Grainger (1953). Do - -.-. 2.00 1,822 Do. Ungava Bay 41.2 30.6 31.4 Do. 0.26 Maar (1949). Do MaSr (1950). Do 0.26 2.31 Do. malma- Clark's Fork R., Mont 61.9 29.9 31.4 1 14.4 12.70 14.44 16.82 19.65 22.3 20.3 25.4 27.9 35.6 2,133 Brunson (1952). fontinalis: (<) (') Hayford and Embody (1930). Do. Do Cooper (1953). 1952 1952 1952 1952 0.0215 0. 0316 0.0506 0.0835 6,907 6,032 .5, 435 4,509 .\llen (1956). Do Do. Do Do. Do Do. Smith (1947). Do Do. Do Do. Do Do. 1 Total length converted to fork length by factor 0.92. ' Summary. 3 Partial duplication of fish in previous total or totals. * First spawning. * Previously spawned. * Values read from published regression curve. Appendix Table 4. — Summary of number of eggs by size of fish for ilorih American Salmonidae All specimens Spe cimens with length data Specimens with weight and length data Species Average number of eggs Number offish Average fork length Mean number of eggs Average number of e«gs per centi- meter Number offish Average weight offish .\verage number of eggs Average number of eggs per kilo- gram Number of fish (hlcorhynchus: 4,772 2,546 2,801 567 1,765 3,597 1478 9,092 15,541 2,140 1,113 2,806 1,233 7,186 3,410 988 4,927 461 314 233 137 37 825 1,262 651 534 537 30 104 844 115 97 29 262 28 403 Cm. I 83.46 I 79.26 73.90 r 66.01 \ 66.68 34.00 / 52. 82 \ 52. 61 f 59. 24 \ 59. U 26.95 74.36 •59.00 45.65 4,772 3,759 2,890 2,812 3,106 567 1,778 1,754 3,600 3,896 445 9,092 (5,541) 2,140 57.2 47.4 39.1 42.6 46.6 17.7 33.7 33.3 60.8 65.9 16.5 122.3 93.9 46.9 314 207 72 134 69 37 702 364 1,258 403 25 .534 562 30 Kg. ktla... 6.465 4.685, 3,759 2,890 581 617 207 72 kisutcli< 3.657 0.499 3,106 567 849 1,136 69 37 gorbuseha rteTka. . 1.746 2.121 2.178 1,754 3.946 3,896 1,005 1,860 1,789 364 514 403 Satmo: 4.999 9,092 1,819 534 darki' i.382 2,140 1,548 30 c. hensliawi ... trulla f 35.66 1 39.17 72. U3 ( 52.94 \ 56.00 30.76 61.90 / 18.84 \ 15. 79 1.233 1,258 7,089 3,410 3,589 988 4,927 461 241 34.6 32.8 98.4 64.4 64.1 32.1 79.6 24.5 15.3 115 37 95 29 23 262 28 403 14 Cristivomer: namai/cush 0.605 4.825 1,285 7,089 2,124 1,468 37 95 Satrelin us: alpinus*. . . , atpinut *. , _ 2.000 0.260 2.310 3,645 998 4,927 1,822 3.838 2,133 21 253 malma i _ 28 fontinatis i 0.044 241 5,477 14 1 In fresh water. ' Anadromous. ' Calculated from egg volumes (Shapovalov and Taft. 19M, tabic .\-17). * Values from regression curve. * In Swedish lakes. 468 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Appendix Table 5. — Summary of length-weight data on North Ameriean Salmonidae [A=anadromous; F = fresh water] Species Oncorhynchus: tshawytscha keta -- kisutch kisutch gorbuscha nerka Salmo: salar __- darki trutttt Cri3tivomer: namaycush Salvelinua: alpinui... - alpinus malma fontinalis Habitat A A A F A A A F F F A F F F Average weight Kg. 6. tt.l 4.685 3.657 0.499 1.746 2.178 4.999 1.382 0.605 4.825 2.000 0.260 2.310 0.044 Average fork length Cm. 79.26 73.90 66.68 34.00 52.61 59.11 74.36 45. 65 39.17 72.03 56.00 30.76 61.90 15.79 Number in sample 207 72 69 37 364 403 534 30 37 95 21 253 2S 14 Logarithm Average weight+2.0 Y 2. 8106 2. 6707 2.5631 1.6981 2.2420 2. 3381 2. 6989 2. 1405 1.7818 2.6835 2.3010 1.4150 2.3636 0. 6435 Average length X 1.8991 1.8686 1.8240 1.5315 1.7211 1.7717 1.8713 1.6594 1.5930 1.8575 1.7482 1.4880 1.7917 1.1984 Appendix Table 6. — Summary of various measures of egg size in North American Salmonidae Group and species Eggs per cm. offish Eggs per kg. offish Egg diameter Weight of sac fry Weight of fry Measured Calculated with yolk absorbed Oncorkynckus: Fluvial anadroraous: 47 39 47 33 66 18 122 94 581 617 849 1,005 1,789 1,136 1.819 Mm. Mm. 6. 3-7. 9 7.4 7.2 Gm. 0.520 Om. 0 509 0. 24-0. 62 0.284 gorbuscka Ca. 0.30 6.3 8.4 6.1 5.5 6.6 0.192 5. 4-6. 8 5.1 Salmo: Fluvial anadromous: 0. 110-0. 180 0 057-0. 144 Fresh water: salaT sebago . . . 47 1,548 4. 3-5. 1 clarki lewisi 4.5 4.2 5. 2-5. 8 trutta 33 98 64 32 2,124 1,469 1,822 3,838 4.9 4. 9-5. 8 5.0 CTistivomeT: Fresh water; namaycusk ~ ._ . 0.075-0.080 Salvelinus: Anadromous' a/pJTiws .. .- Fresh water: alpinus (Sweden) aureolus - - . . 4.7 0.055 80 15 2,133 5,477 fontinalis ' 4.0-4.4 4. 2-4. 6 0. 033-0. 040 ' The small sample available contained only small fish so that these figures are not believed to be representative for eggs per centimeter or per kilogram. o EFFECTS OF UNIALGAL AND BACTERIA-FREE CULTURES OF GYMNODINIUM BREVIS ON FISH, AND NOTES ON RELATED STUDIES WITH BACTERIA By Sammy M. Ray and William B. Wilson, Fishery Research Biologists Association of the dinoflagellate Gymnodinium breiis Davis with the mass mortality of marine animals that occurs sporadically in the Gulf of Mexico is well established (Davis 1948; Galtsoff, 1948 and 1949; Gunter et al., 1948; Wilson and Ray, 1956; Woodcock 1948; and others). In- direct evidence presented in these papers strongly supports the contention that 0. breiis is the cause of fish kills, commonly referred to as red tides,' when its concentration reaches the order of himdreds of thousands to millions of organisms per liter — concentrations as high as 50 to 60 million organisms per liter have been reported. This evidence includes (1) the presence of dead or dying fish in water containing such concentrations of (r. brems, (2) laboratory demonstration that water containing great numbers of G. breiis is toxic to fish, and (3) demonstration that sub- stances toxic to fish may be extracted from water infested with 6. breiis. Further evidence of a more direct nature is provided by the demon- stration that unialgal cultures of 6. brems are toxic to fish (Wilson and Collier, 1955). Development of stock unialgal cultures of G. breiis opened the way for the elucidation of this organism's role in the mass mortality of marine animals by making available an abundant supply of material for controlled experiments. Previous to this development investigators were handi- capped, since the suspected causative agent was ' Tfio term "red tide" is Rpnprally applied to discolored sea water regardless of causes or consequences; tliat is, the causes of the discolorattons may vary from "blooms" of many difTerent microorganisms to nonliving agents such as iron compounds; and the mortality of animals, especially fish, may or may not be associated with such discoloralions. To avoid confusion, we believe it best to refrain from using this popular though nonspecific term In scientific publications. If a popular name is used, we propose that the name "brevUi red tide" be applied to the mass mortality of marine organisms associated with Oymnodinium brevia. Note— Approved for publication, February 8, 1957. Fishery Bulletin 123. unavailable for study except during outbreaks. Even then, their material was limited to raw samples from the infested waters which contained numerous other organisms. In addition, raw samples were probably held under conditions un- suitable for the survival of G. breiis. The next approach to this problem was to obtain bacteria-free or pure cultures of G. breiis. This isolation is necessary to determine whether a cause-and-efFect relation exists between G. brevis and the catastrophic fish kills. Furthermore, studies of such problems as the nutritional require- ments of G. brevis, nature of the toxic substance, role of associated organisms, and effects of physical and chemical factors may be facilitated with the use of bacteria-free cultures since the uncertainty regarding the effects of associated bacteria would be eliminated. These laboratory studies in coordination with field studies, will provide a better understanding of why the mass mortalities occur. Such knowl- edge will be helpful in predicting when and where outbreaks may be expected and in determining the feasibility of control measures. This report presents the results of our studies on the effects of unialgal and bacteria-free cultures of G. breiis on fish as well as the effects of some bacteria isolated from unialgal cultures of this organism and from waters off the west coast of southern Florida.^ Based upon the results of studies with bacteria-free cultures, we conclude that G. brevis produces the toxic substance (s) responsible for the mass mortality of marine animals associated with blooms of this organism in the Gulf of Mexico. ' We are indebted to K. T. Marvin, Alice Kitchel, and Jean Gates for assistance in performing (he experiments reported here and to E. L. Arnold and R. S. Wheeler for Identifying the test fish. \ 469 470 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE PROCEDURES FOR TESTING STERILITY AND ENUMERATING ORGANISMS Bacteria-free cultures were grown in the same medium prescribed by Wilson and Collier (1955) and were carried through several subcultures. After 10 months they showed no apparent dimi- nution in vigor. Several media were used to establish sterihty. All G. brevis cultures origi- nated from a culture obtained from a sample collected in a bloom that occurred near the coast of Florida in September 1953. Details of the procedures for culturing G. breds, and the methods used to obtain bacteria-free cultures will be presented in another paper. STERILITY-TESTING MEDIA Cultiu-es of G. breds used for transferring were tested for sterility in media prepared according to Spencer (1952): (1) peptone sea water (0.5%, bacto-peptone, 0.01%, FePOi dissolved in 75% aged sea water) and (2) peptone sea-water agar (peptone sea water plus 1 .5% bacto-agar) . These media as well as all other sterility-testing media subsequently described were dispensed in screw- cap tubes and autoclaved at 121° C. for 15 minutes. We frequently used four other media similar to those employed by Droop (1954) for routine sterility testing. These media included (1) dis- tilled-water liquid, (2) distilled-water agar, (3) sea- water liquid, and (4) sea-water agar. Our media contained the following substances: 0.5% dex- trose, 0.1% Difco neopeptone, 0.4% bacto-beef extract, 0.5% bacto-yeast extract, 0.015% sodium acetate (NaC2H302.3H20), and soil extract (2.0 ml./lOO ml.). These substances (dextrose was often excluded) with and without bacto-agar (1.5%) were dissolved in both distilled water and 75% aged sea water to give the four combinations. Droop (1954) listed the substances used, but not the quantities. A personal communication (1956) , however, revealed that his formula contained the organic substances in concentrations which were roughly 10 to 15 times less than the quantities we used. Furthermore, he included bacto-tryptone, which was not listed in his paper, whereas we used Difco neopeptone. Subsequent to the comple- tion of the present studies, the absence of bacteria from several G. breds cultures was confirmed with media of Droop's formulation and also with these media diluted to 10 percent. Other media used to supplement the routine tests included (1) the sterility-test medium used by Sweeney (1954) containing 0.05% bacto- peptone, 0.0136% sodium acetate (NaCsHjOz. 3H2O), 0.0202% KNO3, 0.00356% K2HPO4, 0.00016% FeCl3.6H20, and 0.000012% MnC^. 4H2O dissolved in 75% aged sea water with and without bacto-agar (1.5%); (2) Spencer's peptone sea-water media supplemented with 0.1% bacto- yeast extract as employed in medium 2116E (Morita and ZoBell, 1955); (3) semisolid medium composed of 0.075% trypticase (Baltimore Bio- logical Laboratory), 0.075%, bacto-peptone, 0.075% bacto-yeast extract, 0.01% sodium acetate (NaC2H302.3H20), and 0.2%o Difco special (Noble) agar dissolved in aged sea water; (4) 1% bacto- peptone in aged sea water with and without bacto-agar (1.5%), the medium used by Bein (1954) to isolate and cultivate certain chromo- genic bacteria found in Florida waters; and (5) Spencer's (1952) casein sea-water agar com- posed of 0.05% bacto-peptone, 0.05% bacto- isoelectric casein, 0.05% soluble starch, 0.1% (v/v) glycerol, 0.02% K2HPO4, and 1.5% bacto- agar dissolved in 75% sea water. Sterility tests for anaerobic bacteria were con- ducted occasionally with three difi'erent media: (1) Bacto-fluid thioglycollate medium rehydrated with both distilled water and 75% aged sea water; (2) the general anaerobic medium (slightly modified) used for marine bacteria by Morita and ZoBell (1955) containing 0.5%o bacto-peptone, 0.1%, bacto-yeast extract, 0.01% FePO^, 0.1% sodium formaldehydesulphoxylate, and 0.0001% resazurin dissolved in 75% aged sea water with and without bacto-agar (1.5%,); and (3) an anaer- obic medium prepared by adding 0.01% sodium thioglycollate to the semisolid medium described in the previous paragraph. The melted, general, anaerobic agar medium was cooled to 40°-42° C. before addition of the test culture which was mixed by swirling the tube before the agar solidi- fied. After adding the test culture, sterile melted vaspar (50% vaseline and 50% paraffin) was poured into each tube of anaerobic medium, except the fluid thioglycollate medium, to exclude oxvgen. INOCULATION AND INCUBATION All sterility tests, unless otherwise indicated, were made with 1.0 ml. of test cvilture in 10.0 ml. tJNIALGAX, AXD BACTERIA-FREE CUinTRES OF G. BREVIS 471 of medium. The agar media were inoculated in tlie following ways: (1) Pour-plate — mixing test culture in a sterile Petri disli with melted medium cooled to 40°^2° C, (2) streak-plate— streaking 0.1 ml. of test culture on a freslily prepared plate, (3) stah culture — placing 0.1 ml. of test culture into medium in screw-cap tubes (20 mm. x 125 mm.), and then stabbing an inoculating needle to the bottom, and (4) slant cultures — placing the test culture on freshly slanted medium in screw-cap tubes. Slant cultures were generally prepared for most routine tests. The agar plates were sealed with masking tape to prevent desiccation and mold contamination during incubation. Semisolid me- dium was inoculated by stabbing to the bottom with a micropipette and then gradually releasing the inoculum as the pipette was slowly withdrawn. We incubated tlie sterility-test cultures in the dark at 28°-30° C. for a minimum of 6 weeks before discarding them as sterile. This tempera- ture level was selected since some of the bacteria isolated from the unialgal cultures of 6. brevis appear to grow more slowly at 24°-25° C On one occasion the sterility of several cultures was tested in duplicate in various liquid and agar media; the four methods tor inoculating agar cul- tures were used. One set was incubated with illumination (175-300 ft.-c.) and temperatui'e (24°-25° C.) the same as used tor G. brevis cul- tures; the other set was incubated in the dark at 28°-30° C. After 6 weeks none of the cultures showed either visible colonies or cloudiness of any sort except an occasional mold or bacterial colony on the surface of a few streak- and pour-plates. We attribute the occasional appearance of mold or bacterial colonies in our test cultures, especially on the surface at the periphery of streak- and pour-plates, to contamination while the plates were exposed by necessary manipulations. The position of the colonies as well as the appearance of similar colonies on some control plates (unin- oculated agar plates), which were treated in tlie same manner as the test cultures, supports this conclusion. We rarely encountered accidental contamination of sterility-test cultures contained in screw-cap tubes. MISCELLANEOUS CHECKS FOR STERILITY We consider that the medium used for cultiuing G. brevis is unlikely to be suitable for the growth of photosjmthetic bacteria. Nevertheless, a few cultures were checked for such organisms. The checks were made with a medium developed by Dr. T. J. Starr of this laboratory for the isolation of marine nonsulfur purple bacteria. Tiiis medium is composed of 0.2% sodium acetate (XaC2H302. 3H.,0), 0.05% \a2SO3.7HjO,0.01% MgS04.7H20, 0.05% K2HPO4, 0.1%(NH4)2SO4, 0.0001% FeCl,. 6H2O, 2.5% NaCl, 0.01% bacto-yeast extract, 0.01% sodium thioglycollate, and 1.5% Difco special (Noble) agar dissolved in double-distilled water. Just before inoculation the medium was melted, and sterile NaHCOa solution was added aseptically to each tube to give a 0.1% concentra- tion and a final pH of 8.0. This medium was inoculated and treated in the same manner as previously described for the general anaerobic medium. These sterility-test cultures, which were incubated under the same light and temperature conditions used for G. brevis, showed no evidence of growth after 6 weeks. Phase-contrast microscopic examination (X 970) of wet preparations of a few G. brevis cultures that were determined to be pure by cultural methods did not reveal any contaminating organisms. These examinations, conducted several months after the initial establishment of bacteria-free cul- tures, were performed to check for possible con- taminants which might have maintained themselves in G. brevis culture medium after repeated sub- culturing, and yet had not grown in any of the sterility-test media employed. BACTERIA ENUMERATION Aerobic bacterial counts presented in the experi- ments to follow were estimated by plating serial dilutions of the test sample. The dilution water blanks (75% aged sea water) dispensed in 9-ml. amounts in screw-cap tubes were autoclaved at 121° C. for 15 minutes. Serial dilutions of 1:10; 1:100; 1:1,000; 1:10,000; and 1:100,000 were prepared of each sample to be counted. One ml. of each dilution and 10.0 ml. of melted Spencer agar, cooled to 40°-42° C, were mixed in a sterile Petri dish by gentle swirling before the agar hardened. Because of tiu> possibility of low counts a plate was also prepared with 1.0 ml. of undiluted sample. After the agar hardened, the plates were sealed with masking tape and incubated in an inverted position for 4 to 7 days at 28°-30° C. Colonies were enumerated with the aid of a Quebec colonv counter. Plates with either more than 472 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 300 or less than 30 colonies were not used in quan- titative estimates except in a very few instances. The exceptions were in cases where either the most dilute plates contained more than 300 colonies or the undiluted plate contained less than 30 colonies. The bacterial counts most likely represent minimal concentrations, because nutritional and environmental requirements of an entire bacterial population cannot be satisfied with any one me- dium or with a single set of incubation conditions. We made no attempt to enumerate anaerobic bacteria. All colonies except those with the typical appearance of molds were counted, there- fore, any microorganisms producing bacterialike colonies were included in the counts. Since we could not prepare pour-plates of all water samples immediately after collecting them, another possible source of error in the counts should be considered. Both quantitative and qualitative changes in the bacterial population may have occurred before some of the samples were plated, particularly those plated several hours or even days after collection. Immediately after collection all water samples were refrigerated (4° C.) until shortly before preparation of the plates. In most cases the storage period did not exceed 6 hours; however, this period varied con- siderably in some experiments, especially those in which several samples were counted. Conse- quently, we have recorded the extremes of the storage period for each experiment. DINOFLAGELLATE ENUMERATION The concentration of G. hrevis and other dino- flagellates was determined in two steps: (1) a preliminary counting of 1.0-ml., 0.1-ml., and 0.01- ml. aliquots from a sample mixed by gently swirling the tube (vigorous shaking frequently causes many of the organisms to cytolyze) to determine sample size best for counting, and (2) counting 3 to 9 aliquots of the quantity selected in the first step. The latter counts were averaged to obtain the G. brevis concentrations. The counts probably repre- sent minimal levels because the organism tends to disintegrate when manipulated. Because of this tendency, only one aliquot was withdrawn at a time and it was counted immediately. A wide- field stereoscopic microscope with a magnification of X 54 was used in making the counts. EXPERIMENTS WITH UNIALGAL CULTURES OF GYMNODINIUM BREVIS AND OTHER DINOFLAGELLATES Seven experiments testing the toxicity to fish of unialgal cultures of G. brevis and some other dino- flagellates were performed. All of these studies, even those which were preliminary, such as experi- ments 1 through 3, are presented because the de- tails and results vary considerably in some cases. In some experiments only one test fish was used per container because either the available fish were too few or the containers were too small to accommodate more. Moreover, duplicate con- tainers were iiot always used because of limita- tions imposed by insufficient supply of either test fish or test materials. We have taken special care to record all experimental details, some of which may be of no significance, since they could prove of value to others in reviewing our work. EXPERIMENT 1.— A Simple Test of the Toxicity of Gymnodinium brevis Cultures This experiment was conducted to determine whether unialgal G. brevis cultures would kill fish. We used a 3K-week-old culture, replenished with fresh medium three times weekly, that contained 1.8 million organisms per liter. Sea water from a lagoon at the east end of Galveston Island, Texas — the locality where the test fish were collected — served as control material. The test materials were not aerated. One rough silversides {Membras vagrans), 3J^ inches long, and one sailfin molly (Mollienisia latipinna), 2K to 3 inches long, were placed in each of two 1 -liter beakers — one con- taining sea water, the other G. brevis culture. The beakers were covered with polyethylene sheeting. Membras vagrans survived only 4 minutes in the G. brevis culture whereas this species survived 43 minutes in the sea water. M. latipinna died after 85 minutes' exposure to the G. brevis culture and the fish in the sea water was alive when the experiment was discontinued 4 days later. Al- though the lethality of G. brevis cultures to fish is evident from these results, they do not neces- sarily prove that a toxic substance is involved. UNIALGAL AND BACTERIA-FREE CULTURES OF G. BREVIS 473 EXPERIMENTS 2 and 3. — Comparison of Effects of Gymnodinium brevis and Gymnodinium splendens Cultures fish were alive at the close of the stiulv 19 davs The mere presence of mimeioiis dinoflagcllates may have been responsible for the toxicity of the unialgal culture used in experiment 1. To test this possibility, fish were subjected to unialgal G. breris and 6^. splendfus cultures in experiments 2 and 3. Experiment 2 was conducted under the same conditions as experiment 1. A 4-week-old unialgal culture of G. brevis and a 10-week-old unialgal culture of G. splendens that contained 2.1 and 2.8 mdlion organisms per liter, respectively, were tested for toxicity to Mollienina latipinna (2}^ to 3 in. long). Both cultures were replenished with fresh medium three times weekly during the incubation period. Sea water from which the test fish were taken was used as control. One fish was placed in each of three test materials. The fish in the G. brevis culture died after 47 minutes; the later in the G. splendens culture and in sea water. Experiment 3 duplicated experiment 2 in most respects, except that the cultures were a week older. At this time there were 2.0 million G. brevis and 2.6 million G. splendens per liter in the cul- tures. M. latipinna (2^ to 3 in. long) lived only 68 minutes in the G. brevis culture, but they were alive in the G. splendens culture and in sea water 3 days later when the experiment was discontinued. The excellent survival of the fish in G. splendens cultures, in contrast with the lethality of G. brevis cultures, indicates that the latter cultures con- tained a toxic substance(s). Since the cultures of G. brevis were not pure, the toxic substance could have been produced by G. brevis, associated bac- teria, or both. EXPERIMENT 4. — Effects of Unialgal Gymnodinium brevis Cultures and Associated Bacteria A series of experiments (4, 5, 6, and 7) was designed mainlj' to determine whether G. brevis or its associated bacterial flora is responsible for the toxic effects of unialgal cultures to fish. If the bacteria prove nontoxic under the same cultural conditions, one could reasonably assume that G. brevis produces the toxic substance. Much of the value with regard to the original purpose for con- ducting these experiments has been lost subse- quent to the development of mass bacteria-free cultures of G. brevis. Bacteria-free cultures made it possible to demonstrate experimentally that G. brevis produces a fish-killing substance. The de- tails are presented later in this paper. To obtain some of the test materials used in these experiments (4, 5, 6, and 7), 20 liters of cul- ture medium were prepared, 5 liters of which were placed in each of two P3Tex bottles {2}^ gal.) ; another Pyrex bottle (5 gal.) received the remain- ing 10 liters. Each bottle of medium was heated to 75° C. (5 to 6 hours' heating required) on three successive days to reduce the bacterial load. One of the 2K-gallon bottles (No. 1) was inoculated with 10.0 ml. of a 6-week-old unialgal G. breds culture with a bacterial count of 8.1 million per ml. The otlier 2K-galloT) bottle (No. 2) was seeded with 10.0 nd. of G. brevis-hee inoculum in an attempt to culture the associated bacteria. This inoculum, having a bacterial count of 10.3 million per ml., was obtained by heating between 37°-39° C. for 30 minutes a portion of the same culture used to inoculate bottle 1. The 5-gallon bottle (No. 3) containing uninoculated medium was arranged so that bottles 1 and 2 could be replenished from this reservoir when culture materials were removed for toxicity tests and chemical analyses. Samples were taken from the three bottles at irregular intervals during the first 25 days of in- cubation to follow the bacterial growth. The bacterial counts (table 1) of samples taken at 1-, 4-, 14-, and 25-day intervals from the unialgal G. brevis culture (bottle 1) and the G. brevis-iree bac- terial culture (bottle 2) were comparable except for the 4-day samples. The 4-da3^ sample from the G. brevis culture had a bacterial count of 2.7 million per ml. — about 50 percent greater than that of the G. bi-ems-hee bacterial culture. Some bacterial counts are questionable because of pro- longed refrigeration of samples before preparation of the plates. They indicate, however, the rela- tive number of bacteria in the three bottles at the various sampling intervals. Although bottles 1 and 2 apparently were inoculated with the same bacterial flora, we can only presume that the floras that subsequently developed in these bottles were qualitatively comparable. Approximately 6 weeks after bottles 1 and 2 were inoculated, materials from these bottles and the reservoir (bottle 3), in addition to an 11- month-old unialgal G. brevis culture and centri- fuged sea water, were used to conduct experiment 4 52982 O -58 -2 474 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE' Table 1. — Experiments 4, 5, 6, and 7: Bacterial counts of samples taken from study bottles at irregular intervals [Bottles No. 1- -imialgal Gymn odinium hrevis: No. 2 — Gymnodinium brevis- free bacteria; No . 3 — reservoir (uninoculated)! Number of Incubation period bacteria Remarks per ml. (millions) Shortly before in- oculation: Bottle No. 1. 0.000070 Media cooled to room temperature after Bottle No. 2. 0.000090 final heating before samples were col- Bottle No. 3 Shortly after in- oculation: Bottle No. 1. 0.014 Plates prepared shortly after samples were Bottle No. 2. 0.018 collected. Bottle No. 3 1 day: Bottle No. 1- 0.37 Samples refrigerated 5 days before plates Bottle No. 2. 0.41 were prepared. Bottle No. 3. 0.0015 4 days: Bottle No. 1. 2.7 Samples refrigerated 2 days before plates Bottle No. 2- 1.8 were prepared. Bottle No. 3- 0.26 14 days: Bottle No. 1. 2.0 Plates prepared 15 to 20 minutes after sam- Bottle No. 2. 1.8 ples were collected. Bottle No. 3. 0.13 25 days: Bottle No. 1. 1.4 Samples refrigerated 2 days before plates Bottle No. 2. 1.2 were prepared. Bottle No. 3- > 12.2 ' The value of 12 million bacteria per ml. for the reservoir appears to be excessively high when compared with the other counts obtained at either earlier or later intervals. With the exception of the presently considered value, the highest bacterial count obtained from the reservoir was 1.8 million per ml. (table 5, experiment 7). 4. One striped mullet (Mugil cephalus), 2% to 3 inches long, was subjected to 750 ml. of each of the five different nonaerated test materials in 1 -liter beakers. We observed the fish closely and recorded the time at which they began to show imbalance (distress time) and the time at which they showed no visible opercular movement (death time). Bacterial-count samples were ob- tained from each container before the fish was added. These samples were refrigerated 1 hour to 1^ hours before they were plated. The results (table 2) of this 24-hour experiment show that the fish in the two unialgal G. hreeis cultures died in 50 minutes and 2){ hours. Two of the three control fish survived considerably longer, 7K hours in the G. brems-iree bacterial culture and the entire test period in the uninoculated culture medium; however, the fish in the centrifuged sea water died after 58 minutes. The early death of this control fish was, perhaps, due to injury. The bacterial count of 6.0 million per ml. for the G. breins culture (bottle 1) in container 3 was five times greater than that for the G. brevis-hee bacterial culture (bottle 2) in container 4. Prior to this 6-week check the bacterial counts of these two cultures were comparable (table 1). Dis- parity in the bacterial counts for experiment 4 necessitated additional studies in order to deter- mine the toxicity agent in unialgal G. brevis cultures. Table 2. — Experiment 4: Effects on Mugil cephalus of unialgal Gymnodinium brevis cultures and of a G. brevis- free culture presumed to contain bacteria associated with this organism in unialgal cultures Number of Con- Material in Distress Death bacteria Remarks tainer container time' time 2 per ml. (mUlions) No.l.. Centrifuged sea 0:50 0:58 0.0070 water (Galves- ton beach). No. 2.. 11-month-old (7. brevis culture, 0.8 million or- ganisms per liter. 0:05 0:50 10.2 Fresh medium added to G. brevis culture occasionally. No.3-- 6-week-old G. brevis culture (bottle 1), 0.6 million organ- isms per liter. 0:15 2:16 6.0 Do. No.4-_ 6-week-old (7. brevis-tree bac- terial culture (bottle 2). (') (') 1.2 Fresh medium added to bac- terial culture occasionally. No. 5 6-week-old (•) 7:35 0.080 medium from reservoir (bot- tle 3). 1 Time (hr.:min.) required for fish to show first signs of imbalance. ! Time (hr.:min.) of cessation of opercular movement. 3 Distress or death did not occur during the 24-hour test period. * Not known. EXPERIMENT 5. — Effects of Unialgal Gymnodinium brevis Cultures, Associated Bacteria, and Unialgal Proro- centrum sp. Cultures In addition to testing the effects of G. brevis culture (bottle 1) and G. breds-hee bacterial culture (bottle 2), another dinoflagellate, Proro- centrum sp., was tested for toxicity to fish in the second experiment of this series. This organism was isolated from water samples taken in the lagoon, at Galveston, Texas. The materials in bottles 1 and 2 were '3% months old at this time. Freshly collected sea water served as control. The four different materials, 2 liters of each in 4-liter beakers, were tested in duplicate for toxicity to Mugil cephalus (2^ in. long). One fish was tested in each container without aeration. Samples were collected from each container for bacterial counts before the fish was added. These samples were plated after IK to 5 hours' refrigeration. The test fish subjected to the G. breds culture died within an hour (29 and 47 minutes) whereas the fish in the other test materials lived a mini- UNIALGAL AXD BACTERIA-FREE CULTURES OF G. BREVIS 475 mum of 8/3 hours to a maximum of 24 hours — the (hu-atioii of tlie experiment (table 3). The bacterial counts of botli the G. brevis culture (bottle 1) and the G. brenti-iree bacterial culture (bottle 2) had decreased since experiment 4 was conducted. Just as in experiment 4, however, the G. brevis culture had a much higher count — 2.9 to 3.4 million bacteria per ml. in contrast with 0.20 to 0.23 million per ml. for the G. brems- free culture. One of the containers of G. brevis culture (6) used in experiment 5 was employed in a supple- mentary study to determine whether adding several fish to the same culture would affect its toxicity. Another phase of this study was to check the response of fish transferred to sea water after being subjected to G. brevis culture. Im- mediately after the fish in container 6 died (29 minutes after beginning of experiment 5) it was removed and the first of five additional M. cephalus were placed in this container. This fish suc- cumbed after 21 minutes' exposure. After re- moving the dead Hsh, the second one was allowed to remain in container 6 for 15 minutes It was then transferred to sea water (container 1) where it died 12 minutes later. Exposures of 15 minutes and 7 minutes, respectively, in container 6, were required to kill the third and fourth fish. Each fish was removed from the container after it died. After 3 minutes' exposure in container 6, the Table 3. — Experiment 5: Effects on MiiKil cephalus of nnialgal cultures of Gymnodiniirn brevis ami Proro- centru'ii sp. and Gymnodiniu-n brevis-/ree culture pre- sumed to contain bacteria associated with this organism in uniatgal cultures Con- tainer Material In container Death time ' Number of bacteria per ml. (millions) Remarks No. 1.. Freshly collected sea water (Galveston beach) . do.... 3-month-old ProTOCentrum sp. culture, 0.5 million or- ganisms per liter. do Between 10 and 22 hours. do m Between 10 and 22 hours. 0:47 0.0025 0.0030 0.59 0.46 2.9 3.4 0.20 0.23 No. 2.. No. 3. . No. 4.. Fresh medium added to Prorocenlrum culture oc- casionally. No. 5.. 3j2-month-old 0. breris culture (bottle 1), 1.9 million or- ganisms per liter. do SJ-S-month-old G. 6reris-free bacterial cul- ture (bottle 2). do..-.. No. 6.. 0:29 added to 0. brevia culture occasionally. No. 7.. No. 8.. Between 10 and 22 hours. 8:20 Fresh medium added to bacterial culture oc- casionally. 1 Time (hr.rmin.) of cessation of opercular movement. ' Death did not occur during the 24-hour test period. fifth fish was removed to container 1 where it survived for 2% liours. Experiments 4 and 5 (tables 2 and 3) were in- adequately controlled with regard to quantities of bacteria. Experiment 6 was performed in an attempt to correct this shortcoming. EXPERIMENT 6. — Comparison of Toxicity of Unialgal Cultures of Gymnodinium brevis and Prorocentrum sp., and Effects of Heating and Filtration on Toxicity Besides attempting to ascertain the source of the toxic substance in unialgal G. brevis cultures, this experiment included a study of the effects of heating and filtration on the toxicity of such cultures. One month prior to conducting experiment 6 the remaining portion of the G. brevis-iree bac- terial culture (bottle 2) received an inoculum of unialgal Prorocentrum sp., which had proved non- toxic to M. cephalus in experiment 5 (table 3). This step was taken in an attempt to increase the bacterial concentration in bottle 2 to a level com- parable to that in the G. brevis culture (bottle 1). Centrifuged sea water was used in addition to these two bottles of material, which were 4% montiis old at this time. These three materials were tested in duplicate (containers 1 through 6). The test materials in all of these containers, except one container (2) of sea water, were sampled for bacterial counts just before the fish were added. These samples were refrigerated 20 minutes to 2}i hours before the plates were poured. Five containers (7 through 11) of the test ma- terial were used to test the effects of heating and filtration on the toxicity of unialgal G. brevis cul- tures. Bacterial counts were not made for these materials because such information was not needed. A filtrate, which was prepared by passing G. brevis culture througli filter paper (Whatman No. 42), was tested in duplicate. A single con- tainer of another test material consisted of the residues retained by the two filter-paper discs eluted in 2 liters of sea water. Two liters of 0. brevis culture were passed tlirough each disc. 476 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE' Other test materials included single containers of G. brevis cultures that had been heated to 35° and 45° C. Each of the 11 containers (4-liter beakers) re- ceived two common killifish {Fundulus grandis), 3 to 3K inches long. The test materials, 2 liters in each container, were not aerated. Only one of the four fish placed in each of the two control materials, sea water and Prorocentrum culture (bottle 2), failed to survive the 4-hour test period (table 4). On the contrary, the four fish subjected to tlie G. breins culture (bottle 1) died before the end of the test period. The death times were 9, 16, 100, and 130 minutes. Again, however, the G. brevis culture (bottle 1) had a greater bacterial concentration than the material in bottle 2, in spite of the addition of Prorocentrum sp. a month earlier. The count for the former was 2.2 to 2.4 million bacteria per ml. in contrast with 0.19 to 0.20 million per ml. for the latter. The count for the material in bottle 2 is quite similar to that obtained for this bottle in experi- ment 5 (table 3). The fish lived 20 to 100 minutes in the G. brevis culture heated to 35° C. In the culture heated to 45° C. the death times were only 13 and 18 minutes. Three of the four fish exposed to filtrates of a G. breins culture survived the experi- mental period. The two fish subjected to the materials eluted from filter paper through which G. brevis culture had passed died in 23 and 130 minutes. Filtration appears to reduce the toxicity of G. brevis cultures; however, other filtering methods must be tested before this effect can be established as a characteristic of filtration. Table 4. — Experiment 6: Effects on Fundulus grandis of unialgal cvltures of Gymnodinium brevis and Prorocentrum sp., and effects of heating and filtration on toxicity [2 flsh tested in each container] Container Material in container Death time • Number of bacteria per ml. (millions) Remarks No. 1 Centrifuged aged sea water (aged 1 month in dark) do..- , / 3:10 I m (') 1 0:09 1 0:16 f 1:40 1 2:10 / 2:45 1 (=) P) / 2:45 \ m 0:23 2:10 0:20 1:40 0:13 1 0:18 } 0 12 No. 2... No. 3 4>^-month-old 0. brevis culture (bottle 1), 0.7 million organisms per liter. do. ! " } 2.2 } 0.20 0.19 ) Fresh medium added to 0. brevis culture No. 4 occasionally. No. SK l-month-old Prorocentrum sp. culture (bottle 2), 0.9 mil- lion organisms per liter. do. - No. 6 occasionally. No.7< Filtrate (No. 42 Whatman paper) ot9-month-old (1. brevis culture, 1.2 million organisms per liter. do casionally. Filtrate refrigerated 2 days. No. 8 / No. 9 2 liters sea water (same as in containers 1 and 2) plus filter-paper discs used for containers 7 and 8. 2 liters of 0. brevis culture passed through each disc. days, eluted with sea water just before fish No. 10 were added. tainers 7 and 8 heated to 35° C, then cooled to room temperature. Same as container 10, except portion of culture heated to 45° C. No live a. brevis observed after culture was heated. No. 11 heated. ' Time (hr.:min.) of cessation of opercular movement. 2 Death did not occur during the 4-hour test period. 3 The O, brevis-free bacterial culture (bottle 2} was inoculated with unialgal Prorocentrum sp. 1 month prior to e.xperiment. * Another O. brevis culture was used for the heating and filtration studies because of Insufficient culture in bottle 1. EXPERIMENT 7. — Comparison of Toxicity of Unialgal Gymnodinium brevis, Prorocentrum sp., and Gymnodinium sp., and Effects of Filtration on Toxicity The final toxicity study in this series (experi- ments 4-7) compared the effects of unialgal cultures of Gymnodinium brevis, Prorocentrum sp., and Gymnodinium sp. The two latter organisms were isolated from water samples taken in the lagoon at Galveston, Texas. Gymnodinium sp. is morphologically similar to the cultured forms of G. brevis originally isolated from Florida waters. A portion of the experiment was to determine whether passage of G. brevis cultures through a millipore membrane would reduce toxicity as did their passage through filter paper. Striped mullet [Mugil cephalus) and variegated minnows {Cyprinodon variegatus) were used as test fish. The mullet (3 to 4 in. long) were main- tained in aerated aquariums about 24 hours before beginning the experiment. The minnows (about Iji in. long), collected 3 days prior to beginning UNIALGAL AND BACTERIA-FREE CUUTURES OF G. BREVIS 477 of tlic experiment, were kept in a nonaerated aquarium since this species survives well without aeration. Five of the seven different test materials included in this study were tested in duplicate. These materials consisted of two different unialgal G. hreins cultures (containers 1 and 2, 3 and 4) ; of Gymnodinium sp. (containers 5 and 6) ; Proro- cenfrum sp. (containers 7 and 8) ; and of culture medium from bottle 3 (containers 9 and 10). Also included was a filtrate prepared by passing 1 liter of G. brevis culture through a millipore membrane (container 11) and the residues retained with this membrane eluted in 1 liter of culture medium from bottle 3 (container 12). The millipore membrane (HA) retains particles as small as 0.5 micron. Each of the 12 containers (2-liter beakers) received approximately 1 liter of test material that was not aerated. Before adding the fish, bacterial samples were taken from two containers of G. brevis culture (1 and 2) and one container of culture medium (9). These containers were arbitrarily selected in order to compare the bacterial counts in some of the containers before the fish were added with those counts ■ obtained after death of the test fish. The samples for bacterial counts were refrigerated from 3 to 3^3 hours before the prepa- ration of pour-plates. All containers with dino- flagellates (1 through 8) as well as the filtrate of the G. brevis culture (contain": 11) were sampled for counts of these organisms just before the fish were added. These counts were completed within 3 hours after collection of samples. One M. cephalus was placed in each of the 12 containers. Each M. cephalus was removed from its container shortly after death and a C. variegatus was added. The fish were not introduced simul- taneously, since the M. cephalus were rather large for the containers. Immediately after the M. cephalus died in the containers (1, 2, and 9), that were initially sampled for bacterial counts, these containers were again sampled for such counts. Such samples were also taken from one of each of the remaining duplicate test materials (containers 3, 5, and 7) and from the millipore filtrate (con- tainer 11) following death of the M. cephalus. Pour-plates were prepared with these samples after 4 to 6)^ hours' refrigeration, except that the sample from container 7 was stored only l)i hours. The results (table 5) of this 27-hour experiment show that all of the M. cephalus subjected to either G. hrei^s culture or filtrate and residues of such a culture died in less than an hour. The death times varied from 14 to 53 minutes. M. cephalus in the uninoculated culture medium lived approximately 2^ and 3 hours. The bacterial count of 1.8 million per ml. of this culture medium (bottle 3) was higher than the 1.0 to 1.5 million Table 5. — Experiment 7: Effects of unialgal cultures of Gymnodinium brevis, Prorocentrum sp., and Gymnodinium sp. on Mugil cephalus and Cyprinodon variegatus and effects of filtration on toxicity of unialgal cultures Material in container Mu^il cepkdLu9 diprinodon variegatut Number (in millions) of— Container Distress time 1 Death time' Distress time' Death time ' Dinoflagel- lates per liter s Bacteria per ml.