The Trophic Ecology of Two Ommastrephid Squid Species, Ommastrephes Bartramii and Sthenoteuthis Oualaniensis, in the North Pacific Sub-Tropical Gyre

THE TROPHIC ECOLOGY OF TWO OMMASTREPHID SQUID SPECIES, OMMASTREPHES BARTRAMII AND STHENOTEUTHIS OUALANIENSIS, IN THE NORTH PACIFIC SUB-TROPICAL GYRE

A DISSERTATION SUBMITTED TO THE GRADUATE DIVISION OF THE UNIVERSITY OF HAWAIl IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

IN

OCEANOGRAPHY (MARINE BIOLOGY)

MAY 2003

By

Matthew P. Parry

Dissertation Committee:

Richard Young, Chairperson Robert Bidigare Brian Popp Jeffrey Polovina David Greenfield Abstract

This paper examines the trophic ecology ofthe squids Ommastrephes bartramii and Sthenoteuthis oualaniensis, using stomach contents and stable isotopic techniques. Simple energetics models were constructed using some ofthe data collected.

Samples for stomach contents were collected from 1996-2001 and 323 O. bartramii and 302 S. oualaniensis were captured. Fish and cephalopod remains dominated the stomach contents. Myctophids were found most abundantly in both squids, Symbolophorus evermanni was recovered at the highest percentage (7.5%) in 0. bartramii, while M. lychnobium or spinosum, Lobianchia gemellerii, and

Myctophum selenoides were all recovered at similar proportions (~5'10). Gfthe Myctophidae found in S. oualaniensis stomachs, S. evermanni was the most abundant (37%), followed by C. warmingii and H. proximum/rheinhardti (both ~15%), and M. Iychnobium (5%). Beaks from Onychoteuthidae occurred most frequently (14%) in 0. bartramii, while Histioteuthidae, Enoploteuthidae, and unidentified beaks all occurred at similar frequencies (10-12%). In S. oualaniensis, Enoploteuthidae occurred most frequently

(17%) followed by Onychoteuthidae (10%). The diet of 0. bartramii was more general while S.

Dualaniensis diet was more specialized on certain prey groups.

From 1998-2001 samples were taken from captured squids for stable isotope analyses, 143 0. bartramii and 160 S. oualaniensis. SIA was conducted on the mantle muscle ofO. bartramii that were divided into five categories based on mantle length, (1-7 nun) was 6.40/00, (75-100 nun) was 6.90/00, (200­

300 nun) was 11.10/00, (300-400) was 13.3%0, (400-570 nun) was 12.8%0. The olSN values for all 0. bartramii mantle muscle samples showed a logistic increase with mantle length. The mean Ol5N value for

S. oualaniensis sub-adult and adult mantle muscle (128 to 324 nun) was 8.2 0/00 The mean Ii IS N value for paralarvae was 6.20/00. The olsN values for all S. oualaniensis mantle muscle samples showed an exponential increase with mantle length.

Eye lenses, and blood samples were also taken from each squid species and showed similar patterns of(jlsN increase with mantle length respectively, blood was unav~ilable in the smaller size ranges of0. bartramii.

III CHAPTER I: STOMACH CONTENTS ANALYSIS (SCA) 3

INTRODUCTION 3

MATERIALS AND METHODS 6 Bias and difficulties in SCA '" 6 Dietary measurements- 7 Cruises 9 Sample collection 10 Squids (Ommastrephes bartramii and Sthenoteuthis oualaniensis) 10 Potential Prey of O. bartramii and S. oualaniensis 11 Sample manipulation andpreparation 12 Squids 12 Potential Prey of O. bartramii and S. oualaniensis 14

RESULTS 14 General information 14 Otolith data (0. bartramii}- 17 Frequency: 17 Abundance: , 17 Fish eye lenses (0. bartramii)- 17 Otolith data (S. oualaniensis)- 18 Frequency: 18 Abundance: 18 Beak data (0. bartramii)- 18 Frequency: 18 Abundance: '" 19 Beak data (S. oualaniensis)- 19 Frequency: 19 Abundance: 19 Ommastrephid prey remains (0. bartramii and S. oualaniensis)- 20 O. bartramii: 20 S. oualaniensis: 20 Gender comparisons (0. bartramii)- 20

1 Otoliths: 21 Beaks: 21 Gender comparisons (S. oualaniensis)- 22 Otoliths: 22 Beaks: 22 Niche breadth and overlap: 23 Ommastrephes bartramii 24 Sthenoteuthis oualaniensis 24

