Pacific Science (1973), Vol. 27, No.4, p. 305-318 Printed in Great Britain

Size, Growth, and Reproduction of the Sandbar Shark,

Carcharhinus milberti, in Hawaii I

RICHARD C. WASS2

THE SANDBAR SHARK, Carcharhinus milberti usually white, but this coloration is generally (Muller & Henle 1841), is a wide-ranging lost by the time the shark has attained maturity. species found in the subtropical and tropical The teeth in the upper jaw are bladelike and areas of the Pacific, Atlantic, and Indian oceans. broadly triangular. Those in the lower jaw are Investigations by Bigelow and Schroeder broad-based but the recurved cusps are thin and (1948), Springer (1960), and Clark and von spikelike. The tooth formula for 15 Hawaiian Schmidt (1965) have resulted in a fairly thor­ individuals is: ough knowledge of the life history of the sandbar shark population in the western North 14 or 15-1 or 2-14 or 15 Atlantic. Populations in other areas, however, 12 to 14-1-12 to 14 differ in several respects. In a recent paper con­ with cerned with C. milberti in the East China Sea, 14-1-14 Taniuchi (1971) documented differences in fe­ 14-1-14 cundity. The present study reveals significant being the usual case. differences in size at maturity, sex ratio, and fecundity for Hawaiian C. milberti. The Cooperative Shark Research and Con­ trol Program (Tester, unpublished) afforded an METHODS excellent opportunity to gather life history data The Cooperative Shark Research and Con­ on inshore Hawaiian sharks. A total of 1,727 trol Program operated from July 1967 through were captured of which 789 (45.7 percent) were June 1969 and was directed by Dr. Albert L. sandbar sharks. The study species is pictured Tester (unpublished). Objectives of the pro­ in Fig. 1 and can be distinguished from all gram were to study the ecology of sharks other Hawaiian members of the Carcharhinidae present in inshore Hawaiian waters, to deter­ by the relatively large first dorsal situated far mine if shark abundance could be reduced by forward (originating over the axil of the pec­ fishing effort, and to recommend future mea­ toral). It also has a prominent ridge between sures for controlling abundance. the two dorsals and dermal denticles that are The larger sharks were caught on a longline widely spaced and nonoverlapping. It is a set overnight. Seventy-two hooks (size 14/0) fairly heavy-bodied shark at maturity and can were attached to the line at 20-fathom intervals. be various shades of gray or brown dorsally Hooks were generally baited with tunaalthough and creamy white ventrally. The tips and trail­ porpoise, eel, reef fish, cat, squid, or mackerel ing edges of the fins of young individuals are also occasionally were used. Over 200 sets were made parallel to shore at depths of 15 to [ This investigation was supported by National Sci­ 30 fathoms. Thirty-two sets were made at right ence Foundation grant no. GB-6139; by a grant from angles to the depth contours with the shallow the Oceanic Institute, The Oceanic Foundation, Hono­ lulu; and by a special grant from the state of Hawaii. end anchored at about 10 fathoms and the deep This study fulfills in part the requirements for the Ph.D. end anchored at 60 to 100 fathoms. Smaller degree at the University ofHawaii. Manuscript received sharks were captured by handline or by a light 16 January 1973. longline (12 hooks, size 7/0) which was fished 2 Department of Zoology, University of Hawaii, at 10 to 100 fathoms. Fishing methods will be Honolulu, Hawaii 96822. Present address: Marine Re­ sources Division, Trust Territory of the Pacific Islands, detailed in a subsequent paper concerned with P.O. Box B, Ponape, Eastern Caroline Islands 96941. the distribution of Hawaiian C milberti.

305 20-2 306 PACIFIC SCIENCE, Volume 27, October 1973

FIG. 1. A mature female Carcharbinlls milberti. Preeaudallength is 132.1 em;

Live sandbar sharks (47 to 142 cm precaudal into the gill slits for 2 to 5 minutes until the length) were maintained for periods of up to shark ceased struggling. Only once did an indi­ 30 months at the Kewalo Basin branch of the vidual show ill effects from the above pro­ Hawaii Institute of Marine'Biology. When cedure; a mature female failed to recover from the sharks had to be handled for measurement what was probably an overdose of the anes­ or observational purposes, they were first thetic. No other problems were encountered anesthetized with M.S. 222 (tricaine methane­ during h~ndreds of anesthetizations. sulfonate). A variation of the method described by Gilbert and Wood (1957) was utilized. A 1: 1,000 solution of the powdered anesthetic and seawater was prepared and placed in a RESULTS 20-liter container. Sharks up to 90 cm pre­ Size caudal length were netted out of the holding Total length (horizontal distance between the tank and immersed in the solution for'15 to tip of the snout and the tip of the caudal in the '60 seconds. Larger sharks were anesthetized in normal swimming position) is dependent upon the holding tank. The water level was lowered the angle at which the caudal is positioned to a depth of 0.5 m and the shark was wrestled and must be estimated if a portion of the or lured over a net measuring 2.4 x 3.0 m which caudal is missing. For C. milberti, which has hung along the wall of the tank and extended a distinct precaudal pit on the dorsal surface out onto the floor. When the shark was in of the caudal peduncle, precaudal length (hori­ position, the net was lifted, thereby suspending zontal dista!1ce between the tip of the snout and the animal against the wall just above the water­ the deepest part of the precaudal pit) is a more line. M. S. 222 was tl\.en poured and squirted consistent and accurate measure ofsize and will Size, Growth, and Reproduction of the Sandbar Shark-WASS 307