* No. 1 5-week-old G. brevis culture _ _-. do 0:39 0:10 0:10 0:09 0:32 0:22 Not known Not known 2:13 2:54 0:06 0:48 0:49 0:16 0:16 0:17 1:09 0:46 6:52 6:51 2:20 3:05 0:14 0:53 4:38 Not known 2:39 2:42 2:10 2:23 « (') (>) (') 3:26 (') 5:20 7-8 hours 2:42 6:04 2:23 2:42 (') (') (') (') 4:30 (') 1.3 1.3 1.2 1.0 1.4 1.1 1.0 0.9 1.5 No. 2 (1.3) 1.0 1-year-old G. brem culture, fresh medium added 6-8 weeks prior to use. (1.1) No 4 (1.1) No. 5 No 6 do (0.90) No. 7 8-month-old Prorocentrum sp. culture, fresh medium added approximately weekly, do No 8 (3.3) No. 9 • 6-week-old culture medium (uninoculated) from reservoir {bottle 3). do 1.8 No. 10 (2.1) No. 11 Millipore Oltrate of portion of same culture used in containers 3 and 4. Milliiiorc membrane used to obtain filtrate in con- tainer 11 eluted in 1 liter culture medium from bottle 3 (same as used in containers 9 and 10). 0 No. 12 (0.0060) 1 Time (hr.: niin.) rcquirod for fish to show first sipns of imbalance. 'Time (hr.: niia.) of ces.^ul inn of oporcuhir movomont. 'All sampU'S for dinoflapollatf counts were collected from the containers just before introduction of the fish. * Samples for the first bacterial count listed for each container were col- lected just before introduction of the M. cephalus: samples for the second count, in parentheses, were taken immediately after the A/, cephalus died. s Death or distress; did not occur duriuR the 27-hour tost period. • Original supply of medium In reservoir (bottle 3) became exhausted and It was renewed about 6 weeks prior to experiment. 478 FISHERY BULLETIN OF THE FPSH AND WILDLIFE SERVICE per ml. obtained for 6. brevis cultures. M. cephalus in the Prorocentrum culture survived nearly 7 hours. Those exposed to Gymnodinium sp. died after 46 and 69 minutes. Cyprinodon van'egatus survived considerably longer than M. cephaluf! in all test materials. In G. brevis cultures the death times for C. variegatus varied from 2?^ to 7-8 hours. This species survived the 27-hour test period in the two control materials (the culture medium and the Prorocentrum culture). C. variegatus lived about 2}^ and 2% hours in Gymnodinium sp. culture. M. cephalus lived only 14 minutes in the milli- pore filtrate of the G. brevis culture, in contrast with 53 minutes in material eluted from the milli- pore membrane. Likewise the filtrate was more to.xic than the residues to C. variegatus: the fish Hved 4K hours in the filtrate, whereas the fish in the residues survived the test period. RESULTS OF EXPERIMENTS WITH UNIALGAL CULTURES The fish subjected to unialgal cultures of G. brevis, with the exception of one fish, died more rapidly than those exposed to control materials in the seven experiments considered in this sec- tion. The greater survival of the control fish demonstrates that unialgal cultures of this or- ganism are toxic to the five species of fish tested {Membras vagrans, Mugil cephalus, Fundulus grandis, Cyprinodon variegatus, and Mollienisia latipinna). Indeed, the rapidity with which the fish succumbed in some of the cultures emphasizes the toxicity of unialgal G. brevis cultures. For example, the minimum death times for some of these fish were 4 minutes for M. vagrans (only 1 fish tested), 14 to 16 minutes for M. cephalus, and 9 to 16 minutes for 7^. grandis. However, all five species did not show such extremes of sensitivity to G. brevis cultures. The minimum death times for C. variegatus were 2)i to 2% hours. The concentration of G. brevis in the cultures used for the seven experiments varied from 0.6 to 2.1 million per liter. Since these cultures were not free of bacteria, such organisms possibly con- tributed to their to.xicity. The results of a preliminary study suggest that the survival period after exposure to G. brevis de- pends on the length of exposure and that fish subjected for just a few minutes may not recover when transferred to sea water. M. cephalus, sub- jected to unialgal G. brevis for 3 and 15 minutes and then transferred to sea water, lived for 165 and 12 minutes, respectively (experiment 5). Moreover, there is some suggestion that G. brevis cultures may become more toxic with each subsequent addition of test fish. For example, in experiment 5, the death times of each of four M. cephalus added in succession to the same culture decreased progressively from 29 to 7 minutes. We do not know the reason (s) for this apparent increase in toxicity. Since the test materials were not aerated, possiblj^ the o.xygen content was progressively lowered by the test fish and the death times decreased thereby. The toxicity of unialgal G. brevis cultures does not depend on the presence of living organisms. Cultures heated to 35° and 45° C. were no less toxic than the untreated ones. The removal of G. brevis from a culture by millipore filtration did not reduce the toxicity. The relative toxicity of a filtrate appears to be dependent upon the type of filter membrane employed. A paper membrane (Whatman No. 42) apparently retains the more toxic portion of the culture. Attempts to determine whether G. brevis or its associated bacterial flora produces the toxic sub- stance were not entirely successful. G. brevis -iree cultures presumed to contain the bacterial flora associated with unialgal cultures of this organism were not to.xic to the test fish. However, when these experiments were performed, the bacterial counts of these cultures were considerably lower than those of the G. brevis cultures. Nevertheless, a culture of presumed associated bacteria with a count of slightly more than a million per ml. proved nontoxic to M. cephalus. The bacterial count of this culture in experiment 4 (table 2) was comparable to those of the unialgal cultures that were toxic to M. cephalus and C. variegatus in experiment 7 (table 5). Furthermore, in experi- ment 7 uninoculated culture medium containing about 2 million bacteria per ml., which was slightly higher than the counts of two different G. brevis cultures, was not toxic to the test fish. Unialgal cultures of three species of dinoflagel- lates {Gymnodinium splendens, Gymnodinium sp., and Prorocentrum sp.) isolated from the lagoon, Galveston, Texas, were tested for toxicity to fish. Two species proved nontoxic and one toxic. The nonto.xic forms, G. splendens and Prorocentrum sp., were used in concentrations comparable to XJNIALGAL AND BACTERIA-FREE' CUL/TURES OF G. BREVIS 479 aiui, in some cases, even exceeding those of the toxic G. brevis. In fact, M. cephalus survived considerably longer in the Prorocentrum culture than in the control material (uninoculated culture medium) in experiment 7 (table 5). Prorocentrum possibly aided survival of the fish by liberating oxygen since the test materials were not aerated. Gymnodinium sp. was toxic to M. cephalus and C. variegatus (table 5). In view of these results and since this organism in culture is morphologi- cally similar to G. breins, we tentatively consider the Galveston Gymnodirdum to be G. hrenis. The Galveston Gymnodinium was observed only three times (all within the same week) although samples were collected from the lagoon three times weekly during an 11-month period. The concentrations in the lagoon samples varied from 1,000 to 60,000 organisms per liter. EXPERIMENTS WITH BACTERIA- FREE CULTURES OF GYMNODINIUM BREVIS Mass bacteria-free cultures of G. brevis were established following the completion of experi- ment 7. The two experiments, 8 and 9, to follow were the first toxicity tests to be conducted with pure cultures. The importance of these studies lies in the fact that the observed effects can be attributed to G. brevis with certainty since no other organisms were present during the incuba- tion period. Substantiation of the toxicity of unialgal G. brevis with bacteria-free G. brevis should establish the existence of a cause-and- effect relation between blooms of this organism and associated mass mortalit}' of marine animals. The experiments with bacteria-free cultures were more refined in several respects than those with unialgal cultures. In addition to aerating most of the test materials, such factors as temperature, dissolved oxygen, salinity, and pH were deter- mined. EXPERIMENT 8. — Effects of Unialgal and Bacteria- Free Gymnodinium brevis Cultures The first mass bacteria-free cultures of G. brevis were tested for toxicity to striped mullet (Mugil cephalus) and variegated minnows (Cyprivodon variegatus). (\ variegatus (1% to \%, in. long) were maintained in aerated aquariums for 5 days and M. cephalus (1 to 1)4 in. long) were brought into the laboratory the night preceding the ex- periment. One liter of distilled water was placed in each 2-liter beaker, containers for the various test materials, 5 daA's before commencing the experiment so that the aeration equipment could be tested and adjusted. The water was dis- carded and the beakers received no further treatment before introduction of the various test materials. The material in 12 of the 14 containers received gentle aeration continuously from a small aerator; the main air line from the aerator was equipped with a nonabsorbent cotton filter. To preclude possible excessive oxygen demand by G. brevis, light was provided continuously with two fluorescent lamps equipped with two 18-inch, 15-watt daylight tubes. The experimental setup is shown in figure 1. Six different bacteria-free cultures (containers 3, 4, 6, 7, 9, 10, 12, and 13) and four hatches of sterile control material (containers 5, 8, 11, and 14), all of which had incubated for a month, were employed in this study (table 6). The control material consisted of uninoculated culture medium otherwise subjected to the same conditions as the inoculated medium. The sterility of the G. brevis cultures and control materials was established by inoculating routine sterility-test media with sam- ples withdrawn from the culture vessels shortly before these materials were dispensed into the experimental containers. Duplicate containers (3 and 4, 6 and 7) of two of the six bacteria-free cultures were set up to compare the survival of test fish in aerated (containers 3 and 6) and non- aerated (containers 4 and 7) cultures. A year-old unialgal G. brevis culture (container 1), which was replenished with fresh medium about 4 days previously, was used to compare the effects of unialgal and bacteria-free cultures. Aged sea water (container 2) served as control material for the entire experiment. Tlie volume of test ma- terial placed in each container varied from 750 to 1,000 ml. Before adding the fish, samples were taken from four containers (3, 6, 9, and 12) of bacteria-free G. breins culture and two containers (5 and 11) of uninoculated culture medium for bacterial counts. Also at this time the nine containers with G. brevis were sampled for enumeration of this organism. 480 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Figure 1. — Experimental setup used in experiment 8. UNIALGAL AND BACTERIA-FREE CULTURES OF G. BREVIS 481 T.\BLE 6. — Experiment 8: Effects of bacteria-free and unialgal cultures of Gymnodinium brevis on Mugil cephalus and Cyprinodon variegatus ■ [AH containers received 2 flsh of each species eicept No. 2, In which 3 M. cephalua were tested] Material in container > Mutil cephalut Cyprinodon variegattu Number (in mUllons) of— Container Distress time' Death time' Distress time ' Death time • Oymnodin- ium brevit per liter • Bacteria per ml.' No. 1.— No. 2 / 1:06 \ 1:07 (') / 0:13 I 0:19 / 1:34 \ 1:39 1:13 1:13 (') 0:19 0:29 1:44 1:44 5:55 Not known (') 3:10 4:55 5:42 5:44 6:50 (') 7:20 7:22 5:10 6:31 (>) 3:40 4:55 3:21 3:48 6:17 6:47 (') 3:18 5:33 5:47 5:56 29:48 (') 7:47 7:50 6:05 7:02 (') 3:51 4:68 3:44 5:54 (') 3.4 (2.4) 1,000 ml. aged sea water (3.5) (5.0) 0.00025 No. 3 -. 4.1 (3.7) 3.1 (2.0) No. 4 -. 1,000 ml. of same culture used in container 3, but not aerated. 1,000 ml. sterile culture medium, pH 7.6— control for 3 and 4. (0.033) No. 5 (0.075) 0.0010 (') f 1:07 \ 1:38 / 0:28 I 0:28 m f 0:10 I 0:12 f 0:13 I 0:13 (') (') 1:47 1:48 0:34 1:56 (') 0:15 0:25 1:09 1:12 (21. 0) No. 6 4.8 (2.6) 4.0 (3.2) 0.020 No. 7 800 ml. of same culture used in container 6, but not aerated. 800 ml. sterile culture medium, pH 7.6— control for 6 and 7. 1,000 ml. bacteria-free G. brevis culture, pH 7.7 do (0.30) No. 8 — (0.60) No. 9 3.2 (2.7) 3.5 (2.4) 0.0070 No 10 (0. 075) 1,000 ml. sterile culture medium, pH 7.7— control for 9 and 10. No. 11 0 025 (7.0) 0.0050 No. 12 r 4:18 \ 4:32 f 2:16 I 2:19 4:24 4:44 2:26 2:41 (') 6:07 31:27 7:14 7:40 (») 27:17 •31:57 7:58 15:08 3.4 (1.8) 2.3 (1.5) No 13 do (24.0) 750 ml. sterile culture medium— control for 12 and 13. No. 14 ' Determinations of pH made on samples withdrawn directly from culture vessel. All containers aerated except as noted. ' Time (hr.:min.) required for fish to show first signs of imbalance. ' Time (hr.:min.) of cessation of opercular movement. * The first G. brevis and bacterial counts listed for each container were obtained from samples collected before introduction of the fish. Samples for second counts, in parentheses, were taken immediately after the death The samples taken for the initial bacterial counts were refrigerated from 1 hour to nearly 3 hours before preparation of the pour-plates. Most of the G. brems counts were completed within a few minutes to an hour after withdrawal of the samples. A few samples, however, stood for a maximum of approximately 3 hours. Each of the 14 containers received two M. cephalus and two C. variegatus except the con- tainer (2) of sea \i^ater, which received three M. cephalus. Each fish was removed from the container shortly after it died so that test materials would not become excessively fouled by decom- posing fish. After commencing the experiment, samples were again taken for G. brevis and bacterial counts. Seven of the containers with G. brevis were sampled for counts of this organism immediately after the death of the last fish in the container. Two other containers (12 and 13) of G. brevis, in which the last fish in the container survived beyond 8 hours, were sampled about 8 hours after begin- ning the experiment. The second set of samples for bacterial counts was taken either immediately after the death of the last fish in the container of la's! flsh in container, except for two G. brevis counts (containers 12 and 13) and four bacterial counts (containers 2, 5, 11, and 12): the former were taken about 8 hours and the latter 30 to 3U2 hours after start of experiment. ' Distress or death did not occur during the 3Ui-hour test period. • The last fish (C. variegatus) in container 12 died about 30 minutes after close of experiment. or after 30 to 31}^ hours, provided that one fish survived the test period (3 1)2 hours). The 10 containers sampled for bacterial counts included all of those that were sampled for such counts initially (containers 3, 5, 6, 9, 11, and 12), the container (1) of unialgal G. brevis, the container (2) of sea water, and the two containers (4 and 7) of nonaerated G. brevis cultures. These samples were plated after 15 to 90 minutes' refrigeration. Following the collection of bacterial and G. brevis samples, all test materials were sampled for dissolved oxj^gen, pH, and salinity determina- tions (table 7). Aeration of each container (if aerated) was discontinued just before collection of the samples, which were taken cither imme- diately after the death of the last fish in the container or near the end of the experimental period if at least one test fish survived. Seven containers (2, 5, 8, 11, 12, 13, and 14), those in which at least one fish survived beyond 7 hours, were also sampled for pH determinations 7 to 7% hours after start of the study. During this 31}^-hour experiment, room temperature varied from 20° to 24.5° C. 482 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 7. — Experiment 8: Dissolved oxygen, temperature, salinity, and pH data for test materials [See table 6 for materials In containers] Dissolved oxygen Temper- ature ' (°C.) Salinity (0/00) pll' Con- tainer P. p. m. Satura- tion I (%) Time* (hours) No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8 No.9 No. 10.... No. 11.... No. 12.... No. 13.... No. 14.... 6.77 6.10 6.70 3.38 5.69 6.02 4.18 5.59 6.48 6.84 6.62 6.60 6.34 6.10 89.4 85.2 89 3 45.0 77.8 81.7 56.7 76.8 86.1 91.5 90.8 90.6 86.0 83.7 22 8 23.6 21.9 22.1 23. S 23.0 22. S 23.8 .21.7 22.4 23.7 23.9 23.3 23.8 29.84 35.93 32.73 32 72 32,66 33.43 33.86 33.44 33.80 32 71 32.97 32.69 32 71 32.97 7.5 (7. 6) 7. 4 7.3 7.2 (7.2) 7.2 7.3 7.4 (7. 2) 7. 2 7.4 7.4 (7. 3) 7. 3 (7.3) 7.4 (7.3) (7.1)7.2 7 mi 5H 6 30H 7W 7 30?4 5 6 31 31 15 3U4 ' Percentage of saturation in relation to !^ea water of the given temperature and salinity in equilibrium with normal dry atmosphere. 2 Temperature of material in container at time dissolved oxygen was determined. ' Values In parentheses were determined 7 to 7^2 hours after start of experiment. ' The approximate time that elapsed between start of experiment and collection of samples for analyses. Only 1 of the 32 fish subjected to initially bacteria-free cultures survived the 3lK-hour test period; all except 3 (C. variegatus) died within 8 hours (table 6). The lone fish (C. variegatus) surviving the experimental period succumbed about 30 minutes later. On the contrary, only 1 of the 21 fish exposed to control materials of sea water and initially sterile culture medium failed to survive the test period. This fish (C. variegatus) died after nearly 30 hours in culture medium. The death times of M. cephalus, which varied from Yi hour to 4% hours, were considerably less than were those of the C. variegatus, 3K to nearly 32 hours. Fifty percent (16) of the test fish in the bacteria-free cultures died earlier than those (4) in the unialgal culture; the death time in this culture was about 1% hours for M. cephalus and 6)i and 6% hours for C. variegatus. The test fish survived considerably longer in the nonaerated than in the aerated G. hrevis culture in one instance; in the other case the opposite occurred, although the difference was less marked. The bacterial counts of the initially bacteria- free G. hrevis cultures sampled before adding the test fish varied from 250 to 20,000 per ml. The rather high initial bacterial counts are attributed to the prolonged standing (5 days) of the distilled water in the experimental containers while the aeration equipment was being tested and adjusted. The counts obtained from these cultures following the death of the last fish in the container (5 to 8 hours later) varied from 33,000 to 600,000 bacteria per ml. A count of 24 million bacteria per ml. was obtained after 31)^ hours from the initially bacteria-free G. hrevis culture in which one fish survived the test period. The first bacterial counts for two containers of initially sterile culture medium were 1,000 (container 5) and 25,000 (con- tainer 11) per ml. When these containers were sampled again near the end of the 3lK-hour test period the bacterial count had increased to 21 million (container 5) and 7 million (container 11) per ml. The results of experiment 8 show clearly that bacteria-free G. hrevis cultures are toxic to fish. Nevertheless, we desired to confirm these results by using test materials with greatly reduced initial and terminal bacterial counts. EXPERIMENT 9.— Effects of Bacteria- Free and Unialgal Gymnodinium brevis Cultures, and Effects of Filtration on Toxicity The second study with bacteria-free G. brevis differed somewhat from the first one (experiment 8). In experiment 9, the initial bacterial con- tamination of test materials by containers and aeration equipment was reduced, the effects of two methods of filtration on the toxicity of bacteria-free cultures were studied, and the sensi- tivity of two size groups of mullet (Mugil cephalus) to G. brevis cultures was compared. The experi- mental setup was the same as for experiment 8 (fig. 1) except that more containers and a larger air pump were used. We employed three precautions to reduce the initial bacterial contamination of the test ma- terials. One precaution was to heat the experi- mental containers (2-liter beakers) in a hot air oven at 150°-160° C. for 2 hours and allow them to cool overnight in the oven. The containers were removed from the oven shortly before the test materials were added. Secondly, the aera- tion apparatus was autoclaved for 15 minutes at 15 pounds pressure. Before sterilization this ap- paratus was assembled and packaged so that a glass air-delivery tube could be inserted into each of the 18 containers without handling the portion that contacted the test materials. All air pumped UNIALGAL AND BACTERIA-PREE CtTLTURES OF G. BREVIS 483 into the containers passed through a nonabsorbent cotton filter installed in the main air line from the pump. The third safeguard against excessive bacterial contamination was to place the test fish in autoclaved 85-percent aged sea water for several minutes before their transfer to the experi- mental containers. This concentration of sea water is about the same as that in the culture media to which the fish were exposed. Three different, month-old bacteria-free 6. hrevis cultures, two batches of sterile culture medium of the same age, a 10-week-old unialgal G. brevis culture, and autoclaved 85-percent aged sea water constituted the test materials. One liter of test material was placed in each of the 18 containers. Since the test materials were aerated more vigor- ously in this study, the increased air flow made equalizing the degree of aeration in each container difl!icult and the agitation of the test materials probably varied more. One portion of the experiment (containers 1 through 8) was designed to compare the effects of millipore (HA membrane) and paper (Whatman No. 40) filtration on the toxicity of one of the bacteria-free G. brevis cultures. One batch of the sterile culture medium treated in the same manner as the G. brevis culture served as control material. The residues of millipore and paper filtration of both the G. brevis culture and the sterile culture medium were each eluted in 1 liter of sterile culture medium to obtain four of the eight test materials used in this phase of the study. The test materials for the remaining portion of this experiment (con- tainers 9 through 18) consisted of duplicate containers of two different bacteria-free G. brevis cultures, sterile culture medium, autoclaved 85- percent aged sea water, and a unialgal G. brevis culture. The distribution of the test materials and numerical designation of the containers are presented in table 8. Samples were taken for bacterial and G. brevis counts immediately after the test materials were dispensed. All containers of unfiltered bacteria- free G. brevis culture (9, 10, 11, and 12) and of unfiltered culture medium (13 and 14), in addition to a container of sea water (15) and a container of unialgal G. brevis culture (17), were sampled for bacterial counts. The containers of filtrates or residues were not sampled because such counts were not needed. These samples were refrigerated 4)j to 6}^ hours before being plated. Samples for G. brevis counts were obtained from the containers with filtrates of G. brevis culture (1 and 2) and bacteria-free G. brevis (9, 10, 11, and 12). The G. brevis concentration of the unialgal culture used in containers 17 and 18 was ascertained by withdraw- ing a sample from the culture before dispensing it. The G. brevis samples were counted between l}i to 2% hours after collection. Each container received four fish: three small mullet (1 to XYi in. long) and one large mullet {4}i to 5K in. long). The small mullet were main- tained in aerated aquariums overnight and the large mullet were used within a few hours after they were collected. Since the volume (1 liter) of test material was somewhat small for the large mullet, they thrashed about vigorously and pos- sibly injured some of the smaller test fish. Excepting the filtrates of G. brevis cultures (containers 1 and 2), the bacteria and G. brevis were again enumerated for those containers for which such counts were made initially. The second set of bacterial samples from the five containers with G. brevis (9, 10, 11, 12, and 17) was plated after 4% to 5% hours' refrigeration whereas those samples from the containers with culture medium (13 and 14) and sea water (15) were stored only Yz hour to IK hours. The second set of G. brevis counts was completed between 1 hour and 2 hours after the samples were collected. In addition, samples for pH and salinity determinations were taken from all 18 containers. Bacterial, G. brevis, pH, and salinity samples were taken either shortly after the death of the last fish in the container or at the end of the test period (24 hours) provided that one fish in the container survived, except for some pH and salinity samples as noted under Remarks in table 8. The room temperature varied from 22° to 25° C. during the 24-houT test period. The fish subjected to bacteria-free cidtures as well as filtrates and residues of such cidtures died more rapidly, for the most part, than those exposed to the control materials (table 8). The survival rate among the control fish, especially small M. cephalus, was not good. Fish of this size group died quickly in one batch of sterile culture medium (containers 13 and 14). The death times, varying from 25 minutes to 2% hours, for the small mullet in these two containers were comparable to the death times, varying from 15 minutes to 2% hours, for the same-sized fish in unfiltered bacteria-free 484 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 8. — Ezperimenl 9: Effects on two size groups of Mugil cephalus of bacteria-free and unialgal Gymnodinium brevis cultures, and effects of filtration on the toxicity of bacteria-free cultures [1 large mullet and 3 small mullet tested in each container] Container Material in container Distress time ' Death time 2 Number (in millions) Gym- nodinium brevis per liter ' Bacteria per ml.J pH3 Salinity »/oo Remarks Millipore filtrate (HA membrane) of 1 liter of bacteria-free O. brevis culture, 3.3 million organisms per liter. Two membranes used; first one became clogged. Paper filtrate (Whatman No. 40) of 1 liter of the same culture that was used to prepare the filtrate in con- tainer 1. Both millipore membranes used to prepare filtrate in container 1 eluted in 1 liter of sterile culture medium, which was 1 week old with a pH of 7.8. Filter-paper disc used to prepare the filtrate in container 2 eluted in 1 liter of sterile medium from same batch as used for elution in container 3. Millipore filtrate (HA membrane) of 1 liter of sterile culture medium — con- trol for container 1. Paper filtrate (Whatman No. 40) of I liter of sterile medium from same batch as used to prepare filtrate in container 5 — control for container 2. Millipore membrane used to prepare filtrate in container 5 eluted in 1 liter of sterile medium from same batch as used for elution in container 3 — con- trol for container 3. Filter-paper disc used to prepare the filtrate in container 6 eluted in 1 liter of sterile medium from same batch as used for elution in container 3— con- trol for container 4. Bacteria-free G. breviis culture .do. .do. .do. Sterile culture medium— control containers 9, 10, 11, and 12. .do.. for Autoclaved 85% sea water. .do.. 10-week-old unialgal <7. brevis culture that was not replenished with fresh medium. .do.. (0:09) 0:19 0:42 1:10 (0:46) 0:34 0:44 0:55 (Notknown) Not known 5:00 5:05 (0:31) 0:36 0:37 0:38 (') 0:48 2:12 (<) (*) Not known (*) (') (') 0:41 (*) (') (•) (') (<) (') (0:15) 0:16 0:23 0:32 (0:16) 0:32 0:48 0:68 (0:30) 0:42 0:43 0:43 (0:32) 0:35 0:37 0:41 (') 0:18 0:30 0:52 (Notknown) 0:46 0:52 Not known (Notknown) Not known Not known Not known (Notknown) Not known (') (') (0:12) 0:24 0:54 1:48 (0:06) 0:12 0:14 0:24 (0:14) 0:2.5 1:27 1:44 (0:54) 0:44 1:10 8:05 CAVi-WA hr.) 6:00 6:06 5:10 (0:39) 0:42 0:42 0:42 (<) 2:01 2:17 1:13 (') (') (') (') (<) (') (<) (') (<) (') (') (0:23) 0:42 1:45 1:64 (0:28) 1:41 2:14 2:28 (0:41) 0:44 0:45 1:08 (0:40) 0:45 0:45 0:45 (') 0:25 0:33 1 '38 (8^2-13^ hr.) 0:47 0:56 2:39 (14:10) 8H-13}^ hr. 8;--2-13H hr. iVrlZH hr. (8H-13>-2 hr.) 7:16 (*) W (0:14) 0:45 0:56 2:30 (0:10) 0:15 0:20 0:25 2.4 (2.3) 3.1 (2.9) 2.7 (2.8) 3.3 (1.8) 2.6 (2.7) 0 (0. 0050) 0.000008 (0.015) 0.000001 (0. 030) 0.00010 (0. 0044) 0.000002 (in excess of 80 millions) 0.000026 (25 to 50 millions) 0. 000012 (in excess of 80 millions) } - 6,6 5.9 (6.4) 1.3 (1.4) 33.3 33 0 30.7 28.9 33.2 33.4 32.! 32.1 33.0 33.1 33.0 33.2 33.2 Salinity — ?^ hour after last fish died. Some G. brevis passed by the filter paper when it was inadvertently over- flowed. Salinity and pH— after 14 hours. pH— 54 hour, salinity— H4 hours after last fish died. Salinity— H hour after last fish died. Salinity— l?i hours after last fish died. Salinity— IH hours after last fish died. Salinity, first pH, and second bacterial count- after 14 hours. Second pH (6.4)— after 25 hours. Salinity, pH, and second bacterial count— after H'i hours. pH— 1 hour, salinity— 2 hours after last fish died. ' Time (hr.: min.) required for fish to show first signs of imbalance. The first distress time listed tor each container, in parentheses, pertains to large mullet (4J* to 51.4 in. long) and the other three distress times pertain to small mullet (1 to Hi in. long). ' Time (hr.: min.) of cessation of opercular movement. The first death time listed for each container, in parentheses, pertains to large mullet and the other three death times pertain to small mullet. ' The first G. brevis and bacterial counts listed for each container were obtained from samples taken from the containers before addition of the fish except for containers 17 and 18. For these containers, the sample was with- drawn from the culture before portions were dispensed into containers 17 and 18. With the exceptions noted under Remarks, samples for the second 0. brevis and bacterial counts, in parentheses, as well as pH and salinity samples were collected cither shortly after the last fish died if all fish died withm 24 hours or at the end of the 24-hour experimental period if at least 1 fish survived. . I Distress or death did not occur during the 24-hour test period. s The large mullet in container 16 probably died of oxygen deficiency. The fish appeared in good condition after 13^* hours and the material in the con- tainer was being aerated. When noticed again 30 minutes later the fish was dying and no air was being pumped into the container at this time. UNIALGAL AXD BACTERIA-FREE CULTURES OF G. BREVIS 485 and unialgal G. hrems cultures. Howovor, large M. cephalus survived considerably longer in the two containers of sterile culture medium than in the G. hrevifi cultures. One large fisli survived the 24- hour test period and the other lived at least 8K hours. In contrast, the death time for the six large midlet in three different unfiltered G. hrei-is cultures, two of which were bacteria-free, varied from 10 to 41 minutes. All fish in the sea water lived for at least 1% hours and two of them survived the test period. The early deaths of the small mullet in con- tainers 13 and 14 were probably due to the abnormally low pH of the culture medium. The pH of the material in container 14 was 5.9 about 14 hours after beginning of experiment and in- creased to 6.4 after 25 hours. A pH value of 6.6 was obtained for the culture medium in container 13 at the end of the test period. The control fish survived much better in the batch of sterile culture medium (containers 5, 6, 7, and 8) used in the filtration phase of this ex- periment than they did in the batch of medium (containers 13 and 14) used in the nonfiltration phase. All (4) of the large mullet and two-thirds (8) of the small mullet subjected to filtrates and residues of culture medium survived the 24-hour test period. On the contrary, none of the fish (4 large mullet and 12 small mullet) exposed to filtrates and residues of the G. brevls culture sur- vived tlie test period. The difference between the effects of the two methods of filtration on the toxicity of the G. brevis culture was marked. The millipore filtrate was much more toxic to the fish than the residues; in the filtrate the death times varied from 14 to 104 minutes in contrast with a variation of 5 to 13)2 hours in the residues. The toxicity of the filter-paper residues was greater than the filtrate; the test fish lived from 39 to 42 minutes in the residues, whereas they survived from 44 minutes to about 8 hours in the filtrate. The attempt to reduce the initial bacterial con- tamination of the bacteria-free G. brevis cultures and the sterile culture medium by the previously mentioned precautions was successful. The counts for these materials varied from 0 to 100 bacteria per ml. The bacterial counts in the initially bacteria-free G. breris cultures varied from 4,400 to 30,000 per ml. at the time the last fish died in the container (% hour to 2]^ hours later). The initial count in the unialgal culture was 1.3 million bacteria per ml.; when the last fish died 2^2 hours later the count was 1.4 million bacteria per ml. At the end of the test period the bacterial concen- tration was in excess of 80 million per ml. in one container (13) of initially sterile culture medium. After about 14 hours the count was 25 to 50 million bacteria per ml. in the other container (14) of culture medium and in excess of 80 million per ml. in one container of sea water. The pH of the initially sterile culture medium in container 14 increased from 5.9 to 6.4 during the 11 -hour period after the last fish died. Addi- tional fish were subjected to this material to test its toxicity at the higher pH. EXPERIMENT 9a. — Supplementary Toxicity Tests of Some Test Materials Previously Used in Experiment 9 Near the close of experiment 9, we conducted a supplementary study to determine whether the initially sterile culture medium in container 14 was still toxic to small M. cephalus. Five other containers originally a part of experiment 9 — one container (15) of 85-percent, autoclaved sea water and the four containers (9, 10, 11, and 12) of initially bacteria-free G. brevis culture — were in- cluded in this study. The other containers of culture medium (13) and sea water (16) were not used because this part of experiment 9 was still in progress, as they contained one or more live fish. Small mullet from the group used in experiment 9 served as test fish. Four of these fish, which were maintained in 85-percent, autoclaved sea water for 24 hours, were placed in each container. Only two of the four containers of G. breiis culture were aerated in an attempt to determine the effects of agitation and aeration on the toxicity of G. brevis. The culture medium (container 14) and sea water (container 15) were not aerated. The results (table 9) show that G. brevis cultures were still toxic to small mullet, but the fish sur- vived well in the previously toxic culture medium. The 16 fish in the G. brevis cultures died in from 8 to 125 minutes; 10 of them survived less than 1 hour. In the control materials (containers 14 and 15), however, only 1 of the 8 fish failed to survive the 24-hour test period; 1 fish in the sea water succumbed after 5^4 hours. 486 FISHERY BULLETIN OF THE FPSH AND WILDLIFE SERVICE Table 9. — Experiment 9a: Effects on Mugil cephalus of certain test materials initially used in experiment 9 [4 small flsh tested in each container; see table 8 for container contents) Container Distress Death time' times 0:37 1:26 1:02 1:28 1:18 2:05 1:50 2:05 0:05 0:36 0:20 0:55 0:20 1:11 0:20 1:11 Not known 0:08 0:08 0:18 0:10 0:18 0:18 0:34 0:05 0:10 0:05 0:10 0:08 0:10 0:08 0:13 (») (») Not known 5:47 (') (') m (=) (') « Number (in millions) of Oymnodinium brevis per Uter Remarks No. 9.. No. 10 No. 11. No. 12 No. 14 No. 15. 0.9 (sample taken 30 minutes after 4 fish were added). 2.4 (sample taken 25 minutes after 4 fish were added). 1.3 (sample taken 15 minutes after 4 fish were added). 