DISCUSSION 27 General discussion 27 Otolith data 28 Beak data 29 Gender differences 30 O. bartramii- 30 S. oualaniensis- 31 Niche breadth and overlap- 32

CONCLUSIONS 35

REFERENCES: 37

LIST OF TABLES 44

LIST OF FIGURES 45

2 Chapter I: Stomach contents analysis (SCA)

Introduction

This paper examines the trophic ecology of Ommastrephes bartramii and Sthenoteuthis oualaniensis, both speices are members of the order Teuthoidea, suborder Oegosida, family Ommastrephidae. Most oceanic squids in this family are large fast moving species that are key members in the pelagic ecosystem (Roper, 1984). Ommastrephid squids play an important dual role in the pelagic realm as both predator on wide ranging meso- and epipelagic species, and as prey for large top end predators such as swordfish, sharks, tuna, cetaceans and many others (King and Ikehara, 1956; Okutani et aI., 1976; Seki, 1993). O. bartramii is a cosmopolitan squid within the temperate and subtropical regions of the Pacific, Atlantic, and Indian Oceans (Clarke, 1966; Roper et aI., 1984). In the North Pacific, O. bartramii generally occurs in a region between 25 and 50° N latitude. These squid undergo a yearly roundtrip migration from their feeding grounds near the Subarctic Boundary (36-46° N) during the summer/fall and head south to subtropical spawning grounds (20-35° N) during the fall/winter (Murata and Nakamura, 1998). Spawning is known to occur in at least two general geographical areas, southeast of Honshu, Japan, and in waters near the Hawaiian Archipelago (Saito and Kubodera, 1993; Young and Hirota, 1990; Bower, 1996).0. bartramii appears to be a vertical migrator, occupying near surface waters during the night, and descending to depths during the day. Two studies done on the daytime depth distribution of O. bartramii have indicated two depth ranges in the North Pacific, 150-300m in the in the open ocean (Murata and Nakamura, 1998) and 400-700m near the Ogaswara Islands off Japan (Nakamura, 1993), the difference presumed to be due to water clarity at the different latitudes. O. bartramii is a prey item for a host ofpelagic marine mammals, and sperm whales (Physeter macrocephalus) in particular seem to feed heavily on this squid (Kajimura and Loughlin, 1988; Okutani et aI., 1976; Okutani and Satake, 1978). In addition, O. bartramii is a mainstay in the diet of the blue shark (Prionace glauca) and commercially important species such as swordfish (Xiphias gladius), blue marlin (Makaira mazara), striped

3 marlin (Tetrapturus audax), yellowfin tuna (Thunnus albacares), bigeye tuna (T. obesus), and bluefin tuna (T. thynnus) amoung others (Nakamura, 1985; Reintjes and King, 1953; King and Ikehara, 1956; Pinkas et aI., 1971; Matsumoto, 1984). In addition to being forage for commercially valuable species, O. banramii is itself a commercially exploited species. In the North Pacfic, O. banramii was the target of an international drift-net fishing fleet of approximately 500 vessels. Annual landings between 1985-1990 were between 248-378,000 tons (Murata and Nakamura, 1998). Catch rates have been reduced following the moratorium on drift-net fishing in 1992, and 0. bartramii remains an under-utilized resource. Sthenoteuthis oualaniensis ranges from the tropical to sub-tropical waters of the Indian and Pacific oceans, and occasionally extends into temperate waters (Roper, 1984). There are purported to be up to five separate fonns of S. oualaniensis (Nesis, 1993), the "typical" fonn that occupies the Hawaiian waters is the fonn being discussed here. S. oualaniensis is a permanent inhabitant ofHawaiian waters, and apparently spawns throughout the year, with an increase in spawning during the summer months (Hannan and Young, 1985). Adult and sub-adult S. oualaniensis are thought to vertically migrate, inhabiting near-surface waters during the night and descending to depth during the day. Daytime observations of S. oualaniensis have occurred in submersibles at depths of 400 and 1080 m in the Indian Ocean (Moiseev, 1991). Distribution data from around the Hawaiian islands suggest that these squids occur deeper than 650 m during the day (Young and Hirota, 1998). Smaller S. oualaniensis are important prey items for a range of daytime feeding seabirds (Ashmole and Ashmole, 1967; Seki and Harrison, 1989) indicating that the younger squids are still abundant in near-surface waters during the day. S. oualaniensis is also a major component in the diets of commercial fish species such as yellowfin tuna (T. albacares), bigeye tuna (T. obesus), bluefin tuna (T. maccoyii), and skipjack tuna (K. pelamis) (Reintjes and King, 1943; Kornilova, 1981; Brock, 1985; Young et aI., 1997). S. oualaniensis has also been the target of fisheries in Japan and China (Okutani and Tung, 1978) as well as being targeted for food and bait in the Hawaiian tuna handline fishery (Yuen, 1979). There are limited data on the specific feeding habits of both 0. banramii, and S. oualaniensis in the Central North Pacific. Seki (1993) did some general stomach contents