• Unealcified

o Calcified

160

120 (') a • 15 ••• •• ~ .- •• • ::J .... • co'" 80 :T- :. ••••• • 3' .. ~ , ...... ••• 40 • ••• • ...Ia.;a ... •• • • • • •• • •• • •

0 50 60 70 80 90 100 110 120 130

Preeaudal Length (em)

FIG. 2. Relationship between clasper length and prccaudallength.

>100 ______: Cl!LfLQ&t£J..!l~ I 100 00 o • • Not Pregnant .00 • • o 0 00 • 80 o Pregnant o 0'{J ·0 • o 00· • c ~ ~ • . so • • • • • co 60 •• 0 •• l • • :T- .. ..~: • '3 ...., ... .2. 40 • • •• • • .•I,• • ______:L~..__M~~~~_,------. .. 20 Immatures •• I a··. ·fe·•• .,.,• ..I •• ....,..• I I OL...--::....----'_.....:.._...: ...... ---:.~_...ac...... ;:.:;.....:__....=...,..:__....___' " ...... :1. _'__.J...._• ...... '___ 50 60 70 80 90 100 110 120 130 140

Precaudal Length !em)

FIG. 3. Relationship between uterus breadth and precaudal length. 308 PACIFIC SCIENCE, Volume 27, October 1973

G F 140 - I( 130

120

"" 110 m n'" ~ 100 »0 ~

~ m 90 Z ....Cl :J; 8 [ 70

60

50

5 0 N 0 J F M A M J __ J A 5 0 N 0 J F M A M J J A 5 0 N 0 J F M A M J 1968 1969 1970 1971 MONTH

FIG. 4. Growth of captive sharks at the Kewalo Basin facility. Curves A, E, and] are males; curves C, D, F, G, H, and K are females. generally be used in this report. The regression is at least as great as the breadth of a fully con­ equation tracted uterus following parturition. Minimum size at maturity for female C. milberti was esti­ precaudallength = 0.784 (totallength)--3.022 mated by plotting uterus breadth against pre­ can be used if only total length is known. caudal length (Fig. 3). It may be seen from the Shortly before male elasmobranchs attain figure that the breadth of a fully contracted sexual maturity, their claspers begin to lengthen uterus is larger than 20 mm, so it appears that rapidly. Enlargement of the testes to functional females reach maturity at a precaudallength of size is immediately followed by calcification of about 115 cm. Average size at maturity was the clasper cartilages (Springer 1960). Minimum 130.7 cm and the range was from 112 to 146 cm. size at maturity for male C. milberti can be esti­ Of 90 juveniles captured, the smallest had mated by plotting clasper length (distance be­ precaudal lengths of 45, 48, 49, and 49 cm. tween the joint at the proximal end and the Four pups born at the Kewalo facility measured clasper tip) and calcification status against pre­ 47.8, 51.0, 51.0, and 51.2 cm at birth. The caudal length (Fig. 2). It appears that males largest embryos found within the uteri of preg­ mature at an average length of about 110 cm nant females measured 50, 50, 49.5, and 49.5 with a minimum clasper length of 130 mm. cm. Precaudal length at birth, therefore, is Precaudal lengths of mature males averaged generally between 45 and 51 cm. A value of 119.4 cm and ranged from 109 to 132 cm. 47 cm will be used below in calculations of Calcified clasper lengths averaged 157 mm and growth rates. ranged from 124 to 180 mm. The breadth of the flattened uterus at the widest point can be used to determine sexual Growth maturity among female sharks. A mature female Thirteen C. milberti lived for 4 months or is defined as an individual whose uterus breadth longer in the large tank at Kewalo Basin. Eight Size, Growth, and Reproduction of the Sandbar Shark-WASS 309

------140 .----­ ------

",~",,, 120 / , / / ~ "/",, 8 100 ,/ a..c , Q.. / / -Mal•• ~ I ::l I

60

400'----5...00---10...0-0---15...0-0---2..00-0---2""'5""'00---3-0..00---3-5"0-0--4-0"'0-0---450""'-0--5-00"0