1.4 {sample taken 15 minutes after 4 fish were added). Materia] aerated continuously during previous 24-hour period, aerated vigor- ously for last 12 hours and during experimental period. Material not cloudy. Material neither aerated during previous 21-hour period nor during experi- mental period. Material not cloudy. Material not aerated during previous 22-hour period. Vigorous aeration began 18 minutes before flsh were added and continued throughout experiment. Material not cloudy. Material neither aerated during previous 23-hour period not during experi- mental period. Material not cloudy. Material neither aerated during previous 10-hour period nor during experi- mental period. Material was cloudy when fish were added. Two fish developed heavy microbial growth on caudal fin; they died between 30 and 48 hours after study was begun. Two fish aUve after 5 days. Material relatively clear at this time. Material neither aerated during previous 9-hour period nor during experimental period. Material cloudy when fish were added, but relatively clear 5 days later. Three fish aUve after 5 days. ' Time (hr.:min.) required for fish to show first signs of imbalance. ' Time (hr.:min.) of cessation of opercular movement. ' Distress or death did not occur durmg the 24-hour test period. Although the experiment was discontinued after 24 hours, containers 14 and 15 were set aside and observed for 4 more daj^s. Two of the fish in the culture medium died between 30 and 48 hours, and the other two were alive after 5 days. The three fish remaining in the sea water also lived through 5 days. There was some indication that nonaerated G. brevis cultures were more toxic than the aerated. RESULTS OF EXPERIMENTS WITH BACTERIA- FREE CULTURES The results of the experiments with bacteria-free G. brevis confirm that this organism produces a fish-killing substance (s) as indicated by tests with unialgal cultures. Bacteria-free G. brevis was toxic to the two species of fish tested (Cyprinodon variegatus and Mugil cephalus). The minimal death time for C. variegatus was about 3% hours. The mullet were more sensitive, with a minimum death time as lov/ as 15 minutes. Furthermore, small mullet appear to be less sensitive to the substance than the large ones, since the three small fish in each container of G. brevis culture outlived the large one. The con- centration of G. brevis in these bacteria-free cultures varied from 2.3 to 4.8 million organisms per liter. Such concentrations are considerably less than the 10 to 60 million per liter sometimes encountered in areas where dead or dying fish occur. There was good agreement between the results of experiments 8 and 9 with regard to the toxicity of bacteria-free G. brevis cultures to small M. cephalus, which was the only species used in both experiments. Nevertheless, the small mullet sur- vived much better in the control materials in experiment 8 than in experiment 9. In the latter experiment, the small mullet died rapidly in one particular batch of sterile culture medium (con- tainers 13 and 14). Large M. cephalus in this batch of medium, however, survived much better than the small mullet. We attribute the early death of the small midlet in the culture medium placed in containers 13 and 14 of experiment 9 to the abnormally low pH of this particular batch of medium. Moreover, the relatively poor survival of these fish in some of the other control containers of experiment 9 was pos- sibly due to their being damaged by the vigorous thrashing about of the large mullet. The length of some of the large test fish slightly exceeded the inside diameter of the experimental containers. The low pH of the material in container 14 sug- gests that the batch of medium placed in container 14 was abnormal before the fish were added. Ap- proximately 14 hours after experiment 9 began the pH of the medium in container 14 was 5.9 and the bacterial count was between 25 and 50 million per ml. Since the pH of this material increased from 5.9 to 6.4 during the next 11 hours in spite of UNIALGAL AND BACTERIA-FREE CULTURES OF G. BREVIS 487 heavy bacterial growth, the initial pH of this particular hatch of sterile medium possibly was lower than 5.9. The pH of the medium in con- tainer 13 (duplicate for container 14) was 6.6 after 24 hours. The pH values, after 24 hours, for the millipore and paper filtrates of the other batch of sterile culture medium used in experiment 9 were 7.2 and 7.0, respectively. The initial pH of month-old sterile culture medium and bacteria- free G. hreris cultures used in experiment 8 varied from 7.6 to 7.7. No pH values were ascertained initially in experiment 9, excepting a pH of 7.8 for the week-old sterile culture medium used for eluting the millipore and filter-paper residues. The pH of culture medium either with or with- out G. hrevis customarily drops after fish are introduced; this decrease probably results from increased bacterial growth and accumulation of fish waste products. Excepting two containers (13 and 14) of culture medium and a container (4) of filter-paper residues of a bacteria-free G. breris culture eluted in sterile culture medium in experiment 9, a minimal pH value ol 7.0 was recorded after test periods as long as 24 to 30 hours in the two experiments, 8 and 9, conducted with initially bacteria-free G. hrevis cultures and sterile culture medium. Small mullet survived in culture medium with pH values as low as 7.0 and 7.1 in experiments 8 and 9. We do not know why the pH of one particular flask of culture medium used in experiment 9 was so low. It is our surmise that the abnormally low pH resulted from failure to rinse the culture flask after 7-percent, hot nitric acid was used in the cleaning process. By actual test we found that failing to rinse the flask after the nitric-acid treat- ment did significantly lower the pH of 2.0 liters of sea-water-base medium. In this particular test the pH of the autoclaved mediiun stabilized at approximately 6.4. A flask of culture medium, companion to one with the low pH (containers 13 and 14) of experi- ment 9, that was inoculated with G. hrevis failed to support growth of this organism after incubat- ing for 1 month. Unfortunately, the medium in this flask was discarded without checking the pH. Since the two flasks of mediimi were prepared at the same time, we consider that G. hrevis possibly failed to grow because the pH was unfavorable. We have experienced thus far only this one failure of bacteria-free G. hrevis to grow in low-form culture flasks (3-liter). Growth of G. hrevis has not occurred in sea-water-base medium with an initial pH of less than 7.0. Furthermore, preliminary studies indicate that a pH below 7.4 is unsatis- factory for this organism. The results of experiment 9a (table 9) show that the culture medium in container 14 was no longer toxic to small mullet when retested about 24 hours after experiment 9 began. The pH of this medium was 6.4 about 1 hour after start of experiment 9a. All four test fish were alive after 24 hours; two of them were alive after 5 days. The pH of the medium in container 14 was 6.9 at this time. The effects of millipore and paper filtration on the toxicity of bacteria-free G. hrevis cultures were the same as observed when unialgal cultures were so treated. The more toxic portion of the culture passes through the millipore membrane, whereas filter paper retains the more toxic fraction. Bacteria-free G. hrevis cultures proved just as toxic as the unialgal ones in simultaneous tests with comparable concentrations of this organism. For example, in experiment 8 (table 6) the death times for M. cephalus in initially bacteria-free cultures varied from a minimum of 15 minutes to a maximum of 4 hours and 44 minutes. Seven of the 16 M. cephalus died in less time than was required (1 hour and 13 minutes) to kill the two M. cephalus subjected to a unialgal culture. The C. variegatus succumbed in the bacteria-free cul- tiu-es in periods varying from a minimum of 3 hours and 18 minutes to a maximum of 32 hours; 13 of the 16 test fish died within 8 hours. Nine of the 16 C. variegatus in bacteria-free cultures died in less than the minimum time (6% hours) required to kill the two C. variegatus in a unialgal culture. Further data for comparison of the effects of bacteria-free and unialgal cultures are available from experiment 9 (table 8). Two large M. cephalus subjected to unialgal cultures died in 10 and 14 minutes. The four large mullet in the bacteria-free cultures died within 23 to 41 minutes. The death times for six small M. cephalus in the unialgal cultures varied from 15 minutes to 2% hours. The extremes of death times for the 12 small mullet in bacteria-free cultures were quite similar — 25 minutes to 2 hours and 28 minutes. Despite the evidence that bacteria are not directly responsible for the toxic effects of G. hrevis cultures, the possibility that bacteria play a 488 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE significant role in the development of G. brevis blooms should not be overlooked. Such organ- isms maj' contribute appreciably to the nutrition of G. brevis. For example, some of the bacteria isolated from unialgal G. brevis cultures produce vitamin Bij-active substances (Starr et al., 1957). Vitamin Bjj apparently stimulates the growth of G. brevis in sea-water-base medium (Wilson and Collier, 1955). Once fish kills are initiated by blooms of G. brevis, excessive bacterial growth resulting from the increased availability of organic matter may possibly cause the blooms to decline in isolated situations. Bacteria could unfavorably affect G. brevis blooms by producing substances toxic to this organism, by competing for nutritive sub- stances, and by adversely altering the pH. We have frequently observed the failure of unialgal G. brevis to grow in tubes in which the medium became cloudy with bacterial growth. Another role for which bacteria must be con- sidered is that of a detoxicating agent. Shilo and Aschner (1953) found that bacteria decreased the toxicity of cultures of Prymnesium parvum, a marine and brackish water chrysomonad that is toxic to fish. .Similarly, bacterial activity may influence the toxicity of G. brevis in the laboratory and in nature. TOXICITY STUDIES WITH BACTERIA Although the available evidence indicated that Gymnodinium brevis causes the toxicity of unialgal cultures, direct proof was lacking prior to the development of bacteria-free cultures. In addi- tion to the bacterial studies previously considered, tl^e possibility that bacteria may cause all or some of the toxic efi'ects was investigated by test- ing pure cultures of some of the bacteria from unialgal G. brevis cultures. Furthermore, toxicity tests were conducted with pure cultures of an unidentified red-pigment-producing bacterium iso- lated from Florida water and Flavobacterium piscicida Bein. Bein (1954) suggested that F. piscicida, a chromogenic bacterium, possibly was a cause of mass fish mortality associated with discolored water off the west coast of Florida. BACTERIA ISOLATED FROM UNIALGAL GYMNODINIUM BREVIS CULTURES Because of the preliminary nature of these studies and the crudeness of the quantitative bacterial estimates, onty a summary will be pre- sented. Tlie bacleria used have not been identi- fied; presently they are being characterized morphologically and physiologically by Dr. T. J. Starr. The test fish were Gulf killifish (Fundulus similis'). Two test fish per container (1 -liter beaker) were subjected to about 500 ml. of nonaerated test materials. Bacterial suspensions were prepared by adding 16.5 ml. of a 24-hour culture to 500 ml. of filtered Galveston Bay water. Control ma- terials consisted of the same ratio of sterile culture medium and bay water as well as bay water alone. Crude estimates of bacterial concentrations were made by preparing a pour-plate of 0.02 to 0.03 ml. of a sample collected shortly after the fish were added. A second sample was taken either after both fish died or after 23 hours if at least one fish survived this period. The colonies were too numerous to count in most of the plates prepared from the second set of samples. Therefore, not even rough estimates could be made for the bacterial concentrations. The most-abundant colony type isolated from unialgal G. brevis cultures on Spencer's peptone sea-water agar and Bein's peptone agar is gener- ally a convex, white, opaque colony produced by Gram-negative rods. This colony type may represent several different species and physiologi- cal types of bacteria. Two separate isolates of the white, opaque colony did not give evidence of being toxic to F. siinilis. The initial bacterial concentration was in the order of 0.5 to 1 million per ml. A flat, white, transluscent colony with an irides- cent sheen, also produced by Gram-negative rods, is usually the second most-abundant colony type isolated from unialgal G. brevis cultures. An initial concentration of approximately 1 million bacteria per ml. of this colony type gave no evidence of toxicity to F. si7nilis. Chromogenic bacteria constitute only a small portion of the bacterial flora of unialgal cultm-es of G. brevis; however, yellow-pigment-producing bacteria become abundant in cultures treated with dihydrostreptomycin sulfate. They dominate in G. brevis cultures treated with 500 to 1 ,000 Mg. of this antibiotic per ml., and often occur in nearly pure cultm-e. This antibiotic may cnliance the XHSriALGAL AXD BACTERIA-FREE CULTURES OF G. BREVIS 489 growtli of the "yellow bacteria" by inhibiting competing bacteria. Dihydrostreptomyciii sulfate (125 /ig. per ml.) initially lowers the pH of the culture medium by 0.5 to 0.8 of a pH unit. This change in the medium may be a factor favoring the increased growth of the pigment -producing bacteria. Cultures of an isolate from a nontreated and a dihydrostreptomycin-treated unialgal culture, each with an initial count of about 1 million yellow bacteria per ml., had no toxic effects on F. similis. Plates prepared from samples taken 23 hours after the start of the experiment showed no yellpw colonies. The counts of all bacteria were about the same in the initial and 23-hour samples. A CHROMOGENIC BACTERIUM ISOLATED FROM THE WEST COAST OF FLORIDA After Bein (1954) reported the toxic effects of Flavobacterium piscicida to fish, we made a cursory check for chromogenic bacteria in Florida off the Fort Myers-Naples area during November 1954. Gymnodinium brevis was present in the area at that time although the maximum concentrations were usually less than 1 million organisms per liter. Small fish kills, mainly of mullet, were being reported sporadically at that time. During the sampling trips, however, we observed less than 10 dead fish. Surface samples from 15 stations in this area were plated on Spencer's sea-water peptone agar. Pour-plates containing 1 ml. of each sample were prepared within 1 minute after collection to avoid possible changes in the bacterial flora. All except two of the samples from the 15 stations contained 0. brevis, and the counts varied from 7,000 to 0.5 million per liter. A white, opaque colony was the most abundant type in the 15 samples; some samples showed a few lemon-yellow colonies. A total of two red colonies were observed in the 15 plates. One of these colonies was isolated from a sample taken 5 miles west of Wiggin's Pass on November 4, 1954. The G. brevis count for this sample was 8,000 per liter. The red-pigment-producing bac- terium, which has not been identified, is a Gram- negative, motile rod. A 24-hour pure culture of the "red bacterium" was tested for toxicity to Fundulus similis as a part of the studies dealing with bacteria isolated from unialgal G. brens cultures. Two fish were tested in each of the four containers of test material in which the bacterial count varied from approxi- mately 0.5 to 1 million per ml. The bacterial culture gave a pink tint to the test materials. .\11 eight fish died within 2 to 8 hours. After the last fish died in each of the four containers, samples were taken for bacterial counts. The red colonics in the plates prepared with 0.02 to 0.03 ml. of these samples were so abundant that enumeration was impossible. We believe that the minimum concentrations were of the order of 1 to 2 million red bacteria per ml. at the time the last fish died. A 6-month-old unialgal G. brenis culture (replen- ished with fresh medium about three times weekly) that contained 1.3 million organisms per liter killed the test fish less rapidly than cultures of the red bacterium. In the G. brevis culture, one fish died after l)^ hours and the other one died after 10 to 19 hours. These death times appear rela- tively long when compared with the usual death times of fish subjected to other unialgal cultures. These results may mean either that F. similis is less sensitive to unialgal cultures than other fish tested thus far, or that this culture was less toxic than the others. FLAVOBACTERIUM PISCICIDA BEIN A chromogenic bacterium was isolated by Reuben Lasker (Bein 1954) from a pooled water sample collected after the occurrence of a fish kill associated with discolored water in Whitewater Bay on the southwest tip of Florida. Bein found that 24-hour cultures of this bacterium, which he named Flavobacterium piscicida, killed several species of marine fish. He gave no quantitative values concerning the concentration of bacteria used other than that 500 ml. of a 24-hour culture of this species grown in a 0.1 -percent peptone solution in aged sea water were added to 4 gallons of continuously aerated sea water. After 24 hours all fish in the experimental aquariums died and the water exhibited a bright orange-yellow dis- coloration. We attempted to estimate the minimum number of F. piscicida required to kill mullet {Mugil cephalus), since we desired to know what concen- tration of this bacterium might be required to kill fish under natural conditions. Tiie Marine Lab- oratory of the University of Miami provided the stock from which our cultures were derived. Since Bein gave no indication of the amount of inoculum used to seed the medium to obtain the 490 FISHERY BULLETIN OF THE FrSH AND WILDLIFE SERVICE 24-hour test cultures, we could not duplicate his inoculation procedures. In our experiment, 24-hour cultures were ob- tained by inoculating two loops of culture removed from a 24-hour slant culture (1-percent peptone agar) into duplicate Erlenmeyer flasks containing approximately 150 ml. of sterile 0.1 -percent pep- tone solution in aged sea water. One flask was incubated at 30° C. and the other at 25°. The culture incubated at 30° was deep orange after 24 hours and the one incubated at the lower tempera- ture was yellow. The more-intensely pigmented culture was used because it probably contained the greater concentration of bacteria. The ratio of bacterial-culture volume to sea-water volume varied from a maximum of 33.0 ml. of 24-hour culture to 1.0 liter of sea water, which is about the ratio used by Bein, to a minimum of one-hun- dredth of this ratio. Sea water and sterile 0.1- percent peptone solution were used as control materials. Each experimental container (2-liter beaker) received 1 .0 liter of sea water, which was aerated continuously. Two mullet (3 to 4 in. long) were placed in each container. The test fish, which had been maintained in the laboratory for several days, were acclimated in the experi- mental containers for about 24 hours before the beginning of the study. The water, which became cloudy in all containers by the end of the acclima- tion period, was replaced with fresh sea water shortly before the test materials were added. Samples were collected for bacterial counts about 30 minutes after the experiment began. These samples were refrigerated % hour to 2 hours before preparation of the pour-plates. .A.11 fish were alive when the experiment was discontinued 5 days later and had shown no evi- dence of distress. A second set of bacterial- count samples was taken at this time. The samples were plated after being refrigerated ^ to 3 hours. All bacterial plates were prepared with Bein's agar medium, and they were counted after 6, 12, 15, and 21 days' incubation. The number of pigmented subsurface colonies as well as the pigment intensity of such colonies increased after prolonged incubation. After 21 days, how- ever, some colonies began to lose pigment. Bac- terial counts of the test materials used in this ex- periment are listed in table 10. RESULTS OF STUDIES WITH BACTERIA The results of the preliminary experiment with some of the bacteria isolated from unialgal Gymnodinium brevis cultures suggest that cultures of the two dominant colony types, both nonpig- mented, and a sparsely occurring pigmented form are not toxic to fish. On the contrary, a "red bacterium" isolated from 6. brevis-miested water off the west coast of Florida appears to be toxic to fish. Neverthe- less, until this bacterium is found more abun- dantly or its association with fish kills is estab- lished, we shall not consider it of importance as a cause of fish mortality occurring off the west coast of Florida. Further toxicity studies with this organism have been discontinued until such Table 10. — Initial and final counts of bacteria in materials used in experiment testing effect of Flavobacterium piscicida on Mugil cephalus Material in container (in addition to 1.0 liter of sea water) Initial count (in millions) Final count (in millions) after 5 days Container Bacteria per ml. Flavobac- terium pisci- cida per ml.i Bacteria per ml. Orange and yellow-pig rrent- producing bac- teria per ml.' No. 1 None >0,80 •0.46 5.0 3.2 1.2 1.2 0,40 0.32 None None 1.9 0.80 0.42 0.35 0.22 0.020 1.6 1.7 3.6 3.1 2.5 6.7 1.8 3.7 0.050 No. 2 33.0 ml. sterile 0.1% peptone solution 0.060 No. 3 __ 0.010 No. 4 _. 16.6 ml. 24-hour F. piscicida culture and 16.5 ml. sterile 0.1% peptone solution 8.2 ml. 24-hour F. piscicida culture and 24.8 ml. sterile 0.1% peptone solution do No. 6 0.020 No. 6 0.040 0.028 No. 7 3.3 ml. 24-hour F. piscicida culture and 29.7 ml. sterile 0.1% peptone solution _ 0.3 ml. 24-hour F. piscicida culture and 32.7 ml. sterUe 0.1% peptone solution _ __ No. 8 0.060 0.020 ' The number of F. piscicida based on the observed number of deep-orange colonies. 2 Since some of the chromogenic colonies varied from either yellow to orange or orange to yellow at various intervals during the 21 days of incubation, orange and yellow colonies were combined to indicate the maximum ob- Berved number of bacteria in this color range. There were no definitely orange colonies in samples from containers 3 and 8; about one-third of the pigmented colonies in the sample from container 6 and more than one-half of those in samples from all other containers were definitely orange. 3 Approximately one-sixteenth of the bacteria produced yellow pigment. * Approximately one-tenth of the bacteria produced yellow to light-orange pigment. insriALGAL AND BACTERIA-FREE CULTURES OF G. BREVIS 491 time that evidence is obtained to implicate it as a fish-killing agent in nature. A bacterium producing red pigment was isolated b}- Bein (1954) from Indian River on the east coast of Florida at the time of an outbreak of dis- colored water during August 1951. Although the water was discolored when the isolation was made, no fish either dead or alive were observed. Bein found that this bacterium was nontoxic to several species of fish. Howell (1953), studying the same outbreak of discolored water, reported that no great quantity of fish was killed; and that the dis- coloration was caused by a dinoflagellate, which he described as a new species, Gonyaulax monilata. The attempt to determine the minimal lethal concentration of Flavobacterium piscicida failed, since none of the fish died during the experi- mental period. The reason for this lack of toxic effect is not known. Although we followed as nearly as possible the procedures used by Bein (1954), the experimental conditions employed were unfavorable for the bacterium, especially with regard to toxin and pigment production. The bacterial counts given in table 10 indicate that high concentrations of F. piscicida (orange- pigment-producing bacteria) were present ini- tially. The second counts, which were made from samples collected at the end of the 5-day experimental period, however, showed that either this bacterium or its chromogenic characteristic decreased greatly during the intervening period. A similar disappearance of chromogenic bacteria occurred during the toxicity study with yellow bacteria isolated from the unialgal G. brevis cultures. The experimental conditions may have affected the toxicogenic and chromogenic char- acteristics of this bacterium in several possible ways: (1) By killing the organism, (2) by in- hibiting the growth of the organism, or (3) by altering the toxicogenic and chromogenic prop- erties of the organism. With only initial and terminal counts available, the question as to when the decline of F. piscicida occurred — either early or gradually during the experimental period — can not be answered. At no time during the 5-day period did the contents of any container show the bright orange-yellow discoloration observed by Bein after 24 hours. The contents of container 3, which received the maximum amount of F. piscicida culture, ex- hibited a tinge of orange after 24 hours. This slight discoloration gradually became less notice- able and completely disappeared by the fifth day. The numbers of F. piscicida recorded in table 10 represent minimal counts, which may be considerably lower than the actual concentrations, because several of the white subsurface colonies produced the characteristic orange pigment after being transferred to agar slants (1 -percent pep- tone) . All deep-orange colonies in plates prepared with samples that contained F. piscicida were considered to have been produced by this species. A few orange colonies appeared in plates prepared from some of the control samples, particularly in those taken after 5 days. We made no allowance for orange colonies which possibly were not produced by F. piscicida, since tlie counts were considered low even with the inclusion of such colonies. An examination of the initial counts (table 10) obtained for the control containers (1 and 2) and the two containers (7 and 8) receiving the smallest volumes of bacterial cultures suggests that counts of bacteria, exclusive of F. piscicida, varied in the order of 0.3 to 0.8 million per ml. From these values we presume that the actual concentration of F. piscicida was approximately twice as great as the values listed. For example, the initial number of F. piscicida in container 3 was probably about 4 million if one assumes that only about 1 million of the bacteria per ml. were other than F. piscicida,. The initial counts of this bacterium were in relatively good agreement with the various dilutions employed. For example, the material in container 3, which received the greatest volume of bacterial culture, yielded a count of about 2 million chromogenic bacteria per ml., whereas the material in container 8, which received only one-hundredth as much as container 3, gave a count of 20,000 chromogens per ml. Since Bein gave only the ratio of F. piscicida L'ulture and sea water employed, the concentrations of bacteria used in his studies cannot be directly compared with those we used. If the amount of water discoloration is proportional to the abundance of F. piscicida, we presume that the water in Bein's aquariums probably contained in excess of 4 million F. piscicida per ml. at the end of the 24-hour experimental period. This presumption is based on the observation that the sea water in container 3 of our experiment was not appreciably 492 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE discolored after the addition of 2 to 4 million F. piscicida per ml., whereas in Bein's studies the water was bright orange-yellow 24 hours after it received the bacterial cultures. GENERAL DISCUSSION The well-established association of Gym.nodi- nium hrevis with the sporadic mass mortality of fish and other marine animals that has occurred in the Gulf of Mexico since 1947, in conjunction with the clear-cut laboratory demonstration that this dinoflagellate in pure culture is toxic to fish, leaves no reasonable doubt that this organism causes these mortalities. Considering this evi- dence, we propose the name "brevis red tide" for such mortalities instead of the nonspecific term "red tide" that is used commonly in popular and scientific writing. In our opinion, ample character- istics properly identify these mass mortalities occurring in the Gulf of Mexico as a distinct phenomenon. This phenomenon can be diagnosed by the presence of G. hrevis in the waters in which fish and other marine animals are dying. An additional diagnostic characteristic is the odorless, human-respiratory-tract irritant often present in or near mortality areas where droplef.i of sea water become air-borne as a result of wind, wave action, et cetera (Galtsoff 1948; Gunter et al., 1948; Woodcock 1948; Ingle 1954; and others). Failure to find either one or both of the men- tioned diagnostic characteristics in an isolated area of dead or dying fish would not necessarily eliminate G. hrevis as the cause. At least four possible reasons support this statement: (1) The conditions or agents required to make droplets of sea water air-borne may be absent. (2) Dead fish may drift or be carried into an area either unsuit- able for the survival of G. hrevis or removed from the bloom. (3) An isolated mass of water in which G. hrevis is blooming may suddenly become unsuitable for this organism and yet not lose its toxicity to fish until sometime later. Experimental evidence supports this suggestion, since the re- moval of living G. hrevis from cultures by millipore filtration or the killing of the organisms with gentle heat did not inactivate the toxic substance. A specific diagnostic test for the toxic substance (s) produced bj- this organism would be helpful in diagnosing the cause of mortality in such cases. G. hrevis is so delicate that under adverse con- ditions it may die within a matter of minutes, leaving only fragmentary remains that are not readily identifiable. (4) The fish may contact the toxic substance in a bloom of G. hrevis in one area and yet not succumb until it moves into an area where the organism is not flourishing. This possi- bility is based on limited observations that fish, exposed to G. hrevis cultures for short periods and removed before they showed distress, died after being placed in sea water. The death time after removal to sea water appears to decrease as the exposure time is increased. The results of our studies support Galtsoff's (1948) conclusion that fish are not killed by clogging of the gill filaments by masses of G. hrevis. The available evidence makes untenable the view that fish suffocate as a result of mechanical occlusion of gill surfaces by the mere presence of large num- bers of organisms. Results of the studies with both heated and filtered G. brevis cultures, tests with other dinoflagellate cultures, and oxygen analyses of G. hrevis cultures emphasize the exist- ence of a toxic substance(s). G. hrevis cultures heated to 35° and 45° C. did not lose their toxicity although the organisms were disrupted. Like- wise, the removal of this organism (both unialgal and bacteria-free) by millipore filtration did not detoxicate the cultures. Unialgal cultures of G. splendens and Prorocentrum sp. were nontoxic despite the fact that the number of organisms compared to and even, in some cases, exceeded the concentrations in the toxic G. hrevis cultures. Filtrates of G. hrevis cultures are toxic; how- ever, the method of filtration determines whether the more toxic portion of the culture wiU pass through or be retained by the filter membrane. In our studies conducted with both unialgal and bacteria-free cultures, the filter-paper residues eluted in either sea water or culture medium were more toxic to the fish than were the filtrates. The results were reversed when a culture was passed through a millipore membrane under suction: the more toxic portion passed through the mem- brane. The reasons for the different effects of these two methods of filtration are not known. The filter materials differ in composition and size; the miUipore membrane (diameter 47 mm.) is made of cellulose esters, whereas the filter paper UNIALGAL AND BACTERIA-FREE CULTURES OF G. BREVIS 493 (diameter 18.5 cm.) consists of cellulose fibers. Retention of the toxic substance by filter paper may be due to greater adsorptive area and/or to differences in physical and chemical properties. Another possibility is that filtration by gravity flow used with filter paper may result in the re- tention of considerably more intact organisms than in the case of millipore filtration under suc- tion. Assuming that greater numbers of G. brevis were broken up by millipore filtration, more toxin might be released in such a case. However, our preliminary studies showed no apparent increase in the toxicity of G. brevis cultures in which the organisms were cytolyzed by gentle heating. There are no indications that the fish kills caused by 6. brevis result from the great masses of this organism depleting the oxygen in the sea water. Connell and Cross (1950) suggested that anaerobic conditions created by the high bio- chemical oxygen demand of an armored dinoflag- ellate, Gonijaulaj, was the cause of mass mortality of fish associated with discolored water in OfFatts Bayou (Galveston Bay) during the summer of 1949. Gunter et al. (1948) concluded that the 1946-47 incidence of mass mortality of marine animals on the west coast of Florida was not associated with low oxygen. Oxygen deficiency can be excluded as a factor in the death of the fish in the G. brevis cultures that were aerated. In experiment 8 (tables 6 and 7), the dissolved- oxygen content of all aerated test materials ex- ceeded 75-percent saturation and some were as high as 90 percent. With continuous gentle aeration one of the G. brevis cultures was about 90- percent saturated, although the bacterial count was 24 million per ml. at the time the dissolved- oxygen content was determined. The results of an attempt to determine the effects of aeration on the toxicity of bacteria-free G. brevis cultures were contradictory. In experi- ment 8, the fish — especially Mugil cephalus — showed distress and died more rapidly in the aliquot that was aerated (dissolved oxygen, 89- percent saturation) than in the one that was not aerated (dissolved oxj-gen, 45-percent saturation). Tlie fish in another nonaerated aliquot (dissolved oxygen, 57-percent saturation) in this experiment showed distress much sooner than those in the aerated culture (dissolved oxygen, 82-percent saturation). In experiment 9a (table 9), M. cephalus died somewhat faster in the nonaerated aliquot of a G. brevis culture than in the aerated aliquot. It is apparent from these contradictory results that the influence of such factors as aeration cannot be evaluated until standardization of the toxicity tests is more complete. Several factors probably influence the degree of toxicity of G. brevis cultures. Aside from the concentration of G. brevis other factors such as the growth phase of culture, pH of culture during growth and during test period, temperature and salinity of test culture, size and number of test fish, volume and degree of aeration of test culture, and bacterial growth, are among those that must be consider h1 in standardizing the toxicity tests. Shilo and Aschner (1953) found that a number of factors influenced the toxicity of Prymnesium parimm cultures. The toxicity was decreased by oxidizing agents, aeration, adsorbents including pond-bottom soils, bacterial growth, and low pH (below 7.5). They improved the standardization of their toxicity tests b}' using a buffer to control pH of the test culture and streptomycin to sup- press bacterial growth. Furthermore, the test cultures were not aerated. McLaughlin ^ work- ing with the same organism reported that cul- tures grown in an alkaline medium were more toxic than those grown in an acid medium. P. parvum cultures (grown in alkaline media), rendered nontoxic by lowering the pH to 6.0, regain their toxicity when made alkaline (Shilo and Aschner, 1953; McLaughlin '). Some of the factors mentioned above may ac- count for the variable death times obtained in duplicate containers of G. brems cultures. Since on some occasions the pattern of response in one of the duplicate containers was different from that in the other, we consider that factors other than variations of the individual test fish were respon- sible. For example, in experiment 6 (table 4) the fish died in 9 and 16 minutes in one container whereas death occurred after 1% and 2 hours' exposure in another. Both containers (non- aerated) held similar amounts of the same unialgal culture. There are other anomalies that defy explanation at present. One of these concerns the variation in the response of fish subjected to cultures sup- posedly grown under the same conditions, and > McLaughlin. John J. \. The physiology and nutritioniil requirements of .some chrysomonads. tJ.'i pp. Thesis, Ph. D., New York University. 1956. 494 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE treated in the same manner throughout the study. For example, in experiment 8 (tables 6 and 7), container 9 received one of the duplicate pure cultures and container 10 received the other. The G. brevis counts of the material in the two containers and such measured factors as pH, dissolved oxj'gen, and salinity were comparable. The similarity between the two containers is further emphasized by the distress times for Mugil cephalus — 10 and 12 minutes in container 9, and 13 minutes for each of the two fish in container 10. Despite all these similarities, M. cephalus in container 9 died in 15 and 25 minutes in contrast with 69 and 72 minutes in container 10. However, Cyprinodon variegatus, a less sensitive fish than M. cephalus, showed more similarity of death times: 3 hours 51 minutes and 4 hours 58 minutes in container 9, and 3 hours 44 minutes and 5 hours 54 minutes in container 10. Another anomaly in experiment 8 is the case in which one pure culture (container 12) with a G. brevis count of 3.4 million per liter was less toxic to both M. cephahis and C. variegatus than the duplicate culture (container 13) in which the count was 2.3 million organisms per liter. The distress times (about 4)^ hours) and death times (about 4K hours) for M. cephalus in the more- concentrated G. brevis culture (container 12) were approximately 2 hours greater than such times in the less-concentrated duplicate culture (con- tainer 13). The distress times and death times for C. variegatus in each culture were not as uniform as in the case of M. cephalus. Nevertheless, they show that the culture in container 13 was more toxic to C. variegatus than the one in container 12: the fish died after about 8 and 15 hours in the former container and after about 27 and 32'hours in the latter. We realize that due to variations in the condi- tion of the individual test fish some will survive longer than others when subjected to toxic agents. It is our opinion, however, that the over-all uniformity in the response of the test fish within each individual container, especially in experi- ment 8, is a strong indication of other subtle variables of which we have no knowledge. Despite the evidence that our toxicity studies require more standardizing, the results of all experiments reported here indicate that the sensi- tivity of fish to G. brevis cultures is variable. Our tests included six species of fish as follows: Membras vagrans, Mugil cephalus, Fundulus gran- dis, Mollienisia latipinna, Fundulus similis, and Cyprinodon variegatus. Possibly the most sensi- tive of these fish is A/, vagrans; the only individual tested died in 4 minutes. M. cephalus, the species used in the greatest number of experiments (5), showed death times varying from a minimum of 8 minutes to a maximum of 4% hours. The great majority of them died within an hour. Small M. cephalus (1 to 1^4 in. long) are possibly slightly less sensitive than large M. cephalus (iYi to dYi in. long) to G. brevis cultures. This possibility is suggested by results of experiment 9 (table 8); the large M. cephalus in each container of unfiltered G. brevis culture, without exception, died before any of the three accompanying small M. cephalus. In some cases the first small mullet died within 3 to 5 minutes after the large mullet; in other cases the first small mullet died 20 to 70 minutes later. F. grandis showed about the same degree of sensitivity as M. cephalus — minimum death time 9 minutes, maximum deatli time 2 hours and 10 minutes. The M. latipinna died in a minimum of 47 minutes and a maximum of 85 minutes. C. variegatus is probably the least sensitive of the six species tested. Its minimum death time was about 2)^ hours, the maximum was 32 hours. The sensitivity of F. simiHs is possibly comparable to that of C. variegatus. Two F. similis died in 7)^ and 10 to 19 hours. A chromogenic bacterium, Flavobacterium pis- cicida, has been suggested as a possible cause of the mass fish kills and associated sea-water dis- colorations occurring along the west coast of Florida (Bein 1954). Bein found that this bac- terium was toxic to several species of fish although he did not indicate the bacterial concentrations employed. In our tests Mugil cephalus were not affected by initial concentrations of 2 million or more F. piscicida per ml. Contrary to Bein's experience, F. piscicida apparently did not grow in our experiment and possibly lived only a short time after being added to the test medium (sea water). We could scarcely detect this bacterium at the end of the 5-day test period. A "red bacterium" that we isolated from G. breds-in- fested water off the west coast of Florida appears to be toxic to fish. Concentrations of the order of 0.5 to 2 million bacteria per ml. were toxic to Fundulus similis. The "red bacterium" has not UNIALGAL AND BACTERIA-FREE CULTURES OF G. BREVIS 495 beon encountered in sufficient abuiulaiice to implicate it as a fish-killing agent. Thus far, neither the association of chromogenic bacteria witli extensive fish kills nor the natural existence of toxic concentrations of such bacteria has been established. On the otlier hand, Gymno- dinium breiis has been clearlj' implicated on both these counts. SUMMARY AND CONCLUSIONS 1. Since 1947, blooms of the dinoflagellate, Gymnodinium brevis, have been associated with sporadic mass mortalities of marine animals and discolored water in the Gulf of Mexico. Extensive laboratory studies conducted with unialgal and bacteria-free cultures, as well as related bacterial studies, offer overwhelming evidence that blooms of this organism are the direct cause of the associated mortalities. 2. Bacteria-free cidtures of G. brevifi with con- centrations varying from 2.3 to 4.8 million or- ganisms per liter were toxic to two species of test fish. Five species of fish were killed when sub- jected to unialgal G. brevis cultures containing 0.6 to 2.1 million organisms per liter. The numbers of G. breds in areas of natural fish kills often greatly exceed these toxic laboratory concentrations. 3. Bacteria apparently do not produce or di- rectly contribute to the production of the toxic substance present in G. brevis cultures. Bacteria- free cultures were just as toxic to fish as the unialgal ones. 4. The toxicity of G. brevis does not depend on the presence of the living organisms. Removing the organisms from culture by millipore filtration or killing them with gentle heat did not appear to alter the toxicity. The high dissolved-oxygen content of aerated G. brevis cultures eliminates oxygen deficiency as a factor. 5. The toxic substance produced by G. brevis readily passes through a millipore membrane, but for the most part it is retained by filter paper. 6. Studies with bacteria-free and unialgal cultures indicate that the sLx species of test fish are difTerentially sensitive to G. brevis cultures. The test fish, listed in order of decreasing sensi- tivity, were Membras vagrans, Mugil cephalus, Fundulus grandis, Mollienisia latipinna, Fundulus similis, and Cyprinodon variegatus . 7. Some chromogenic bacteria isolated from the Gulf of Mexico or adjoining bays are toxic to fish in laboratory tests. However, association of such bacteria with mass fish kills in the Gulf of Mexico has not been established. Flavobac- terium piscicida, previously found to be toxic to several species of fish in undetermined concen- trations, was not toxic to Mugil cejjhalus at an initial concentration of about 2 million organisms per ml. A red bacterium isolated from G. brevis- infested waters appears to be toxic to fish at con- centrations in the order of 0.5 to 2 million per ml. This bacterium was uncommon during our survey since only two colonies were obtained from 15 samples of 1.0 ml. each. LITERATURE CITED Bein, Selwyn Jack. 1954. A study of certain chromogenic bacteria iso- lated from "red tide" water with a description of a new species. Bull. Marine Science Gulf and Caribbean 4 (2): 110-119. [Contribution No. 121, Marine Lab., University of Miami.] CoNNELL, Cecil H., and Joy Barnes Cro.ss. 1950. Mass mortality of fish associated with the protozoan Gonyaulax in the Gulf of Mexico. Science 112 (2909): 359-363. Davis, Charles C. 1948 Gymnodinium brevis sp. nov., a cause of dis- colored water and animal mortality in the Gulf of Mexico. Botanical Gazette 109 (3): 358-360. [Contribution No. 17, Marine Lab., University of Miami.] Droop, M. R. 1954. A note on the isolation of small marine algae and flagellates for pure cultures. Jour. Marine Biological Assoc. United Kingdom 33 (2): 511- 514. Galtsofp, Paul S. 1948. Red tide. Progress report on the investiga- tions of the cause of the mortality of fish along the west coast of Florida conducted by the U. S. Fish and Wildlife Service and cooperating organi- zations. U. S. Department of the Interior, Fish and Wildlife Service, Special Scientific Rept. No. 46: 1-44. 1949. The mystery of the red tide. Scientific Monthly 68 (2): 108-117. 496 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE GuNTER, Gordon, Robert H. Williams, Charles C. Davis, and F. G. Walton Smith. 