4 analyses on O. bartramii, but did not categorize the prey items past family in most cases. Shchetinnikov (1992) performed a detailed analysis on S. oualaniensis in the Eastern Pacific and found heavy predation by S. oualaniensis on young Dosidicus gigas which is a member of the same subfamily. Because O. bartramii and S. oulaniensis are also members of the same subfamily, heavy predation of one on the other was expected at the inception of this study. In addition, few studies have looked at the feeding habits of both of these species during the same study in the same area (Dunning and Brandt, 1985), and none have done so in the Central North Pacific. These two species overlap in habitat during the winter when O. bartramii migrate south to spawn. Much ofthe area of overlap appears to be a marginal habitat for S. oualaniensis during the winter, as females there do not seem to mature properly (Young and Hirota, 1998). Interactions of closely related squid species, such as O. bartarmii and S. oualaniensis, may have profound impacts on shaping the life histories of these animals. As potential competitors, or even predator and prey, the impacts that these species have on each other may be reflected in differing overall life history strategies. Direct effects such as competition for resources or predator prey interactions would only occur in areas oftemporal and spatial overlap such as in near Hawaiian waters during the winter. Young and Hirota (1998) found that, in Hawaiian waters, the mean size of S. oualaniensis commonly caught at jigging stations was about 40% of the mantle length (ML) of the smallest O. bartramii females. This size difference places much of the S. oualaniensis population well within the size range ofpotential squid prey for O. bartramii. Studying O. bartramii and S. oualaniensis where their populations overlap may reveal interactions such as competition for food resources or predator prey interactions that affect of determine habitat and life history patterns for these squids. Studying the feeding ecology of these squids will add to our understanding their effect on the ecosystem since they are viewed as major predators of mesopelagic fauna. Investigations of the feeding ecology of O. bartramii and S. oualaniensis can elucidate their impact on energy transfer in the oceanic ecosystem.

5 Materials and methods

Bias and difficulties in SeA In order to use stomach contents to analyze diet of squids, the squid must first be captured. Due to the ban on driftnets, the sampling methods available for capturing squids for this study were few, and pole and line capture was the most effective method available although the method introduces some bias. This method used light attractants, that alter normal environmental conditions and can cause aberrant feeding behavior. In addition, using a sampling method that relies on attraction generally captures only hungry or partially sated animals. The sampling milieu of pole and line jigging for squids is highly restricted to the near-ship environment and is limited for practical purposes to a few tens of meters in depth. The frequency ofoccunence, in which the number of stomachs containing specific prey is recorded as a percent of all stomachs, is a simple and direct measurement of the feeding spectrum of predators (Frost, 1946; Hunt and Carbine, 1951). Identification ofprey was determined for this study using some of the hard indigestible parts of prey that resist physical destruction by the squid predators. This method does not provide relative contribution of prey to the whole diet because if a species occurs once or multiple times in a stomach, it is still only recorded as being present in the stomach. The percent abundance of prey refers to the percentage of each prey type out of all prey items counted (Crisp, et aI., 1978; Ikusemiju and Olaniyan, 1977). Detennining abundance relies on the ability of the investigator to record prey items as discrete units, which can be difficult in squid that finely chop their prey with their beaks. As in the frequency of OCCUlTence method, otoliths and beaks ofprey that are resistant to physical destruction were used. Percent abundance tends to overestimate the energetic contribution of large numbers of small items of prey. Another problem with this technique is that episodic feeding by a few individuals has the same percentage impact as Jow level feeding by many individuals, but these instances are very different ecologically. Some of these problems can be ameliorated when this technique is coupled with data on frequency of occurrence ofprey items. Otoliths and beaks were used as the basis for much of the analyses done in this study. Otolith morphology has been shown to be highly species specific in many cases