Age (days) FIG. 5. Von Bertalanffy growth curves for captive sharks at the Kewalo Basin facility. For males: x = 139.4 (1-0.6629e-°.OOlloot), where x is precaudallength (cm) and t is age (days). For females: x = 149.0 (1-0.6845e-o.oOlo26t). of these died during the period of study, but for nine of the animals in Fig. 4. The four all deaths occurred from unnatural causes growth curves which do not appear in the (overdose of anesthetic, copper sulfate poison­ figure are almost identical to curves C and D ing, or jumping out of the tank), and the ani­ but have not been plotted because of space mals appeared to be in good health at the time limitations. of their death. In fact, most lived longer than Von Bertalanffy curves were fitted to the 8 months, the length of time listed by Clark data using the method of Fabens (1965). The (1963) as being the longest known period of equation for males, based on 49 growth incre­ captivity for the species. ments of four sharks is: The feeding program was determined arbi­ x = 139.4 (1- 0.6629e-°.OOllOOt), trarily since food requirements in the natural environment were unknown. The sharks were where x is precaudallength (cm) and t is time fed every 2 to 3 days but, instead of being fed since birth (days). The curve is plotted in a set amount each period, they were simply Fig. 5. The calculated value for asymptotic thrown morsels of food until their rate of size is 139.4 cm, this value being a little larger feeding decreased markedly. They were usually than the length of the largest male captured in fed cut pieces of frozen skipjack (Katsuwonus the field (132 cm). Comparison of the theore­ pe/amis) or akule (Trachurops crumenoptha/mus), tical average length at the end of 1 year (77 cm) but frequently they also were fed fresh or with the interpol~ted lengths for actual data at frozen red snapper, jack, eel, small reef fish, the end of 1 year (assuming each individual smelt, octopus, and squid. The sharks seemed was born at a length of 47 cm) for curves E to prefer frozen smelt, a fish which does not (78 cm) and J (77 cm) in Fig. 4 indicates a occur in Hawaii, above all else. remarkably close fit. It appears, therefore, that At 1- to 8-week intervals the sharks were the von Bertalanffy equation adequately de­ anesthetized and their precaudal lengths mea­ scribes the growth of captive males at the sured. The growth data obtained are plotted Kewalo Basin facility. 310 PACIFIC SCIENCE, Volume 27, October 1973

FIG. 7. Squirrelfish from stomach of Carcharhil1t1s milberti. Top, fresh; bottom, after 48 hours' digestion.