1948 Catastrophic mass mortality of marine animals and coincident phytoplankton bloom on the west coast of Florida, November 1946 to August 1947. Ecological Monographs 18 (3): 309-324. [Con- tribution No. 14, Marine Lab., University of Miami.] Howell, John F. 1953. Gonyaulax monilala, sp. nov., the causative dinoflagellate of a red tide on the east coast of Florida in August-September, 1951. Trans. American Microscopical Soc. 72 (2): 153-156. Ingle, Robert M. 1954. Irritant gases associated with red tide. Uni- versity of Miami, Marine Laboratory, Special Service Bull. No. 9: 1-4. [Mimeographed.] MoRiTA, Richard Y., and Claude E. Zobell. 1955. Occurrence of bacteria in pelagic sediments collected during the Mid-Pacific Expedition. Deep-Sea Research 3 (1): 66-73. Pergamon Press, London and New York. [Contribution from Scripps Institution of Oceanography, Univ- ersity of California, La JoUa, New Series No. 808.] Shilo [Shelubsky], M., and M. Aschner. 1953. Factors governing the toxicity of cultures containing the phytoflagellate Prymnesium par- vum Carter. Jour. General Microbiology 8 (3): 333-343. Spencer, C. P. 1952. On the use of antibiotics for isolating bacteria- free cultures of marine phytoplankton organisms Jour. Marine Biological Assoc. United Kingdom 31 (1): 97-106. Starr, Theodore J., Mary E. Jones, and Domingo INIartinez. 1957. The production of vitamin B,2-active substances by marine bacteria. Limnology and Oceanography 2 (2): 114-119. Sweeney, Beatrice M. 1954. Gymnodinium splendens, a marine dinoflagel- late requiring vitamin B12. American Jour. Botany 41 (10): 821-824. [Contribution from Scripps Institution of Oceanography, University of California, La Jolla, New Series No. 728.] Wilson, William B., and Albert Collier. 1955. Preliminary notes on the culturing of Gymno- dininmbrevis Davis. Science 121 (3142): 394-395. Wilson, W. B., and S. M. Ray. 1956. The occurrence of Gymnodinium hrevis in the western Gulf of Mexico. Ecology 37 (2) : 388. Woodcock, Alfred H. 1948. Note concerning human respiratory irritation associated with high concentrations of plankton and mass mortality of marine organisms. Jour. Marine Research 7 (1): 56-62. [Contribution No. 389 of the Woods Hole Oceanographic In- stitution.] I'. S. GOVERNMENT PRINTING OFFICE : 1958 O -432982 OBSERVATIONS ON THE SPEARFISHES OF THE CENTRAL PACIFIC By WILLIAM F. ROYCE, Fishery Research Biologist Since 1950 the Pacific Oceanic Fishery Investi- gations (POFI), U. S. Fish and Wildlife Service, has been investigating the high-seas fishery resources of the tropical and subtropical Pacific Ocean. These investigations have shown that several kinds of tunas, particularly j^ellowfin (Neothunnus macropterus), skipjack (Kats7twonus pelamis), and albacore {Germo alalunga), form the most promising fishery resources in this area.' Moreover, these tunas are commonly found asso- ciated with two other gi'oups of large fishes, the spearfishes, principally the marlins, and certain species of sharks. An understanding of the role of both of the latter gi-oups is important because they compete with the tunas for food, but the marlins, in particular, are objects of gi-eat interest in themselves because of their value for sport along the coast of the Americas and for food along the coast of Asia. Despite the interest in and value of the marlins, these spectacular fish are little known to Ameri- cans. Their habits, their wide distribution on the high seas, even the number of their species, have not been known. Much information has appeared in Japanese literature during the last two decades, but little has been wTitten in English, and even if it had been, the uncertainty about the species occurring on both sides of the Pacific might have prevented associating the species of the western Pacific with those found off the Americas. The spearfish problems discussed in this paper include (1) a study of diagnostic characters and morphological comparisons of the species, (2) a decision as to the correct names, and (3) observa- tions on distribution, abundance, and habits. For the latter we shall use our observations and refer extensively to the Japanese literature. We shall not attempt a monograph, however. Observa- tions on the spearfishes in all parts of the world are ' Reports of the tuna studies, tORCther with detailed tables of the catch and fishing localities, may be found In Murphy and Shomura (IB.'iSa. 1953b, 1955). Note— Approved for publication, October 1!. igsfi. Fishery Bulletin 124. being added to the literature so rapidly and so little is known that such a treatment would be premature. Most of our data have been obtained from spear- fishes caught on longline fishing gear from POFI vessels. This gear has been described by Niska (1953) but, briefly, it consists of a series of baited hooks 15 to 30 fathoms apart suspended from a line at depths of about 200 to 400 feet. On all cruises made after July 1952 (table 1) records were kept of the species of spearfishes caught at each station (fig. 1), and in man)' instances morphometric measurements were made, together with observations on sex, sexual condition, and food in the stomach. Such observations were not as complete as might be desired because the primary assignment on each cruise was to obtain information on the tunas, and observations on the spearfishes were made as time permitted. Table 1. — The longline fishing cruises of POFI vessels on which spearfish data have been collected, 1952-64 Vessel and cruise Cruise period Locality and west longitude John R. Manning: No. 12... 8/16-9/16/52 No. 13...- 10/16-12/fi/.'i2 Equatorial area. LM" and 170°. No. 14 1/22-3/25/53 Equatorial area. 140° and 160°. No. 16 4/28-6/16/53 Equatorial area. 150° and 170°. No. 16 7/24-9/2/53... Equatorial area around Line Island?. 15.1° and 160°. No. 17 10/15-11/8/53... Around Christmas Island. No. 18 11/21-12/19/53 Equatorial area. 155° and 155° to 159°. No. 19 l/I5-l/17/541p„., , 1/19-2/6/54 /•^"ril 2/16-3/10/54 Part 2 North of Hawaiian Islands. 160°. North of Hawaiian Islands. 155° and 147°. No. 20 5/11-6/23/54 Equatorial area around Line Islands. 157° to 163°. Oharles H. Gilbert: No. 16 2/1S-4/20/54. Equatorial area. 110° to 120°, and 155°. Hugh M. Smith: No. 18 10/7-11/22/52.. Equatorial arm. 120° and 130°. No. 19 1/8-2/12/53... Equatorial area around Line Islands. 157° to 162°. Copalieri. 8/13-9/27/52.. ... Equatorial area. 140°. In the collection of data, assistance was rendered by many members of the POFI staff, including the officers and crews of the vessels who had the problem of handling these large and troublesome 497 498 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 180° 170° 160° 150° 140° -30° 20° H A IV , ^^^ o O ^f^^ -XJ- 130° -I0°- ^ s St^'p fjl PHOENIX > IS ^ -10° •t .*';, IS -10° 180° I 170° _l 160° _J 150° I 140° I 130° I 120° I ,10° I Figure 1. — Position of longline fishing stations where spearfish data were obtained. fish. Many scientific staff members have made observations and those who measured the fish are hsted in the appendix. Some people made very special contributions: Vernon Brock, of the Divi- sion of Fish and Game, Board of Agriculture and Forestry, of the Territory of Hawaii, in addition to his many lielpful suggestions, made available to us observations on the spearfishes recorded by his division, and critically read this manuscript; Wilvan G. Van Campen, Japanese translator for POFI, brought to our attention and translated various Japanese publications on the spearfislies, which added so greatly to our knowledge of this group; and Daniel T. Yamashita and Dorothy D. Stewart most carefully brought together the ob- servations obtained on the longline cruises and assisted notably in the computations. I am also indebted to Carl L. Hubbs, James E. Morrow, Hiroshi Nakamura, Luis R. Rivas, and Robert L. Wisner, for their critical reading of the manuscript. SPECIES OF SPEARFISHES IN THE CENTRAL PACIFIC The separation and naming of tlie species of spearfishes has been a problem of particular diffi- culty, because the original descriptions of most of the species are so poor and some of the species are so similar and variable that it is impossible to identify them immediately from the original descriptions. It has been necessary for us to start witli identifications made by our fishermen, most of whom are experienced longline fishermen and have seen many marlins. We also have had the benefit of the key to Hawaiian fishes by Brock (1950), which was based on observations of the marlins landed in the Hawaiian market. The fishermen of Hawaii recognize six species of spearfishes to which they have given the English names of black marlin, silver marlin, striped marlin, Indian spearfish, sailfish, and broadbill swordfish. After seeing several hundred speci- SPEARFISHES OF THE CENTRAL PACIFIC 499 mens which included all of these spearfishes, we concur with the fishermen tliat these are six clearly distinct and easily recognizable species. All of them are fishes of tlie liigh seas, and seem to be the same in Hawaiian waters as along the Pacific Equator, where we have caught them in considerable numbers. They fit so well tlie descriptions given by Nakamura (1949) that there seems little doubt that they occur also in tlie western Pacific from Japan to Australia. Further- more, the description of the marlin fishing off Acapulco, Mexico, given by Gabrielson and La- Monte (1950) indicates that the Acapulco "black" marlin is similar to the Hawaiian "black" marlin and tliat the "silver" marlin and "striped" marlin of Hawaii also occur in Mexican waters. South of the Equator off Peru, Chile, New Zealand, and Australia, there appear to be two common species of marlin: a "striped" marlin comparable to the "striped" of Hawaii, and a "black" marlin, similar to the one called "silver" in Hawaii and "white" in Japan. A third marlin, similar to the "black" marlin of Hawaii and Japan has been described from New Zealand by Griffin (1927) and from Australia by Whitley (1954), but ap- parently it is not as common as the other marlins in the Southern Hemisphere. In the discussion to follow, a single common name will be used for eacli species to avoid confusion. The following key ' is based on the subsequent analysis of characters, distribution, and synonomy. Line drawings of spearfishes of different sizes which will aid in identifications are given in figures 2 and 3. KEY TO THE SPEARFISHES OF THE CENTRAL PACIFIC la. Snout broad, flattened and long, pelvic fins absent, one pair of keels on caudal peduncle Xiphiidae. Broadbill Swordflsh Xiphias gtadius Linnaeus. lb. Snout shorter, nearly circular in cross section, pelvic fins present, two pairs of keels on caudal peduncle Istiophoridae 2 2a. First dorsal fin very high with middle rays longest, about as long as head. Sailflsh Istiophorus orienlalis (Temminck and Schlegel). 2b. First dorsal fin moderate with anterior rays longest, middle rays much shorter than head 3 3a. Snout short, tip to anterior edge of orbit about equal to length of mandible. Body slender; greatest depth less than 13 percent of fork length. Not striped on sides. Rarely weighs more than 100 pounds. Shortxose Spearfish Tetraplurus angustirostris Tanaka. 3b. Snout longer, tip to anterior edge of orbit more than 1.3 times length of mandible. Body stouter, greatest depth more than 13 percent of fork length. Striped or not on sides. Commonly weighs more than 100 pounds .... 4 4a. Pectoral fin rigid, cannot be folded flat against side. Height of first dorsal less than 80 percent of greatest body depth, averaging about 60 percent. Pelvic fins 18 to 31 cm., average 26 cm. in fish over 150 pounds. Rarely striped on sides; stripes never conspicuous after death. Black Marlin ' Istiompax marlina (Jordan and Hill). 4b. Pectoral fin turns and folds flat against side. Height of first dorsal usually more than 70 percent of greatest body depth. Pelvic fins 22-42 cm., average about 33 cm. in fish over 30 pounds. Stripes on sides usually visible for a few hours after death 5 5a. Height of first dorsal lobe less, usually much less, than greatest body depth. Height of first anal fin more than 76 percent height of first dorsal, average 86 percent. Height of 20th ray of first dorsal 3-9 cm., average 6 cm. above fin sheath in fish more than 2 m. fork length. Body stouter, more cylindrical. Stripes usually present, but seldom conspicuous after death. Pacific Blue Marlin • Makaira ampla (Poey). 5b. Height of first dorsal lobe more than 90 percent of greatest body depth. Height of first anal fin less than 76 percent of height of first dorsal, average 66 percent. Height of 20th ray of first dorsal 7-14 cm., average 10 cm. above fin sheath in fish more than 2 ni. fork length. Body more slender, compressed, and tapered. Stripes usually conspicuous after death. Striped Marlin Makaira audax (Philippi). ' Refer also to the complete discussions referring to specimens weiehlng less than 50 pounds. ' Wliite marlin of Japan, silver marlin of Hawaii, black marlin of South America, Australia, and New Zealand. ' Blue marlin of .Atlantic Ocean, blaclt marlin of Hawaii an WILDLIFE SERVICE I I I I I L 2 _i i_] lJ I I I I I I I I METERS Figure 2. — The body proportions at 50 pounds of (a) Tet- raplurus angustirosfris, (b) Makaira audai, and (c) Mak- aira ampla; and at 200 pounds of (d) Isiiompax marlina, (e) Makaira andax, and (f) Makaira ampla. ANALYSIS OF DIAGNOSTIC CHARACTERS It is obvious from an examination of the literature on marlins and from study of a few specimens that a proper designation of the species can be made only after a suitable account of the variation in diagnostic characters. All too fre- quently casts or photographs of single specimens have been used to describe new species and sub- species. The danger of such a practice has been shown by Conrad and LaMonte (1937) and Gregory and Conrad (1939), who measured numerous specimens of three species from re- stricted localities and found marked variation in body proportions in each species. Furthermore, since Shapiro (1938) and Morrow (1952a) demon- strated marked changes in certain proportions due to allometric growth, it is dangerous to use ratios to describe the size of body parts. Of the spearfishes, the marlins are the species of most concern, and the numerous authors who have considered them have tried to recognize their differences with a great variety of external characters. These characters have included the proportions of the head with its unique sword, body proportions, length or height of certain fins, character of the lateral line, color patterns, and in a few instances, the number of rays in certain fins. Also, it has been observed repeatedly that the pectoral fin of certain marlins cannot be folded against the body, whereas the pectorals of other marlins fold readily. The work of Nakamura (1938) has shown that considerable differences in bone structure account for this variation in flexibility. SOURCE OF THE DATA There is now available a considerable amount of material for morphological comparison which includes the 12 sets of measurements of IsHompax marlina and the 30 of Makaira audax. from New Zealand and Australian waters recorded by Gregory and Conrad (1939); also the 23 sets of measurements of the Atlantic blue marlin, Ma- kaira ampla, obtained at Bimini, Bahama Islands, in July 1937 and reported by Conrad and LaMonte (1937). Morrow (1952a) gave a few measure- ments for 49 audax from New Zealand. From our POFI collection, we have measurements of 11 marlina, 68 ampla, 25 audax, 6 hliophorus orien- talis, and 8 Teirapturus angustirostris (appendix tables 1-A to I-E, p. 541). Almost all of these spearfishes are from the central equatorial Pacific waters. In addition, Vernon Brock of the Ha- waiian Division of Fish and Game (DFG) has made available to us certain measurements from 5 marlina, 27 ampla, 30 audax, and 2 angustirostris (appendix tables 2-A to2-D, p. 548), obtained from SPEARFISHES OF THE CENTRAL PACIFIC 501 0 12 3 1 I I I I I I I I I I I I I I I I I I I I I I I I METERS Figure 3. — The body proportions at 800 pounds of (a) Istiompax marlina and (b) Makaira ampla. fish landed at the Honolulu market and which, undoubtedly, were caught within 200 miles of the Hawaiian Islands. From many of the POFI specimens we obtained notes on food and sexual condition, which are summarized in the discussions under the species. Also, for several specimens not listed in the ap- pendix, length and weight data were obtained which have been used together with listed observa- tions to compute the length-weight relation. Considerable material on the weight of spear- fishes landed at tlie Honohdu auction market has also been made available by the Hawaiian Division of Fish and Game. This consists of weights of individual fish identified and recorded by dealers who allowed their records to be copied. These data show the range of sizes, seasonal trends, and modal sizes landed in Honolulu. These weights are slightly less than live weights, however, because the swords, pectorals in marlina, and sometimes the lobes of the tail are removed before delivery to the market. Also large fish are frequently cut in two or more pieces so that they have lost body fluids. All measurements taken by POFI and by the Hawaiian Division of Fish and Game have been obtained with sliding calipers read to the nearest millimeter. All measurements are the shortest straight line between the points specified. No attempt was made to obtain offset measure- ments parallel to the midline of the body. The fish to be measured were laid on their sides in as natm-al a position as possible with the jaws shut and with the snout propped up so that the sword was an extension of the midline of the body. The POFI measirrements were taken by people accustomed to measuring tunas according to the methods of Marr and Schaefer (1949). Where applicable, the same methods were followed in measuring the spearfishes, but certain morpho- logical differences required special definition. The orbit was measured instead of the uis and it was measured parallel to the midline of tlie body. The depth of the head was measured from the supra- occipital (which may be felt easily) to the throat on a line perpendicular to the midline of the body. The heights of the first anal and fiist dorsal fins were measured from the top of tlic fin sheath, and the posterior end of the fin was considered to be the end of the fin groove. The lengtli of the mandible was measured from the tip to the posterior end of the mandibular bone at the joint. 502 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE which can be found easily by moving the jaw. The body width was usually measured when the fisli was balanced on its belly but occasionally when the fish was on its side. Having in mind the difficulty in sexing Xiphias gladius reported by LaMonte and Marcy (1941), we expected that the marlins also might be troublesome. We have, however, encountered large numbers of marlins in which the eggs or milt were unmistakable, and on examination of the mature gonads of these fish we found difl^erences that make it possible to determine the sex with assurance. The most obvious difference is the presence of a firm, connective-tissue sheath around the ovary that is lacking in the testis. The in- active testes superficially resemble the fatty tissue of mammals. They are usually approximately cylindrical, but when bent can be seen to be dis- tinctly lobed and without a sheath. On the other hand, the inactive ovaries are also roughly cjdindrical but have a definite sheath and no evi- dence of lobes. When an ovary is cut, the interior is usually orange in color and appears distinctly granular to the naked eye due to ova in early stages of development. We have noticed no ex- ternal sexual differences, except that in marlina and ampla all specimens of more than 322 pounds have been females. DETERMINATION OF ALLOMETRIC GROWTH In view of the known allometric growth ^ in some parts of marlins it is desirable to examine each diagnostic measurement to determine if allometry exists. If so, it will be feasible to com- pare samples only at specified body sizes, which usually is done from regression equations. If the growth is isometric we can use ratios. In addi- tion, it will be shown that the size of certain parts is completely unrelated to the size of fish (within the range of fish sizes studied) and that it is pos- sible to compare samples by use of the simple length frequency and mean size. ' We follow wlmt we believe to be the intent of Huxley and Teissier (1936), who proposed that allometry be used in place of other terms to denote growth of a part at a rate difTerent from that of the whole. This they defined to he the case where the relative growth could be expressed by a formula of the type y = bi'- with ap^l, in which y is the part, i the .standard or whole, and a and h are constants. When a = l, growth would be considered to be isometric. We have used a growth equation of the type !i=a+hx, and have considered growth to be allometric when aj^n, and the ratio of part to whole changes with size of the whole. When a=0, the ratio is constant and the growth is considered to be isometric. This is consistent with the proposal of Huxley and Teissier because, if a^^O and the line is extrapolated from the data to the zero point, a curve results, and if the formula j/=6r» is applied, then a^\. A determination of allometric growth suffi- ciently accurate for our purposes can be had from a plot of each character on graph paper. When the points are in place, it is a simple matter to fit by eye a trend line (curved if need be) and then draw two other lines from the origin representing constant ratios near the upper and lower bound- aries of the distribution. It is convenient if the boundary lines are drawn to represent even per- centages of the abscissal character. Now, if growth is isometric the trend line will be straight, pass through the origin, and approximately bisect the angle of the outer lines. If growth is not isometric, the trend line will curve or cross one or both of the outer lines and it is possible to judge approximately how much the ratio changes over the range of the data. In the marlin data, we found it easy to judge when the trend line changed over the range of the data more than about one- third of the difference between the boundary lines. When the change was greater we used straight-line regression analysis. Wlien the trend line was curved we omitted part of the data and used only that from the straight portion. Such approximations are adequate for our pur- poses for two reasons: (1) We are concerned here principally with differences among species and not the minutiae of racial or subspecific differences, and (2) some of the marlin measurements show curvilinear relationships which our samples arc not adequate to describe precisely and which cannot be dealt with easily through the loga- rithmic growth equation. An example of the method is the plotting of the length of the pectoral fin against the fork length, using the data from the POFI collections (fig. 4). Use of this character is appropriate because MoiTow (1952a) found a slight, although not statistically significant, negative allometry in this character. We notice in our plot which includes small specimens of audax and ampla that the growth is probably curvilinear in both of these species. But if we omit the specimens of less than 200 cm. fork length, the evidence of allo- metric growth is very small indeed. There is a suggestion that the length of the pectoral in audax increases or shows a slight positive allom- etry (contrary to Morrow's finding), whereas in ampla and marlina the allometry appears to be trivial. However, if we omit the small speci- mens, the trend in any one species changes only SPEARFISHES OF THE CENTRAL PACIFIC 503 80 70 60 50 o < q: O H a uj 40 I en 30 20 ■• MAKAIRA AMPLA • o MAKAIRA AUDAX -i ISTIOMPAX MARLINA o TETRAPTURUS ANGUSTIROSTRIS 200 FORK LENGTH (CM.) FiniRE 4. — Relation of length of pectoral fin to fork length. (Measurements by POFI have been supplemented by measurements of the Hawaiian Division of Fish and Game (DFG) on specimens between 150 and 200 centimeter fork length.) about one-fourth of tlic spread of the distribution. Therefore, we consider that it is satisfactory to compare pectoral fins by using the ratio, or percentage, of fin length to fork length for speci- mens of more than 200 cm. fork length. (Figiure 4 demonstrates, however, that this character is of no value for separation of species.) An example, in which a considerable amount of allometry is to be found, is that of the greatest depth of body plotted against fork length, again from the data collected by POFI (fig. 5). There is quite obviously a considerable positive allom- etry in atulax and amjda — which is as expected from the observations that these species tend to become more humpbacked in the larger iiuli- viduals. We, therefore, conclude tliat if we use the relative depth of the body we must use regression analysis. Straight-line regressions are satisfactory, for there is no visible curvUineaiuty within the range of oiu- data. Another obvious 435062 O— 58 2 conclusion is that other measurements may be compared to the depth of the body in a simple ratio only if they happen to grow proportionately to it. Using the graphic technique, we have decided that the following body-part relationships are sufficiently isometric over the range of our samples to permit the use of simple ratios for comparing species: (1) Tip of the snout to the anterior edge of the orbit in relation to the length of the head; (2) height of the anterior lobe of the first dorsal to fork length; (3) length of pectoral to fork length; (4) caudal spread to fork length; and (5) height of the anterior lobe of the first anal to the height of the anterior lobe of the first dorsal. It is necessary to use regression analysis for the relation between the greatest depth of the body and the fork length, the head length and the fork length, the height of the anterior lobe of the first dorsal and the greatest body depth, and 504 80 70 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE O Q. LlI Q 60 - >- Q O 50 CD 40 - 30 - (f) S 20 (T 10 - 1 1 1 1 20% y^ o MAKAIRA AUDAX /-'^' yK — i ISTIOMPAX MARLINA o TETRAPTURUS ANGUSTIROSTRIS X>'^'^ — - Xlo\ / / 13% _ — XM%^^ — X , Vi'V*^'^^^^^ BOUNDARY LINES | - y jy°y° ^^-""'^ / y^ yj^^^ REGRESSION LINES — y^ ^y^^<^ ° TO P 0 F 1 DATA ^ -^ 1 1 1 1 100 200 FORK LENGTH (CM.) 300 400 Figure 5. — Relation of greatest depth of body to fork length. (Measurements by POFI supplemented by measure- ments of the Hawaiian Division of Fish and Game (DFG) on specimens between 150 and 200 centimeter fork length.) tlie length of tlie mandible and the length of the snout, measuring the snout from its tip to the anterior edge of the orbit. Another method must be used to compare the lengths of the pelvic fins (fig. 6). We find no relation between the length of the pelvic fin and the length of the fish, even in the case of the POFI data on ampla with specimens ranging from 28 to 1,002 pounds. Thus, our samples may be com- pared simply bv the mean lengths of the pelvic fin. Finally, in a comparison of the length of the 20th ray (about the middle) of the first dorsal fin (fig. 7) to the length of the fish there is clear evidence of negative growth. This ray is the longest in small (25-pound) specimens of audax, but it becomes not merely relatively but actually shorter as the fish increases in size. A similar but not so pronounced a trend is evident in ampla. We have compared samples with regard to this character by averaging the length of the 20th rays in fish over 200 cm. fork length, since the curves (fig. 7) level off above this size. COMPARISON OF DATA The type of growth will determine how the data may be compared. In the case both of isometric growth, where we have used ratios, and of charac- ters not related to total length, which can be com- pared on the basis of mean lengths, we shall use the graphical method described by Hubbs and Hubbs (1953). This consists of plotting the mean, one standard deviation on either side of the mean, and the range of the observations. We will not use the additional feature of plotting two standard errors on either side of the mean because we shall not be concerned with tests of significance. On the other hand, the characters exhibiting allometric growth will require the use of regres- sion analysis as discussed by Marr (1955). From the regression equations we will compute the mean size of a character for given sizes of fish and the SPEARFISHES OF THE CENTRAL PACIFIC 505 60 — 1 1 1 • MAKAIRA AMPLA ^,^ o MAKAIRA AUDAX 5 50 o — a ISTIOMPAX MARLINA — o > 40 _J — • O — LU Q. •. . • ••• <6 • o • fe30 I H O — • o a A o A • • • • — g 20 — i — _i 10 — 1 100 300 200 FORK LENGTH (CM.) Figure 6. — Relation of length of pelvic fin to fork length. (Measurements by POFI.) 400 20 < CO q: o o I- cn >- < o OJ X o • MAKAIRA AMPLA o MAKAIRA AUDAX fi ISTIOMPAX MARLINA 10 EYE FITTED TREND LINES 0 100 300 200 FORK LENGTH (CM.) Figure 7. — Relation of length of 20th ray of first dorsal fin to fork length. (Measurements by POFI and DFG.) 400 506 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE standard deviation from regression. These will be substituted for the mean and standard devia- tion in the graphical method of Hubbs and Hubbs. (The range around a point on the regression line is usually not available.) Unfortunately, some samples are so small and the allometric growth so marked that it is necessary to consider some characters at only a single size and others merely from the plotted points on the graph. CONVERSION OF LENGTHS Nearly all of the measurements must be con- sidered in relation to other measurements. The best standard is usually length of the fish, but here a difficulty arises. Conrad and LaMonte (1937), Gregory and Conrad (1939), and Morrow (1952a) used body length, measured from the snout to the base of the tail (standard length). Brock, who measured the fish in the Hawaiian market where the snouts are almost always cut ofiF, measured the body length from the naris to the fork of the tail. Measurement from the posterior edge of the orbit to the fork of the tail has been commonly used by Japanese scientists. Thus, a preliminary requirement for examining the characters is to be able to convert from one length to another. We have done this by regres- sion analysis for the three species of marlins, audax, marlina, and ampla, on the basis of POFI measurements. In each case we converted the measurement given to fork length, which is defined as the straight-line distance from the tip of the snout to the tip of the center rays of the tail. These conversions have been made from regres- sion equations (appendix tables 3-A to 3-E, p. 550) on the assumption that straight-line relationships exist between the length measurements. Plots of all measurements for each species have sub- stantiated this assumption. CHARACTERS Weight The general tendency for certain species of the marlins to look heavier than others suggested that it might be possible to separate tlie species on the basis of the length-weight relation. Nichols and LaMonte (1941) attempted this for the Pacific marlins and they stated that for a given length their striped marlin (audax) tended to weigh the least, their silver marlin (marlina) more, and their black marlin (ampla) most. When the relation is plotted (fig. 8) for the POFI measurements from the central Pacific,' it is obvious that audax weighs less than the other two which are much alike, and that the length-weight relation might indeed be useful for distinguishing individuals of less than 150 pounds. At lengths of about 300 cm. and weights of around 300 pounds, however, there is a great deal of overlap, as the weight of audax for a given length then approaches that of marlina and ampla. In the larger sizes, all three species are so alike that it is impossible to distinguish individuals on the basis of the length-weight relation. A comparison of POFI data with DFG material and the published data (Gregory and Conrad, 1939; Conrad and LaMonte, 1937; Morrow 1952a) in figure 9 shows that audax from all areas is lighter at a given length than the other two species. There is, however, a slightly greater overlap between species at the 300-cm. size, especially for the POFI material in which the specimens of audax were slightly heavier at a given length than were those from the other areas. Greatest body depth When this measurement is plotted against fork length a marked positive allometry is obvious (fig. 5) . Both figures 2 and 1 0, in which all samples are compared for given lengths, show that marlina is deepest bodied, ampla intermediate, and audax the most slender, but there is considerable overlap between the species. The species marlina and audax usually can be separated on the basis of body depth, but ampla cannot clearly be distinguished from either. Thus, the character is of httle value for taxonomic purposes. Within each species there is quite close agreement of the means; and the relative position of the means is almost the same as the mean weights of figure 9, which indicates that the local populations that are heavier for a giveiv length are also deeper bodied. Head length Head length has not been used to separate the species of marlins, but Gregory and Conrad (1939, fig. 1) showed that ampla has a mean head length of 36 percent of the body length, whereas this ratio in audax is about 39 and in marlina about 38. Such a difference suggests some possibility of separating the species with this character, and also because most head parts are compared with head length, it is desirable to examine our data for allometric • The data used for this graph include a few specimens not listed in the appendix. SPEARFISHES OF THE CENTRAL PACIFIC 507 1000 500 C/) Q Z) 2 100 X IxJ 50 10 • MAKAIRA AMPLA o MAKAIRA AUDAX A ISTIOMPAX MARLINA o TETRAPTURUS ANGUSTIROSTRIS * ISTIOPHORUS ORIENTALIS 2 3 4 5 LENGTH (METERS) Figure 8. — Length-weight relations. \o regression line has been computed for the few observations on T. angustiroslris. (Measurements by POFI.) 508 FISHERY BULLETEST OF THE FISH AND WILDLIFE SERVICE ISTIOMPAX MARLINA NEW ZEALAND -AUSTRALIA 1 1 1 1 1 1 1 1 (GREGORY a CONRAD) N = 12 CENTRAL PACIFIC 1 V- ^ 1 1 (PO F 1 ) N = 6 HAWAII 1 1 1 (DFG) N=5 MAKAIRA AUDAX NEW ZEALAND -AUSTRALIA (GREGORY a CONRAD) N= 27 NEW ZEALAND \ 1 1 1 n 1 (MORROW) N = 48 CENTRAL PACIFIC 1 1 1 1 1 ( PO Fl) N = 13 HAWAII 1 1 (DFG) N = 30 MAKAIRA AM PL A BIMINI (CONRAD a LAMONTE) N = 23 CENTRAL PACIFIC 1 1 1 1 1 1 1 1 1 1 (P 0 F 1 ) N = 56 HAWAII 1 1 1 1 1 1 ( DFG ) N = 27 100 200 400 500 Figure 9. 300 WEIGHT ( POUNDS) -Mean and standard deviation from regression of the weight of marlins at 250 cm. (left) and 300 cm. (right) fork lengths. (Names in parentheses indicate source of data in the literature.) growth. In the POFI data, head length plotted against fork length shows slight positive allometry in amj)la and slight negative allometry in audax. The condition in marlina is intermediate, but too few measurements are available to be conclusive. Therefore, regression methods are indicated for all species. When we compare the POFI data with those published by Gregory and Conrad (1939) and Conrad and LaMonte (1937), we find good agree- ment between samples of the same species except that marlina from the central Pacific have some- what longer heads than from the New Zealand- Australia sample (fig. 11). However, the number of samples is so small and the overlap is so great that we consider this difference to be only racial. The differences between species, too, are so slight that the character is almost useless for diagnostic purposes. Length of snout Length of snout from front of orbit was used by Jordan and Evermann (1926) as well as by Nichols and LaMonte (1941) in an attempt to separate these species of fish, no doubt because of the gen- eral impression that marlina has the shorter and stouter spear and audax and ampla the longer and slenderer ones. When snout length was compared with head length we found no evidence of allo- metric growth; hence, we can compare snout lengths by simple ratio. When this is done (fig. 12) for the published data and the POFI data we find that appearances as to snout length are misleading, for all samples of all three species show remarkably similar ratios with the overlap among species and between samples almost complete in all cases. Spear stoutness was not investigated because of the small amount of data. Also, measuring the breadth and width at the tip of the mandible, as we did, is not satisfactory because of the allometric growth of tlie mandible in ampla (see next section). Length of mandible When this character is plotted from our POFI measurements (fig. 13), we find a strikingly SPEARFISHES OF THE CENTRAL PACIFIC 509 ISTIOMPAX MARLINA NEW ZEALAND- AUSTRALIA 1 1 1 1 1 (GREGORY a CONRAD) N=I2 CENTRAL PACIFIC 1 1 1 1 (P 0 F 1) N= 7 HAWAII 1 1 1 (D FG) N= 5 MAKAIRA AUDAX NEW ZEALAND -AUSTRALIA 1 1 1 III (GREGORY a CONRAD) N = 30 NEW ZEALAND 1 1 (MORROW) N = 46 CENTRAL PACIFIC 1 1 1 1 1 1 ( P 0 F 1) N = 21 HAWAII 1 1 (D FG) N = 28 MAKAIRA AM PL A BIMINI 1 1 1 1 1 1 (CONRAD a LAMONTE) N = 23 CENTRAL PACIFIC 1 1 1 1 1 1 1 (P 0 F 1 ) N = 6! HAWAII 1 1 1 1 (D FG) N = 27 30 70 40 50 60 GREATEST BODY DEPTH (CM.) Figure 10. — Mean and standard deviation from regression of the greatest body aepth of marlins at 250 cm. (left) and 300 cm. (right) fork lengths. ISTIOMPAX MARLINA NEW ZEALAND- AUSTRALIA 1 1 1 1 1 (GREGORY a CONRAD) N = 12 CENTRAL PACIFIC f 1 1 1 (P 0 F 1) N = 9 MAKAIRA AUDAX NEW ZEALAND- AUSTRALIA 1 1 [ 1 1 1 (GREGORY a CONRAD) N = 30 CENTRAL PACIFIC 1 1 1 1 1 1 (PO F 1 ) N = 20 MAKAIRA AMPLA BIMINI 1 1 1 1 1 (CONRAD a LAMONTE) N = 23 CENTRAL PACIFIC 1 , 1 i 1 1 1 ( P 0 F 1 ) N = 58 80 90 120 100 110 HEAD LENGTH (CM.) FiciiTRE U. — Mean and standard deviation from regression of the liead length of marlins at fork lengths of 250 cm. (left) and 300 cm. (right). 510 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE ISTIOMPAX MARL /N A NEW ZEALAND-AUSTRALIA 1 1 (GREGORY a CONRAD) N = 12 CENTRAL PACIFIC 1 1 1 ( P 0 F 1) N= 7 MAKAIRA AUDAX NEW ZEALAND-AUSTRALIA 1 1 (GREGORY a CONRAD) N = 30 CENTRAL PACIFIC 1 1 1 (PO F 1 ) N = 9 MAKAIRA AM PL A BIMINI (CONRAD a LAMONTE) N=23 CENTRAL PACIFIC 1 1 1 1 1 1 1 ( P 0 F 1 ) N = 26 55 60 65 70 SNOUT LENGTH AS PERCENT OF HEAD LENGTH Figure 12. — Mean, standard deviation, and range of ratio of snout length to head length. 75 70 60 - 50 o CD O 40 30 O 20 10 — 1 1 1 1 1 1 1 1 1 70% — . a h/l Al< A ir? A Ah/lPI A ^y^ : o MAKAIRA AUDAX a ISTIOMPAX MARLINA ° TETRAPTURUS ANGUSTIROSTRIS , 507o~ ^Z^^;^:!-/^^^,--^ REGRESSION LINES - ^""^^^^^^^^^ - ^ 1 1 1 1 1 1 1 1 1 — 10 20 30 70 80 90 40 50 60 SNOUT TO ORBIT (CM.) Figure 13. — Relation of length of mandible (to the joint) to snout from tip to orbit. (Measurements by POFL) 100 different type of growth in ampla than in the other two species. The mandible of ampla tends to become markedly shorter in relation to the snout as the fish grows, whereas in the other two species the gi'owth is nearly isometric. A similar relation is apparent when regression lines are fitted to the published data (fig. 14) of Gregory and Conrad (1939) and Conrad and LaMonte (1937). Their data cover a much smaller range than the POFI data but the same SPEARFISHES OF THE CENTRAL PACIFIC 511 60 < O 40 - 60% MAKAIRA MAKAIRA AMPLA AUDAX ■- a ISTIOMPAX MARLINA Figure 14. 40 50 60 SNOUT TO ORBIT (CM.) -Relation of length of mandible (tip to angle of jaw) to snout from tip to orbit, and Conrad, 1939, and Conrad and LaMonte, 1937.) 