6 (Scott, 1903; Fitch and Brownell, 1968; Schmidt, 1968; Morrow, 1979; Clarke, 1986; and others), although some studies have shown that intra-specific variation can be substantial enough to cause uncertainties in identification in some species (Norden, 1961; Casteel, 1974). Cephalopod beaks also have been used effectively for species identification (Clarke, 1966). Using otoliths and beaks found in the stomachs and not fresh material, such as tissue, to identify prey items helps to reduce bias from aberrant feeding around the ship. Presumably only the freshest material in the stomachs has been subject to the effects of the ship and attraction lights. Free otoliths and beaks must have been retained in the stomach for some time in order to be released from their surrounding tissue. Erosion of otoliths in a fairly acidic gastric environment can cause difficulties in identification. Because cephalopod digestive tracts are nearly neutral (Bidder, 1966; Wallace et al., 1981) calcareous structures such as otoliths retain the details of morphology. Another possible bias concerns otolith and beak accumulation in the stomachs of these squids, representing multiple meals (see Pitcher, 1980; Hunter 1983). In cephalopods, indigestible material apparently does not remain in the stomach for long, although the exact residence time is unknown (Bidder, 1966; Wallace et a\., 1981; Andrews and Tansey, 1983). Ofcourse, for an otolith or beak to be present in the stomach, the head or buccal mass of the prey must be consumed. Squids do not always consume the heads of the fishes that they eat (Bradbury and Aldrich, 1969). Porteiro et al. (1990) found that Loligo forbesi consumed the heads of smaller fishes while rejecting the heads of larger fish, biasing stomach results based on otoliths towards smaller fishes. Also, squids may not have the stomach capacity to consume all of a large prey that is captured. Finally, using hard parts that resist digestion to determine prey types may introduce bias against prey species that lack hard parts and are rapidly digested (Pierce et a\., 1994).

Dietary measurements- In this study niche breadth is used to quantify the degree of specialization to which a squid utilizes available resources (see Krebs, 1989; Cortes et aI., 1996; Heithaus, 2001). Resource states, based on prey taxa, were used in this study. Levins' measure of

7 niche breadth is used to measure the uniformity in the distribution of prey items among all the prey categOlies (Levins, 1968) and is calculated as: B = _1_ ~p/

B = Levins' measure of niche breadth Pi =Proportion of items in the diet that are of food category j (~ Pi =1.0)

B can range from 1 (minimum niche breadth, specialized diet) to a maximum of n (maximum niche breadth, all categories used equally) where n is the total number of resource categories. For ease of comparisons B can be standardized to a scale from 0 (minimum niche breadth) to I (maximum niche breadth) according to Hurlbert (1978).

Levins' standardized niche breadth is calculated as:

B A = Levins' standardized niche breadth B = Levins' measure of niche breadth n =number of possible resource categories

This measure of niche breadth ignores the possibility (and probability) that resources vary in their abundance within the ecosystem and across geographical regions. Varying degrees of rarity were not considered in this project owing to the extreme difficulty in assessing relative abundance of micronekton species in the pelagic environment. MacArthur and Levins' measure is one of the first, and simplest, methods to calculate resource overlap between species. MacArthur and Levins' measure is calculated as:

Mjk =L" Pii!1ik ~p/

Mjk = MacArthur and Levins' niche overlap of species k on species j pij, Pik = Proportion that resource i is of the total resources used by species j and k

8 n = The total number of resource categories

In essence this equation takes the niche breadth of one species and divides it by a combined niche breadth calculated using the proportions of prey items from both species. By applying this equation to two species using the same resource category matrix the equation yields overlap of one species resource usage onto another species usage. This measure is not symmetrical between species. Most other measures of overlap yield a symmetrical value of overlap of two species, representing the actual space of overlap between two species. Using a value of asymmetrical resource overlap to estimate impact of one species on another may be more indicative of most actual species interactions. MacArthur and Levins' measure has been used effectively to determine resource overlap (e.g. Macpherson, 1981; Ellis et al., 1996; Heithaus, 2002) and is used here since initial assumptions were that there would by asymmetrical overlap between O. bartramii and S. oualaniensis. There are several measures of niche overlap that have been developed over the years, and there is debate over which measure is best (Hurlbert, 1978; Linton et al., 1981; Krebs, 1989). Many of the measures of resource overlap contain varying degrees of bias to differing numbers of resource categories, differences in sample sizes between species being compared, and resource evenness (Krebs, 1989).