mature size (115 cm) in a similar length of time (1,070 days). If the rate of digestion is known, then the feeding frequency of sharks in the field can be roughly estimated from the percentage of stomachs containing food. To this end, three captive C. milberti of approximately the same size (95 to 101 cm precaudal length) were FIG. 6. Smelt from stomachs of Carcharhinus milberti. Top, fresh; middle, after 24 hours' digestion; bottom, starved for 4 days and then were fed smelt or after 48 hours' digestion. squirrelfish until they were satiated. The smelt -narrow-bodied fish with thin skin and soft Growth data from captive females also fit flesh-were about t5 cm long. The squirrelfish the model well. The equation, based on 57 were of similar length; but these fish are much growth increments of nine sharks, is: deeper-bodied and have large and heavy scales and firm flesh. x = 149.0 (1-, 0.6845ro.OOl026t). Two C. milberti which had been fed smelt This curve also is plotted in Fig. 5. Again, the were killed 24 and 48 hours after being fed and theoretical asymptotic size (149.0 cm) is slightly their stomach contents were examined (Fig. 6). larger than the largest individual captured in At 24 hours small chunks of flesh and bones the field (146 cm). Theoretical length after 1 were recognizable amidst a gray mush of oil year's growth is 78 cm, a figure which closely droplets and flocculated particles. At 48 hours approximates the interpolated lengths for actual the mush was darker in color and more con­ data after 1 year's growth for curves C (77 cm), densed. .1\ few oil droplets were still visible but D (79 cm), and H (80 cm) in Fig. 4. ." all flesh and bones had been broken down. Field data indicate that mature females are Only one shark was fed squirrelfish. When larger than mature males. Likewise, the theo­ its stomach contents were examined 48 hours retical asymptotic size derived from the von later (Fig. 7), flesh, bones, and eye lenses were Bertalanffy equation is larger for females than still present in the light-colored mush of oil for males. Fig. 5 indicates that males reach droplets and flocculated material. their minimum mature size (110 cm) at 1,050 The death of several captive sharks as a days (about 3 years) and that females attain result of copper sulfate poisoning yielded fur- Size, Growth, and Reproduction of the Sandbar Shark-WAss 311 ther data on digestive rates. The animals had (1948) found that the teeth of the smooth dog­ been fed whole (25 cm long) and cut akule fish (Mustdis canis) are replaced at the rate of about 30 hours earlier. When their stomach one row every 10 to 12 days. Markel and contents were examined, it was found that Laubier (1969) concluded that a tooth genera­ some of the akule were still whole and most of tion is replaced every 5 weeks in the cat shark the flesh remained on the bones, the implica­ (Sryliorhinus canicula). The most comprehensive tion being that digestion had only just begun. study of tooth replacement is that by Moss Apparently, food is retained within the stom­ (1967) on young lemon sharks (49.5 to 80.2cm achs of C. milberti for about 2 days if the prey in body length). After studying a group of is small and soft. Three to 4 days are required healthy animals that were regularly fed, he cal­ for average-sized prey to pass through the culated a mean functional tooth life of 7.8 days stomach, and more than 4 days are probably for teeth in the upper jaw and 8.2 days for required for large and hard-to-digest items. teeth in the lower jaw. The present investiga­ Captive sharks at Kewalo probably always had tion was patterned after this study. food in their stomachs because they were fed Captive sharks were anesthetized and the every 2 or 3 days. cusps of alternate functional teeth of the first Only 45 percent of the C. milberti captured eight series on either side of the symphysis of during the Cooperative Shark Research and the upper jaw were clipped with wire cutters. Control Program had food in their stomachs. Care was taken not to damage nerves and blood It may be argued that sharks with empty vessels in the pulp cavity or in the surrounding stomachs would be more likely to take a baited gum tissue. The procedure resulted in eight hook, so this value may be biased and, there­ recognizable teeth without seriously impairing fore, too low. However, the data do indicate the shark's ability to use its teeth for grasping that large numbers ofsandbar sharks had empty and slicing food. At subsequent 4- to la-day stomachs. When food was present, the amount intervals, marked teeth still in position were was often much less than stomach capacity. It counted. appears, therefore, that captive sharks at Kewalo The imbricated and overlapping pattern of were fed more than they would normally have tooth position in the jaws of Carcharhinus mil­ acquired in their natural environment and at berti requires that the teeth be lost individually. shorter intervals. Consequently, the growth Consequently, the measure of replacement rate rates determined above are probably greater is termed mean functional tooth life and is than normal, and Hawaiian sandbar sharks defined as the mean number of days a tooth probably require more than 3 years to reach remains in the functional position. Among the maturity in the field. carcharhinids, only a single tooth of each Moss (1967) used a rather indirect method to series in the upper jaw is functional. determine growth rates for lemon sharks The mean functional tooth life for each shark (Negaprion brevirostris). Rate of tooth replace­ was determined from regression of the number ment and size difference between functional of marked teeth remaining in position plotted and replacement teeth are the basic data re­ against the number of days elapsed since the quired. Age at maturity can be calculated if one teeth were marked, with the number of teeth knows the lengths at birth and maturity, the originally marked being the origin. The X number of tooth generations between birth and intercept (number of days when zero marks are maturity, and the mean length of time each present) is the estimate ofmean functional tooth generation is functional. life. The relationship between tooth life and Functional teeth along the outer border of precaudal length is plotted in Fig. 8. Though the jaw are constantly being lost and replaced the data are highly variable, it is plain that by younger teeth through a process which has tooth life increases with size. Previous investi­ been summarized by Moss (1967). Rate of gations have not shown this because they in­ replacement appears to be relatively constant volved only juveniles or individuals of the on a short-term basis and is independent of same size. Mean tooth life ranged from 18 days wear or injury to individual teeth. Ifft and Zinn for very young individuals (50 cm precaudal 312 PACIFIC SCIENCE, Volume 27, October 1973

38 •

36

34 •

2' 32 :;) ~ o' • :;) 30 e.. ~ !? 28 • ::r • .- ~26 n:- 0 '< • ~24 • • • 22 , • • 20 • 18 • •

60 70 80 90 100 110 120 130 140 Precaudal Length (cm)

FIG. 8. Relationship between mean functional tooth life and precaudal length. Y = O.215X+7.63, where Y is functional tooth life and X is precaudallength. length) to 36 days for mature sharks (130 cm The age at maturity was calculated by solving precaudal length). the regression in Fig. 8 for every precaudal The procedure for determining the number length corresponding to a vertical tie line ob­ of tooth generations between birth and matu­ tained from the foregoing method. This yielded rity has been described by Moss (1967) as a tooth life for each generation. These were follows: summed to get age at maturity, the calculations yielding estimates of 10.2 years for males and Strasburg (1963) demonstrated that tooth growth in e1asmobranchs could be considered as occurring in in­ 13.1 years for females. Data variability and crements ofinstantaneous growth, each increment occur­ questionable assumptions reduce the estimates ring at the time of replacement. Thus, tooth width to only rough approximations. One assump­ plotted against body length provides a stepped type of tion, which is almost certainly violated, is that curve for any growing shark. By reason of the fact that tooth replacement rate is the same for captive the teeth in the first replacement row are larger than the older, functional ones, the effective growth increments sharks at Kewalo as it is for animals in the field. of the teeth at the time of replacement can be measured. Replacement rate is inversely correlated with By plotting regressions of functional tooth width and growth rate, and growth at Kewalo was prob­ replacement tooth width against body length, a set of ably more rapid than in the field. It was nearly parallel lines is produced. These are then con­ nected by horizontal and vertical tie lines. assumed that there is no difference in rate of tooth replacement for males and females of The number of vertical tie lines between size equal size, but, if their growth rates do differ, at birth and size at maturity represents the replacementrates may also differ. A final assump­ number of tooth generations preceding tion is that the relationship between func­ maturity. tional tooth life and body length is lineali. It Size, Growth, and Reproduction of the Sandbar Shark-WAss 313