100 (Measurements from Gregory divergence among species is apparant; audax and marlina show sliglitly positive allometric growtli of the mandible in relation to the snout, whereas ampla shows a slightly negative allometric growth. Unfortunately, the POFI measurements of the mandible (to the joint) are not comparable to the measurement used h\ these authors, so compari- sons between areas are not possible. Obviously, here is a character that is useless for separating the species among the intermediate sizes, but the divergence among the very large specimens suggests that, in them, it may be useful for distinguishing ampla from marlina. The length of the mandible to the angle of the jaw, as measured by Gregory and Conrad and by Conrad and LaMonte, is preferred to the measurement used by POFI; also, it may be measured with considerable precision from photographs. The plots of the published data suggest that specimens of more than about 600 pounds in which the mandible is more than 48 percent of the snout (that is, goes into the snout less than 2.1 times) will be marlina, whereas those in which the length of the mandible is less than 48 percent of the snout should probably be considered to be ampla. If we apply this criterion to the type photograph of marlina (Jordan and Evormann, 1920: pi. 17; which weighed only 509 pounds), and to all of 435062 O— 58 3 Farrington's (1953) photographs of black marlin of more than 600 pounds in which the characters can be measured, we find that the length of the mandible is contained in the snout 1.5 to 1.9 times, with an average of 1.76. On the other hand, in the photographs of ampla of more than 400 pounds, shown by Farrington (1937), the length of the mandible is contained in the snout from 1.9 to 2.4 times, with an average of 2.09. Here is a char- acter that may well be useful in distinguishing ampla from marlina, when the unequivocal character of the pectoral fin has not been recorded ; but additional measurements of large specimens are needed to establish the difference. Clearly, too, this difference in the lower jaw is the reason for the apparent differences that have been observed in the snout. When the lower jaw is very short, as in large ampla, the snout seems extremely long and slender, whereas the snout seems shorter when the lower jaw is long, as in marlina. Length of pelvic fin In our previous discussion of allometric growth, we pointed out tliat there was almost no change in the length of the pelvic fin witii size of the fish in any of the three species examined by POFI. Consequently, we may compare these on 512 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE the basis of the average length of the fin and disregard the size of the fish. We find good agreement between the samples of the same species, but marlina has markedly shorter fins on the average than either audax or ampla (fig. 15). Of the 19 measurements available . for marlina, the average is approximately 26 cm. and only 1 measurement is more than 30 cm. This is in contrast with the other two species in which the pelvic fins average about 33 cm. and in which we find only 19 out of the 95 measure- ments less than 30 cm. In most of the samples, the range extends farther from the mean on the lower side than on the upper and we suspect that some of the smaller measurements may be due to broken fins. If a careful watch is kept for broken fins, this character may then be useful to separate marlina from the other two species when other characters are not available. Any marlins with pelvic fins longer than 30 cm. are probably not marlina. Length of pectoral fin Length of pectoral fin also was discussed in the section on allometric growth and it was pointed out that while small specimens appeared to have slightly smaller pectoral fins in relation to fork length, specimens of more than 200 cm. fork length had pectoral fins which grew almost isometrically. When pectoral fins are compared (fig. 16), it is apparent that they show almost as much variation within species as between species and that the character is useless for distinguishing one species from the other. The means vary from only 18.2 percent in ampla from Hawaii to 19.4 in audax from New Zealand and Australia. Height of first dorsal fin Heiglit of the first dorsal fin appears to be one of the best means of distinguishing the three species of marlins. Nichols and LaMonte (1941) com- pared tlie anterior lobe witli head length, Jordan and Evermann (1926) usually compared it with the length of the pectoral fin, and Nakamura (1949) with the greatest depth of the body. When we plotted height of the first dorsal in relation to fork length, we found a negligible amount of allo- metric growth and, hence, we can use it as a ratio. The comparison of all samples (fig. 17) shows that marlina has the lowest fin, ampla intermediate, and audax the highest. The averages are approxi- mately 12, 13.5, and 17 percent, respectively; however, the separation between species is not complete, as there is considerable overlap between audax and ampla and between ampla and marlina. The samples show close agreement within species except for marlina, in which there is a suggestion of a clinal difference. The specimens from New Zealand and Australia have the highest first ISTIOMPAX MARLINA NEW ZEALAND -AUSTRALIA (GREGORY a CONRAD) N= If CENTRAL PACIFIC 1 1 1 1 1 1 1 (P 0 F 1 ) N = 9 MAKAIRA AUDAX NEW ZEALAND -AUSTRALIA (GREGORY a CONRAD) N = 30 CENTRAL PACIFIC 1 1 1 1 1 1 1 1 (P 0 F 1) N =12 MAKAIRA AMPLA BIMINI 1 1 (CONRAD a LAMONTE) N=2J > 1 1 CENTRAL PACIFIC 1 1 ( P 0 F 1) N = 33 20 25 30 35 PELVIC LENGTH (CM.) 40 45 FiouRB 15. — Mean, standard deviation, and range of the length of the pelvic fin. SPEARFISHES OF THE CENTRAL PACIFIC 513 ■ \ ISTIOMPAX MAR LIN. 4 ALIA NEW ZEALAND- AUSTR 1 1 (GREGORY a CONRAD) N= 12 CENTRAL PACIFIC 1 1 ( P 0 F 1 ) N = II MAKAIRA AUDAX NEW ZEALAND- AUSTRALIA t i (GREGORY a CONRAD) t M=30 NEW ZEALAND 1 1 (MORROW) N = 47 CENTRAL PACIFIC 1 1 1 ( P 0 F 1) N = 19 HAWAII 1 1 (D F G) N = 25 MAKAIRA AM PL A BIMINI 1 1 (CONRAD a LAMONTE ) N = 23 CENTRAL PACIFIC 1 1 ( P 0 F 1) N = 57 HAWAII 1 1 1 1 ( D F G ) N = 26 10 15 20 PECTORAL LENGTH AS PERCENT OF FORK LENGTH 25 Figure 16. — Mean, standard deviation, and range of the ratio of length of pectoral fin to fork length in fish of more than 200 cm. fork length. dorsal fins, those from the equatorial Pacific- lower fins, and those from Hawaii the lowest. When height of the first dorsal fin is compared with the greatest depth of the body we find a marked allometric relationship (fig. 18). We found no ampla in which the height of the first dorsal was greater than the greatest body depth and only one audax in which the height of the first dorsal was less than 90 percent of the greatest body depth. The trend lines are such, however, that it is evident that in very small ampla the first dorsal may exceed the greatest body depth and in very large audax may be less than 90 percent. In any event, there is a considerable overlap of specimens in which the anterior lobe is between 90 and 100 percent of the greatest bod> depth. The difficulty presented by allometric growth and most of the overlap between ampla and audax is eliminated if, instead of comparing the height of the first dorsal with the greatest deptli of the body, we compare it witii the heigiit of the first anal (fig. 19). Plots of this relationship indicated no allometric growth and hence the comparison by ratios is valid. This comparison shows that the anal fin in audax averages 66 percent of tlie height of the first dorsal, whereas in ampla it averages 86 percent. If we accept 76 percent as a dividing line between the species, we find only a single overlapping specimen of audax with a greater value. If separate is not easily determined uidess they happen to be in separate fin slots. We consider the interspace between the aiials and dorsals and tlie number of free spines to be of very doubtful value as taxonomic characters. 518 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE OBSERVATIONS ON SPEARFISHES OF THE CENTRAL PACIFIC Having decided which characters are of diag- nostic value, it is now possible to consider our observations on the spearfishes of the central Pacific together with the considerable literature on the group from the several parts of this ocean. This we have done in the following discussion, with the assumption that many of the minor differences reported in body proportions will prove to be individual variation, or at most, varietal differences. During this study we were fortunate to liave ready access to Japanese literature through our translator, W. G. Van Campen. He located many papers including several which were pub- lished entirely in Japanese. Many of these papers were translated and others were summarized. Further, we corresponded extensively with Japa- nese workers on the spearfishes and feel that we quite completely covered the recent Japanese literature on the spearfishes. Xiphias gladius Linnaeus Swordfish, Broadbill Tsun, Shutome, or Mekajiki (Japan) Our catches of the swordfish have been so small that we can add little of significance; however, it seems worthwhile to discuss it here and give a few brief notes from recent Japanese publications. The truly pelagic nature of the swordfish is indicated in Kikawa's (1954) review of the Japanese fisliery. He reported that at the begin- ning of the season in late summer the highest catch rates are to be found northeast of Japan, north of FlouRE 22. — Distribution of POFI catches of swordfish, Xiphias gladius. Fractions indicate stations at which catches were reported out of the total fished; decimals indicate average catch per 100 hooks per day. SPEARFISHES OF THE CENTRAL PACIFIC 519 40° N. latitude, and between 150° and 170° E. longitude. Fishing is carried on at this time north to 45° but, with the advent of winter weather, the fishery moves south to the vicinity of 30° wlicre good fishing is found in December and January. In addition to tiiis offshore fishery there are inshore fisheries around southern Japan and the Bonin Islands, some of which are produc- tive the year round. The swordfish is generally considered to be an inhabitant of warm seas throughout the world, but its distribution in the western Pacific suggests that the adults prefer the cooler waters. Kikawa (1954) noted that they are only sporadically cap- tured in tropic seas, and this is in agreement with POFI experience (fig. 22) and with the results of the Japanese tuna mothership expeditions to the Caroline Islands area in 1950-51. In the latter. Van Campen (1952) reported that the average catch rate of swordfish for all expeditions was less than .01 per 100 hooks, whereas catch rates off northeastern Japan average nearly 1.0 per 100 hooks (Kikawa 1954). Nakamura et al. (1951) think that the tropics are the spawning grounds of the swordfish, and noted therefrom the capture of juveniles less than 30 mm. in length and numerous larvae in the stomachs of other fish. They also reported that the longline catches in the equatorial area are pre- dominantly fish from 50 to 100 cm. in length (orbit to fork of tail). In addition, all of the fish in the northern fishery have undeveloped gonads. The presence of small fish in the equatorial area is substantiated by the reports from the Japanese mothership expeditions to the Caroline Islands where, according to Ego and Otsu (1952), the weight of the swordfish captured during each of the first six expeditions ranged from 58 to 102 pounds. Such catches of large and small swordfish are in accord with the limited POFI experience, for the largest of the three taken in tropical waters on which size data are available was only 80 pounds. The other two were very small, each only 92 cm. long. We also have notes on two small swordfish taken from the stomachs of Makaira ampla: one of 35 cm. taken on May 18, 1954, at 6°02' N., 162°28' W. and another of 38 cm. taken on May 28, 1954, at 6°02' N., 159°34' W. On the other hand, the two specimens taken north of Hawaii 453062 O — 58 4 were each large, more than 300 cm. total length. Swordfish landed in the Honolulu market (table 2) ranged from 75 to 1,061 pounds, according to the records collected during 1949 and 1950 by the Hawaiian Division of Fish and Game. There was no pronounced mode in this weight distribu- tion. Additional insight into the habits of the sword- fish is available from Kikawa's (1954) account of the methods used in the Japanese fisherj'. Most of the swordfish are taken by vessels specializing in the fishery that use longline gear similar to tuna gear. The principal difference is in the mode of operation, for these swordfish vessels fish at night when the catch rate is approximately twice what it is in the daytime. Such a difference in the habits of the broadbill makes it difficult to com- pare these catches with the abundance in other parts of the Pacific, where swordfish are taken by tuna vessels that fish entirely during the day. Table 2. — Weight frequency of swordfish, Xiphias gladius, landed at the Honolulu market during 1949 and 1950 [Data collected by the Hawaiian Division of Fish and Qame] Weight group (pounds) Number weighed in— Total 1949 19S0 60-fi9_- ._ 70-79 1 1 80-89 - . .. 90-99 2 2 2 2 4 3 1 1 i 2 100-109 _ 2 110-119 2 120-129 3 130-139 . 4 140-149 3 150-159 1 160-169 1 170-179 180-189 - 1 2 2 1 2 1 "3 2 ' 3' i 1 2 190-199 2 200-219 S 220-239 210-259 2 260-279 2 280-299 1 2 4 3 1 1 300-319 5 320-3.39 4 340-359 2 360-379 3 380-399 2 400-419 1 420-439 440-l.')9 460^79 2 1 2 1 5 3 1 3 2 1 1 2 2 1 1 2 3 480-499 2 50O-,'>49 4 550-599 . 3 600-619 6 650-6(19 4 700-719 1 7.50-799 2 >800 4 Number .. 80 Maximum weight (pounds) 863 1.061 520 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Tetrapturus angustirostris Tanaka Shortnose Spearfish Furaikajiki (Japan) ; Indian Spearfish (Hawaii) Distinguishing characteristics This little spearfish is characterized by a snout that extends onlj- slightly beyond the lower jaw, a long, slender, compressed body, the greatest depth of which is less than 13 percent of the fork length, relatively short pectoral fins that are less than 14 percent of the fork length, and an easily visible, single lateral line. It lacks the stripes of most of the other species. Unfortunately, we lack data on a sufficient number of specimens of less than 30 pounds of the other species to separate them clearly from Tetrapturus. In very small Makaira ampla the snout is scarcely longer than the mandible, but the body is heavier, rounder, and the middle of the first dorsal fin is probably less than two-thirds of the height of the anterior lobe. In small M. audax, the middle of the first dorsal fin approaches the height of the anterior lobe as it does in Tetrap- turus, but audax may be distinguished by the presence of stripes and a snout markedly longer than the mandible. We have seen no very small specimens of Istiompax marliva, but presumably they may be distinguished unequivocally by the stiff pectoral fin, which in Tetrapturus is flexible. The close resemblance of the shortnose spear- fish to the young of the other marlins has led some to suspect (LaMonte and Marcy 1941:21) that it is merely a juvenile form. This view, however, was effectively disproved by the work of Naka- mura (1937) who figured the eggs, ovaries, and testes and described one ripe female taken in November which was 152 cm. in fork length and 27 pounds in weight. This fish was taken along with several others with enlarged ovaries. A female with running-ripe ovaries (specimen No. 5 in appendix table 1-A, p. 541) that we captured on March 18, 1954, was 164 cm. fork length. It was not weighed but the weights of other specimens of similar length suggest that its weight should have been about 40 pounds. Such sizes are far below the sizes at which the other species commonly occur and appear to mature. Jordan and Evermann (1926) listed four Pacific and one Indian Ocean species of this genus, but it appears probable that there is but a single species in this whole area. Two of the species, ectenes and bredrostris, obviously do not belong to the genus Tetrapturus. Two other species, illingworthi and kraussi, were described as new from Hawaii and were separated from the Japanese species angus- tirostris because the latter was described as having a dorsal lobe longer than the pectoral. In the 9 specimens from the central Pacific on which we have these measurements we find 8 in which the dorsal lobe is very slightly longer than the pec- toral, and 1 (from Hawaii) in which the opposite is true but the variation in these two characters is such that this comparison of fins is not a good specific character. These authors also distin- guished illingworthi and kraussi on the basis of the separation of the dorsal fins — a character we find highly variable in our specimens. We, therefore, place both species in synonymy with angustirostris. There appears to be no valid reason to retain the genus Pseudohistiophorus as proposed by De Buen (1950:170-171). He established this genus because the previous attempts to place hetero- geneous species in Tetrapturus suggested to him that Tetrapturus was the synonym of Makaira. We cannot accept this view because the redescription of Tetrapturus helone Rafinesque given by Cuvier and Valenciennes (1831:205 — the earliest descrip- tion available to us) is excellent and obviously represents a species extremely close to, if not iden- tical with, our Pacific species. If Tetrapturus should prove to be monotypic, the species name helone described from a Mediter- ranean specimen would have priority. Jordan and Evermann (1926) separated it on the basis of the short, stiff spines in the interspace between the dorsal and anal fins. We have seen none of these spines in the Pacific form, but in some specimens of M. ampla we have noticed that the first dorsal fin may continue almost to the second dorsal or may be broken up into separate spines, sometimes even in separate fin slots. This appears to be a matter of individual variation, and further ex- amination of the species from several areas is needed to determine whether this is a constant character. Color Immediately after death T. angustirostris is a brilliant, deep metallic blue on the back and first dorsal with silvery gray on the sides and white on the belly. In about an hour this rapidly fades SPEARFISHES OF THE CENTRAL PACIFIC 521 to a dark, slate gray on the back and to black on the first dorsal. We have seen no evidence of stripes and, according to Nakamura (1949), it never has them. Distribution in the Pacific According to Nakamura (1951), this pelagic species does not enter coastal or enclosed seas. Off Japan it occm-s south of 35° N. latitude and rather densely in the waters east of Formosa and the Philippines from November to January. Nowhere is it abundant enough to be of impor- tance to the fishery. In our POFI fishing we have taken only the 8 specimens recorded in appendix table lA; their distribution is indicated in figure 23. In the Hawaiian fishery it is one of the miscellaneous spearfishes that comprises onlj' a small fraction of the total spearfish catch. On the first six Japanese mothership expeditions to the vicinity of the Caroline Islands (Ego and Otsu, 1952) it was combined with the saiLlsh in the statistics, and on each of these trips the catch of the two species together averaged only from .02 to .07 per 100 hooks. Size This is the smallest of the spearfishes and, according to Nakamura (1949), attains a weight of only 44 pounds, but the POFI specimens which we have weighed from the central Pacific ranged from 18 to 51 pounds. Based on the data ob- tained from the Honolulu market by the Hawaiian Division of Fish and Game (table 3), the maxi- mum weight found in 177 specimens was 114 pounds. However, the modal size was approx- imately 38 pounds. FinvRE 215. — Distribution uf POFI catches of shortiio.so spearfish, Tetraplurus angusliroslris. Fractions indicate stations at which catches were reported out of the total fished; decimals indicate average catch per 100 hooks per day. 522 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Table 3. — Weight frequency of shorlnose spearjish, Tet- rapturus angustirostris, from the Honolulu market in 1961 [Data collected by the Hawaiian Division of Fish and Qame] Weight group (pounds) Number offish 15-19 20-24 --- 6 25-29 - 11 30-34 23 35-39 32 40-44 25 45-49 23 60-64 10 68-59 14 60-64 15 65-69 3 Weight group (pounds) Number offish 70-74 n 75-79 4 80-84 -- 2 85-89 1 90-94 95-99 - . - >100 2 Number Maximum (pounds) -- 'weight 177 114 Food We have notes on the contents of 6 stomachs of which 2 were empty and the other 4 contained squid. Tliree stomachs contained fish of which only bramids were identified. Synonymy and references Tetrapturus angustirostris, Tanaka, 1914:324 (Japan); Nakamura, 1937 (Formosa); Nakamura, 1938:24 (Formosa); LaMonte and Marcy, 1941:2 (Japan); Nakamura, 1942 (Formosa); Hirasaka and Naka- mura, 1947:11 (Formosa); Nakamura, 1949:56 (Ja- pan); Rosa, 1950:159 (Japan); Nakamura, 1951:35 (northwestern Pacific). Tetrapturus illingworthi, Jordan and Evermann, 1926:32, pi. 8 THawaii) ; LaMonte and Marcy, 1941:2; Brock, 1950 (Hawaii); Rosa, 1950:161 (Hawaii). Tetrapturus kraussi, Jordan and Evermann, 1926 "33, pi. 9 (Hawaii). \ Tetraptxirus brevirostris, De Beaufort and Chapman, 1951:238 (850-mm. specimenl; Fowler, 1928:136 (Hawaii). Not Histiophorus brevirostris, Playfair and Ounther, 1866:53, 145 (Indian Ocean). Not Tetrapturus ectenes, Jordan and Evermann, 1926:34, pi. 11, fig. 2 (Haw.aii). Istiophorus orientalis (Temminck and Schlegel) Sailfish Bashokajiki (Japan) Distinguishing characteristics This genus is effectively distinguished from all other spearfishes by its very high first dorsal fin. It also has a slenderer, more greatly compressed body and much longer pelvic fins. Problems of identification arise within the genus because so many species have been described. Those listed by Jordan and Evermann (1926) are differentiated mostly on the basis of the inter- space between the dorsal fins, whether or not that space has spines, the shape of the first dorsal, the color, the length of the pectoral, the length of the spear, or the relative size of the second dorsal and second anal fins. We have seen only a few sailfish, but most of these characters are so variable in the other spearfishes that they have little value for identifying species. There seems little doubt that the species occur- ring in the central Pacific should be orientalis, which most authors have used. On the basis of a cast in the Bishop Museum,' Jordan and Ball, in Jordan and Evermann (1926), also describe eriquius from Hawaii in which the first dorsal fin is subtruncated behind with only 34 dorsal spines. The photo in Jordan and Ever- mann (p. 101) suggests that the posterior part of the dorsal fin was missing from the cast. Further, there are no reports from Hawaiian fishermen of two species of sailfish. We, therefore, regard eriquius as a synonym of orientalis. Distribution in the Pacific Nakamura (1949) gave the distribution of the sailfish as extending from the northeastern coast of Japan south and noted that it is comparatively abundant in the Kinan Sea area. He also stated that this species often enters coastal waters. It is, however, widespread in the tropical Pacific. It was taken in small quantities by the Japanese mothership expeditions near the Caroline Islands in 1951 and 1952 (Ego and Otsu, 1952), and 20 specimens were taken during the POFI longline fishing, as indicated in figure 24. Some of the POFI specimens were taken many hundreds of miles from the nearest land. Spawning Spawning sailfish were taken on July 10 and 12 off Hainan Island, according to Nakamura (1940), along with several juveniles of less than 10 mm. He also reported that a spawning female caught on the hook was followed by a companion fish, presumably a male. He (1949) noted that they spawn in Formosan waters from April to August. We can add nothing to the information on spawn- ing because none of the POFI specimens examined had ripening gonads. ' Not Included In the current list of specimens In the museum. SPEARFISHES OF THE CENTRAL PACIFIC 523 Figure 24. — Distribution of POFI catches of sailfish, Isliophorus orientalis. Fractions indicate stations at which catches were reported out of the total fished; decimals indicate average catch per 100 hooks per day. Size Specimens which we liave weighed in tlie POFI catches ranged from 26 to 106 pounds. Weights of 1 1 Honolulu market specimens recorded in July and August 1950 by the Hawaiian Division of Fish and Game ranged from 25 to 114 pounds with all but 1 weighing less than 45 pounds. Xakamura (1949) stated that sailfish attain a weight of 132 pounds. Food Probably these fish are broadly carnivorous like the other spearfishes but perhaps it is significant that 8 of the 9 stomachs examined contained squid, which usually was the predominant food. The other food items included octopus, nautilus, Alepisaurus, one bramid, and one pilot fish. Synonymy and references Isliophorus (Hisliophorus) orientalis, Temminck and Schlegel, in Siebold, 1844:103, pi. 55 (Japan); Jordan and Evermann, 1926:46, pi. 15, fig. 1 (Japan); Fowler, 1928:136 (Hawaii); Nakamura, 1938:25 (Formosa); Nakamura, 1940 (South China Sea); LaMonte and Marcy, 1941:2 (Hawaii, Japan); Nakamura, 1942 (Formosa); Hirasaka and Nakamura 1947:12, pi. 1, fig. 2 (Formosa); Fowler, 1949:74 (Tahiti); Naka- mura, 1949:58 (from northea.stern Japan south); Brock, 1950:146 (Hawaii); Ro.sa, 1950:151 (western Pacific from Indonesia to Vladivostok, Hawaii); De Beaufort and Chapman, 1951:241 (Singapore, Java, Japan, Siam, Hawaii); Yabe, 1953 (Japan); Murphy and Otsu, 1954 (Caroline Islands). Isliophorus eriquius, Jordan and Ball, in Jordan and Evermann, 1926:48, pi. 15, fig. 2. Isliophorus brookei, Fowler, 1934:400 (Tahiti). Bashokajiki, sailfish, Xakamura, 1944b. 524 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Istiompax marlina (Jordan and Hill) Black Marlin Shirokajiki "White Marlin" (Japan); Silver Marlin (Hawaii) Distinguishing characteristics Nakamura (1938) has described the anatomical differences between marlina on the one hand and Makaira audax and M. ampla on the other, which differences were subsequently used by Hirasaka and Nakamura (1947) to propose the genus Marlina. The principal differences are that (1) the shoulder girdle in marlina is considerably broader than in the other species, and the articula- tion with the pectoral fin restricts its movement; (2) the pelvic girdle in marlina has the two sides fused together and difficult to separate, whereas in the other two species the two sides of the girdle are separated by a broad space and they can be easily separated; and (3) the air bladder in marlina consists of several layers of small cham- bers, whereas it has only a single layer of chambers in ampla and audax. We believe the differences noted here warrant the retention of marlina in a separate genus; however, the generic name Marlina cannot be applied to this genus. In the first place, its use is prevented by Zane Grey's introduction of the name Marlina mitsukurii in 1928. Since he used the name solely in this combination and prior to 1928 when such a proposal was permitted, mitsukurii is the haplotype of Marlina Grey. In the second place, Whitley (1931:18) proposed the genus Istiompax for /. australis, new species, recognized as a synonym of Makiara marlina Jordan and Hill. Therefore, the generic name Istiompax has precedence over Marlina Hirasaka and Nakamura (nan Grey) . The most distinctive external characters of marlina, in addition to the rigid pectoral, are the short ventral fins which range in length from 18 to 31 cm., with an average of 26 in our speci- mens, and the very low first dorsal, which in its anterior lobe averages about 60 percent of the greatest body depth, but may range from 50 to 80 percent. Many other subtle differences aid in recognizing marlina at a glance. The body seems compressed more than in ampla, and it appears markedly heavier in the pectoral region than either audax or ampla because of the larger hump on the back. Although marlina has been reported by Nakamura to differ from ampla by having a single, simple lateral line, the lateral fine is a poor field character because it is difficult to see in fresh specimens. Marlina and ampla are the only marlins that appear to surpass 1,000 pounds in weight. When near this size, marlina is readily distinguishable because the lower jaw from tip to corner of the mouth is at least half the length of the snout from tip to orbit. In ampla the lower jaw recedes with growth, and in very large specimens the snout has the appearance of being much longer and more slender than in marlina. The name marlina may lack priority if adequate descriptions of marlins from the type localities of Tetrapturus herscheli Gray (South Africa), or Histiophorus breinrostris Playfair (Zanzibar) be- come available. Gray's (1838) description of herscheli agrees well with marlina in most char- acters. In sizes estimated from his figure (pi. X), the relation of the height of the anterior lobe of the first dorsal to the fork length (13 percent), the height of the first anal to the height of the first dorsal (77 percent), the length of the ventral fins (23 cm.), all agree with our measurements of marlina. The height of the 20th ray of the first dorsal (9 cm.) is slightly greater in herscheli than marlina but the difference is not unreasonable if we assume that the fin slot in herscheli may have shrunk during preservation. The relation of the height of the first dorsal to body depth in herscheli is not like marlina, but the drawing is from a stuffed specimen which may have been distorted. Playf air's (1866) description of brevirostris could also have been taken from a slender marlina. The height of first dorsal, color, and length of pelvic fin, all fit marlina but the body depth is comparable to that of audax. We do not suggest changing the name marlina, however, until better evidence is available. Color The name "white marlin," a literal translation of the Japanese, shirokajiki, probably arises from the appearance of the fish — sometimes a milky white when freshly hooked. We have been amazed at the whiteness of some of these huge fish as they swam near the boat before they had fought hard on the line. When near death and immediately after death the milkiness is replaced by shades of SPEARFISHES OF THE CENTRAL PACIFIC 525 metallic bluish gray (hence the name silver) rang- ing from deep color on the back to almost white on the belly. Usually at this time there is smooth gradation in color from the back to the belly, though in a few specimens a sharp line separates side and belly color. A few hours after death the color of the back deepens to a dark lead gray, when it is reasonable to call these fish black marlins. In Japanese, Hawaiian, and central equatorial Pacific waters the absence of stripes usually distinguishes marlina from audax and ampla, but Nakamura (1938) stated that stripes may some- times be detected after death and complete re- moval of the slime. We have seen only one speci- men with faint stripes immediately after death. Whitley (1954) described a stiff -finned marlin that had pale blue bars when first caught, and J. E. Morrow, in a personal communication, stated that marlina are commonly striped in Peruvian waters when alive. We suspect that the stripes and the white color in life may be more noticeable among the smaller sizes. Distribution in the Pacific In the literature reviewed by Rosa (1950), marlina has been reported in the eastern Pacific from California to Peru, in New Zealand, Aus- tralia, Hawaii, and Tahiti. Nakamura (1949) reports tliat it occurs widely in the warm seas of the Pacific and Indian Oceans and north off the coast of Japan to about 41° N. latitude. No verification has been obtained for rumors of the occurrence of "black marlin" in California, though both marlina and ampla appear to occur occa- sionally off western Mexico (Carl L. Hubbs and Robert L. Wisner, personal communication). In certain of these peripheral areas marlina is apparently one of the abundant marlins, certainly in the sport-fishing centers off Peru, New Zealand, and Australia. Nakamura (1951) believes that the center of its distribution is in the extreme western Pacific and in the adjacent waters of the South China Sea, Sulu Sea, and Celebes Sea. He (1942) calls it the most abundant marlin off Formosa and (1951) reported the catch by species in Formosa for 1943 which, in numbers of fish tak- en, ranks sailfish, marlina, ampla, and audax in decreasing order. He (1951) also reported that marlina is the most abundant spearfish off Okinawa. Off Hawaii and in the equatorial Pacific from south of Baja (California to the Caroline Islands it is much less abundant than ampla. The scattered POFI catches (fig. 25) occurred mostly in the vicinity of the Line Islands and north of the Marquesas, but nowhere was marlina numerous. In the Hawaiian fishery marlina is so scarce that Otsu (1954) lumped it in the catch data with sailfish, shortnosc spearfish, and broadbill swordfish, which together comprised less than a tenth of the total spearfish landings in an average year. The nine Japanese mothership expeditions to the vicinity of the Caroline Islands in 1950 and 1951 (Van Campen 1952) had a com- bined catch rate of less than .01 fish per 100 hooks for marlina, which may be contrasted with the catch rate of .53 for ampla. Despite this general scarcity, marlina has been taken in sufficient num- bers in the open Pacific to establish the strong probability that its distribution is continuous from America to Asia but that the concentrations are peripheral off the coasts of the Americas, Asia, and Australia. The intervening distances, together with ana- tomical evidence, suggest that these concentrations may be isolated enough for local varieties to be evolving. A difference in color between marlina from the central Pacific and from Peruvian areas has been noted. Furthermore, the difference in head length, length of pelvic, and height of the anterior lobe of the first dorsal is somewhat greater between samples of marlina than between samples of the other species. Indeed, the overlap in the height of the anterior lobe of the fii-st dorsal (fig. 17) between the samples of marlina from Hawaii, New Zealand, and Australia is beyond the com- monly accepted level of subspecific differentiation. However, the samples are small and a sample from the equatorial Pacific is intermediate in this dimension, so we shall consider the differences as merely varietal. Size This is one of the largest species of bony fishes. Nakamura (1949) stated that marlina attains a body length of 350 cm. and a weight of 570 kg. (1,250 lb.). The world's record angling catch taken off Peru on August 4, 1953, weighed 1,560 pounds (Farrington 1953). The previous record, 1,352 pounds, was cauglit only 6 days earlier. Farrington also reports that the fu-st 25 "black marlin" caught by angling off Peru averaged 817 pounds with many weighing more than 1,000 pounds. It apparently reaches similar sizes oft' 526 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Figure 25. — Distribution of POFI catches of black marlin, Istiompax marlina. Fractions indicate stations at which catches were reported out of the total fished; decimals indicate average catch per 100 hooks per day. Australia where one weighing 1,226 pounds was stranded in April 1938, according to Gregory and Conrad (1939), and also off Hawaii where the largest of 77 weighed in the market was 1,100 pounds (table 4). Such record fish are always females; the largest males have been much smaller. Nakamura (1951) reports a ma.ximum weight for the males of 287 pounds. The largest male in the POFI collection of six in which the sex was determined was 270 pounds and in those reported by Gregory and Conrad from off New Zealand and Australia was 322 pounds. Data on the size composition of a large catch are given for the Formosan fishery by Nakamura (1944a). He reports that in the 1943 landings at Takao, only 104 marlina of 2,542 weighed were more than 440 pounds and the modal size was 90 to 110 pounds. At Suo, of 4,448 weighed, 74 were more than 440 pounds; and there was a broad Table 4. — Weight frequency of Istiompax marlina landed at Hawaiian markets in 1950 and 1951 [Data from the Hawaiian Division of Flsb and Gamel Weight group (pounds) Total Weight group (pounds) Total fiO-79 440-159 3 80-99 5 7 8 12 5 4 2 2 2 460-479 100-119 480-499 1 120-139 600-549 5 140-159 550-599 3 160-179 600-649 . . 180-199 650-699 - 200-219 700-749 - . 220-2.39 7,50-799 1 240-259- 800-849 3 260-279 850-899 1 280-299 900-949 .__ - 300-319 950-999 . . . 320-339 2 3 1 3 >1,000 - 2 340-359 Number Maximum weight (pounds)- 360-379 - - 77 380-399 400-419 420-439. -. 1 1,100 mode at 1 10 to 200 pounds with a secondary mode at 250 to 270 pounds, in February and March. At Takao, the largest percentage of marlina SPEARFISHES OF THE CENTRAL PACIFIC 527 weipjhing; loss tlian 110 pounds was laiidod in Oc- tober and Xovombcr. In the Hawaiian market data (tal)Ie 4) one modal group from 100 to 150 pounds occurs, but too few data arc available to show other modes. Food No specific food studies of marlina have ap- peared but Nakamura (1949) in his general dis- cussion of the food of marlins indicates that they feed on live food but will take dead bait or arti- ficial lures and do not seek food on the bottom. June (1951) has recorded the gluttony of one specimen which contained a 158-pound bigeye tuna.* Of our 10 specimens on which we have food notes, 3 contained remains of the sunfish, Mola, and 2 contained tunas, 1 a 30-cm. Katsu- woruis (1 lb.), and 1 a 94-cm. Germo (40 lb.). An- other contained vertebrae and fin rays which were evidently from a fairly large fish, since the verte- brae were 5 cm. long in the centrum and the fin rays were about 20 cm. in length. This marlin is probably as broadly carnivorous as the other species of marlins which eat a great variety of fish and squid. Certainly, if they can capture tunas few other animals would be fast enough to escape them. Spawning Nakamura (1949) stated that he had no con- crete data on the spawning habits of marlina but suspected from some data on the condition of the gonads and the relative abundance of males and females that it spawns off Formosa around August to October. None of the POFI specimens had ripening gonads. Seasonal occurrence The season for marlina in Formosa is from Octo- ber through April, according to Nakamura (1938, table 9; 1951, table 43) who gave catch statistics for the Suo fish market. Off Cabo Blanco, Peru, the sport fishermen have taken it throughout the year, according to a personal communication from » ArttT this manuscript was wTitten, two similar records were obtained. Joseph E. KinR of the I'OKI .staff refwrted that on April 4, 1955. a marlina 402 cm. fork lenRlh was taken on a lonijline at 1°49' N'.latitudeand 157°38' W. longitude. It contained a yellowfin tuna 1.54 cm. fork length which was esti- mated from lencth-weiglit curves to weigh 157 pounds. The marlina ,ip- parently had taken the dead herring halt after eating the tuna tH'cau.s<> the tuna showed no signs of being hooked nor d id it have l>ait In Its stomach. The marlina was l)ooke200 1 3 i 2 1 Number ., 268 1,387 647 605 480 351 116 16 22 186 276 450 Maximum weight (pounds) 308 239 227 227 203 215 205 434 SPEARFISHES OF THE CENTRAL PACIFIC 531 Table 5. — Weight frequency of striped marlin, Makaira audax, from the Honolulu market, January 1949-FebruaTy 1952 — Continued [Data from the Hawaiian Division of Fish and QameJ Weight group (pounds) Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Year 1950: 10-19 2 70 186 81 10 28 88 149 123 69 30 11 4 9 4 21 228 217 46 37 105 138 125 77 36 11 8 3 4 56 63 9 6 20 34 128 188 137 101 47 13 14 6 7 2 2 1 1 2 4 4 3 5 25 51 50 33 14 15 4 2 10 97 10 5 7 30 82 138 141 88 53 17 9 2 2 5 20-29 45 34 11 6 13 30 32 31 35 8 6 4 1 2 25 74 45 17 34 73 99 53 21 17 6 3 4 2 1 2 18 31 22 13 15 15 44 67 62 29 15 14 12 5 6 2 2 2 1 10 42 12 2 8 28 49 37 28 21 9 2 1 3 15 17 6 2 1 4 1 3 3 2 2 3 13 7 2 4 22 SO 1 4 28 102 172 160 107 61 30 8 30-39 40-49 50-59 _ 60-69 70-79...- - 80-89 90-99 100-109 - 110-119 120-129 130-139 3 2 140-149 ... 15(V-159 1 3 160-169... _ 1 1 1 170-179 1 3 1 i" 180-189 1 190-199...- - >200 - — 1 1 3 2 1 Number 256 858 1,058 479 375 837 256 62 36 215 691 7S7 Maximum weight (pounds) 310 297 239 211 229 226 233 Year 1951: 10-19 1 5 7 6 3 13 30 51 61 34 15 10 3 1 6 1 1 3 7 16 U 15 9 6 2 9 1 2 3 6 3 .. 13 29 37 29 12 6 11 1 3 2' 20-29 6 10 4 59" 94 22 10 46 88 102 63 32 10 \ 1 1 1 1 1 17 42 32 7 18 62 51 43 23 14 7 3 10 5 .. 2 1 8 8 6 10 6 19 59 71 83 42 68 37 18 19 7 4 2 1 3 2 15 5 5 3 8 34 77 122 78 44 21 22 6 6 4 1 6 5 6 2 5 7 17 26 14 7 12 4 1 4 30-39 1 1 3 9 6 3 .. 2 1 1 3 1 3 3 3 .. 8 1 40-49 50-59 60-69 -.. 4 16 31 24 17 9 4 1 8 28 27 25 9 6 80-89 90-99 100-109 110-119 120-129 3 1 1 6 2 140-149 150-159 2 160-169 170-179 __ 180-189 1 190-199 1 1 1 3 1 1 >200 1 1 1 Number 126 241 539 329 461 454 114 34 17 93 161 129 Maximum weight (poimds) ... . 200 216 231 210 219 304 226 214 10-19 _ 20-29 7 22 10 1 6 19 30 23 10 5 1 2 1 1 13 15 6 10 53 84 45 25 6 1 2 1 3 1 30-39... 40-49 50-59 60-69... 70-79 80-89 90-99 100-109. 110-119 120-129 130-139 140-149 150-159 160-169 170-179 180-189 190-199 >200 Number 138 266 Maximum weight (pounds) ' 1 latitudinal distribution appears to exist in the central Pacific, for the POFI catches in tlie equatorial area iiu^luded only a few striped marlin of less tiian 100 pounds, whereas in the Honolulu market (table 5) about half of the fish weighed less than 60 pounds each. In the latter case, the distribution is very definitely and characteristi- cally bimodal in most months of the year. In the 532 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE winter months, the position of the modes cor- responds quite closely to those given by Ueyanagi for the 10° to 20° latitudinal zone. If we follow the progi-ession of the modes from month to month in table 5, two rather striking things may be noted. First, after the period of low catches in August and September, the fishery resumes on striped marlin of very different weight composition than existed in early summer. Second- ly, between about November and April the smaller mode progresses with reasonable smoothness from about 30 to 50 pounds, and from about October to Jidy the larger mode progresses, again with reasonable smoothness, from about 80 to 105 pounds. If we assume that the fishery has been fishing the same, stock of fish through these months it would appear that such a progression might be due to growth and, hence, an annual increment of about 30 pounds can be estimated. Spawning In the South China Sea near Formosa, spawn- ing seems to be at its peak from April to May according to Nakamura (1949). He also stated that audax is known to spawn near the Ogasa- wara Islands around May and June. That spawning occurs at this season is sug- gested by the scanty POFI observations 'from the central equatorial Pacific. Two males with milt in the testes were taken during March, and two females with enlarged ovaries were taken in February and March. Synonymy and references Hisliophorus audax, Philippi, 1887:35-38, pi. 8, figs. 2 and 3 (Chile) . Isltophorus audax, Fowler, 1944:499 (Tarapacd, Iquique, Valdivia). Marlina audax, Smith, 1956a:30 (South Africa). Makaira audax, Smith, 1956b:758 (South Africa). Tetra-pturus milsukurii, Jordan and Snyder, 1901:303, pi. 16, fig. 5 (Japan); Fowler, 1928:136 (Hawaii, tropical Pacific). Makaira milsukurii, Jordan and Evermann, 1926:61, pi. 18 (Japan, Hawaii, California); Griffin, 1927:143, pi. 14 (New Zealand); Walford, 1937:47 (California, Pacific coast of Mexico); Nakamura, 1938:27 (For- mosa) ; Gregory and Conrad, 1939: 443 (New Zea- land, Australia); Nichols and LaMonte, 1941:8, fig. 2; LaMonte and Marcy, 1941:2 (Japan, Hawaii, California, Chile, New Zealand, Australia) ; Naka- mura, 1942 (Formosa); Farrington, 1949:150 (Chile, Peru, Ecuador, Pacific coast of Panama and Mexico, California, Hawaii, New Zealand, Australia, Marianas, Japan); Brock, 1950:147 (Hawaii); Gabrielson and LaMonte, 1950:28 (California, west coast of Mexico and Panama, Ecuador, Peru, Chile [S. to Caldera], Australia, New Zealand, Hawaii); Rosa, 1950:132 (Americas from California to Caldera, Chile, New Zealand, Australia, Philippines, Japan, Hawaii) ; Nakamura, 1951:36 (warm seas of western Pacific); Morrow, 1952a:53 (New Zealand); Morrow, 1952b: 143 (New Zealand); Murphy and Otsu, 1954 (Caroline Islands); Morrow, 1954:819 (East Africa); LaMonte, 1955:333, pi. 7, pi. 8(2), and 346, pi. 12 (2) thought to be a young one. Marlina milsukurii. Grey, 1928:47 (New Zealand). Makaira grammalica, Jordan and Evermann, 1926:55, pi. 16 (Hawaii). Makaira hold, Jordan and Evermann, 1926:63, pi. 19, fig. 1 (Pacifi.c coast of Mexico). Makaira zelandica, Jordan and Evermann, 1926:65, pi. 19, fig. 2 (New Zealand). Telraplurus eclenes, Jordan and Evermann, 1926:34, pi. 11, fig. 2 (Hawaii). Kajikia milsukurii. Hirasaka and Nakamura, 1947:14, pi. 2, fig. 1 (Formosa); Nakamura, 1949:60 (south from northeastern Honshu, Japan) ; Nakamura, Yabuta, and Ueyanagi, 1953 (Japan) ; Ueyanagi, 1954a (northwestern Pacific from Equator to 42° N.); Ueyanagi, 1954b (Western Pacific from Japan to Australia, Indian Ocean off Sumatra). Kajikia formosana, Hirasaka and Nakamura, 1947:13 (Formosa); Nakamura, 1949:61 (Philippine Sea to Japan). Kajiki, makajiki, akakajiki, striped marlin, Nakamura, 1944b (Formosa); Van Campen, 1952 (Caroline Islands); Yabuta, 1953 (Bonin Islands); Nakamura, 1954b (northwestern Pacific, 14° to 30° N. latitude); Farrington, 1953 (Chile, Peru, Ecuador, Pacific coast of Mexico, California, Hawaii, New Zealand, Australia) . Probable synonym Isliophorus ktdibundus, Whitley, 1933:83 (New South Wales) . Makaira ampla (Poey) Blue Marlin Black Marlin (Hawaii) ; Kurokajiki (Japan) Distinguishing characteristics This is the giant marlin with the flexible pectoral fin that can be folded flat against the body, with the more nearly cylindrical body and, in very large sizes, with the relatively long snout. There is less of a hump on the back than in marlina, more than in audax. The anterior lobe of the first dorsal is higher than in marlina, but lower than in audax. The anterior lobe of the first anal fin, on the contrary, is higher in ampla than in either audax or marlina and the relation between the first anal SPEARFISHES OF THE CENTRAL PACIFIC 533 and first dorsal is the best clmracter we liave found for distinguishing ampla from audai. In ampla, the height of the first anal averaged 86 percent of the height of the first dorsal with a range of 76 to 100 percent; in audax, the range was from 50 to 76 percent %vith an average of 66 percent. The center of the first dorsal fin is low and in our specimens there is a suggestion of an actual decrease in the average length of the 20th ray with the growth of the fish up to 200 cm. ; but in the specimens of more than 200 cm. fork length the length of the 20th ray is nearly constant. Tlie average lengtli of the ray in ampla is appro.xi- mately 6 cm. with the range in our specimens from 3 to 9 cm. ; in audax, which has a similar growth pattern, the range is from 8 to 14 cm. with an average of 10 cm. The length of the pelvic fin is comparable to that of audax and longer than that of marlina, averaging about 34 cm. in our speci- mens with no change in size according to length of fish. This species appears to be unique among the marlins in the growth relation of mandible and snout (fig. 13).' In audax and marlina the snout and mandible grow approximately isometrically, whereas in ampla the mandibular growth definitely is negatively allometric. As a result the snout appears long in very large individuals. The lumping of the Atlantic and Pacific forms of this marlin in the single species ampla will no doubt be contested by people who automatically consider that such geographic separation indicates distinct species. However, in none of the char- acters considered in the preceding pages do we find a difference that even approaches the sub- specific level. Until morphological differences can be found it seems preferable to consider both forms as belonging to the same species. Color In the living specimens of ampla that we have seen in the Pacific, the predominant color of tlie upper parts is a brilliant, deep metallic blue which fades rapidly after death to a lead-gray color mixed with browns wherever the fish has been rubbed or scraped. Stripes usually are present on the sides immediately after death but are rarely conspicuous, and generally some are so faint that ' Dr. Hiroshi Xakamura, in a [HTSonal ctimniunication, jwinted out that allometric p-owth of the snout occurs in Istiophonts oritntaliti. At about 140 mm. the snout is extraordinarily long in relation to body lenpth and, as the fish grows, the length of .snout in relation to body length decreases. it is difficult to count them. They may be absent or remain conspicuous after death and cannot be relied on to distinguish the fish from either marlina or audax. Distribution in the Pacific This is the predominant marlin of the central tropical Pacific, having been taken in all of the tropical areas fished by POFI, from 110° W. longi- tude to 180° longitude, with catch rates up to 0.35 per 100 hooks (fig. 27). West, along the Equator in the Marshall and Caroline Islands area, the Japanese mothership expeditions of 1950 and 1951 found it even more abundant, for they had an average catch rate of 0.53 per 100 hooks. Off Formosa it is taken in lesser quantities than marlina and orientali.'^ (Xakamura 1951 , table 1 14). Northward from the Equator its abundance de- clines with latitude and, according to Nakamura (1951, fig. 31), for the zone from 143° to 150° E. longitude just off the coast of Japan ampla be- comes less abundant than audax at about 15° N. latitude, but moderate quantities are caught as far north as 40°. It has been reported recently off Australia by Whitley (1954). LaMonte (in Gabrielson and LaMonte 1950, p. 515) showed a photograph of a black marlin from off Acapulco which almost certainly is of this species because the fin is folded against the side, and the body shape, height of first dorsal, and the very short mandible are typical of ampla. The localities where ampla has been taken by the Japanese longline fishery are shown in the atlas prepared by the Nankai Regional Research Laboratory (Yabuta 1954). Catches are reported from the South China Sea off Hainan Island, from the Celebes Sea just east of the Philippines off northern New Guinea, and then almost continously along the Equator east to 155° W. longitude. The best catches were made during the summer months at 10° to 15° N. latitude north of the Caroline Islands. A few were taken during winter months in the albacore fishery along 30° N. latitude, east of Japan as far as 175° W. longitude, north of Midway. They also were taken at fishing stations in December and January in the Indian Ocean along the coasts of Java, Sumatra, and in the vicinity of the Nicobar Islands. Special concen- trations were found during February 1952 off northwest Australia at about 15° S. latitude, 118° E. longitude, and in the vicinity of the Solomon Islands. In the Hawaiian longline fishery, ampla 534 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE FiouRE 27. — Distribution of POFI catches of the blue marlin, Makaira ampla. Fractions indicate stations at which catches were reported out of the total fished; decimals indicate average catch per 100 hooks per day. is the most abundant spearfish by weight, and the annual landings ranged from a low of 512,000 pounds to a high of 679,000 pounds during the period 1948 to 1952 (Otsu 1954). Food All reports indicate that this species is broadly carnivorous on fish and cephalopods of the open ocean. Nakamura (1942) tabulated the food contents of 163 stomachs from fish taken in the east Philippine Sea. Of these, 53 stomachs con- tained squid, 11 Leiognathus, 12 BaliMes, 11 Auxis, and lesser amounts of some 9 other genera of fish. One contained a species of shrimp. In the POFI catches, 36 stomachs contained food, of which 34 contained fish and 16 cephalopods. The com- monest fishes in the stomachs were the tunalike fishes, particularly Katsuwonus in the larger individuals. Most of the cephalopods were squid. In two additional stomachs from POFI catches the complete contents were not noted but speci- ems of Xiphias gladius were preserved for later examination (see p. 519). Seasonal occurrence Yabuta (1954) presented data which show that the catch rate varies little thoughout the year in the tropical seas in the vicinity of the Caroline and Marshall Islands. Murphy and Otsu (1954) noted that the catches of ampla by the nine Japanese mother-ship expeditions in this same area showed a minor peak in February and another in October 1951, but that the catch rate during the summer months of 1951 was only about half that during the summer months of 1950. North of the Carolines, however, there is evi- dence of a seasonal migration and the peak abun- dance which occurs in May at 12° to 16° N. lati- tude, moves farther north with the season until the peak is at 24° to 28° N. in September. Farther east SPEARFISHES OF THE CENTRAL PACIFIC 535 off the Marshall Islands the principal fishery is from 8° to 12° N., and here the seasonal abundance gradually increases until July, then slowly de- clines. Off Formosa, Nakamura (1942) reported that arnpla is plentiful on the Pacific side during the summer, and he (1949) stated that they are extremely rare in the Kuroshio Current region from October to April. Off Hawaii, Otsu (1954) showed that ampla reaches the peak of abundance and is the principal spearfish in the catch from July through November, whereas during the other months of the year audai is the principal species in terms of pounds landed. Thus, north of the equatorial area the seasonal occurrence suggests a summer movement of ampla northward followed by a return south in the late autumn. Size The reports from the Japanese and Hawaiian fish- eries indicate that ampla rivals and may even sur- pass matiina in maximum size reached. Nakamura (1949) reported that ampla attains 1,100 pounds, but a weight of 1,450 pounds has been recorded from the Hawaiian fishery (table 6), and fishermen recall weighing specimens of more than 1,600 pounds.'*' The POFI specimens include one of 1,002 pounds from Hawaii and another nearly as large from the equatorial area, which was partly eaten by sharks (Nos. 67 and 68 in appendix table 1-E, p. 548). As in marlina, the large ampla are alwaj'S females. The largest male weighed by POFI was 218 pounds. Ueyanagi (1953) and Nakamura et al. (1953) both reported that males do not exceed 200 cm. (orbit to fork), which is equivalent to about 255 pounds. Yabuta (1954), in the atlas of Japanese longline fishing, summarized data on size composition by sex of ampla from several areas and all length frequencies showed a mode at about '" ThP maximum sizes of ampla attained in the Pacific are much greater than reported from the Atlantic where the angling record is 742 pounds (official 1955 list of the International Game Fish Association). Table 6. — Weight frequency of blue marlin, Makaira ampla, from the Honolulu market, January 1949-Fehruary 1952 [Data collected by the Hawaiian Division of Fish and Game] Weight group (pounds) Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Year 1949: 10-19.... 20-29. 1 30-39 1 1 1 40-49 \ 1 1 1 50-59 2 2 3 6 3 3 5 3 4 9 10 13 10 3 6 9 3 5 8 9 4 2 5 7 2 3 1 3 2 7 3 5 2 2 r fiO-69... 2 1 1 .- 1 8 17 32 22 21 28 19 14 12 6 4 1 .. 1 1 ? 10 13 19 14 8 13 7 2 1 1 1 1 1 1 70-79 ._ 1 1 80-89... 90-99 2 1 2 100-109...- 1 110-119 120-129 2 4 5 2 1 3 2 2 3 3 3 3 2 2 3 1 -- 4 3 2 3 1 4 8 8 5 5 4 3 3 2 3 1 1 2 1 1 1 1 3 6 5 7 4 2 2 5 5 5 4' 3 3 3 2 4 1 4 2 6 2 1 5 10 16 11 8 8 2 3 10 3 1 1 2 2 5 2 6 2 4 5 3 2 5 2 5 2 6 4 1 130-139 2 1 1 1 6 1 3 6 3 5 10 I 5 9 5 6 4 6 1 3 3 3 4 7 4 2 1 1 6 5 4 2 5 2 3 1 5 5 4 6 3 6 4 3 4 4 3 4 1 2 3 2 140-149 2 2 4 3 5 2 2 6 3 6 5 3 9 4 3 4 4 4 1 1 1 4 2 1 150-159 1 ] 160-169 170-179 i 1 1 180-189 190-199. 200-219 2 220-239 240-259 260-279 2 280-299 4 6 5 3 1 4 6 2 7 8 5 9 8 6 2 3 1 8 4 2 2 2 4 4 1 5 2 8 11 5 2 2 1 300-319 4 320-339 3 34ft-359 3 360-379 4 380-399 2 2 2 400-119 4 420-439 4 440-159 460-479 2 3 480-499 2 500-549. 3 2 1 s 550-599 1 600-649 1 650-699.. _ 700-749..-- 1 1 1 750-799 1 1 800-849 1 1 1 1 850-899 1 1 900-949 950-999 1 1 1 >1.000 1 1 Xumber. . 6 S6 88 102 93 85 142 270 168 173 89 St Maximum weight (pounds) 1.015 1,011 536 FISHERY BULLETIN OF THE FISH AND "WILDLIFE SERVICE Table 6. — W eight frequency o} hhie marlin, Makaira ampla, /rom the Honolulu market, January 1949-February 195Z — Con. [Data collected by the Hawaiian Division of Fish and Oame) Weight group (pounds) Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Year 1960: 10-19 -. 20-29 --- 2 1 30-39 - 1 40-49 - 2 1 2 1 5 6 4 4 1 4 8 11 13 9 11 5 6 12 4 5 4 6 2 3 3 6 1 .- 3 2 3 6 1 5 2 1 2 2 50-59 1 2 60-69 -- - 1 70-79 -- 1 1 4 5 4 4 2 7 25 38 28 25 28 15 20 5 9 9 7 8 4 5 3 4 3 2 6 3 5 8 7 3 ' 2' 80-89 90-99 --- . - 1 1 1 3 1 7 2 9 15 13 7 5 5 7 1 3 3 2 1 1 7 3 5 3 2 1 2 2 3 1 1 2 4' 6 9 19 24 21 14 25 20 10 15 9 5 5 2 2 2 1 5 3 4 5 4 6 2 11 9 4 3 1 1 1 4' 7 15 32 47 26 31 16 18 12 9 8 5 1 3 6 3 3 3 6 7 7 7 8 14 7 5 3 3 1 1 1 2 1 100-109 3 110-119 2 1 5 -- 3 1 1 4 3 2 1 4 2' 4 2 2" 3 2 .- 120-129 130-139 1 1 1 3 3 1 2 1 140-149 .- -- -- 1 1 2 2 2 1 2 3 150-159 3 1 160-169 -- 3 1 2 1 1 2 5 2 1 2 1 3 2 2 1 1 3 180-189 4 190-199 1 4 4 1 1 1 1 2 2 1 1 1 2 3 3 1 1 1 3 3 2 1 2 5 1 1 1 3 1 3 200-219 2 1 2 1 2 1 1 2 1 4 4 4 3 4 2 3 6 2 1 2 3" 3 2 6 240-259 -- 5 1 280-299 4 2 320-339 - . - - 1 2 360-379 - 1 380-399 1 1 2 2 400-119 -- 4 3 440-159 2 1 2 3 2 4 2 -- 1 480-499 1 500-549 2 1 3 550-599 - 2 1 1 1 650-699 2 1 1 3 700-749 - 750-799 1 850-899 1 i 1 1 950-999 1 2 .- 1 >1,000 1 1 1 1 Number - 26 39 42 51 46 60 123 261 323 304 164 67 1,002 1,030 1,066 1,001 1,058 1,287 Year 1961: 50-59 1 3 2 2 3 -- 8 14 11 17 20 16 11 11 6 4 7 14 6 7 7 4 10 5 1 8 5 2 6 8 1 1 70-79 80-89 1 1 1 2 4 4 5 15 17 25 40 46 27 24 8 17 10 8 8 4 4 3 3 8 9 4 6 14 5 5 19 8 3 2 3 90-99 2 1 3 5 18 29 36 37 30 35 15 17 4 6 7 1 6 1 6 7 9 4 3 6 6 9 7 16 7 8 5 I 110-119 - 3 6 8 16 13 5 6 8 6 6 5 2 5 2 3 3 5 3 4 3 4' 3 1 10 7 5 3 5 6 15 24 39 34 31 24 20 17 6 7 3 8 9 7 5 7 6 9 5 11 10 4 11 23 10 15 9 5 1 1 2 2 130-139 2 1 2 1 1 1 1 1 -- 1 150-159 3 1 1 1 3 6 1 3 5 7 3 2 6 1 4 5 3 4 3 2 8 6 8 1 2 8 3 2 2 1 9 2 6 3 7 3 1 4 3 6 4 2 1 1 6 4 4 2 1 1 2 170-179 --. 7 1 1 2" .. 2 2 1 2 1 2 3 4 4 2 3 1 -- i 2 2 1 4 1 1 1 10 1 3 1 1 190-199 7 2 7 220-239 3 1 2 260-279 - 1 2 4 3 3 3 300-319 3 320-339 1 1 2 340-359 2 360-379..-- - 5 380-399 1 2 2 2 420-439 3 440-469 . . 2 3 460-479 1 2 500-649 3 1 1 1 3 2 2 600-649..-- -. 1 650-699 2 1 700-749...- 2 SPEARFISHES OF THE CENTRAL PACIFIC 537 Table 6.-Weighl frequency of blue marlin, Makaira ampla./rom the Honolulu market. January 1949-February 195Z-Con. ll'ata collected by the Hawaiian Division of FUh and Oame) Weight group (pounds) Year 1951— Continued 750-799 . 800-849... 850-899 900-949 950-999 >1.000.... Number Maximum weight (pounds). Jan. Feb. Mar, 1,450 Year ig.W: 10-19 20-29. 30-39 40-49 50-59 60-69 70-79 80-89 90-99 100-109 110-119 120-129 130-1.39.. . . 140-149.. 1.50-159 100-169 170-179 180-189 190-199.. 200-219 220-2.39 240-259 260-279 280-299 300-319 320-339 340-359 360-379 380-399... 400-419 420^39 440-159.... 460-479... 480-199.. 500-549 550-599 600-649 650-699.. 700-749 750-799.... 800-849.... 850-899.. 900-949.. 9.50-999 >1,000 Number Maximum weight (pounds). 1,003 160 cm. (12.5 lb.) with most of the males ranging from 140 cm. (85 lb.) to 180 cm. (175 lb.). The ampla of less than about 200 pounds, which some of the Japanese authors consider to be mostly males (Ueyanagi 1953; Nakamura et al., 1953; Yabuta 1954), appear in the fishery in quantity only during the summer months. Tliis is thought to indicate a segregation by sex during migrations. A similar phenomenon exists in Hawaii (table 6), where the catch of ampla from July to October contains a large modal group from 130 to 220 pounds which may be males. There is also at this Apr. May June July Aug. 1,107 389 Sept. Oct. Nov. 345 362 224 Dec. 71 time some increase in the catch of larger fish, but not nearly as great an increase as in the modal group. Spawning Among the ampla specimens examined from the POFI catches, we found no ripe females but did find males with freely flowing milt in the gonads from February through October (and captured only three between November and January). So it is likely that at least some of the males may be ready to spawn at almost any time of the year in the equatorial Pacific. 538 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Nakamura (1942) thought that, like the rest of the spearfishes, ampla spawns over long periods of time in wide areas of ocean, and he suspected that the great increase in the proportion of males in the catch off Formosa during May is indicative of the spawning season. He also stated (1951) that ampla spawns east of Luzon from May to July. Synonymy and references No attempt has been made to include a comprehensive list of reference.s to the Atlantic form. Tetrapturus amplus, Poey, 1860: 243, tab. 15, fig. 2, (Cuba). Tetrapturus mazara, Jordan and Snyder, 1901:305 (Japan); Fowler, 1934:400 (Japan, Hawaii). Makaira mazara, Jordan and Evermann, 1926:53, pi. 11, fig. 2 (Japan, Hawaii); Griffin, 1927:141, pi. 13 (New Zealand); Nakamura, 1938:28 (Formosa); Nakamura, 1941 (Philippine Sea); Nakamura, 1942 (Formosa); Brock, 1950:146 (Hawaii); Nakamura, 1951:37 (north- ern tropical Pacific) ; Murphy and Otsu, 1954 (Caro- line Islands); LaMonte, 1955:336 (in part). Makaira nigricans ampla, Conrad and LaMonte, 1937:207 (Bahamas); Shapiro, 1938 (Bahamas); Gregory and Conrad, 1939, pi. V (Bahamas); Gabriel- son and LaMonte, 1950:29; Rosa, 1950:145 (north- western Atlantic, Caribbean Sea to New England). Makaira ampla, LaMonte, 1955:344. Makaira ampla mazara, LaMonte and Marcy, 1941:2 (Japan); Nichols and LaMonte, 1941:8 (Japan). Makaira nigricans mazara, Rosa, 1950:141 (Pacific coast of Mexico, California, Hawaii, Japan, Australia, New Zealand) . Makaira ampla ampla, LaMonte and Marcy, 1941:2 (Cuba to North Carolina). Eumakaira nigra, Hirasaka and Nakamura, 1947:16, pi. 2, fig. 2 (Formosa); Nakamura, 1949:65 (warm seas of Pacific and Indian Oceans) ; Nakamura, Yabuta, and Ueyanagi, 1953 (Japan); Yabuta, 1954 (western Pacific, Japan to Australia and ea.st to Line Islands, Indian Ocean off Sumatra). IsHompax howardi, Whitley, 1954:58, pi. 3, fig. 3 (Aus- tralia) . Acapulco black marlin, Gabrielson and LaMonte, 1950:515 (Pacific coast of Mexico). Kurokajiki, black marlin. Van Campen, 1952 (Caroline Islands); Yabuta, 1953 (Bonin Islands). Blue marlin, Farrington, 1937; Farrington, 1949:153 (Cuba to New England, Caribbean Sea). Silver marlin, Farrington, 1953:160 (Hawaii). LITERATURE CITED Brock, Veknon E. 1950. Keys to Hawaiian fishes. Board of Commis- sioners of Agriculture and Forestry, Honolulu, Territory of Hawaii, Division of Fish and Game, Special Bull. 4, 191 pp. Conrad, G. Miles, and Francesca LaMonte. 1937. Observations on the body form of the blue marlin (Makaira nigricans ampla Poey). Bull. American Museum Natural History 74 (4): 207-220. CuviER, Georges, and Achilles Valenciennes. 1831. Histoire naturelle des poissons. Tome 8:205- 226, Paris. Day, Francis. 1878-1888. The fishes of India. Text and atlas in 4 parts. (Text 778 pp., 198 pis.) London. De Beaufort, L. F., and W. M. Chapman. 1951. The fishes of the Indo-Australian archipelago. Vol. 9, 484 pp. E. J. Brill, Leiden. De Buen, Fernando. 1950. Contribuciones a la Ictiologla. III. La familia Istiophoridae y desoripci6n de una especie uruguaya (Makaira perezi de Buen). Publicaciones Cien- tificas 5:163-178, Servicio Oceanografico y de Pesca, Montivideo, Uruguay. Noviembre. Ego, Kenji, and Tamio Otsit. 1952. Japanese tuna-mothership expeditions in the western equatorial Pacific Ocean (June 1950 to June 1951). U. S. Fish and Wildlife Service, Commercial Fisheries Review 14 (6):1-19. June. Farrington, Selwyn Kip, jr. 1937. Atlantic game fishing. 298 pp. Kennedy Bros., New York. [Published in 1939 by Garden City Publishing Co., New York.] 1949. Spearfish of the world. In Game Fish of the World, edited by B. Vesey-Fitzgerald and Francesca LaMonte, pp. 144-155. Harper and Brothers, New York. 1953. Fishing the Pacific, offshore and on. 297 pp. Coward-McCann, Inc., New York. Fowler, Henry W. 1928. The fishes of Oceania. Memoirs Bernice P. Bishop Museum 10:540 pp. Honolulu. 1934. The fishes of Oceania. Supplement 2. Mem- oirs Bernice P. Bishop Museum n(6):385-466. Honolulu. 1944. The fishes. In Results of the Fifth George Vanderbilt Expedition (1941). Monographs, No. 6, p. 499, Academy Natural Science, Philadelphia. 1949. The fishes of Oceania. Supplement 3. Mem- oirs Bernice P. Bishop Museum 12 (2):37-186. Honolulu. Gabrielson, Ira N., and Francesca R. LaMonte. 1950. The fisherman's encyclopedia. 698 pp. Stack- pole & Heck, New York. Gray, John E. 1838. Description of a new species of Tetrapturus from the Cape of Good Hope. Annals Natural History 1(36):313, pi. 10. London. Gregory, William K., and G. Miles Conkad. 1939. Body-forms of the black marlin (Makaira nigricans marlina) and striped marlin (Makaira milsukurii) of New Zealand and Australia. Bull. American Museum Natural History 76 (8):443-456. Grey, Zane. 1928. Big game fishing in New Zealand seas. Natural History 28 (l):46-52. SPEARFISHES OF THE CENTRAL PACIFIC 539 Griffin L. T. 1927. Additions to tho fish fauna of New Zealand. Trans, and Proceed. New Zealand Institute 58 (parts 1 and 2): 136-150. HiRASAKA, K., and H. Nakamura. 1947. On the Formosan spearfishes. Bull. Oceano- graphic Institute of Taiwan 3:9-24, 6 pi. HuBBS, Carl L., and Clark Hubbs. 1953. An improved graphical analysis and comparison of series of samples. Systematic Zoology 2(2): 49-56, 92. HuBBs, Carl L., and Robert L. Wisner. 1953. Food of inarlin in 1951 ofT San Diego, California. California Fish and Game 39(1):127-131. January. Huxley, J. S., and G. Teissier. 1936. Terminology of relative growth. Nature 137: 780-781. London. Jordan, David Starr, and Barton Warren Evermann. 1926. A review of the giant mackerel-like fishes, tunnies, spearfishes and swordfishes. Occasional Papers California Academy Sciences 12:1-113. Sept. 30. Jordan, David Starr, and J. O. Snyder. 1901. Description of nine new species of fishes con- tained in museums of Japan. Jour. College of Science, Imperial University, Tokyo, I5(2):30I-311. June, Fred C. 1951. Note on the feeding habits of the giant white marlin of the Pacific. Pacific Science 5 (3): 287. July. KiKAWA, Shoji. 1954. Mekajiki, Xiphias gladius Linneaus. Average year's fishing condition of tuna longline fisheries for 1952. Edited by Nankai Regional Fisheries Research Lab., Kochi; published by Japanese Tuna Boat Owners Assoc, Tokyo. (In Japanese with English titles.) LaMontb, Francesca R. 1955. A review and revision of the marlins, genus Makaira. Bull. American Museum Natural History 107(3) :323-358. New York. LaMonte, Francesca R., and Donald E. Marcy. 1941. Swordfish, sailfish, marlin, and spearfish. Ich- thylogical Contributions of the International Game Fish Assoc. 1(2): 1-24. Mabr, John C. 1955. The use of morphometric data in systematic, racial and relative growth studies in fishes. Copeia 1955(1):23-31. January. Marr, John C, and Milner B. Schaefer. 1949. Definitions of body dimensions used in describ- ing tunas. U. S. Fish and Wildlife Service, Fishery Bull. 51(47):241-244. Morrow, James E., Jr. 1952a. Allometric growth in the striped marlin, Makaira mitxukurii, from New Zealand. Pacific Science 6(l):53-58. January. 1952b. Food of the striped marlin, Makaira milsukurii, from New Zealand. Copeia 1952 (3): 143-145. 1954. Fishes from East Africa, with new records and descriptions of two new .species. Annals and Maga- zine of Natural History, 12th Ser., 7 (83): 797-820. October. MiRPHY, Garth I., and Tamio Otsu. 1954. Analysis of catches of nine Japanese tuna long- line expeditions to the Western Pacific Ocean. IJ. S. Fish and Wildlife Service, Special Sci. Rept.: Fish- eries 128:1-46. Murphy, Garth I., and Richard S. Shomura. 1953a. Longline fishing for deep-swimming tunas in the Central Pacific, 1950-51. U. S. Fish and Wildlife Service, Special Sci. Rept.: Fisheries 98:1-47. 1953b. Longline fishing for deep-swimming tunas in the Central Pacific, January-June 1952. U. S. Fish and Wildlife Service, Special Sci. Rept.: Fisheries 108:1-32. 1955. Longline fishing for deep-swimming tunas in the Central Pacific, August-November 1952. I'. S. Fish and Wildlife Service, Special Sci. Rept.: Fisheries 137:1-42. Nakamura, Hiroshi. 1937. On the habits of some istiophorid fishes of Taiwan waters especially the short-nosed spearfish. Zool. Mag. (Dobutsugaku zasshi) 49(6):233-238. (In Japanese.) 1938. Report of an investigation of the spearfishes of Formosan waters. Reports of the Taiwan Gov- ernment-General Fishery Experiment Station 1937(10):l-34, 15 pi. (English translation: U. S. Fish and Wildlife Service, Special Sci. Rept.: Fisher- ies 153:1-46. 1955.) 1940. On the spawning habits of the sailfish. Zool. Mag. (Dobutsugaku zasshi) 52(8):296-297. (In Japanese.) 1941. On the habits of the istiophorid fishes, particularly the spawning of the black marlin, Makaira mazara. Zool. Mag. (Dobutsugaku zasshi) 53(1): 17-21. (In Japanese.) 1942. On the ecology of the istiophorid fishes of Taiwan waters. Suisan Gakkai Ho 9(1):45-51. (In Japa- nese.) 1944a. Studies on the istiophorid fishes of Taiwan waters. 8. Seasonal variations in size. B. White marlin. Trans. Natural History Society Taiwan 34(251) :293-297. (In Japanese.) 1944b. Studies on the istiophorid fishes of Taiwan waters. 8. Seasonal variations in size. C. Other species. Trans. Natural History Society Taiwan 34(252) :338-341. (In Japanese.) 1949. The tunas and their fisheries. Takeuchi Shobo, Tokyo, 1949. (English translation: U. S. Fish and Wildlife Service, Special Sci. Rept.: Fisheries 82: 1-115. 1952.) 1951. Tuna longline fishery and fishing grounds. As- sociation of Japanese Tuna Fishing Cooperatives, Tokyo, 1951. (English translation: U. S. Fish and Wildlife Service, Special Sci. Rept.: Fisheries 112:1- 168. 1954.) 1954a. Ocean currents and tuna fishing grounds. Suisan kagaku (14):9-17. (In Japanese.) 1954b. The two-year cycle in the size composition of tuna and spearfishes. Suisan jiho 6(57):53-55. (In Japanese.) 540 FISHERY BULLETIN OF THE FISH AND' WILDLIFE SERVICE Nakamura, Hiroshi, et al. 1951. Notes on the life-history of the sword-fish, Xiphias gladius Linnaeus. Japanese Journal Ich- thyology 1(4): 264-271. 4 figs. Nakamtjra, Hiroshi, Y. Yabuta, and S. Ueyanagi. 1953. Relation between the spawning season and the sex ratio of some fishes of the family Istiophoridae. Contributions Nankai Regional Fisheries Research Lab. 1(13): 1-8. (In Japanese.) Nichols, J. T., and F. R. LaMonte. 1935a. The Tahitian black marlin, or silver marlin swordfish. American Museum Novitates 807:1-2. 1935b. How many marlins are there? Natural History 36(4): 327-330. 1941. Differences in marlins based on weights and measurements. Ichth. Contributions of the Interna- tional Game Fish Assoc. 1(1): 1-8. New York. NiSKA, Edwin L. 1953. Construction details of tuna long-line gear used by Pacific Oceanic Fishery Investigations. U. S. Fish and Wildlife Service, Commercial Fisheries Review 15(6); 1-6. June. Otsu, Tamio. 1954. Analysis ot the Hawaiian long-line fishery, 1948- 52. U. S. Fish and Wildlife Service, Commercial Fisheries Review 16(9): 1-17. September. Philippi, Rudolfo Amandus. 1887. Sobre los tiburones y algunos otros peoes de Chile. Anales Universidad Chile 7.1:1-42, pis. 1-8. Playfair, R. Lambert, and Albert C. L. G. GiJnther. 1866. The fishes of Zanzibar. 153 pp., 21 pis. London. POEY, PhILIPE. 1860. Memorias sobre la historia natural de la Isla de Cuba. 11:243-244, tab. 15 (2). Ramsay, E. P. 1881. Notes on Histiophorus gladius. Proceed. Lin- nean Soc. New South Wales 5:295-297. Rosa, Horacio, Jr. 1950. Scientific and common names applied to tunas, mackerels, and spearfishes of the world, with notes on their geographic distribution. Food and Agriculture Organization of the United Nations, 235 pp. Wash- ington, D. C. Shapiro, Sydney. 1938. A study of proportional changes during the post larval growth of the blue marlin [Makaira nigricans ampla Poey). American Museum Novitates, No. 995, 20 pp. Siebold, Ph[ilipp] Fr[anzJ von. 1844. Fauna japonica, p. 103, pi. 55. Smith, James L. B. 1950. The sea fishes of southern Africa. 550 pp. Haf- ner Publishing Co., New York. 1956 a. Swordfish, marlin and sailfish in South and East Africa. Ichthyological Bull. No. 2:1-34, 2 pis. De- partment Ichthyology, Rhodes University, Grahams- town, South Africa. 1966 b. The striped marlin, (Makaira audax Philippi) in South Africa. Nature 177:758. Snedecor, George W. 1946. Statistical methods applied to experiments in agriculture and biology. 485 pp. Iowa State College Press, Ames, Iowa. Tanaka, Shigeho. 1911-14. Figures and descriptions of the fishes of Japan. Vols. 16-30, p. 324, fig. 285. 2d. ed. rev. (1935). Daichi Shoin, Tokyo. Ueyanagi, Shoji. 1953. The sexual difference of the sizes of marlins. Contributions Nankai Regional Fisheries Research Lab. 1 (14):l-5. (In Japanese.) 1954a. Annual variations found in the fishing condi- tions and size compositions of Makajiki, Kajikia mitsukiirii (Jordan and Snyder). Bull. Japanese Soc. Scientific Fisheries 19 (11):1100-1108. (In Japanese with English summary.) 1954b. Makajiki, striped marlin, Kajikia miisukurii. Average year's fishing condition of tuna longline fisheries for 1952. Edited by Nankai Regional Fisheries Research Lab., Koohi; published by Japa- nese Tiina Boat Owners Assoc, Tokyo. (In Japanese with English titles.) Van Campen, Wilvan G. 1952. Japanese mothership-type tuna-fishing opera- tions in the western equatorial Pacific, June-October 1951. U. S. Fish and Wildlife Service, Commercial Fisheries Review 14 (11): 1-9. November. Walford, Lionel A. 1937. Marine game fishes of the Pacific Coast from Alaska to the Equator. 205 pp. University Cali- fornia Press, Berkeley. Whitley, Gilbert P. 1931. Studies in ichthyology. No. 5. Records Austra- lian Museum 18 (4): 138-160. 1933. Studies in ichthyology. No. 7. Records Austra- lian Museum 19 (1):60-112. 1954. More new fish names and records. Australian Zoologist 12 (1): 57-62. 1955. The Australian Museum's marlins. Australian Museum Magazine 11 (9):292-297. Yabe, Hiroshi. 1053. On the larvae of sailfish (Istiophorus orientalis) collected in the southwestern sea of Japan. Con- tributions Nankai Regional Fisheries Research Lab. 1 (6): 1-10. (In Japanese.) Yabuta, Yoichi. 1953. On the stomach contents of tuna and marlin from the adjacent seas of Bonin Islands. Contri- butions Nankai Regional Fisheries Research Lab. 1 (15):l-6. (In Japanese.) 1954. Kurokajiki (black marlin) Eumakaira nigra. Average year's fishing condition of tuna longline fisheries for 1952. Edited by Nankai Regional Fisheries Research Lab., Koehi; published by Japa- nese Tuna Boat Owners Assoc, Tokyo. (In Japanese with English titles.) APPENDIX Because it is necessary to compare the spear- nshes of the world by means of measurements, our original data and some computations are presented here. The original observations were obtained by members of the POFI scientific staff in addition to their regular observations on tunas. These members were: Donald K. F. Ching, Thomas S. Hida, Isaac I. Ikehara, Edwin S. Iversen, Joseph E. King, Walter M. Matsumoto, Sueto Murai, Garth I. Murphy, Tamio Otsu, Thomas J. Roseberry, William F. Royce, Richard S. Shomura, Wilvan G. Van Campen, and Henny S. H. Yuen. Appendix Table 1-A. — Original data and morphometric measurements of 8 specimens o/Tetrapturus angustirostris, by POFI IMeasurements in millimeters] Item No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8 Latitude 2°56' N. 150''08'W. 2-17-53 24 1,470 7°00'S. 169°59' W. 5-28-53 •> 18 1,509 1,415 327 387 406 387 215 38 134 1,306 1,256 216 15 2°57' S. 169''49' W. 11-21-52 Male 21 1,530 15°46' N. 154''13' W. 1-26-53 33 1,589 g-oi' S. 131°24' W. 3-18-54 Female ' WOO' N. 151°02'W. 5-3-53 5i ID'OO- N. 151°02'W. 5-3-53 9 15=46' N Longitude .,. 154° 13' W 1 26-52 .Sex Weight (pounds)--- _ 46 Tip .snout to fork tail 1,638 1,537 346 415 435 408 234 33 141 1.428 1,371 225 17 12 124 179 88 160 140 210 177 1,017 50 61 114 215 204 332 490 0 100 Yes 0 0 (?) 1,791 Tip snout to inside 1st dorsal Tip snout to inside pectoral. -_ Tip snout to posterior edge opercle 359 379 403 437 Tip snout to anterior edge orbit ___ _ 225 Orbit diameter 29 34 31 43 34 Posterior edge orbit to posterior edge opercle 1,296 1,526 1,463 234 14 1,577 Length of mandible 239 Sword width opposite tip mandible _ ._. 9 17 Depthofhead _. 110 151 70 124 165 161 196 217 Body width at tip pectoral 135 196 179 898 44 58 108 164 174 330 409 33 114 Yes 1st dor.sal height longest anterior ray . 190 168 900 45 212 225 172 955 53 222 204 998 56 59 134 213 217 351 503 87 123 Yes 0 0 233 1st dorsal height 20th ray 171 1st dorsal length base 1 067 56 1st anal height 100 128 132 143 Pectoral length 180 308 205 402' 197 •353 214 Pelvic length 366 Interspace 1st and 2d dorsals .. - 23 81 Yes 0 0 IS 48 Inlcrspiice 1st and 2d anals Yes 0 0 Number stripes on sides 0 Number free spines between dorsals 0 1 Running ripe; ovaries 4 cm. diameter, mature ova about 1 mm. diameter. ' Snout broken. • Shark-eaten. * Approximate: broken parts. 541 542 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Appendix Table 1-B. -Original data and morphometric measurements of 6 specimens of Istiophorus orientalis, by POFI [Measurements in millimeters] Item No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 Latitude Longitude... Date talven. Sex. 7°39' N. 131°20' W. 11-14-52 Weight (pounds) Tip snout to (ork tail Tip snout to upper tall notch Tip snout to inside 1st dorsal Tip snout to inside pectoral Tip snout to inside pelvic Tip snout to posterior edge opercle Tip snout to anterior edge orbit Orbit diameter -.- -- Posterior edge orbit to posterior edge opercle. Naris to foric of tail Posterior edge orbit to fork of tall Length of mandible Sword width opposite tip mandible Sword depth opposite tip mandible Depth of head Greatest body depth. Body width at tip pectoral... Body depth at vent Ventral groove to inside anal 1st dorsal height longest anterior ray 1st dorsal height 20th ray 1st dorsal length base 2(1 dorsal height .-. 2d dorsal length base 1st anal height 1st anal length base Pectoral length ._ Pelvic length Caudal spread Interspace !st and 2d dorsals Interspace 1st and 2d anals — Pectoral fin folds against side Number stripes on sides 1,530 7°10' N. 152°14' W. 10-25-52 Female 33 1,809 9°05'N. 131''40' W. 11-15-51 30 ,142 30 'i,l79 39 1,560 7''20' N. 110°2n' W. 3-4-.'>4 Female 64 2,232 2,090 666 713 787 735 513 43 167 1,727 1,664 l°Ofl' S. 140°05' W. 3-13-53 Male 77 C) 187 229 274 183 293 119 200 323 '678 393 759 1,161 81 90 201 253 338 581 764 13 119 Yes 20 33 1,784 3 825 1.173 104 209 398 623 36 8°59' N. U0°n9' W. 3-3-54 Female 86 J 2, 516 '- 2. 336 2 725 >808 >830 »790 ! 535 S3 202 2,001 335 124 109 474 •607 1,320 101 107 249 271 500 536 867 SO 155 Yes ? • Snout broken. 3 Includes estimated 50 mm. for broken snout. ' Longest ray (#19) was 842 cm. ' Longest ray (#18) was 777 cm. Appendix Table I-C. — Original data and morphometric measurements of 11 specimens of Istiompax marlina, fct/ POFI [Measurements in millimeters] Item Latitude Longitude.- Date taken. Sex. Weight (pounds) Tip snout to fork tall Tip snout to upper tail notch Tip snout to inside 1st dorsal Tip snout to inside pectoral. Tip snout to inside pelvic Tip snout to posterior edge opercle Tip snout to anterior edge orbit Orbit diameter.- Posterior edge orbit to posterior edge opercle. Naris to fork of tail. Posterior edge orbit to fork of tail Length of mandible Sword width opposite tip mandible Sword depth opposite tip mandible Depth of head Greatest body depth Body width tip pectoral Body depth at vent \'entral groove to inside anal 1st dorsal height of longest anterior ray 1st dorsal height of 20th ray.. 1st dorsal length base 2d dorsal height 2d dorsal length base 1st anal height 1st anal length base. Pectoral length Pelvic length Caudal spread... Interspace 1st and 2d dorsals Interspace 1st and 2d anals Pectoral fin folds against side Number of stripes on sides Number of free spines between dorsals ... No. 0°03' N. 156°1.5' W. 12-4-53 ? 182 2,379 2,209 y.so 823 870 825 530 51 244 1,861 1,798 343 45 230 284 74 898 87 115 231 260 413 232 880 278 139 No 0 3(?) 1°48' S. 139° 59' W. 3-11-53 Male 183 2,507 928 "36 1.904 "41 455 280 71 1,153 86 243 467 279 92 168 No 0 0 No. 3 2<'34' S. 155°23' W. 8-19-53 2,568 2,375 823 920 970 915 613 51 251 1,983 1,904 378 34 210 311 60 95 123 443 256 90S 113 No. 4 2° 1.3' N. 155° 1.5' W. 4-15-54 Male 270 2.716 2,490 858 1.000 1,045 981 647 57 277 2,084 2,012 404 44 35 281 491 267 449 96 280 45 1,216 98 146 281 314 481 186 980 93 116 No 30 0 No. 5 2°15' N. 151° 19' W. 10-30-52 Female '■ 2. 749 37 '2,149 505 97 274 574 No. 6 4°58' S. 149°57' W. ,V15-53 ? 1291 2,836 2,6,35 895 999 1,068 665 58 2,198 2.113 414 50 2.54 542 220 141 335 78 1.345 94 136 274 290 .535 270 960 61 146 No 0 See footnotes at end of t.able SPEARFISHES OF THE CENTRAL PACIFIC 543 Ai'PE.NDix T.4BLE 1-C. — Original data and morphometric measurements of 1 1 specimens of Istiompax marlina, by POFI — Con. Item No. 7 No. 8 No. 9 No. ID No. 11 Latitude 3°22' N. 160°24' W. 8-23-53 ? 293 2.842 2.6as 9.52 1,005 1,085 1.039 715 52 272 2.160 2.075 405 SO 5">20'S. 179<'5.5' \V. 2-22-52 Female 2,999 4°36' N. 154<'41' W. 4-20-54 Female 418 3.027 2.788 940 1.054 1.135 1.090 703 M 323 2.3Kf 2.260 440 S6 43 310 .591 339 505 52 386 67 1.425 121 148 308 336 566 255 1.120 66 ISO No 0°08' N. 154051' W. 8-13-53 7 3.214' 3°52'S Longitude _ .,. . _ I55°I3' W Sex AVeight (pounds) 587 Tip snout to fork tail 3,4«7 3 210 Tip snout to upper tail notch 975 l.i45" 1,024 Tip snout to inside pectoral ,., 1, 151 i.i4i 743 61 337 2. 508 2,410 Tip snout to posterior edge opercle . 1 2.35 Tip snout to anterior edge orbit ■ 808 Orbit diameter.-- - Posterior edge orbit to posterior edge opercle 364 2.693 2,596 518 Length of mandible . Sword width opposite tip mandible 61 51 Depth of head _ _ 230 277 386 Greatest bodv depth 568 647 Body width tip pectoral .. 355 Bodv depth at vent - , _ _ S3S 387 69 1.316 107 163 297 367 596 295 194 1st dorsal height of longest anterior ray__ 308 72 1.231 102 142 278 294 534 304 1.023 127 148 No 0 0 348 393 72 1st dorsal length base . _ . 1 501 2d dorsal height , _., . 2d dorsal length base _ 1st anal height 295 1st an:il length base _ . 336 531 644 Pelvic length 255 Caudal spread 1.112 1, 166 243 134 No 0 0 Interspace 1st and 2d anals 210 No 0 Number of free spines between dorsals 0 ' Excluding stomach contents. ' Approximate measurement; tip of snout broken. ' About 12 stripes faintly showing when first caught. Appendix Table 1-D. — Original data and morphometric measurements of 25 specimens of Makaira audax, by POFI [Measurements in millimeters] Item No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. V No. 8 No. 9 Latitude S^SS'N. i6ni' w. 5-21-54 •> 2i 1.423 1.315 495 547 560 548 383 35 130 1.049 1.005 219 14 8°07' N. 149° 57' W. 8-11-52 7 52 1 1,869 6''36' S. 120°26' W. 10-29-52 7 8°07'N. 149°57'W. 8-11-52 7 75 1.968 4'>18' N. 130°11'W. 11-12-52 7 2«23' N. 130°25' W. 11-10-52 7 86 2,160 3°23' N. 130°29' W. I1-11-.52 7 92 2.165 1°59' S. 120°03' W. 10-29-52 3''23' N Longitude 130<'29' W Date taken Se.x 7 Weight (pounds) ..- 99 Tip snout to fork tail 1.891 2,036 2.282 2 294 Tip snout to upper tail notch . Tip snout to inside 1st dorsal . .. 689 732 712 764 745 801 818 Tip snout to inside pectoral 808 740 833 753 837 867 868 781 914 829 935 Tip snout to posterior edge opercle 848 Tip snout to anterior edge orbit.. Orbit diameter 36 35 36 36 36 37 36 37 Posterior edge orbit to posterior edge opercle. 1,402 1,417 1,464 1,550 1.649 1,670 1.742 1.720 Length of mandible. Sword width opposite tip mandible.. Depth of head 109 259 279 309 293 305 321 338 329 Body width tip pectoral.. Bodv depth at vent-- Ventral groove to inside anal 1st dorsal height longest anterior ray 1st dorsal height 20th ray 232 151 665 55 51 146 200 204 297 455 17 44 Yes 16 340 374 354 359 390 405 413 307 2d dorsal height., 84 82 97 96 87 94 Isl anal height 201 254 217 230 243 266 258 262 Pectoral length .. _ 330 376 371 387 398 429 434 438 Interspace 1st and 2d dorsals. Number stripes on sides See footnotes at end of table. 544 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Appendix Table 1-D. — Original data and morphometric measurements of Z5 specimens of Makaira audax, by POFI — Con. [Measurements in millimeters] Item No. 10 No. 11 No. 12 No. 13 No. 14 No. 15 No. 16 No. 17 Latitude 5°15' N. 110°17'W. 3-5-54 Male 0''59' S. lll'>28' W. 3-9-54 Female 167 2.518 2,333 813 895 914 878 587 53 238 1,959 1,878 398 29 19 220 365 216 274 88 436 115 1,245 8-59' N. 110°09' W. 3-3-54 Male 145 2.545 2.342 798 873 900 859 562 53 244 2,012 1,930 396 28 18 222 369 200 273 77 438 82 3°19' S. 112°11'W. 3-10-54 Female 205 2.548 2.362 835 923 964 915 606 55 255 1,969 1.888 418 36 20 205 390 260 336 57 411 109 1,237 100 93 277 358 463 220 844 49 61 Yes 1°I1'N. 130°15'W. 11-9-52 Female 1°04' N. 151°05' W. 2-10-53 1 2°34' S. 165°23' W. 8-15-53 ? 245 2.736 2.500 909 1.002 1.025 1,002 664 67 271 2,094 2,005 446 39 0°51' N. 168''53' W 6-10-54 Sex ? 2 180 12,443 2,571 2,622 2,747 2,530 877 929 1,028 963 884 i.'ooi' 1,065 1,015 700 Orbit (^liameter - -- 53 225 1,904 1,827 361 30 20 208 365 195 265 63 399 86 1,210 91 84 241 321 448 379 875 20 83 Yes U6 0 43 42 58 257 2,019 2,076 1,989 Length of mandible. - 441 32 20 194 197 387 405 363 243 461 448 89 1,227 107 481 127 416 109 1,271 119 96 289 268 525 387 962 103 123 106 293 362 536 361 1,046 72 Yes 16 0 89 99 290 340 488 319 916 305 304 280 348 490 527 348 531 357 901 53 63 105 Yes 21 0 75 Yes !16 0 75 Yes Ca. 15 0 10 0 0 Item No. 18 No. 19 No. 20 No. 21 No. 22 No. 23 No. 24 No. 25 Latitude 2°39' S. 179°.'>4' E. 2-20-62 Female 1<'20'S. 169O00' W. 3-8-52 Female 6°07' N. 154°47' W. 4-22-54 Female 290 2,842 9''57' N. 155°06' W. 7-29-63 1°20'N. 155°03' W. 2-3-52 Female 6°47' N. 162'>06' W. 1-26-63 ? 8°39' N. 164°57' W. 7-30-63 ? 314 3.039 2,792 969 1.066 1,097 1,071 691 67 313 2,386 2,281 473 36 1°47' N. 168°16'W. Date taken _ .- 6-9-64 ? 2280 2,757 2,798 12,889 2,911 2,933 3.101 2,870 978 1.010 946 1,024 1,051 1,027 655 67 305 2,223 2,120 467 40 22 280 490 281 396 63 450 77 1,346 111 113 288 388 589 281 1,075 82 94 Yes 16 0 1.000 1,120 1,107 1,022 1,154 1,050 1,173 1,028 1,083 1,096 715 Orbit diameter 47 48 65 47 67 314 2,100 2,115 2,256 2,247 2,413 2,166 2,319 490 35 25 Depth of head 238 243 Greatest body depth_ 435 422 545 474 476 283 60 1st dorsal heiEht longest anterior ray 460 500 486 673 486 110 1.288 70 466 99 1.364 136 117 294 366 640 343 475 101 1.405 121 111 118 302 327 259 326 443 Pectoral length 549 592 522 566 416 251 1,053 Interspace 1st and 2d dorsals 163 126 Yes 16 0 160 165 Yes 12 1 134 75 Yes Yes 12 0 ' Approximate: tip of snout broken. 