100 • •

80

~ E "- 60 (1) c. V) -£ 3 V') (1) "0 ~

20

o

J F M A M J J A S o N D 4 o 14 12 11 35 7 7 10 1 4 3

Month and Number Examined

FIG. 9. Monthly percentages of mature males with sperm in seminal vesicles. The line is a running average of the three nearest values. may actually be curvilinear if there is a direct Reproduction relationship between growth rate and tooth life. The sex ratio of embryos does not deviate The two methods used to estimate age at significantly from 1: 1. Of 308 embryos from maturity yield widely disparate results. The 55 litters of Hawaiian C. llJilberti, 152 (49.4 real value probably lies somewhere in between percent) were males. Apparently, the sex ratio but will likely never be known unless a success­ for mature individuals is also 1: 1. Forty-five ful tag-and-recapture study is conducted. adult females and 47 adult males were captured Eighty-six C. llJilberti were tagged and released on 32 longline sets which were made at right around Oahu and Niihau during the present angles to the depth contours and which ex­ investigation, but only one was recaptured. It tended from a depth of about 10 fathoms out was a mature male which showed no growth to a depth of about 80 fathoms. This fishing during its 595-day period of liberty. procedure exerted approximately equal effort 314 PACIFIC SCIENCE, Volume 27, October 1973 over almost the entire depth range of the the embryos grow. Following parturition they species. contract to a breadth of 20 to 40 mm. Initially The mating season for carcharhinids can be the embryos derive their nourishment solely ascertained in several ways. Among the males, from the stored yolk. Later, they may also vascular congestion about the cloacal area at absorb nutrients from fluids in the oviducts. the base of the claspers and the presence of When the embryo reaches a length of about sperm in the seminal vesicles indicate the ar­ 30 cm, a pseudoplacenta is formed from the rival of the mating season. Mature ovarian eggs yolk sac which allows the exchange of gases, and fresh mating wounds are indicative of the nutrients, and waste products between the mating season for females. Eggs in the uteri embryo and the mother for the remainder of which have not yet begun development are the gestation period (Springer 1960). further evidence. Both uteri are equally functional. Of 499 If the percentage of mature male C. milberti embryos collected from 90 pregnant females, with sperm in their seminal vesicles is plotted 245 (49.1 percent) were in the right uterus. against the month of observation (Fig. 9), the Average embryo size is plotted against date data tend to peak during the summer. The of measurement in Fig. 10 for 80 C. milberti highest percentages were found in July and litters. The points intersect with the zero or­ September. Vascular congestionabout the clasp­ dinate during the months of July, August, and ers was also noted during the summer months. September, indicating that this is when develop­ Data from mature females also indicate a ment began. These data agree with the time-of­ summer mating season. The average diameter mating estimates already discussed. Total of the four largest ovarian eggs was generally lengths at birth averaged 64 cm (47 cm pre­ small (10 to 15 mm) during the fall but began caudal length). Because embryos attained this to increase during the spring when the eggs size and were born during July, August, and assume a reddish yellow coloration. Eggs at­ September, they must have had a gestation tained the size at which ovulation occurs (35 to period of about 12 months. Further credence 40 mm) during June, July, and August. In is lent to the theory of a summer pupping addition, 11 females were captured between season by the fact that the smallest free-swim­ 23 July and 18 August that had recently ovu­ ming individuals were caught during the late lated eggs in their uteri. Fresh mating wounds summer and early fall. Corroborative evidence were also observed on several of these indi­ was also furnished by the two sandbar sharks viduals. which gave birth after being hauled aboard the Springer (1960) discussed the development fishing vessel during the middle of September of C. milberti; and TeWinkel (1950), Amoroso and by the numerous spent females with en­ (1960), and Schlernitzauer and Gilbert (1966) larged, flabby uteri captured during the months have provided comprehensive discussions of of September and October. the embryology of live-bearing sharks. The Forty-two percent of the mature female C. embryology of C. milberti in Hawaii is similar milberti captured in Hawaii were pregnant. to that of other carcharhinids and will be The average number of embryos in 91.litters briefly summarized. examined was 5.5 (range: 1 to 8). Females Only the right ovary is functional. Approxi­ holding full-term embryos and those that had mately 8 months before ovulation, several ova recently given birth did not possess ripe ova. begin enlarging and turn reddish yellow as yolk If a 1-year gestation period is assumed, a mini­ is deposited. Upon reaching a diameter of35 to mum of 2 years would then be required for the 40 mm, about five or six ova are released from reproductive cycle. A 2-year cycle is probably the ovary and proceed through the body cavity, the rule because, as mentioned, about half ostium, and right or left shell gland to the (42 percent) of the mature females were preg­ uteri. Prior to maturity the uteri are less than nant. Completely healed mating scars under­ 20 mm in breadth but, at maturity, they enlarge lying fresh bite wounds found on some re­ and reach breadths of about 60 mm at ovula­ cently impregnated females were evidence that tion. They continue to expand and thicken as more than a single litter may be produced. The Size, Growth, and Reproduction of the Sandbar Shark-WASS 315 • • ••

60 • .' •• • ~ 50 • • . • • • • • 40 • •

E $ ...c • 0>c 30 C]) ...J • 0 ~ ••

20 -. • • • • •••• 10 } •• I ... ,. : . • • O __&....,-_&....,---JI....---JI....---I_--"_---L_---L_...... _ ...... _--"-_...... L._...... L._... A s o N o FM A M A s Month FIG. 10. Monthly average total lengths of embryos in letters of Hawaiian Carcbarbinlls milberti. total number of litters borne by the average hanging along the wall of the main tank. She female, however, cannot be estimated because struggled for 3 to 4 minutes before succumbing the life-span is unknown. to the anesthetic and was then laid on a low A sandbar shark captured on 27 December table in the tank while length measurements 1968 and held at the Kewalo facility gave birth (138.6 cm precaudal length) and tooth replace­ to four viable pups on 24 September 1969. ment data were taken. Sandy was rather heavy­ Bequse I observed the births and because I bodied when captured 9 months earlier and her have found no account ofthe actual process in girth had increased considerably since she was the literature, I will describe the event here. placed in the tank. On this day I noticed vas­ At 1600 hou m, I guided "Sandy" into a net cular congestion and edema about the cloacal 316 PACIFIC SCIENCE, Volume 27, October 1973 region for the first time. When I stuck my ceased her frequent turning maneuvers. The finger through the cloacal aperture, I felt the umbilical cords were still trailing from her tail of an embryo. I must have ruptured the cloaca. On the following day they were found shell membrane at this time because fluid on the bottom of the tank along with the spilled from the opening when I withdrew my remains of the pseudoplacentas and egg cases. finger. When the observations were completed, The pups were measured the day following the anesthetic was flushed from Sandy's system their birth. The lengths of the live pups were: by "walking" her around the tank for about female, 67 cm total length, 51.2 cm precaudal 5 minutes until she began swimming on her length; female, 67 cm total length, 51.0 cm own. About 10 minutes later, the tail of an precaudal length; male, 67 cm total length, embryo was protruding from her cloaca. I 51.0 cm precaudallength; and male, 63 cm total pulled the embryo out, and placed it in a length, 47.8 cm ·precaudal length. The pups smaller observation tank where it soon began were all larger than 47 cm precaudal length swimming normally. which is the estimate of average length at birth By 1800 Sandy had recovered from the derived from field data. This is probably be­ anesthetic but was swimming more rapidly than cause they were also born later in the season was her usual habit. The tails of two more than is usual for C. milberti. The conditions of embryos were protruding about 15 cm from confinement may have prolonged their period her cloaca, swinging back and forth in unison of prenatal development. with the caudal fin and peduncle area of the Total lengths of the stillborns were: male, mother. At 1845 I saw a newborn pup swim­ 33 cm; male, 33 cm; female, 32 cm; and female, ming awkwardly around the tank and, because 32 cm. It is believed that stillborns are rare the posterior half of the third embryo was among Hawaiian C. milberti inasmuch as dead visible, another birth appeared imminent. I embryos were seldom found in the uteri of reentered the tank, which was now about 1 m captured specimens. Those occurring in the deep, to observe the birth. Sandy was swim­ tank were probably the result of an injury ming at twice her normal rate and making sustained by the mother during capture or frequent sharp turns. When more than half of confinement. the embryo was protruding, she made a quick Neither the pups nor their mother showed 1800 turn and left the pup at her pivot point. interest when food was offered about 4 hours It sank slowly to the bottom making feeble, after birth. On the following evening, how­ but uncoordinated, swimming movements. It ever, Sandy fed well. Four days after birth at lay on its side for 15 to 20 seconds, opening and least one pup was observed feeding, and all closing its mouth and presumably pumping took food on the 5th day. water over its gills. Then it wriggled rapidly At parturition a 1-cm remnant of the um­ but awkwardly to the surface for 4 or 5 seconds bilicus protruded through a slit in the belly of before sinking slowly to the bottom again. each pup. This was absorbed after 2 weeks and Bursts of swimming toward the surface fol­ the slit closed a week later, leaving an obvious lowed by decreasing periods of rest on the scar. During this 