2 Without viscera. ' 5 intermediate stripes. SPEARFISHES OF THE CENTRAL PACIFIC 545 Appendix Table 1-E. — Original data and morphometric measurements of 68 specimens of Makaira ampla, by POFI IMeasurements In mUllmetors] Item No.l No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8 No. 9 No. 10 Latitiido 5''03' S. 150''05' W. 5-15-53 3°36' S. 149055' W. 5-14-53 1.788 1.648 529 597 604 574 379 47 148 1,438 1,362 227 28 2°15' S. 169° 58' W. G-l-.W Male 93 1,824 1, 673 533 588 600 588 383 41 164 1.463 1,400 241 25 ,5°5.r N. 16ri5' w. 1-27-.53 Male 90 1,897 2°56' S. 150°08' \V, 2-17-53 Male 110 1,985 2°5)' N. 1.W<'19' W. 2-3-53 4°03' S. 179°.58' E. 2-21-52 Female 6''47' S. 180° 2-24-52 Male 5°03' S. 1.50°05' W. 5-15-.53 Male 118 2,086 1,917 685 721 729 723 491 52 180 1.612 1.543 283 31 8°.'i9' N. 110°09' W. Date taken .V3-S4 Sex 28 1.350 1.233 338 375 388 372 214 34 124 1,151 1,102 185 16 136 1,989 2,011 2.019 2,126 Tip snout to upper tall notch... Tip snout to inside 1st dorsal . Tip snout to inside pectoral Tip snout to Inside pelvic Tip snout to posterior edge opercle Tip snout to anterior edge orbit. 1 971 659 634 641 707 734 675 736 670 708 574 677 656 699 466 33 37 37 33 40 60 Posterior edpe orbit to posterior 183 Narls to fork of tall 1,558 1.560 1,593 1,590 1,599 1,680 Posterior edge orbit to fork of 1,610 274 Sword width opposite tip 28 30 33 33 Sword depth opposite tip 23 117 189 119 170 147 300 199 132 330 209 196 290 328 a49 323 321 359 Body width tip pectoral 230 323 Ventral groove to inside anal. . . 1st dorsal height longest an- ter ior rav - - , , - 1st dorsal height 20th ray 28 171 84 716 51 67 138 215 187 310 470 9 40 Yes (') 0 44 233 71 887 65 76 195 275 313 355 684 30 38 Yes 40 272 63 975 80 83 241 320 382 300 .802 34 .59 Yes 12 0 61 235 80 890 02 74 189 286 32.5 371 690 23 44 Yes 12 0 226 SO 909 68 244 66 g.w 70 294 63 1.005 68 260 275 245 51 1.056 79 85 288 206 247 227 233 245 335 332 352 350 363 408 394 369 396 377 Pelvic leneth 301 762 758 Interspace 1st and 2d dorsals... Interspace 1st and 2d anals Pectoral fin folds against slde = .. Number stripes on sides Number free spines between 4S 55 Yes 13 0 37 37 20 58 Yes IS 0 0 0 0 Item No. 11 N'O. 12 No. 13 No. 14 No. 15 No. 16 No. 17 No. 18 No. 19 No. 20 Latitude ri'ia' s. 150"'05' W. 5-16-53 Male 180 2.168 2.000 644 703 724 716 440 61 215 1,762 1.667 294 35 6°47' S. 180° 2-24 .52 Male 0'>02' N. 179°48' E. 2-18-52 Female r43' N. 169''59' W. 6-4-.'i3 Male 123 2.232 2 072 720 778 785 767 514 57 196 1.745 1.661 285 37 5'>41' S. 169" 44' W. 5-29-53 Malo 153 2.244 2.071 689 746 760 746 495 56 195 1.769 1.693 294 29 9°01' S. 131°24' W. 3-18-54 Male 176 2,262 3°36' S. 149°55' W. 5-14-53 Male 1.30 2.288 2.108 730 781 803 784 527 55 202 1.779 1.706 .305 30 6°40' S. 169°03' W. 3-4-52 Male 0°01' N. IfiO'eZ" W. 3-9-52 Male 1°13'S. 150°11'W. Date taken 2-12-53 Sex Male 145 2.184 2,219 >2,293 '2.302 2.329 Tip snout to upper tail notch... Tip snout to inside 1st dorsaK.. Tip snout to inside pectoral Tip snout to inside pelvic Tip snout to posterio.- edge 711 721 786 732 790 736 797 Tip ,';nout to anterior edge orbit. 41 38 39 42 Posterior edge orbit to posterior Xarls to fork of tall 1,743 1,753 1,787 1.830 1,814 1.834 Posterior edge orbit to fork of tail Sword width opposite tip 45 Sword depth opposite tip 203 394 247 158 301 203 185 361 247 167 384 351 397 245 362 59 360 364 370 Body width tip pectoral... . Ventral groove to Inside anal... 1st dorsal height longest an- 41 306 70 1,090 78 92 2.i7 350 412 51 289 42 1.0.17 75 89 230 324 412 310 790 53 63 Ve.s 42 298 70 1.094 81 91 255 363 400 357 855 43 52 Yes 295 305 61 1.088 90 101 261 322 447 362 875 26? 72 Yes 14 0 314 297 2f^ 1st dorsal height 2flth ray 46 1.118 2d dorsal height 75 77 84 71 251 257 2S7 253 242 392 412 435 389 292 810 30 64 Yes 8 0 SS3 30 Interspace 1st and 2d anals. Pectoral fin folds ag:iinst side. . Xum her stripes on sides Xuniher free spines between ? 0 0 n See footnote."! at end of table. 546 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Appendix Table 1-E. — Original data and morphometric measurements of 68 specimens of Makaira ampla, by POFI — Con. [Measurements in millimeters] Item No. 21 No. 22 No. 23 No. 24 No. 25 No. 26 No. 27 No. 28 No. 29 No. 30 15°30' S. 149°30' W. 2-23-53 Male 156 2,330 7''06' N. 152''11'W. 10-25-52 Male 2,350' 5°43' N. 150°06' W. 8-13-52 Male 170 2,374 8°07' N. 149°67' W. 8-11-52 Male 176 2,377 e^is' N. 131°00' W. 11-13-52 ? 0°02' N. 179°48' E. 2-18-52 9°62'N. 151°12'W. 5-3-63 7 175 2,417 2,236 762 837 852 836 662 57 226 1,880 1,808 320 38 6°13' N. 131°00' W. 11-13-52 ? 2,422" 2° 12' S. 150°20' W. 8-21-52 ? 185 2,435 3°ir S. Longitude 130°17' W. 11-6-52 Sex -.- Male Tip snout to fork tail 2,382 2,402 2,438 736 785 821 776 783 782 816 823 Tip snout to inside pelvic 780 733 882 823 903 832 843 788 874 789 840 792 963 864 913 Tip snout to posterior edge 776 839 Tip snout to anterior edgeorbit. 41 60 47 38 39 38 39 44 Posterior edge orbit to posterior Posterior edge orbit to fork of 1,853 1,908 1,853 1,842 1,889 1,899 1,930 1,904 1,921 33 Sword depth opposite tip 160 383 401 371 414 389 408 373 ,. 426 448 1st dorsal height longest an- terior ray 1st dorsal height 20th ray 1st dorsal length base .- 298 58 1,116 80 337 302 294 348 300 288 50 1,062 82 106 259 346 428 257 850 110 99 Yes 0 0 348 296 341 80 97 78 87 242 280 271 257 296 274 308 226 285 415 364 446 421 437 440 421 470 494 431 905 876 Interspace 1st and 2d dorsals. .. Interspace 1st and 2d anals Pectoral fin folds against side... Number stripes on sides Number free spines between 66 85 Yes 13-14 0 Item No. 31 No. 32 No. 33 No. 34 No. 35 No. 36 No. 37 No. 38 No. 39 No. 40 Latitude. . Longitude _. Date taken .. 6°40' S. 169'03'W. 3-4-52 Male 8°07'N. 149°57' W. 8-11-52 Male 205 2 2, 477 3''23' N. 130">29' W. 11-11-52 ? 184 2,479 8°14'N. 120°32'W. 10-19-52 } Hawaii 1-23-53 f 9°20' S. \120''53'W. 3-15-54 Male 207 2,538 2,349 789 861 5''62' N. m-ir w. 10-22-52 Female 2,546 9°01' S. 131°24' \V. 3-18-54 Male 183 2,550 9°20'S. 120°63' \V. 3-15-54 Male 173 2,579 2,396 838 927 924 899 600 61 238 2,010 1,918 360 35 23 205 392 228 334 50 308 58 1,192 78 109 253 367 466 284 883 51 92 Yes Ca. 13 0 S'OS' S. 150°05' W. 5-16-53 Sex Male(?) Weight (pounds) 218 2,465 2,517 2,628 2,596 Tip snout to upper tail notch 2,404 826 856 872 790 835 926 Tip snout to Inside pelvic Tip snout to posterior edge 889 853 910 860 926 864 879 813 937 867 867 669 58 240 1,994 1,911 918 Tip snout to anterior edge orbit. 608 43 45 47 42 42 40 62 Posterior edge orbit to posterior edge opercle 248 Naris to fork of tail 1,938 1,957 1,940 1,972 1,986 2,039 1,998 2,025 Posterior edge orbit to fork of tail 1,926 Length of mandible 360 Sword width opposite tip mandible 35 36 25 213 404 269 366 58 344 69 1,224 87 105 289 380 452 382 906 40 74 Yes Ca. 14 0 37 Sword depth opposite tip mandible 224 416 404 398 419 396 448 407 231 332 68 333 1st dorsal height longest an- terior ray 332 311 319 343 63 1,193 87 307 60 1st dorsal length base 1,199 95 102 81 96 94 2d dorsal length base 99 293 262 282 272 296 270 1st anal length base 364 466 431 427 429 282 513 473 Pelvie length 297 894 873 972 923 65 69 61 Pectoral fin fol2,«53 3°23' N. Longitude ._. 130°29' W. Date taken _ 11-11-52 Sex Female Weight (pounds) 367 2,853 Tip snout to fork tail 2,602 805 2,607 2,414 820 894 924 893 595 51 247 2,034 1,961 329 40 27 233 462 280 397 61 367 61 1,262 91 110 324 406 528 418 992 28 47 Yes 11 2.695 2,856 2,650 907 956 953 974 640 63 271 2,244 2,153 362 45 2,881 2,984 Tip snout to upper tall notch... Tip snout to inside 1st dorsal.. 9S8 965 972 1,024 Tip snout to inside pelvic Tip snout to posterior edge 875 815 43 2,124 1,032 982 1,136 1,016 1,032 977 1,109 938 1,025 Tip snout to anterior edge orbit. 60 58 42 70 294 2,332 2,252 380 49 48 Posterior edge orbit to posterior edge opercle. . 2,086 2,180 2,230 2,264 2,332 Posterior edge orbit to fork of Sword width opposite tip 225 513 309 292 209 542 339 423 447 501 554 497 526 59 347 35 1,350 85 120 296 41 399 68 1st dorsal height longest an- 311 381 376 325 401 373 80 1,373 100 110 285 398 556 370 1,062 80 104 Yes ' 14 0 404 1st dorsal heieht 20th rav 72 102 85 239 317 267 341 377 413 545 344 342 Pectoral length 477 485 582 681 546 621 496 Pelvic length 1,074 44 1,020 101 Yes 13 0 Yes Number free spines between 0 Item No. 51 No. 52 No. 53 No. 64 No. 55 No. 56 No. 67 No. 58 No. 59 Latitude 4<'00'N. 162° 20' W. 10-28-52 Female 2°10'N. 151°45' W. 10-30-62 Female 6''29' S. 149°60'W. 5-16-53 Female 376 3,075 2,858 992 1,089 1,110 1,084 729 64 291 2,371 2,282 386 50 8°59' N. 110°09'W. 3-3-54 Female 361 3,088 2,846 1,012 1,075 1,100 1,092 7.W 60 274 2,357 2,270 379 50 31 273 517 342 458 61 417 60 1,4,17 96 128 374 474 574 388 1,164 52 84 Yes 10 0 7''57' N. 169°48' W. 6-9-53 Female 7°57' N. 169''48' W. 6-9-53 Female 5-30' N. 149°58' \V. 5-6-53 4°32' N. 170°02'W. 6-7-53 Female 2°38' N. 169°59' W. Date taken 6-5-53 Sex 605 3,236 2,997 1.069 1.154 1,170 1.184 806 70 308 2. 455 2.360 424 53 Tip snout to fork tail 3,005 3,005 3.152 2,891 979 1,050 1.050 1,061 691 70 300 2.472 2,391 408 46 3,182 2,932 987 1.102 1,135 1,101 735 67 299 2.474 2.380 406 50 •3,251 3,308 3,060 839 1,018 962 Tip snout to inside pectoral 1,071 905 864 1,104 1,029 1,071 1,075 ""m 311 2.590 2.499 402 50 697 41 43 65 Posterior edge orbit to posterior edge opercle 313 2,462 2.347 2.640 Posterior cdpc orbit to fork of tail 2, 546 Length of mandible 425 52 258 524 315 263 573 337 296 581 384 253 280 502 330 265 Oreatcst body dpjjth 420 478 593 368 71 388 68 1,462 105 121 348 425 5C>4 318 1,096 .M 136 Yes 66 439 61 1,551 106 123 392 466 624 327 1,136 31 91 Yes .57 430 45 1.540 106 131 375 478 585 381 1.118 .W 119 Yes 77 460 71 1.641 % 128 400 441 600 384 1.210 126 119 Yes 10 0 69 1st dorsal height longest anterior ray 335 446 412 62 1.423 ins 132 357 386 .578 320 1.194 198 139 Yes 0 0 413 85 Ist dorsal length basr 1.641 2d dorsal height _ 2'\ dorsal length biise 102 115 104 165 1st anal height . 388 387 1st anal length base 470 Pectoral length . 608 610 556 271 Caudal spread 998 1,149 1.138 52 Interspace 1st and 2d anals . . 119 Yes 0 0 0 0 See footnotes at end of table. 548 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Appendix Table 1-E. — Original data and morphometric measurements of 68 specimens of Makaira ampla, by POFI — Con. [Measurements in millimeters] Item No. 60 No. 61 No. 62 No. 63 No. 64 No. 65 No. 66 No. 67 No. 68 7''67' N. 169°48' W. 6-9-63 Female 2°38' N, 169''59' W. 6-5-53 Female 640 3,402 3,152 1,083 1,170 1,207 1,188 784 70 334 2,609 2,648 423 55 3"'06' N. 160°12' W. 8-16-52 7 605 3,419 1°00' s. 120°13' W. 10-27-62 ? 3°19' S. 112°ir W. 3-10-54 Female 540 3,621 3,290 1,162 1,238 1,277 1,250 868 65 317 2,682 2,688 442 64 40 272 585 364 524 82 436 53 1,631 108 140 400 517 637 280 1,232 80 129 Yes Ca. 15 5°15' N. 149'56' W. 2-1-53 2°42' S. 166°05' W. 2-6-52 Female 4°10' N. 168°30' W. 3-12-62 7 } Hawaii 11-13-54 Female 1,002 3,342 3,082 1,037 1,128 1,139 1,128 761 59 318 2,620 2,532 396 60 3,445 3,665 2 3,766 4,012 3,714 1,126 1,139 1,237 1,357 1,234 1,146 1,241 1,150 1,383 1,198 1,362 921 69 45 41 42 41 69 Posterior edEe orbit to posterior edge opcrcle_ 382 2,690 2,720 2,812 2,824 3,096 3,131 3,032 Leng^th of mandible 472 58 62 48 276 603 318 631 286 355 597 608 665 683 740 723 382 692 73 453 68 1,664 110 129 393 607 618 396 1,087 68 106 Yes 69 407 60 « 1,641 108 125 408 456 632 368 1,264 56 146 Yes (<) 104 1st dorsal height longest anterior ray. 432 462 369 55 1,656 99 458 506 512 70 1,961 120 128 193 408 388 443 478 453 620 Pectoral length -^ 583 635 717 272 597 696 686 345 1.283 1,460 1.458 162 226 Yes 54 106 Yes 12 0 0 0 ' Not visible 2 hours after death. 2 Approximate: tip of snout broken. s Includes estimate of 30 mm. for broken snout. ' About 14 stripes appeared faintly about 1 hr. after death— these were dark or brown on the sides and lighter on the back. * Includes estimate of 10 mm. for broken snout. 8 Base of 1st dorsal includes 2 disconnected spines. Appendix Table 2-A. — Original data and morphometric measurements of two specimens of Tetrapturus angusti- rostris taken in Hawaiian waters [Measurements by the Hawaiian Division of Fish and Game; in millimeters] [Measurements by the Hawaiian Division of Fish and Game; in millimeters) Appendix Table 2-B. — Original data and morphometric measurements of five specimens of Istiompax marlina taken in Hawaiian waters Item Date taken Sex Weight (pounds) Tip snout to fork tail Orbit diameter Naris to fork tail Greatest body depth 1st dorsal height longest anterior ray 1st dorsal height 20th ray 1st dorsal length base Pectoral length No. 2 3-18-50 ? 50 1,857 40 1,645 219 229 174 1,160 211 Item No. 1 No. 2 No. 3 No. 4 No. 5 Date taken 6-6-50 7 270 (2, 562) 44 1,974 626 286 56 1,116 4-12-60 ? 341 (2, 835) 46 2,195 570 316 68 1,146 4-11-50 ? 305 (2, 970) 42 2,306 523 353 60 1,310 680 3-29-50 7 468 (3, 149) 47 2,450 643 324 65 1,383 4-14-50 Sex -, - - Female Weight (pounds) 617 Tip snout to fork tail > (3,220) 46 Naris to fork tail - -. _ - 2,507 G reatest bod y depth 1st dorsal height longest 616 329 1st dorsal height 20th ray.,.. 1st dorsjil length base 60 1,410 ' Measurements in parentheses estimated from regression data of table 3-E. SPEARFISHES OF THE CENTRAL PACIFIC 549 Appendix Table 2-C. — Original data and morphometric measurements of SO specimens of Makaira audax taken in Hawaiian waters [Measurements by tbe Hawaiian Division of Flsb and Qame; In mlllimetersl Item No. 1 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8 No. 9 No. 10 Date taken - . . Sox Weight (pounds) Tip snout to fork tall ' Orbit diameter N'aris to fork of tail (Ireate.-il body depth, _ 1st dorsal heigbl longest anterior ray 1st dorsal height 20tb ray 1st dorsal length base Pectoral length 3-24-50 ? 32 (1,665) 34 1,240 223 284 147 803 282 3-17-50 t 41 (1, 717) 36 1,282 235 299 167 793 310 3-24-50 7 42 (1, 792) 39 1,343 236 299 155 866 292 3-22-50 ? 72 (1,985) 42 1,499 279 332 105 974 374 3-20-50 t 58 (1,996) 37 1.508 260 312 180 933 361 (2,003) 37 1,514 105 916 376 3-X-bO 9 si (2, 105) 41 1,596 304 370 106 967 382 7-5-50 Male 78 (2, 109) 40 1,600 295 297 129 977 376 7 80 (2, 134) 40 1,620 138 975 420 6-6-50 9 94 (2, 149) 42 1,632 306 367 107 1,009 413 Item No. 11 No. 12 No. 13 No. 14 No. 15 No. 16 No. 17 No. 18 No. 19 No. 20 Date taken - - Sex Weight (pounds) — Tip snout to fork tail' _ Orbit diameter - Xaris to fork of tail Greatest body depth 1st dorsal height longest anterior ray 1st dorsal height 20th ray... 1st dorsal length base Peetoral length 6-6-50 0 85 (2, 160) 43 1,641 303 346 91 1,010 390 3-22-50 Female 95 (2, 184) 43 1,660 325 379 982 415 7-5-50 Male 90 (2, 211) 43 1,682 310 354 105 1,000 395 3-23-50 ? 110 (2, 227) 41 1,695 338 361 108 993 452 6-6-50 7 89 (2, 250) 41 1,714 296 338 88 1,090 359 6-6-50 ? 104 (2, 275) 41 1,734 319 387 112 1,068 390 6-22-50 94 (2, 292) 42 1,748 315 336 110 1,062 397 4-12-50 ? 110 (2, 302) 41 1,756 345 413 116 1,088 487 6-6-50 Female 110 (2, 307) 44 1,760 338 424 124 1,134 434 6-5-50 7 100 (2,312) 44 1,764 328 388 117 1,116 443 Item No. 21 No. 22 No. 23 No. 24 No. 25 No. 26 No. 27 No. 28 No. 29 No. 30 Date taken Sex Weight (pounds) Tip snout to fork tail' Orbit diameter N'aris to fork of tail Greatest body depth 1st dorsal height longest anterior ray, 1st dorsal height 20th ray 1st dorsal length base Pectoral length 6-5-50 ? Ill (2, 319) 45 1,770 342 411 102 1,065 459 161 (2,344) 48 1,790 369 389 102 1,177 454 6-6-50 Female 125 (2, 366) 45 I, 358 356 86 ,080 349 7-5-50 Male 124 (2, 405) 46 1,839 333 373 102 1,038 433 6-6-50 7 110 (2, 412) 46 1,845 318 J280 94 1,107 443 7-5-50 Female 124 (2,463) 45 1,886 350 379 118 1,117 470 7-5-50 Female 129 (2, 470) 42 1,892 366 396 100 1.159 464 6-5-50 Female 107 (2, 479) 47 1, 342 383 109 ,156 439 3-23-50 7 147 (2, 512) 46 1.926 369 429 105 1,169 435 3-31-50 Female 164 (2, 528) 46 1,939 409 435 82 1,201 530 ' Immature. ' Measurements in parentheses estimated from regression data of table 3-E. ' Questionable measurement. .\PPEND1X Table 2-D. — Original data and morphometric measurements of 27 specimens of Makaira ampla taken in Hawaiian waters [Measurements by the Hawaiian Division of Fish and Game; in millimeters] Item No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No. 7 No. 8 No. 9 Date taken.. Sex Weight (pounds) Tip snout to fork tail ' Orbit diameter... Xaris to fork of tail Greatest body depth 1st dorsal height longest anterior ray 1st dorsal height 20th ray 1st dorsal length base Pectoral length _ 4-10-50 7 58 (1,731) 34 1.385 359? 217 77 860 272? 7-5-50 Male 147 (2, 315) 45 1,829 363 258 53 1,085 391 4-10-50 7 170 (2, 458) 51 I, 938 380 334 61 1,213 441 4-17-50 Female 220 (2, 549) 43 2,007 412 332 64 1,252 479 5-8-50 Female 260 (2, 592) 44 2.040 438 349 63 1.166 490 6-6-50 Female 207 (2.609) 45 2,053 424 368 66 1,253 469 4-17-50 Female 2.% (2, 633) 45 2,071 472 389 54 1,268 515 5-8-50 Female 330 (2, 754) 46 2,163 502 382 58 4-18-50 Female 297 (2,756) 46 2.165 473 413 56 1.301 540 Item Date taken Sev... Weight (pounds) Tip snout to fork tail ' (libit diameter Naris to fork of tail. (ii'i'atest body di'pth. 1st dorsal lieight longest anterior ray, l.st dorsal li.-ipht 20th ray 1st dorsal leiiglli base Pectoral length No. 10 4-27-50 Female 342 (2, 800) 40 2, 198 4.W 365 39 1,320 505 No. 11 4-18-50 Female .304 (2,821) 48 2,214 454 412 64 1.3.12 522 No. 12 No. 13 4-27-50 Female :«2 (2. 839) 46 2,228 490 357 52 1,220 525 6-5-50 Female 448 (2, 9.32) 49 2,299 541 4.33 61 1,356 542 No. 14 4-28-50 Female 431 (2. 944) 45 2,308 566 410 60 1,399 574 No. 16 No. 17 6-21-50 Female 426 (2. 9ti2) 48 2,322 537 415 62 1,413 544 4-14-50 Female 290 (2. 969) 48 2.327 513 386 41 1.274 536 4-17-.'iO Female 469 (3. 024) 47 2,369 539 378 43 1,325 526 No. 18 4-17-50 Female 433 (3. 026) 50 2,370 555 409 46 1,391 551 See footnote at end of table. 550 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Appendix Table 2-D. — Original data and morphomelric measurements of S7 specimens of MaksLira. ampla, taken in Hawaiian waters — Continued [Measurements by the Hawaiian Division of Fish and Game; in millimeters) Item Date taken . . _ Sex Weight Tip .snout to fork tail ' Orbit diameter Naris to fork of tail Greatest body depth. - 1st dorsal height longest anterior ray 1st dorsal height 20th ray 1st dorsal length base Pectoral length _._ No. 19 5-4-50 ? 458 {3, 127) 50 2,447 559 412 54 1,362 565 No. 20 8-1-50 Female 572 (3, 220) 51 2,518 627 364 60 1,472 617 6-8-50 Female 408 (3, 244) 48 2,536 533 361 52 1.451 495 4-11-50 Female 564 (3, 245) 50 2,537 610 383 43 1,475 560 4-5-60 Female 508 (3, 250) 60 2,541 570 425 33 552 No. 24 4-14-60 Female 570 (3, 320) 50 2.594 579 443 54 1,531 591 No. 25 5-12-50 Female 553 (3, 363) 47 2,627 692 410 54 1,495 617 No. 26 4-10-50 Female 791 (3, 629) 62 2,829 685 494 58 1,735 671 No. 27 5-8-60 Female 701 (3. 681) 49 2,869 609 480 47 1.683 621 ■ Measurements in parentheses estimated from regression data of table 3-E. Appendix Table 3-A. — Reduced regression statistics for various morphometric relationships, by species [Symbols follow Snedecor (1946); -Y=log total length in cm.; K=log weight in pounds] Location Source of data N sx SY SX' SY' SXY Sx' Sy' /. marlina: New Zealand-Australia Gregory and Conrad (1939)"-.. P0FI3... - . 12 6 5 27 48 13 30 23 56 27 5.271 14.686 2.341 12.338 21. 460 30. 672 10 294 79. 075 133. 496 12. 434 5.708 14.810 2.831 10.504 112.531 26. 773 58.886 55. 947 127.410 68.580 2. 380077 35. 963524 1. 102288 5. 647082 9. 627610 72. 4,59840 3. 594782 271.926081 318. 581533 5. 841042 3.691712 36. 757096 1. 661375 4. 255982 264. 261669 66. 429689 116. 392004 136. 977437 293. 195804 175. 673666 2. 756802 36. 307817 1. 342524 4. 832820 .50. 410703 63. 510290 20. 421341 192. 575952 304. 770010 31. 9849.50 0. 064790 .017091 . 006231 .009073 . 033202 . 902796 . 062567 . 062793 .350909 .114956 0. 976607 , 201079 Hawaii M. auttax: New Zealand-Australia New' Zealand Hawaiian Division Fish and and Game.' ^ Gregory and Conrad (1939)"... Morrow' {1952a)' . 058463 . 169537 . 444462 Central Pacific POFI s 1. 291725 Hawaii M. ampla: Bimini .._ Central Pacific- Hawaiian Division Fish and Game.' Conrad and LaMonte (1937)'.. POFI! . . .806637 . 887576 3.315303 Hawaii Hawaiian Division Fish and Game.' 1. 480366 Location Source of data Sxy I V 6 a « '250 X y:.x /. marlhta: New Zealand-Australia Central Pacific Gregory and Conrad (1939)' ». . POFn . 0. 249663 . 057874 . 017050 .032881 .099968 . 342486 . 215592 . 227734 1.046047 402.'i9O 2.439 3.448 2.468 2.467 2.447 2. 3.69 2.343 2.438 2.384 2 461 2.476 2.468 2.566 2.389 2.344 2.059 1.963 2.432 2.275 2 .540 3. 85188 3. 38622 2. 73631 3. 62405 3.01090 3. 69078 3. 44678 3. 62674 2,97811 3 .'i02I5 -6.919 -9. 208 -4. 187 -6. 615 -5. 024 -6.648 -6. 110 -6. 410 -4.825 — 6 079 0. 0391 .0357 .0627 .0449 .0568 .0502 .0477 .0642 .0613 0.531 208.0 199. 1 149.7 "" 159.2' 142.3 193.7 207.5 208 9 418.8 369 0 Hawaiian Division Fish and Game. '2 Gregory and Conrad (1939)' 2... Morrow (1952a) ' 390 0 M. audax: New Zealand-Australia New Zealand 289.8 271 7 Central Pacific Hawaii M. ampla: Bimini ... POFIi Hawaiian Division Fish and Game.' Conrad and LaMonte (1937)'.. POFn.. 311.9 266. 1 374 2 Central Pacific . - 356.5 Hawaii Hawaiian Division Fish and Game.' 394 5 ' Sums of X, A'!, and A' V computed in log of meters. 2 Sums of Y, y2, and A' Y computed in log of weight in hundreds of pounds. s Sums of X, A'2, and A'K computed in log of centimeters. < Sums of A', A'', and XY computed in log of millimeters. SPEARFISHES OF THE CENTRAL PACIFIC Appendix Table 3-B. — Reduced regression statistics for various morphomelric relationships, by species (Symbols follow Snedecor (1946); X=tork length in cm.; y=head length In cm.) 551 Location Source of data N- sx SY SX' SY> SXY Sr» Sv' 7. maTlina: Now ZoaIand-.\ustralia Central Pacific Gregory and Conrad (1939) POFI 12 9 30 20 23 58 3347,5 2571. 9 8463.0 5069.1 6353.3 15165. 9 1149.5 917.8 3037.5 1843.0 2161.3 5126. 3 960824.67 744995. 49 2400373.04 1318922. 51 1779512. 43 4123041.91 112924.59 94846.98 309132.89 173745. 38 205521. 07 474395. 33 329095.76 265754. 50 861104.33 478411.45 604409.38 1397338. 84 27011.6492 10032.2000 12960.7400 34133. 7695 24537. 6087 157446. 6891 2812 0692 1251.7756 1586.0150 3912 9300 ,\/. audai: .New Zealand-Australia Central Pacific Gregory and Conrad (1939) .\/, ample: Bimini Conrad and LaMonte (1937).. POFI 2424 6487 Central Pacific 21309 9561 Location Source of data Sri/ X 5 b a 9 y,.z 'mx /. marlina: New Zealand-Australia Gregory and Conrad (1939).... Gregory and Conrad (1939) POFI 8433. 1559 3477. 8534 4225.5800 11293.8850 7392.5214 56908. 6129 278 958 285. 766 282. 100 253. 455 276. 230 261.481 95. 792 101. 978 101. 250 92.150 93.970 88.384 0.25582 .34667 .32603 .33087 . 30127 . 16145 -17.438 2.912 9 277 8 289 10.750 -6 128 2.928 2.566 2.728 3.128 3.067 3.6.16 86.75 89 58 90.79 91.01 86.07 84 24 102.36 ^f. audaj: .New Zealand-Australia 107.09 .\/. ampla: Bimini . Conrad and LaMonte (1937).. POFI 101 13 102 31 Appendix Table 3-C. — Reduced regression statistics for various morphometric relationships, by species (Symbols follow Snedecor (1946); Jf=snout to orbit in cm.; y=length of mandible in cm.) Location Source of data N SX SY SX' SY' SXY Sx' Sy= /. marlina: Xew Zealand-Australia Cenlral Pacific Gregorj' and Conrad (1939).... POFI 12 7 29 9 21 25 744.1 468 1 1960.5 556.2 1314. 5 1531.8 382.3 290.2 1056.8 374.8 599.9 865.7 47201. 65 31758. 01 133262. 19 35217. 54 83237.27 100666.60 12664. 85 12210. 94 38963. 40 16132. 20 17326. 69 31273. 67 24337.68 19676.41 71877.41 23808.67 37917. 87 55941. 10 1061. 2492 455. 4943 725. 6297 844.3800 955. 8296 6810. 1504 485. 4092 180 0771 ?.I. audax: .New Zealand-Australia Central Pacific Gregory and Conrad (1939).... POFI 452. 1504 523 8622 A/, ampla: Bimini Conrad and LaMonte (1937) . . POFI 189 5467 Central Pacific 1296. 2104 Location Source of data Sry z » 6 a t Kx Kx /. marlina: New Zealand. .Australia Gregory and Conrad (1939) . . . 631. 8942 270. 3214 434. 0859 646.0300 366.9867 2897. 9296 62.008 66.871 67.603 61.800 62. 595 61.272 31.858 41. 457 36.441 41.644 28.567 34.628 0. 59543 .59347 .59822 .76509 .38395 .42553 -5.063 1.771 -4.000 -5.639 4.534 8.555 3.304 1.982 2.670 2.056 1.600 1.656 27.69 34.41 28.90 36.44 25.65 31. 959 36.62 43 31 M. audas: New Zealand-Australia Gregory and Conrad (1939).... POFI 37.88 47 92 M. ampla: Conrad and LaMonte (1937) . . POFI 31 41 Central Pacific 38.342 Appendix Table 3-D. — Reduced regression statistics for various morphometric relationships^ by species [Symbols follow Snedecor (1946); Ar=fork length in cm.; F=greatest body depth in cm.] Location Source of data N SX sr SX' SY' SXY Sr» Sy' /. marlina: New Zealand- Australia Central Pacific Gregory and Conrad (1939).... 12 7 30 46 21 28 23 61 27 3347.5 2030.1 1473. 7 8463.0 12866.1 5200.1 6233.6 6353.3 16047.7 7879 3 647.1 379 9 287.7 1342.2 2014.3 791.9 891.1 1112.5 2709.1 1384.1 960824.67 594384. 21 437125. 11 2400373. 04 3612137. 27 1314184. 11 1401800. 78 1779512. 43 4389237. 69 2343583.77 36748.28 20874.09 16669 59 60649. 36 88823.75 31007. 63 28865.65 55274. 25 128055.79 72774. 59 187423. 86 111351.36 85231.60 381105.30 565496.86 201192.09 200862.70 312813.85 747114.63 412177.08 27011.6492 5626.2086 2766. 7720 12960.7400 13517. 0698 26515. 5381 14023. 3172 24537. 6087 167456. 1379 44199. 7519 1853. 4125 256.3743 Hawaiian Division Fish and Game. Gregorj- and Conrad (1939).... Morrow (1952a) IIS. 3320 A/, audai: New Zealand-Australia New Zealand 599.3320 619 3046 Cenlral Pacific. POFI 1145.4581 Hawaii Hawaiian Division Fish and Game. Conrad and LaMonte (1937). . POFI... 506.3925 M. ampla: 1463. 1087 Central Pacific 7740. 6620 Hawaii. . . Hawaiian Division Fish and Game. 1821. 5230 552 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Appendix Table 3-D. — Reduced regression statistics for various mor-phometric relationships, by species — Continued (Symbols follow Snedecor (1946); A'=fork length in cm.; K=greatest body depth in cm.) Location Source of data Szy I. marlina: New Zealand-Australia Central Pacific Hawaii M. audax: New Zealand-Australia New Zealand _ . . _ Central Pacific Hawaii AI. ampla: Bimini Central Pacific -.- Hawaii Gregory and Conrad (1939).-.- POFI Hawaiian Division Fish and Game. Gregory and Conrad (1939)... Morrow (1952a) POFI Hawaiian Division Fish and Game. Conrad and LaMonte (1937) . . POFI Hawaiian Division Fish and Game. 6909. 9225 1174.9329 434. 9020 2470. 6800 2101. 5289 6098. 7952 2478. 3800 5507. 4915 34412. 4961 8260. 8160 278. 9.58 290.014 294. 740 282. 100 279. 698 247. 624 222. 629 276. 230 263. 077 291.826 53. 925 54. 271 57.540 44.740 43, 789 37. 710 31.825 48. 370 44.411 51.263 0. 26582 .20883 . 15719 . 19063 . 15547 . 19229 . 17673 . 22445 .20560 . 18690 -17.438 -6.293 11.210 -9.037 .304 -9.906 -7.620 -13.630 -9.651 -3. 279 2.928 1.484 3.967 2. 141 2.579 2.947 1.622 3.287 3.367 3.332 46.52 45.92 50.61 38.62 39.17 38.67 36.66 42.48 41.72 43.45 59.31 56.36 58.37 48.15 46.95 47.78 45.50 63.71 52.00 52.79 Appendix Table 3-E. — Reduced regression statistics for various morphometric relationships, by species [Symbols follow Snedecor (1946); measurements In cm.: specimens from POFI collection in Central Pacific] Relationship and species N SX SY SX' SY2 SXY Si' Sy' .Y= greatest body depth y=height 1st dorsal 1. marlina . - M. audax M. ampla A'= tip of snout to upper tail notch y=fork length /. marlina M. aiLdar M. ampla- _. __ A^=naris to fork of tail y=fork length I. marlina M. audar M. ampla . X= posterior edge orbit to fork of tail y=fork length /. marlina M. audar... _,_ M. ampla 329.4 791.9 2623.1 1834.6 2168.3 6096.8 2211. 7 3971. 4 12430. 7 1716.8 1741. 6 5100.8 202.2 898.5 2010.2 1983. 5 2349,9 6594. 8 2856.5 5184. 8 16793. 2 230.49 2349.9 6794. 7 18323. 84 31007, 63 126396, 29 486997, 19 538835, 07 1556682.98 496707. 37 773558, 78 2616040. 45 373314. 74 348776. 21 1045728. 88 6935. 74 39045, 69 72748. 00 569335. 99 632949. 73 1820646. 10 825424, 45 1314676. 26 4233130. 96 672633. 95 632949. 73 1860606. 11 11266. 11 34640. 32 96316. 50 626564. 53 583992. 11 1683480. 40 639618. 84 1008316. 16 3327267. 42 501082. 16 469790. 74 1394686. 89 239. 7800 1146.4581 7763. 3313 6227. 1643 16443.4156 70331. 8744 6545.0810 22510. 3068 82887. 9502 4889. 4600 11795. 9600 45030. 3939 121.6000 602. 7257 3077. 2407 7297. 0972 19390. 8400 80990. 6184 10036. 4250 34573. 8296 144193. 8086 8563. 4487 19390. 8400 84915.7989 Relationship and species Sirj depth depth A'=great<>st body depth y=height 1st dorsal /. marlina ._ M. audax , At. ampla jr=tlp of snoi't to upper tail notch y=fork length I. marlina JV/. audax AI. ampla X=naris to fork of tail y=fork length /. marlina Af. audax -- M. ampla Ar=posterior edge orbit to fork of tail >'=fork length /. marlina Af. audax M. ampla _ 164.3300 758.3129 4402. 4721 6736.8514 17848.9800 75457. 1264 8067. OO-W 27796,4015 108888. 0554 6450.6200 16085. 0800 61670. 2839 54.900 37. 710 45.226 262. 071 240. 922 243. 832 221. 170 189.114 203, 782 214.600 193.500 196. 185 33.700 42. 786 34. 659 283. 357 261. 100 263. 792 286. 6.60 246. 895 258.905 288.113 261, 100 261,3.35 0. 68534 .66202 .56709 1.08169 1,08648 1. 07287 1. 23255 1. 23483 1.31368 1. 31929 1. 27884 1, 36953 -3. 925 17. 821 9.012 -.123 -.416 2.192 12. 947 13, 371 -8. 799 4.993 13.644 -7.346 1.498 2 302 3.220 1.483 1.618 1.232 3.397 3.627 4.414 2.978 3.769 4.361 23. 19 44.30 31.70 40.62 64.23 40.20 SPEARFISHES OF THE CENTRAL PACIFIC 553 Appendix Table 4. — Reduced statistics for ratios and mean lengths of varioxis parts, by species [Symbols used (ollow Snedccor (1946)) Species and location Source of data N Mini- mum Maxi- mum SX SX' Sr« I > snout to orbit head length /. marlina: New Zealand- Australia Gregory and Conrad (1939)..., POFI.. 12 7 30 9 23 26 11 10 5 30 25 28 23 63 27 12 11 30 47 19 25 23 57 26 12 8 30 9 23 39 11 8 30 22 23 58 11 9 30 12 22 33 9 6 11 24 35 27 0.698 .642 .626 .638 .637 .575 .121 .103 .102 .147 .151 .116 .118 111 111 .176 .173 .166 .175 .178 .148 .168 .166 .153 .331 .336 .284 .320 .327 .325 .681 .751 .502 .591 .775 .764 212 186 230 220 242 257 45 56 77 82 35 33 0.683 .688 .701 .699 .694 .694 .134 .128 .119 .185 .198 .184 .161 .148 .150 .204 .209 .216 .220 .212 .212 .198 .207 .196 .373 .371 .379 .382 .395 .386 .766 1.004 .811 .686 .963 1.002 283 304 373 387 427 418 78 68 127 138 85 77 7.783 4.610 20.057 5.971 16.230 17. 193 1.397 1.158 .547 5.048 4.308 4.551 3.246 8.116 3.560 2.235 2.031 5.832 9.113 3.665 4.627 4.209 10.458 4.728 4.183 2.861 10.018 3.179 8.211 14.213 8.658 6.712 20.328 14.211 19. 621 60.233 277.1 233.2 968.8 395. 9 765.7 1,111.6 60.8 30.9 110.4 255.6 217.7 148.5 6. 054625 3.037550 13.416037 3.964641 10.089880 11.382779 . 177657 . 134552 .060039 .853002 .746056 .744995 .460900 1.060914 . 472408 . 416825 . 376996 1. 137588 1. 772563 .704853 . 861269 . 771899 1. 924614 .862264 1. 469817 1. 024495 3. 360392 1. 127497 2. 936609 5. 188813 6. 674676 5. 678072 13. 878122 9. 195089 16. 789827 43. 65%23 7, 036. 01 6,141.88 30,933.88 13, 433. 21 27, 023. 19 38,052.96 418.44 191.85 1,131.88 2, 762. 72 1, 401. 39 840.31 0.006601 .001536 .006595 .003214 .004971 .013577 .000238 .000456 .000197 .003592 .003701 .006295 . 002791 .005367 .003016 .000566 .000999 .003847 . 006610 .001746 .004894 . 001662 .006846 . 002485 .001693 .001330 .015048 .004604 .005282 .009083 .016552 .046704 . 103869 .015429 .051408 . 153515 55. 6091 99.4089 290.6320 371.8092 373. 3495 608.8825 7. 7022 .8880 23.8655 40.5800 47.2960 23.5600 0.6486 .6586 .6686 .6634 .6622 .6613 .1270 .1158 .1094 .1683 .1723 .1625 .1411 .1288 .1319 .1863 .1846 .1944 .1939 .1924 .1851 .1830 .1835 .1818 .3486 .3576 .3339 .3532 .3570 .3644 .7780 .8390 .6776 .6460 .8531 .8661 25. 191 2,^.911 31.960 32.992 34.805 33.685 6.756 6.180 100.36 106.50 6.220 5.500 0 02450 M. audax: New Zealand-Australia... Gregory and Conrad (1939) 01508 Central Pacific 02143 M. ampla: Conrad and La Monte (1937).. POFI Central Pacific 02378 height 1st dorsal fork length /. marlina: Gregory and Conrad (1939).... POFI Central Pacific 00755 Hawaii Hawaiian Division Fish and Game. Gregory and Conrad (1939)..-. POFI 00810 M. audax: Central Pacific 01269 Hawaii - Hawaiian Division Fish and Game. Conrad and LaMonte (1937).. POFI Hawaiian Division Fish and Game. Gregory and Conrad (1939)..- POFI 01400 M. ampla: Bimini Central Pacific --. .01126 00930 Hawaii 01098 pectoral length fork length /. marlina: .00711 Central Pacific 01064 A/, audax: Gregory and Conrad (1939) Morrow (1952a) .01152 01104 Central Pacific POFI. .00985 Hawaii Hawaiian Division Fish and Game. Conrad and LaMonte (1937).. POFI 01428 M. ampla: Bimini . _ . 00867 Central Pacific .01022 Hawaiian Division Fish and Game. Gregory and Conrad (1939).... POFI .00997 caudal spread fork length M. marlina: .01241 Central Pacific .01489 A/, audax: Gregory and Conrad (1939).... POFI .02278 Central Pacific .02565 Af. ampla: Conrad and LaMonte (1937)... POFI - - .01549 .01567 ,, height 1st anal height 1st dorsal /. marlina: Gregory and Conrad (1939) POFI... Gregory and Conrad (1939).... POFI .04068 Central Pacific .08168 A/, audax: New Zealand-Australia .05985 Central Pacific . 02778 At. ampla: Bimini Conrad and LaMonte (1937).. POFI - .04834 Central Pacific .05236 A'= pelvic length (cm.) /. marlina: Gregory and Conrad (1939).... POFI 2.358 Central Pacific 3.525 A/, audax: Gregory and Conrad (1939) POFI 3.166 6.098 Af. ampla: Bimini Conrad and LaMonte (1937).. POFI 4.216 Central Pacific 4 432 X^= length 20th ray of 1st dorsal (cm.) /. marlina: Central Pacific POFI . Hawaiian Division Fish and Game. POFI Hawaiian Division Fisli and Game. POFI .9812 Hawaii .544 Af. audax: Central Pacific . 1.628 Hawaii 1.328 Af. ampla: Central Pacific 1.197 Hawaii Hawaiian Division Fish and Game. .971 554 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Appendix Table 5. — Summary of stomach contents and sexual condition of the POFI specimens Specimen Istiompai marlina: No. 1... No. 2... No.3-- No. 4.-- No.5..- No.6..- No 7-.- No.8... No. 9... No. 10-- No. 11-- Maktira audai: No. 1. No. 2.. No. 3-. No. 4.. No. S.- No. 6.- No. 10. No. 11. No. 12. No. 13. No. 14. No. 15. No. 16. No. 17. No. 18. No. 19. No. 20. No. 21. No. 22. No. 23. No. 24. No. 25. Makaha ampta: No. 2- No. 3. No. 4-. No. 5. - Xo. 6. . No. 7.. No. 8.. No. 9. . No. No. No. 12. No. 13. No. 14. No. 16. Stomach contents Empty... 1 aku (30 em.) Empty 1 Mola (18").-. No data 2 Mola; 1 squid (8 cm.). 1 Mola (51b.) Albacx)re (94 cm.) Fish remains Empty Large fish skeleton; 6 vertebrae (5 cm.); fin rays (8"). 1 squid; 2 tunas (juven- ile), 50% fish; 50% squid Anns and bramid 1 fish (100 lb.) Empty 2 fish remains; 1 tuna- like fish. 2 tunalike fish (35 and 32 cm.) 1 Auiis (33 cm.); 1 squid with 8 cm. mantle; fish hones. Fish remams No data.. Squid and seombrid fish. Empty do Flying fish remains; car- apace of large shrimp. Empty No data Afola (45 cm.); fish re- mains. 1 small seombrid No data 40% fish; 60% squid.... Empty. do 2 squid (5 cm.); remains 1 small Mola. 20% squid: 80% fish; re- mains 1 surgeonfish (5 em.). 45% fish: 55% squid 100% fish Empty.- Bait; squld;2scombrids (6"). Tunicates - 3 squid (6 cm. each); 3 fish (10-15 cm. each) well digested. 1 tunalike fish (20 cm.)-- 1 gem)iylid (30 cm.); 2 fish (4 cm. each). No data Fish vertebrae: squid; 2 (4") flsh. Bait Fish remams, including 2 Mola (1 about 30 cm.). Sex Male ? Male Female- Female-. do— Female- Male Female-. Male Female- . do-.. Female.. do-. do-- ? Female- ? ? 7 Male., .do., -do-. ?- Female- Male--- ---.do- .do. .do- Female. Male-.- - do- Sexual condition No data. Maturing. Immature. Not active. No data. Gonads very thin. Immature. No data. Not active. Immature. Not active. Immature. No data. Very immature. Immature. No data. Do. Testes (8 mm. di- ameter) with milt. Not active (ovary 1 X 4 X 25 cm.). Small testes (1 cm. di- ameter) with milt and fatty appearance. Ovaries enlarged, but not near spawning (6x3 cm. diameter). Immature. No data. Immature. Do. Not quite mature. Immature. Ova not visible to naked eye. Immature. No data. Mature. No data. Do. Immature. Milt m center gonad. Young. Little milt. No data. Immature. Maturing. Milt spurts from duct. No data. Milt in lumen of testes. Maturing. No data. Do. Testes pmk with fatty appearance, milt in central duct. Specimen Makaira ampla — Continued No. 16... No. 17. No. 18. No. 19. No. 20. No. 21. No. 22- No. 23- No. 24. No. 25. No. 26. No. 27. No. 28. No. 29. No. 30. No. 31. No. 32. No. 33- No. 34- No. 36. No. 37-- No. 38-. No. 39- - No. 40- No. 41- No. 42- No. 44- No. 45. No. 46. No. 47- No. 48. No. 49. No. 50. No. 51. No. 52. No. 53. No. 54.. No. 55- No. 56. No. 58. No. 60. No. 61. No. 62. No. 63. No. 64. No. 66. No. 67. No. 68. Stomach contents Fish remains, including 1 Coryphaena (about 30 cm.). 4 small squid (6 cm.) ; re- mains of 1 flsh (3 cm.). Empty 1 small unidentified fish. 1 unidentified flsh (10 cm.). Almost empty; frag- ments of fish; squid remams. No data 3 unidentified flsh (about 15 cm. each). 80% fish; 20% squid Nodata Iflsh (15 cm.) Empty do - 1 Katsuwonus (?) (2 Ib.). Empty Fish vertebrae 90% fish; 5% squid; 5% crustaceans. Empty ,- 1 small Coryphaena 50% cephalopods; 50% f)sh, including 1 Cory- phaena (15 cm.). Empty Octopus (H lb.) - 90% fish; 10% cephalo- pods. Empty 2 baits 5 Auxis (33 cm., avg.)- 2 pieces bait; 1 small xmidentified fish. Empty 2 Ka tsuwonus (1 about 65 cm ., other d igested) . Empty—, -.--do-- do... No data. ---do.-- Empty... ----do.... 1 tun? (42 em.)- No data. ----do 1 bait -- Nodata-- Empty ---do Fish bones and squid . . 1 Katsuwonus (66 cm.). 1 Kalsuwonus (28 em.) . 1 Katsuiomius (81b.) Nodata Sex Male., -do. -do- -. -do—, -do.... -do—. -do.— -do— -do... Male... do- do- ? •? Male.. Female- Male ----do.. do... do-.. Female- Female- -do- Female-. ....do... .---do... ....do... ....do... -do- do- do-. -do-, -do- .do.. ?. ?. Female. -do- Female.. Sexual condition Milt runnmg from cut testes. Milt m testes. No data. Do. Milt in testes. Do. Do. Mature. No data. Do. Very immature. No data. Immature. No data. Milt in testes. No data. Slightly running milt. Immature. Very immature. Milt flow-s freely when cut; testes en- larged ( 6 cm. di- ameter) . Immature. Milt in testes. Testes with ruiming milt when cut (about 4 cm. di- ameter) . No milt visible. Milt in testes. Not active (ovary 3 cm. diameter). No data. Do. Gonad with thick wall: mside like nonspawnmg tuna ovary. No data. Do. Do. Mature. Not ripe. No data. Gonad enclosed in heavy connective tissues; semicylindrical; does not appear fatty as males. Not active (ovary about 2^*2 cm. di- ameter) . Eggs visible to naked eye. No data. Do. Do. Do. Do. Immature. Not active (ovary 2x 4cm. diameter). No data. Spent. Not active. U.S. GOVERN MENT PRINTING OFF ICE : 1 958 O 435062 I TREATMENT OF SULFONAMIDE-RESISTANT FURUNCULOSIS IN TROUT AND DETERMINATION OF DRUG SENSITIVITY By S. F. SNIESZKO, Bacteriologist, and G. L. Bullock, Fishery Aid During tlic past 2 years, tlie Microbiological Laboratory, at Leetown, W. Va.,' has received an increasing number of reports of outbreaks of furunculosis that failed to respond to treatment with sulfonamides. Tlie problems resulting from furunculosis, a widespread, bacterial disease of salmonids, and control of the disease have been reviewed in detail by McCraw (1952) and Snieszko (1954a). The latter investigator recommended the use of antibiotics for diseased fish that failed to respond to treatment with sulfonamides. In some of the recent incidences of furunculosis at the Lee- town station the disease did not respond to sulfon- amide therapy, but control was effected with chloramphenicol. Failure of sulfonamides to con- trol furunculosis suggested that the causal organ- ism, Aeninwnas mlinoiiicida, was resistant to the drugs usetl. Several methods are widely used in the routine determination in vitro of the sensitivity of bac- teria to sulfonamides and antibiotics. The results obtained with antibiotics are generally accepted as reliable aids in the selection of the most prom- ising treatment. Such, however, is not the case with sulfonamides, ami "The relation of in vitro sensitivity test to clinical effectiveness is still con- troversial" (Burdette, Plank, and Clapper, 1955). The experiments presented in this paper followed three lines of endeavor: (1) A comparison of the effectiveness of sulfonamide and antibiotic therapy in strains of trout which were either susceptible or resistant to furunculosis. (2) Comparison of the therapy of furunculosis with sulfonamides and chloramphenicol in trout wliich were suffering from furunculosis caused by sulfonamide-resistant, but chloramphenicol-sensitive, strains of A. sal- monicida. (3) The development of a standard method for laboratory and field use of determin- ing in vitro the sensitivity of A. salmonicida to sulfonamides and antibiotics. 1 Po.st Office, Kearnevsville, W. Va. Note— Approved for publication, March 12, 1957. Fishery Bulletin 125. Many factors are capable of affecting in-vitro tests of microbial sensitivity to sulfonamides; therefore, correct interpretation and application of the results of such tests will be reliable if a proved method of obtaining reproducible results is used. As the result of the findings presented in this study and practical experience gained in the treatment of this disease in trout hatcheries over the entire country, a revised procedure has been evolved and describetl for the treatment of this disease. The authors wish to express their thanks to Dr. K. E. Wolf for his assistance during the prepara- tion of the manuscript and for furnishing the strains of ^1. salmonicida isolated in Iowa, Utah, Minnesota, and Wisconsin; and to Dr. R. E. Lennon for supplying the fingerling brook trout from Erwin, Tenn. MATERIALS AND METHODS Treatment with sulfonamides and chloramphenicol Experiments on tlie therapy of furunculosis were carried out with two strains of fingerling brook trout {Salrelinux fontinalis) and a strain of brown trout [Salmo trutta). A strain of brook trout from Bellefonte, Pa., was selected because it was known to be resistant to furunculosis (Wolf 1954; Snieszko 1954b), and a strain from Erwin, Tenn., was used because it was suspected to be suscep- tible. The brown trout were from Cortland, X. Y. In general, brown trout are known to be more resistant to furunculosis than are most strains of brook trout (McCraw 1952; Wales and Berrian, 1937). Equal weights of the trout were distributed among stainless steel troughs in which the water was maintained at temperatures of 12° to 13° C. (54°-55° F.). Infection of the fish was initiated by adding fresh cultures of AeromonaK salmonicida to the diet, and treatment was started when the first mortality due to furunculosis occurred. 555 556 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE Sulfonamides were administered at the rate of For the inoculation of the solid media, 24-hour- 200 milligrams per kilogram of trout per day, or old cultures in nutrient broth were used. These 9gramsper 100 pounds of fish. Chloramphenicol^ were streaked with cotton swabs on 10-cm.- was given at the rate of 75 milligrams of pure diameter petri plates, each containing 20 ml. of antibiotic activity per kilogram of fish per day, medium, and the disks were placed in position. or 3.4 grams per 100 pounds of fish. Treatment The quantitative sensitivity tests were run in was continued to the end of the observations liquid media. In order to select the most satis- recorded in tables 1 to 3. Drugs were mixed factory substrate, the following media were tested: with the quantities of food calculated from the Leetown standard furunculosis medium with- Cortland Hatchery feeding charts (Deuel et al., out agar. 1952). A daily record of mortality was kept and Vitamin-free casamino acids (Difco), 1 per- most of the dead trout were examined bacteriologi- cent, pally. Tryptic digest of casein, 1 percent. Nutrient broth (Difco). Drug sensitivity tests in vitro Nutrient broth (Difco) buffered with 0.5 The drug sensitivity tests were performed with percent of secondary sodium phosphate 84 cultures isolated from the experiment on j^^j adjusted to desired pH. therapy here described or procured during the Proteose peptone No. 3 (Difco), 1 percent, past years from other experiments or received N-Z-Case (Sheffield Chemical Co.), 1 percent, from other trout hatcheries (table 4). The rou- Nutrient broth buft'ered with sodium phosphate tine qualitative sulfonamide and antibiotic sensi- .^^^s selected as most promising for our purposes, tivity tests were run with commercial multiple Because most of the sulfonamides are only sensitivity paper disks ^ on solid bacteriological slightly soluble in water, stock solutions of sodium media. Sensitivity to sulfonamides has also been g^itg were prepared by dissolving a known quan- determined in a quantitative manner by prepara- jity of sulfonamide in diluted sodium hydroxide, tion of serial dilutions of these drugs in liquid Excess alkali over that needed for keeping sulfo- bacteriological media. namides in solution was neutralized with hydro- The following solid bacteriological media were chloric acid. Stock solutions were sterilized by used with the paper disks: filtration, stored under refrigeration, and used 1. Leetown standard furunculosis medium within a week. Serial dilutions of 1:25 or 1:50 having the following composition: stock solutions were carried out in 5 ml. of media. Tryptic digest of casein 10 grams. The first tube contained double-strength medium. Yeast e.xtract 5 grams. and after the serial dilutions were made 5 ml. Sodium chloride 2.5 grams. were discarded from the last tube. Inocula con- Agar 15.0 grams. gjg^p^j ^f q^ j^j of ^ 24-hour-old culture diluted • 1 : 5 with sterile water. All cultures were mcubated prl D.o— /.U. - , at 20° C. and results were read as soon as abundant lA. The same, but buffered with 0.5 percent ^^.^^^j,^ developed, usuallv within 24 to 48 liours. of secondary sodium phosphate and adjusted to pH 8.0. RESULTS 2. Sensitivity Test medium C (Case I^abora- Treatment with sulfonamides and chloramphenicol *°"''^''' The response to treatment with sulfonamides 2A. The same, but buffered aiul adjusted as ^nd chloramphenicol of experimental furunculosis ^o- 2. in two strains of brook trout and one strain of 3. Mueller Hinton medium (Bacto). brown trout is presented in tables 1 to 3. Sul- 4. Trypticase Soy agar (Baltimore Biological fonamide treatment of experimental fiuunculosis Laboratory). in the Erwin brook trout failed completely (table 1). Losses were high because the trout were ' Racemic Chloromycetin, Parke. Davi.sA Co.. was used at a cloiibli. rate SUSCCptible tO furunculosis and tlu' Strain of ^1. because the manufacturer advise.l that it ha.s about .10 percent the thera- .■.almonicida USed tO produce the disease developed peutic value of the d-rotatory i.somer . ■ t i • > Muiiidisits, Case Laboratories, Chicago, 111. rcsistaucc to Sulfonamides. When tlic experuueiit DRUG RESISTA^•CE OF FURUNCULOSIS IX TROUT 557 was repcati'd with the same strains of trout and bacteria, but witli treatment with chlorampheni- eol, the response was rapid and favorable. A similar experiment was run concurrently with a strain of brook trout from Bellefonte, Pa., tliat is known to be resistant to furunculosis. The fish failed to respond to sulfonamide treatment and bacteria isolated from dead trout were found also to be mostly resistant to sulfonamides. Trout surviving the sulfonamide treatment were given chloramplienicol and losses stopped within the first o-day period (table 2). The overall fish losses were much lower in the resistant Bellefonte strain than in the susceptible Erwin strain. The results with brown trout (table 3) were very similar to those obtained with Bellefonte brook trout since they are known to be more re- sistant to furunculosis than most of the strains of brook trout. Table I. — Mortality of a strain of eastern brook trout {Erwin, Tenn.), susceptible to furunculosis, foUowiny treatment with sulfonamides and chloramphenicol A. TREATMENT WITH SULFONAMIDES ' tinltial weight of fish per lot (trough), 1.500 gm.; mean weight per fish. 19.8 gm.; mean number of fish per lot (trough), 76; fish infected orally July 29 and .\ug. 1. 1955. and treatment started .\ug. 8; mortality e.^pressed as percent of fish at beginning of each period] Mortality under treatment with— Period Sulfamerazlne Sulflsoxazole (Qantrisin ') controls Lot A LotB Lot A LotB Lot A LotB Aug. 7-11 7.8 12.6 37.0 33.3 6.5 16.6 63.3 82.0 6.8 22.0 49.0 48.1 2.6 39.1 75.5 63.6 9.1 26.0 48.0 41.0 Aug. 12-16 55 0 Aug. 17-21... .\ug. 22-24 80 0 Total mortality.. 66.2 95.0 81.0 94.7 79.2 97.3 -\verage. total mortality 80.6 87.9 88.2 B. TREATMENT WITH CHLORAMPHENICOL [Initial weight of fish per lot (trough), I.-IOO gm.: mean weight i)er fish, 30 gm.; mean number of fish per lot (trough), 50; fish infected orally Sept. 8 and 9, 1955. and treatment started Sept. 16; mortality expressed as percent of fish at beginning of each period] Period Mortality under treatment in — Mortality in controls Lot A LotB Lot A Lot B Sept. 16-20 4.1 11.0 0 0 14.0 16.0 0 0 9.1 32.0 38.2 50.0 Sept. 21-25.. Sept. 26-30 67 0 Oct. 1-5... Total mortality... 14.2 27.5 87.2 Average, total mortality... 20.8 93.5 Table 2. — Mortality of a strain of eastern brook trout (Bellefonte, Pa.), resistant to furunculosis, following treatment with sulfonamides and chloramphenicol A. TREATMENT WrPH SILFONAMIDES ' llnitiul weight of fish per lot (trough). l.iiOO gm.; mean weight per fish. 27 6 gm.: mean luimber of fish per lot (trough). 54; fish infecteil orally July 29 luid .\ug. 1. 19,5.5, and treatment started .\ug. 19: mortalitv expressed a.s percent of fish at beginning of each period) Mortality under treatment with— Mortality in Period Sulfamerazlne Sulflsoxazole Lot A LotB LotC Lot D LotE LotF Aug. 17-21 Aug. 22-26 . . Aug. 27-31 Sept. 1-5 0 0 3.5 1.8 1,8 7.5 4.0 10.6 0 1.6 1.6 10.1 4.0 0 0 0 3.6 7.5 8.1 4.4 0 4.0 4.1 0 Total mortality... 5.3 22.2 13.1 4.0 21.8 8.0 .Average, total mortality 13.7 8.5 15.0 B. TREATMENT WITH CHLORAMPHENICOL (Fish surviving treatment with sulfonamides (2-.\) used in thi.s experiment; weight of fish per lot (trough). 2,000 gm.; mean weight per fish. 43 gm.; mean number of fish per lot (trough), 46; no additional infection, with treatment started Sept. 7. 1955; mortality expressed as percent of flsh at beginning of each period] Period Mortality under treat- ment in — Mortality in controls Lot A LotC LotE LotB Lot D LotF Sept. 6-10 9.4 0 0 0 0 7.5 0 0 0 0 0 0 0 0 0 12.0 10.8 3.0 3.1 0 0 0 2.1 0 2.1 4 3 Sept. 11-15... Sept. 16-20 Sept. 21-25 Sept. 26-30 0 2.2 Total mortality... 9.4 7.5 0 26.1 4.2 8.7 Average, total mortality 5.7 10.8 ' Strains of .4. salmonicida isolated from dead fish were resistant to sulfon- amides. Table 3. — Mortality of brown trout following treatment with sulfonamides and chloramphenicol A. TREATMENT WITH SULFONAMIDES ' (Initial weight of fish per lot (trough). l.,500 gm.: mean weight per flsh, 22.0 gm.; mean number of flsh per lot (trough), (58: fish infected orally July 29 and Aug. 1, 1955. and treatment started .Aug. 12; mortality expressed as percent of flsh at beginning of each period] ' Strains otA.salmonicida isolated from dead trout were resistant to sulfona- mides. ' Oantrisin used in these studies was supplied free of cost by HolTmann La Roche, Nutley, N. J. Mortality under treatment with Period Sulfamerazlne with sulfa- guanldine Sulflsoxazole Mortality in controls Lot A LotB LotC Lot D LotE LotF Aug. 12-16 Aug. 17-21 Aug. 22-26 3.1 0 4.8 10.0 5.6 3.0 I.S 3.1 8.0 10.3 1.3 0 2.7 5.5 8.8 1.7 3.4 0 7.1 3.8 1.4 0 1.4 8.7 4.8 6.6 5.2 13.0 Aug. 27-31 4.2 Sept. 1-5 2.2 Total mortality... 22.0 25.7 17.3 17. S 17.8 29.0 I strains of .4. fonamides. salmonicida Isolated from dead flsh were resistant to sul- 558 FISHERY BULLETIN OF THE FISH ANID WILDLIFE SERVICE Table 3. — Morlalily of brown troul foUnwing treatment with sulfonamides and chloramphenicol — Con. B. TREATMENT WITH CHLORAMPHENICOL [Fish surviving treatment with sulfonamides (3-A) used in this experiment; weight of fish per lot (trough). 1,500 gm.; mean weight per fish, 31 gm.; mean number of fish per lot (trough), 48; no additional infection, with treatment started Sept. 7, 1955; mortality expressed as percent of fish at beginning of each period) Period Sept. 6-10 Sept. 11-15 Sept. 16-20 Sept. 21-26 Sept. 26-30 Total mortality Average, total mortality Mortality under treat- ment in— Lot A Lot C Lot E 10.0 2.2 0 0 0 12.0 1.6 0 0 1.6 1.6 4.8 3.3 0 0 0 0 Mortality in controls Lot B Lot D Lot r 13 4 8.9 0 0 0 21.0 0 6.1 6 4 0 2.3 6.7 6.1 29.5 While the results obtained in the treatment of furunculosis caused by sulfonamide-resistant but chloramphenicol -sensitive strains of A. salmonicida are of considerable interest in their own merit, they are presented in this paper chiefly as back- ground information for the history of the majority of the strains of this bacterium which were used in the in-vitro sensitivity tests. Summary Table 4. — Sensitivity of 84 strains of A. salmonicitia to stdfonamides tested with sodium sulfadiazine in buffered nutrient broth Place of isolation Number of strains that were— Date and source of culture Sensi- tive' Inter- mediate Resist- ant 2 Total 1956: Brook trout: Not treated Treated with sulfona- mides. Treated with chloram- phenicol. Brown trout: Leetown do do do 0 0 0 1 2 2 1 0 5 0 12 9 6 1 9 14 10 6 7 Treated with several sulfonamides. do 11 3 8 36 47 Leetown Iowa Utah Minnesota. . Wisconsin... 1956: Brown trout: Not treated. Rainbow trout: Not treated . Do Brown and brook trout: Treatment (?). Brook trout: Treatment (?) . 13 2 0 4 3 2 0 0 0 0 2 0 2 0 0 17 2 2 4 3 Total 22 2 4 28 Leetown Before 1965: Miscellaneous stock cul- tures stored in lyophi- lized form since 1962. 4 2 3 9 Grand total 29 12 43 84 ' Cultures in which there was no growth within 24 hours in buffered nu- trient broth having a pH of 8.0 and containing sodium sulfadiazine in the concentration of 1:1,600. ' Cultures in which there was undiminished growth in nutrient broth containing sodium sulfadiazine in the concentration of 1:100. information on the cultures used for testing of the drug sensitivity is presented in table 4. Sensitivity tests with antibiotics Tests of the sensitivity of Aeromonas salmonicida to various antibiotics were performed with 47 strains of the pathogen that were isolated during the summer of 1955 from fingerling brook trout known to be either susceptible or resistant to furunculosis and from fingerling brown trout. Trout from wliicli the bacteria were isolated liad been treated with sulfonamides or chloramphe- nicol, or had served as controls (table 4). Cul- tures of tliese strains were grown on the Leetown standard furunculosis medium on which 6-tipped sensitivity disks containing various antibiotics had been placed (table 5 and fig. 1). All strains of A. salmonicida that were tested were found to be uniform in their response to disks containing antibiotics. No correlation between the sensitiv- ity to antibiotics and sulfonamides was apparent. Table 5. — Sensitivity of .\eromonas salmonicida to selected antibiotics as determined with MuUidisks Chloromycetin ' Terramycin 2__. Tetracycline... Aureomycin 3.. Erythromycin.. Streptomycin. . Neomycin Penicillin Carbomycin Bacitracin Polymyxin B. . Viomycin Furadantin *.,. Quantity of antibiotic in disk 10 ng do ... do... do... do... do... do ... 1.5 units.. 6 »ig 5 units .. 10 units.. 10 Mg 50 (ig Width of zones of growth inhibition (mm.) 14-19 12-14 10-13 10-12 5- 8 4- 6 1- 2 0- 1 0 0 0 0 8-10 ' Chloramphenicol. 2 Oxvtetracveline. 3 Chlortctraivi'line. < Nitri.fur;inhHn; related to furacin, which has possibilities in the treat- ment of furuiu-ulosis (Outsell 1948). Sensitivity tests with sulfonamides Tests were run in liquid and solid media. The first liquid substrate tested was the standard furunculosis medium containing serial dilutions of sodium sulfamerazine. The pH of the medium was 6.8 to 7.0 and the sulfonamide concentrations ranged from 1 : 100 to 1 : 10,000. Thirty resistant and 10 sensitive strains of A. salmonicida were used. The results were not satisfactory because sensitive strains of A. salmonicida grew as well as the resistant strains. Crystals of sulfamerazine appeared in test tubes with the higher drug con- centrations. Therefore, other liquid media such DRUG RESISTANCK OF FURUNCULOSIS IN TROUT 559 Tift Figure 1. — Sensitivity of Aeromonas salmonicida in vitro to various antibiotics as determined with three media: Trypticase Soy agar (TSA), Mueller Hinton agar (MH), and standard furunculosis agar (#4). The antibiotic and quantity used on Multidisks, reading clockwise, were as follows: Upper: Tetracycline (TE), 10/ig.; erythromycin (E), 10 /ig.; terramycin (T), 10 /ig. ; car- boniycin (CA), 5 /ig.; bacitracin (B), 5 units; aureomycin (A),10Atg. Lower: Viomycin (V), 10 /jg. ; Chloromycetin (C), 10 ^g. ; polymyxin B, (PB), 10 units; penicillin (V), 1.5 units; streptomycin (dihydro) (S), 10 /ig- i neomycin (X), 10 ^g. The presence of a clear zone indicates sensitivity. as proteose peptone #3, vitamin-free casamino aciils, nutrient brotli, and N-Z-Case were tested. Surprisingly, the most promising results were obtained witli nutrient broth, and this medium was used in all subsequent experiments unless other- wise inratories. ' A mixture of sulfamerazine, sulfamethazine, and sulfadiazine. the test was repeated with single disks, prepared at the Leetown laboratory, which contained 0.1, 0.25, and 0.5 milligram of sodium sulfadiazine. The results presented in table 8 show that the lack of growth on 10-cm. petri i)lates using multiple ilisks was due to tiie great width of zones produced by disks containing more than 0.1 milligram of sul- fonamide. Therefore use of multiple disks with liigher concentrations of sulfonamides and with Mueller Hinton and Trypticase Soy agar media in 10-cm. petri plates is not advisable. Table 8. — Determining sulfonamide sensitivity of A. salmoiiicida using two media and single disks Quantity of sodium sulfadiazine per disk Width of clear zone (mm.) on— O.I milligram. 0.25 milligram 0.5 milligram. Mueller Hinton medium (Dlfco) Mm. 12 13-15 15-16 Trypticase Soy agar (BBL) A/m. 7-8 10 12-13 The sulfonamide-sensitivity test in which filter paper disks containing drugs were employed could be much more conveniently performed under field conditions if petri plates were replaced by test tubes. Therefore, the buffered and alkalized standard furunculosis medium, Mueller Hinton medium (Difco), and Trypticase Soy agar (BBL) were used as agar slants. Media were inoculated by making longitudinal streaks with a loop con- taining a suspension of .4. salmunicida, and disks were deposited in the center of the slant. Excel- lent clear zones appeared (fig. 2) with sulfa- sensitive strains. Another approach to the performance of the sulfonamide-sensitivity test under simulated field conditions was made with agar media containing sodium sulfadiazine added to the bufferetl and alkalized standard furunculosis medium. These solutions were sterilized by filtration or were incorporatetl in the medium before sterilization in the autoclave. Li all cases, clear differences were noticed in the growth of sulfonamide- sensitive and sulfonamide-resistant strains of A. salmonicida. The optimum concentrations of sodium sulfadiazine in the medium were 1:500 and 1:1,000. It is expected that extensive field trials based on the results presented here will permit selection 562 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE I t S 51 s-r> ?.j-. i-r. S-S) 5-S; * T.^s uft TiO N'H lOo .'^C ica ■?»0 i'ce C Figure 2. — Sulfonamide-sensitivity test performed with a sensitive culture of A. salmonicida using Trypticase Soy agar (left) and Mueller Hinton agar (right). Lab- oratory-prepared disks contained 0.1, 0.25, and 0.5 milligram of sodium sulfadiazine. of the most convenient method for the determina- tion of the sensitivity of fisli-pathogenic bacteria to sulfonamides, antibiotics, and perhaps, to other therapeutic agents. DISCUSSION Micro-organisms appear to be capable of developing resistance to the majority of drugs which are employed clinically. In view of the cross-resistance shown between different groups of drugs, it is obviously important that the clinician should have information as to which groups of drugs may be usefully tried in infections with resistant strains (Work and Work, 1948). It has been found, as coukl be expected, that with the wide use of sulfonamides for treatment of fish diseases, there is an increasing number of outbreaks of furunculosis which are refractory to treatment with these drugs. Since "once resist- ance is liighly developed it is apparently perma- nent" (Northey 1948), it is increasingly important to have means of rapid and reliable determination of such resistance. In this paper, observations are described on the treatment of experimentally induced furunculosis employing strains of trout whicli are susceptible or resistant to this disease. The outbreaks were caused by strains of the pathogen which were found to be either sensitive or resistant to sulfonamides. Wlien it was found that the disease was refractory to treatment with sulfonamides, chloramphenicol was used with good results (tables 1 to 3). Oxy- tetracycline (terramycin), which is also effective in the treatment of furunculosis and ulcer disease (Snieszko et al., 1952), was not used in this investi- gation because strains of A. salmonicida resistant to this antibiotic have recently been isolated. It is evident from the results described here that in addition to chloramphenicol and o.xytetra- oycline other antibiotics and sulfonamides arrest the growth of A. salmonicida in vitro. The results of these tests caiuiot be applied indiscriminately for therapeutic use. While some of the sulfon- amides gave promising results in vitro, their use in the treatment of fish diseases may be limited by their to.xicity to some species of fish or by the arresting of the fish's growth during treatment. Other drugs may not be absorbed from the in- testinal tract and are therefore useless in systemic infections by micro-organisms sensitive to them in vitro (Johnson and Brice, 1953; Snieszko and Wood, 1955; Snieszko and Friddle, 1951; Snieszko and Griffin, 1955). It is interesting that chlor- tetracycline (aureomycin) which lias been found to be effective in vitro against A. salmonicida and Hemophilus piscium is entirely ineffective in the treatment of trout suffering from furunculosis and ulcer disease (Snieszko, Griffin, and Friddle, 1952). Therefore, the in-vitro tests, while very convenient because of the speed and ease witli which they can be performed, must be supplemented by exact treatment trials with several species of fishes, before tiie therapeutic value of drugs can be determined. The greatest value of the in-vitro test is in the determination of the acquired drug resistance of the pathogen concerned. This can be explained best by using an example. It is known that furuncu- losis usually can be effectively treated with stil- fonamides. It has also been shown in this paper that the growth of sensitive strains of A. salmoni- cida can be arrested in vitro with tliese drugs. Therefore, if it is found that a particular strain of A. salmonicida is resistant to sulfonamides in vitro, an outbreak of furunculosis caused by that strain will most likely be refractory to a treatment with sulfonamides. It does not mean, however, that if the organism is found to be sensitive to a certain drug in vitro, the treatment with that drug of fish infected with such an organism will also be effec- DRUG RESISTANCE OF FURtTXCULOSIS IX TROUT 563 live. Tlic experience with cliloitetracycline pre- viously cited may serve as an example. It is well to remember the limitations of the in-vitro test for a correct interpretation of the results. The results of our studies show that in the in-vitro test sulfadiazine, sulfamerazine, and sulfisoxazole, and probably other sulfonamides, can be used equally well to determine sulfonamide sensitivity of ^4. salmonicida. In the therapeutic and feeding experiments, however, sulfadiazine has been found to be more toxic to the fish than has sulfamerazine, and sulfisoxazole has been found to be free from any detectable untoward effects even to brown trout (Snieszko and Wood, 1955). Unfortunately, the experiment on the use of sulfisoxazole for the treatment of furunculosis was unsuccessful because the strains of A. salmonicida isolated from the outbreak were resistant to sulfonamides ^ (tables 1 to 3). Testing of the sensitivity of A. salmonicu/a to sulfonamides and antibiotics can be carried out m several ways. The most practical method seems to be the use of drug-containing, filter-paper disks. Selection of a proper medium is very im- portant when testing the sensitivity of the organ- ism to sulfonamides. Trypticase Soy agar, special furunculosis agar medium buffered and alkalized, and Mueller Hinton agar, are especially recom- mended. In the first two media, the brown pig- ment so characteristic of A. salmonicida is pro- duced within 48 hours and the paraphenylenedia- mine test of Griffin (1952) can be performed as soon as growth is visible to the unaided eye. With the first two media, disks should contain 0.5 milli- gram of the drug; with Mueller Hinton medium, only 0.1 milligram. The test can be equally well performed using either petri plates or agar slants — the latter are more suitable for fieldwork. Also media in which sulfonamides have been incorpo- rated may be used provided that the control medium is free of sulfonamides and that both sulfonamide-resistant and sulfonamide-sensitive strains oi A. salmonicida are available for controls. RECOMMENDATIONS On the basis of the results reported in this paper and past experience in the treatmetit of furuncu- losis, we wish to make the following recommenda- tions : ^ One experiment of this series with sulfisoxazole was performed at Leetowii in 1955 by Bo Svenonius, a visiting fishery biologist from Sweden. 1. As soon as a disease suspected to be furuncu- losis breaks out, start treatment with sulfona- mides. Sulfamerazine is still cotisidered the drug of clioice, but sulfametlmzitie or a combination of sulfamerazine with sulfaguanidine can also be used (Snieszko 1954a). Sulfisoxazole is a drug of great promise, but it still has to be evaluated experimentally. 2. The diagnosis of furunculosis should be con- firmed as soon as possible by bacteriological examination. This should include determination in vitro of the sensitivity of the pathogen to the drug. 3. The dosage of sulfonamides is 8 to 10 grams per 100 pounds of fish per day. Treatment should last for 10 to 20 days and should result in a com- plete stoppage of mortalities due to furunculosis. Recurrences are particularly likely in disease- susceptible strains of trout. 4. Sulfonamides should never be used at lower levels or treatments repeated at short intervals. Such practice is the surest way to produce sulfa- resistant strains of the pathogen. 5. If the response to the treatment with sulfona- mides is not rapid, if recent experience has shown that treatment with sulfonamides is not effective, or if the results of laboratory examination show that the organism is sulfa-resistant but sensitive to oxA'tetracycline or chloramphenicol, either antibiotic should be used at a rate of 2.5 to 3.0 grams (of the antibiotic activity) per 100 pounds of fish per day. Chloramphenicol is somewhat better than oxytetracycline. 6. One should never rely on any drug for a permanent control of furunculosis, or of atiy other disease. There is no drug in existence which would permit elimination of any animal or plant disease. The best a drug can do is to reduce losses tempo- rarily. Long-lasting control of furunculosis, or any other infectious disease, is only possible by elimination of the source of infection, good sani- tation, and introduction or development by selec- tive breeding, of a strain of fish with greater disease resistance. (iood hatchery practices, avoidance of crowding, and balancetl nutrition are very important factors in keeping fish healthy. SUMMARY Therapeutic studies were performed with three sulfoiuunides and one antibiotic (chloramphenicol) with two strains of fingerling brook trout — one 564 FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVIC5E susceptible to furunculosis; the other resistant to this disease — and one strain of brown trout. Since it was found that the disease was caused by a sulfa-resistant type of the pathogen, only the treatment with the antibiotic was effective. Testing for di'ug resistance has shown that many strains of A. salmonicida were resistant to sulfona- mides, but sensitive to antibiotics. Therefore, studies were made in order to develop a rapid and reliable field method for the determination of drug sensitivity of A. salmonicida. Of all media tested, only the Trypticase Soy agar and the modified Leetown standard furuncu- losis medium can be used for the determination of sensitivity of A. salmonicida to sulfonamides and antibiotics and at the same time for the rapid presumptive test of Griffin. Mueller Hinton agar is excellent for the determination of drug sensi- tivity, but not for the presumptive test of Griffin involving the production of brown pigment. A method of the determination of drug sensi- tivity employing media in the form of agar slants and single disks containing the drug is recom- mended for field use. A revised and up-to-date method of treatment and control of furunculosis is described. LITERATURE CITED BuRDETTE, Robin I., LeRoy E. Plank, and William E. Clapper. 1955. A comparison of four methods of testing sul- fonamide sensitivity and the relation to clinical response in urinary tract infections. Antibiotics and Chemotherapy, vol. 5, pp. 392-397. Deuel, Charles R., David C. Haskell, Donald R. Brockway, and O. R. Kingsbury. 1952. The New York State fish hatchery feeding chart. Fisheries Research Bulletin No. 3, third edition, 61 pp. New York State Conservation Department, Albany, N. Y. Griffin, Philip J. 1952. A rapid presumptive test for furunculosis in fish. Progressive Fish-Culturist, vol. 14, pp. 74-75. U. S. Fish and Wildlife Service, Washington, D. C. Griffin, Philip J., S. F. Snieszko, and S. B. Friddle. 1953. A more comprehensive description of Bade rium salmonicida. Trans. American Fisheries Soc, vol. 82 (1952), pp. 129-138. Outsell, James S. 1948. The value of certain drugs, especially sulfa drugs, in the treatment of furunculosis in brook trout, Saheliniis fontinalis. Trans, .\merican Fish- eries Soc, vol. 75 (1945), pp. 186-199. Johnson, H. E,, and R. F. Brice. 1953. Mortality of silver salmon from treatments with sulfamerazine. Progressive Fish-Culturist, vol. 15, pp. 31-32. U. S. Fish and Wildlife Service, Washington, D. C. McCraw, Bruce M. 1952. Furunculosis of fish. Special Scientific Re- port: Fisheries No. 84, 87 pp. U. S. Fish and Wild- life Service, Washington, D. C. Northey, Elmore H. 1948. The sulfonamides and allied compounds, xxvii and 660 pp. Reinhold Publishing Co., New York. Snieszko, S. F. 1954a. Therapy of bacterial fish diseases. Sym- posium, Research on Fish Diseases. Trans. Ameri- can Fisheries Soc, vol. 83 (1953), pp. 313-330. 1954b. Advances in the studies of infectious fish diseases. Trans. Meeting Southeastern Fish and Game Commissioners, 3 pp. New Orleans. Snieszko, S. F., and S. B. Friddle. 1951. Tissue levels of various sulfonamides in trout. Trans. Ameiican Fisheries Soc, vol. 80 (1950), pp. 240-250. Snieszko, S. F., and P. J. Griffin. 1955. Kidney disease in brook trout and its treat- ment. Progressive Fish-Culturist, vol. 17, pp. 3-13. U. S. Fish and Wildlife Service, Washing- ton, D. C. Snieszko, S. F., P. J. Griffin, and S. B. Friddle. 1952. Antibiotic treatment of ulcer disease and fu- runculosis in trout. Trans. Seventh North .Amer- ican Wildlife Conference, pp. 197-213. Snieszko, S. F., and E. M. Wood. 1955. The effect of some sulfonamides on the growth of brook trout, brown trout and rainbow trout. Trans. American Fisheries Soc, vol. 84 (1954), pp. 86-92. Wales, J. H., and William Berrian. 1937. The relative susceptibility of various strains of trout to furunculosis. California Fish and Game, vol. 23, pp. 147-148. Wolf, Louis E. 1954. Development of disease-resistant strains of fish. Symposium, Research on Fish Diseases. Trans. American Fisheries Soc, vol. 83 (1953), pp. 342-349. Work, Thomas S., and Elizabeth Work. 1948. The basis of chemotherapy, xvi and 435 pp. Interscience Publishers, New York. U. S. GOVERNMENT PRINTJNG OFFICE ; 1957 O — 433457 VALIDITY OF AGE DETERMINATION FROM SCALES, AND GROWTH OF MARKED LAKE MICHIGAN LAKE TROUT By LOUELLA E. CABLE FISHERY BULLETIN 107 From Fishery Bulletin of the Fish and Wildlife Service VOLUME 57 UNITED STATES DEPARTMENT OF THE INTERIOR Fish and Wildlife Service ABSTRACT Scales of 1,603 lake trout, Salveliniis n. namaycush, presumably recoveries from plantings of juveniles previously marked by clipping certain fins, were studied. Their characteristics are described and illustrated. Age readings from the scales are highly reliable. In addition to the annuli, an accessory mark designated the O-mark, was found in the field of first-year growth. Annulus formation occurs over a period of several months. Data presented show that most, probably all, of the 102 specimens caught in the southern part of Lake Michigan and a considerable number (86) of those captured in the northern part actually were from wild stock. In general, the lake trout caught in large-mesh nets were larger fish than those taken in small-mesh nets. The fin-clipped lake trout caught in all nets were slightly larger than the wild stock of the same year classes, but wild lake trout of the earlier year classes, inhabitants of the lake before sea lampreys became numerous, were even larger and faster growing than the fin-clipped fish. The larger fish of the several year classes, caught in nets of either mesh- size, were taken from the northern part of Lake Michigan and smaller fish from the southern part of the lake. Ui INDEX TO VOLUME 57 Page Aeromonas salmonicida . 558 Aetidcidae 355 Age determination of trout in Lake Michigan 1-5!) annular markings 11 Agonostomiis monlicola, fresh-water mullet 415 description of species 417 locality — Bahamas and South Atlantic. 410 occurrence of larval and juvenile forms, . 424 spawning area 424 Aeromonas salmonicida resistant to drugs 555, 564 Alutera ventralis 337 Anderson, William W.: Early development, SPAWNING, GROWTH, AND OCCURRENCE OF THE SILVER MULLET {M ugil curema) along the South Atlantic coast of the United States 397-414 .Anderson, William W.: Larval forms of the fresh-water mullet (Agonostomus monlicola) FROM THE open OCEAN OFF THE BAHAMAS AND South Atlantic Coast of the United States 415-425 Annotated bibliography of tuna and tuna fisher- ies ^_ _ 173-249 Annular markings on lake trout 11 characteristics of lake trout scales, Michigan. 11 arcnaria. Mya _ .279-292 artedi, Leucichthys 87 Atlantic round herring, Etrumeus sadina 335 Atlantic lizard fish (Synodus foeten^) 336 abundance in southern Xew England 336 Atlantic sailfish 139-169 Antibiotics. Sensitivity tests on 555 Aujis thazard _ __ _ _ 336 Bacteria, notes on 469-496 Isolated from Gymnodinium brevis cultures.. 488 List of control fishes in unialgal cultures 478 Bahamas and South Atlantic Coast fresh-water mullet 415 Bare Lake, Alaska, salmon fecundity study 457 Bedroom Cove, soft clam growth in 279 Bibliography of tunas and tuna fisheries, 1930- 1953 " 173 Bigelow, Henry B., Clyde C. Taylor, and Herbert W. Graham: Climatic trends .\nd the dis- tribution of marine animals in Xew Eng- land 293-345 Blue marlin, Makaira ampla 169,532-538 Boothbay Harbor, Maine, Mean water tempera- tures . - - 344 Boston otter trawlers, Georges Bank. 266 Bradyidius Giesbrecht 355-358 Page Bradyidius arnoldi, new species 355 i3r«D(s, red tide 475^92 Bullock, G. L., and S. F. Snieszko: Treatment OF sulfonamide-resistant furunculosis in trout and determination of drug sensi- tivity 555-564 Butterfly ray {Grjmnura altavela), south of Cape Cod . 335 Cable, Louella E.: Validity of age determina- tion FROM scales, and GROWTH OF MARKED Lake Michigan lake trout , caerulea, Sardinops, Pacific sardine. - Calanoida - Calanoid copcpods of families Aetidcidae, Euchae- tidae, and Stephidae from the Gulf of Mexico. Carcinides maenas, green crab Carnegie (1929) zooplankton collections, Central Pacific Central Equatorial Pacific, Yellowfin tuna spawn- ing Capelin, Pacific Coast, Mallohis caiervarius calervarius, Mallolits, capelin of Pacific Coast Central Pacific, food of bigcye and yellowfin tuna. Central Pacific, Observations on the spearfishes. Central Pacific, Zooplankton abundance ... Central Pacific, Zooplankton studies, 1951-54 — Cenlropristes slriatus, sea bass, Corea, Maine cephalus, Mugil Charlevoix (Mich.), U.S. Fish Hatchery. Chloramphenicol, Treatment with — Clam, growth of Mya arenaria, soft. Climatic trends and the distribution of marine animals in Xew England, by Clyde C. Taylor, Henry B. Bigelow, .nid llrrhcrt W. Graham 293-345 Clupea pallasi, Pacific herring spawning com- pared wit h sardinops 443 relative abundance of eggs of herring 451 Colhjbus drachmc, porafret, food item of tuna — 66, 75. 79 Comi"ar.\tive study of food of bigeye and yellowfin tuna in the central P.vcific, by .Joseph E. King and Isaac I. Ikehara Copepoda calanoida Copepods, Xew calanoid Coregonits (Artedi) Linnaeus, Great Lakes in Europe Coryphaena hippurus, common dolphin, northern range Crab, green {Carcinides maenas}- curema. Mugil, silver mullet Cyprinodon variegalus, experimental. . 5G5 1-39 127 449 347 355-363 340 3li". 251-264 444 444 64.82 497-554 365 365-395 337 415-421 2 555 279. 292 61-85 347 -354 355 87 87 377 340 397 -414 486 495 566 FISHERY BULLETIN OF THE FISH AND WILDLIFE SOCIETY Page Depth zones. Distribution of different species in_ 274 Development of silver mullet, larval to juvenile stage 402 Dolphin, common (Coryphaena hippurus) 377 Drug resistance 556 furunculosis in trout 555-564 Early development, spawning, growth, and OCCTKRENCE OF THE SILVER MULLET {Mugil curema) along the south Atlantic coast of THE United States, by William W. Anderson. 397-414 Ecological notes on sailfish 167 Effects of environment and heredity on GROWTH OF THE SOFT CLAM {Myo arenario), by Harlan S. Spear and John B. Glude 279-292 Effects of unialgal and bacteria-free cul- tures OF Gymnodinium brems on fish, and notes on related studies with bacteria, by Sammy M. Ray and William B. Wilson 469-496 Elbow Cay, Bahamas, fresh-water mullet collec- tions 415 Estimating abundance of groundfish on Georges Bank 265-278 Elvumeun sadina, herring, northern range _ 335 Euchaeta paraconcinna, new species 358 Euchaetidae 358-361 Eulachon, Thaleichthys paciAcus, (Pacific Coast) sixiwning 443 Euthyiinus aUetteraliis, little tuna, northern range. 336 pelamis, striped bonito, northern range 336 Fecuxdity of Xorth American Salmonidae, by George A. Bounsefell 451-468 Fecundity of the Pacific sardine {sardinops caerulea), by John S. MacGregor 427-449 Filefish (Alutera ventralis) from Tortugas region. 337 Fin regeneration, Lake Michigan trout 7 Fishing of otter trawlers in each depth zone 270 Flarobacterium piscicida, bacterial counts 490, 491 Fleminger, Abraham: Xew calanoid copepods OF the families Aetideidae, Euchaetidae, AND StEPHIDAE FROM THE GuLF OF MeXICO-. 355-363 Fleminger, Abraham: Xew genus and two new SPECIES OF Tharybidae (Copepoda calanoida) FROM the Gulf of Mexico, with remarks on THE STATUS OP THE FAMILY 347-354 Florida Coast M/V Theodore N. Gill collections of mullet 415, 416 Food habits, sailfish 164 Food of yellowfin and bigeye tuna (14, 82 Fraser River, Harrison Rapids, salmon fry 459 Fresh-water mullet (Agonoslomus monlicola) 415-425 Description of larval forms _ . 417 Occurrence and length _ 421, 424 S|)a wiling 424 Frigate mackerel (Auxis (hazard) at Point Judith, Rhode Island 336 Fundiilus mnjnlis, killifish, ncirlhrrn r;iiigr 336 Page Furunculosis in trout _ _ . _ 555-564 drug sensitivity 556 Ciehringer, Jack W.: Observations on the de- velopment OF THE Atlantic sailfish, Isliophorus americanus (cuvier), with notes on an unidentified species of istiophorid. 139-171 Geinpylus serpens, snake mackerel, food important to tuna 64, 67, 75, 79 Georges Bank, estimating groundfish population. 265 Gcrino alahinga, skipjack 497 Cilude, John B., and Harlan S. Spear: Effects OF environment and heredity on growth OF the soft clam Mya arenaria 279-292 Gonyaulax monilata in Indian River, Fla 491 Graham, Herbert W., Clyde C. Taylor, and Henry B. Bigelow: Climatic trends and the DISTRIBI'TION OF MARINE ANIMALS IN X'eW England 293-345 Great Lakes (Mich.), tagging trout 2 Green Bay lake herring (see lake herring) 88 economic importance of fishery 90 map of Green Bay 89 water level (10 Groundfish, Georges Bank, estimating abun- dance 265-278 Growth and development of the sailfish 150 Growth of marked Lake Michigan trout.. . _ 36 in length _ 39-48 in weight _ _ _ 48-56 Growth of silver mullet _ _ _ 410 habitat of juvenile forms 413 Growth of soft clam 285 Growth rate of Lake herring 105, 107, 109 weight changes 104, 116 tlulf of Maine offshore water temperatures 302-307 temperature fluctuations in relation to fishes. 293-345 Gulf of ;\Iexico, new calanoid copepods of the families Aetideidae, Euchaetidae, and Steph- idae 355-363 Gulf of Mexico, new genus and two new species of Tharibidae, with remarks on the status of the family 347-354 Gulf of Mexico, mortality of marine animals in association with Gymnodinium. brevis 469 Gymnodinium brevis with unialgal cultures 472-496 red tide 469-496 Gymnura altavela, butterfly ray, northern range.. 335 Habitat and habits of larval forms of silver mullet 413 llereflity in soft shell clam, Xew England 279 Hida, Thomas S., and Joseph E. King:ZooPLANK- t(ln abundance in the central pacific, Part ii 365-395 Hippocampus hudsonius, sea horse, northern range 336 Hoven, Earl E., and Wilvan G. Van Canipen: Tunas and tuna fisheries of the world: An annotated bibliography, 1930-53 _ _ 173-249 INDKX TO \OLrMl!: 57 507 Ih/pomcsus pretiosus, silver siiult sp.iw nini;. Pa- cific Coast Ikohara, Isaac I., and Joseph E. King: Com- parative STrOY OF FOOD OF BIGEVE AND YELLOWFIN TUNA IN THE CENTRAL PACIFIC Istiophoiids of the western north Atlantic possible identity of specimens list of specimens studied sailfishes Isliophorus orirnlalis fxlinmpax iiinrliiin, black marliii- I'aKC 444 61-85 131) 169 170 139-171 522-023 .524 June, Fred C, and Ileeny S. II. Vucu: ^ ^^.(jw- fjn TUNA SPAWNING IN THE CENTRAL EglA- TORIAL Pacific 251-264 Juvenile sailfish 150 457 34(1 499 365-395 336-337 452 87 1-59 0 398 Karluk Lake and Ui\cr _ 456- Key to Tharybidae family . - Key to spearfishes of Central Pacific KiiiK. Joseph E., and Isaac I. Ikehara: Com- TARATIVE STUDY OF FOOD OF BIGEYE AND YELLOWFIN TUNA IN THE CENTRAL PaCIFIC__ 6185 KinK. Joseph E., and Thomas S. Hida: Zoo- i'lankton abundance in the central pacific, Part ii Kiiin mackerel (Scomberomorus regatis) _ Klainalh lii\er_ I.akt' Michiuaii lake herring in Green Bay Lake Michigan lake trout Age determination from scales . Larvae of silver mullet in plankton tows Larval forms of the fresh-water mullet (Agonos(omus monticola) from the open ocean off the Bahamas and South Atlantic Coast OK THE United States, by William W. Anderson . 415- 425 T,(ii«th-weight relation in growth of lake trout ._ 36 I.I iicichlhi/n arledi (LeSueur), Lake Michigan herring 87 Liiicichthtjs ascribed in Europe to Argyrosonuts 87 Life history op lake herring of Greex Bay, Lake Michigan, by Stanford H. Smith 87-138 Little tuna (Euthynnus tilletteratus) in Barnstable.. 336 Li/ard fish (Synodus foelens) 336 Lobster, Changes in abundance, 325 I'.uctuations in landings^ 32.5 Loliginidae, food of tuna 66 MacGregor. John S.: Fecundity of the Pai iik SAUDINK iSiirdinops curnilcn) ^ _ 427-449 Mackerel, fluctuations in laiidinti^- 324 Mdkdira albida (Poey) . I