3-week period, the average bottom were repeated several times. Gradually, increase in precaudallength of three pups (one the rate of swimming slowed and became more was sacrificed at birth) was 1.2 cm. Umbilical coordinated. Fifteen to 20 minutes after its scars had completely disappeared at 6 months birth the pup was swimming continuously with of age, during which time the average increase a coordinated but slightly more rapid motion in precaudallength was 15.8 cm. than is normal for older individuals and in a more sinuous fashion-probably because the vertebral and caudal cartilages had not yet DISCUSSION AND SUMMARY calcified. A fourth embryo was born live and four Mature Hawaiian C. milberti are smaller than others were stillborn. By 1945 Sandy had their conspecies in other areas. Hawaiian males slowed to her normal swimming rate and had mature at precaudal lengths of about 110 cm, Size, Growth, and Reproduction of the Sandbar Shark-WASS 317 the largest male measuring 132 cm. Females that growth from birth to maturity takes about mature at about 115 cm and the largest mea­ 2 years in the Atlantic, though he admits this sured 146 cm. Springer (1960) found that mini­ estimate is based on little real evidence. mum lengths at maturity in the western North Tooth replacement rates calculated for Atlantic correspond to precaudal lengths of Hawaiian e. milberti indicate a mean functional 138 cm for males and 140 cm for females and tooth life of 18 days for young sharks to 36 that the largest measured 174 and 180 cm, days for mature individuals. Replacement rates respectively. Wheeler (1962) found that males have not been calculated for sandbar sharks in in the Mauritius-Seychelles area mature at a other areas, but the above figures are within total length of 170 cm, a length which corre­ the range ofvalues determined for other species. sponds to a precaudal length of 130 cm; and At birth the sex ratio of Hawaiian e. IIlilberti that females as large as 128 cm precaudallength is 1: 1. Springer (1960) found the same ratio were still immature. J. Bass (personal com­ (50.0 percent males) for 65 litters as did Clark munication) reported that male e. milberti from and von Schmidt (1965) for 21 litters (47 South Africa mature at a precaudal length of percent males) and Taniuchi (1971) for 91 115 cm and attain a maximum precaudallength litters (46.6 percent males). The sex ratio for of 140 to 145 cm. mature Hawaiian sandbar sharks caught on Size differences may result from different right-angle sets during this study is also 1: 1. water temperatures. Hawaiian surface tempera­ Springer (1960), however, found a sex ratio of tures measured during the Cooperative Shark about five females to one male in the western Research and Control Program were uniformly North Atlantic. Likewise, Clark and von high. Quarterly averages were: spring, 25.2° C; Schmidt (1965) concluded that the sex ratio for summer, 26.9° C; fall, 26.9° C; and winter, e. milberti in the central Gulf Coast region of 24.3° C. Surface temperatures in the western Florida was 6: 1. Springer feels that courtship North Atlantic and off the coast of South behavior may be responsible for the uneven Africa are generally cooler and more variable. sex ratio. He believes that males do not feed In addition, sandbar sharks in these areas mi­ during this period but females are not so in­ grate to cooler waters during the summer hibited and may fatally injure the males. months. Apparently, this is not the case in Hawaii. Size at birth is about the same in both the Data from several sources indicate that the Pacific and Atlantic oceans. Springer (1960) mating season for Hawaiian e. milberti peaks reported a total length corresponding to a pre­ during July and August. Springer (1960) found caudal length of 45 cm for Atlantic sharks. an earlier mating season in the waters off Precaudal lengths for Hawaiian individuals southeastern Florida and concluded that June averaged 47 cm. Taniuchi's (1971) estimate for is the period of maximum mating activity. e. milberti in the East China Sea corresponds to Taniuchi (1971) believes that sandbar sharks in precaudal lengths between 48 and 56 cm. the East China Sea mate during June and July. e. milberti held captive at Kewalo Basin for Hawaiian e. milberti are also born during the periods of 4 to 30 months showed growth rates summer, so their gestation period would be indicating a birth-to-maturity span of about about 12 months. Springer (1960) estimated a 3 years. They showed an average growth of gestation period of9 months with limits of 8 to about 31 cm the 1st year, 21 cm the 2nd year, 12 months. Taniuchi (1971) reported a gesta­ and 16 cm the 3rd year. Females grew slightly tion period of 10 to 12 months. faster than did males. Hawaiian litters averaged 5.5 embryos com­ The procedure based on rate and number of pared to an average of nine embryos per litter tooth replacements yielded age-at-maturity in Florida (Springer 1960, Clark and von estimates of about 10 years for males and 13 Schmidt 1965). Wheeler(1962) found anaverage years for females. Both the 3-year and the 10­ of 8.3 embryos and Taniuchi (1971) found an to 13-year estimates are subject to large sources average of 6.0 embryos. When the relationship of error. The true value probably lies some­ between litter size and length of the mother where in between. Springer (1960) believes was examined for all pregnant e. milberti 318 PACIFIC SCIENCE, Volume 27, October 1973 caught in Hawaii, a highly significant, positive CLARK, E. 1963. The maintenance of sharkscin regression was found. Hawaiian sandbar sharks captivity, with a report on their instrumental probably have smaller litters than their con­ conditioning. Pages 115-149 in P. W. Gilbert, species in other areas because the females are ed. Sharks and survival. D. C. Heath & Co., smaller when they give birth. Boston. In Hawaii 42 percent of the mature females CLARK, E., and K. VON SCHMIDT. 1965. Sharks were pregnant. This is considerably higher than of the central gulf coast of Florida. Bull. the values of 17 percent, 18 percent, and Mar. Sci. 15(1): 13-83. "substantially less than 1/3" calculated by FABENS, A. J. 1965. Properties and fitting of Springer (1960) for C. mi/berti off the east coast the von Bertalanffy growth curve. Growth of Florida and the value of 27 percent deter­ 29: 265-289. mined by Clark and von Schmidt (1965) for the GILBERT, P. W., and F. G. WOOD, JR. 1957. central Gulf Coast of Florida. Springer attri­ Method of anesthetizing large sharks and butes the low percentages in Florida to a rays safely and rapidly. Science 126(3266): scarcity of males. Seven of the 15 mature fe­ 212-213. males Wheeler (1962) caught in the Mauritius­ IFFT, J. D., and D. J. ZINN. 1948. Tooth suc­ Seychelles area were pregnant. cession in the smooth dogfish, Muste/us canis. BioI. Bull., Woods Hole 95(1): 100-106. MARKEL, V. K., and L. LAUBIER. 1969. Zum ACKNOWLEDGMENTS Zahnersatz bei Elasmobranchiern. Zool. Beitr., Berl., N.F. 15(1): 41-44. I am deeply grateful for the guidance offered Moss, S. A. 1967. Tooth replacement in the by Dr. Albert L. Tester throughout the course lemon shark, Negaprion brevirostris. Pages of this study and for his critical review of the 319-329 in P. W. Gilbert, R. F. Mathewson, manuscript. Without his direction, I would never have attempted such a project. I would and D. P. Rail, eds. Sharks, skates and rays. also like to express my sincere appreciation to Johns Hopkins Press, Baltimore. SCHLERNITZAUER, D. A., and P. W. GILBERT. the crew of the A/ika and especially to the 1966. Placentation and associated aspects of skipper, the late Mr. John Kuahiwinui. They were skilled fishermen and showed a lot of gestation in the bonnethead shark, Sphryna patience in working with the sometimes seasick tiburo. J. Morph. 120(3): 219-231. SPRINGER, S. 1960. Natural history of the "scientists" who had little or no practical sandbar shark, Eu/amia milberti. Fish. Bull. knowledge of boats and fishing. The author is U.S. 61(178): 1-38. also grateful to graduate research assistants STRASBURG, D. W. 1963. The diet and den­ Mr. Gerald R. Allen, Mr. John J. Naughton, tition of Isistius brasiliensis, with remarks on Mr. Robert H. Snider, and Mr. John H. Teeter tooth replacement in other sharks. Copeia who put in long hours at sea collecting data. 1963(1): 33-40. Finally, special thanks are due to Mr. Victor R. TANIUCHI, T. 1971. Reproduction of the sand­ Faughnan for his help in capturing, handling, bar shark (Carcharhinus mi/berti) in the East and maintaining captive sharks at the Kewalo research facility. China Sea. Jap. J. Ichthyol. 18(2): 94-98. TEWINKEL, L. E. 1950. Notes on ovulation, ova, and early development in the smooth LITERATURE CITED dogfish, Mustelus canis. BioI. Bull., Woods Hole 99(3): 474-486. AMOROSO, E. C. 1960. Viviparity in fishes. WHEELER, J. F. G. 1962. Notes on the three Symp. Zool. Soc. Lond. 1: 153-181. common species of sharks in the Mauritius­ BIGELOW, H. B., and W. C. SCHROEDER. 1948. Seychelles area. Proc. Roy. Soc. Arts Sci. Sharks. Pages 59-576 in Fishes of the west­ Mauritius 2(2): 146-160. ern North Atlantic. Mem. Sears Found. Mar. Res. 1(1).