Cruises Most specimens were collected during six cruises that were conducted aboard the fishery training vessel HOKUSEI MARU during the years of 1996 through 2001. The length of each cruise was 15 days and each cruise started on February 5th and ended on th February 20 • This study evolved as a subset of the original study on the general ecology of Ommastrephes bartramii. Cruise tracks differed per year but were generally located between 172° Wand 158° W longitude and 18° Nand 32° N latitude (Fig. 1). In most years the cruise tracks had to be altered to some extent during the cruise because of prevailing weather conditions. The general plan of each cruise was to proceed north to the beginning of the Subtropical Frontal Zone (usually located near 30° N) where individuals of O. bartramii are abundantly caught, and then move south towards the main islands of the Hawaiian archipelago where O. bartramii is caught in less abundance but

9 where S. oualaniensis is caught more abundantly. Additional specimens of O. bartramii were captured in August 2000 on a separate cruise aboard the HOKUSEI MARU at 41 0

N, 155 0 E. These specimens were stored aboard the HOKUSEI MARU at -200 C and were off-loaded with the regular specimens after the 2001 cruise. Additional samples were collected during two cruises aboard the R/V Townsend Cromwell conducted in November and January of 1999 in connection with the National Marine Fisheries Service (NMFS) Honolulu laboratory. In the context of this study these were cruises of opportunity and samples were collected on a non-interference basis. These cruises involved a substantial portion ofthe total cruise times centered over Cross Seamount.

Sample collection

Squids (Ommastrephes bartramii and Sthenoteuthis oualaniensis)

Squids were caught using four different capture methods, pole and line jigging, automatic jigging, handlines and a submerged driftline method. Fishing operations began roughly around 2100 hrs and continued usually until 0400 hrs depending on weather and sea conditions. The use of each method of capture also was dependent on weather and sea conditions with the most reliable and frequently used method being pole and line jigging. Each fishing station was conducted as a drift station with the ship's engine shut down in order to minimize disturbance. All station positions reflect the latitude and longitude at the point where the fishing operations began. Pole and line jigging consisted of between three and eight people fishing per station using fishing poles rigged with 20 Ib test line and a variety of sizes of squid jigs ranging from small to large (5-15 cm). Jigs consisted of a fusiform plastic or straight metal lure with two rows of hooks located opposite from the end of attachment to the line (Fig. 2). Large squid jigs had barbed hooks while smaller sized jigs had straight barbless hooks. Three automated jigging machines were available onboard but their use was highly dependent on conditions. Automatic jigging machines consisted of a central

10 motorized unit with drums of spooled line with interconnected jigs located on either side (Fig. 3). Roughly 30-40 jigs were set on each line with the end of the line fixed to a 10 kg weight that allowed the line to spool freely. The lines were automatically reeled in and out to a depth of about 40-50 m on average. The drums were elliptical which imparted a slightly erratic motion to the jigs as they were reeled in or out. As the squids were hooked the upward pull of the machine kept the squids on the jig (they were not barbed) and as they were brought out of the water and over the rollers that guide the line, their weight caused them to drop off the jig into a receiving basket. This method ofcapture, although convenient because of its automation, lacked the finesse of hand jigging and would often result in the arms and tentacles ofthe squids being pulled free and the squid dropping back into the water. Handlines consisted of about 50 m of line spooled around an H shaped wooden frame. A large squid jig was attached at the end of the line along with a 2 to 4 kg weight. The jig was fished between 30-40 m and tied off to the ship. The jig was allowed to soak for up to several hours with periodic checks to determine if a squid had been caught. A submerged drift-line was also used (Fig. 4). The drift-line consisted of roughly 300 m of black polypropylene line. At one end of the main line a large weight and a 50 m line with a buoy were attached. Buoy and weight were put overboard and the main line was allowed to sink/drift away from the ship. Long-line clips with three m of monofilament line ending in a barbed squid jig were attached to the main line at IO m increments. After five clips with jigs were placed on the main line a long line clip with a light source was attached. In between every 10 jigs a longline clip with a weight was clipped to the main line. This pattern was repeated until 20 jigs had been rigged on the main line. Once all the jigs were attached to the main line another large weight was attached and the shipboard side of the drift-line was submerged. The drift-line was usually deployed once or twice a night, weather permitting, and was soaked for two to three hours each deployment. No recording devices were attached to the line.

Potential Prey of 0, bartramii and S. oualaniensis

11 Potential prey of O. bartramii and S. oualaniensis were obtained on all six of the HOKUSEI MARU cruises. During years 1996-1999 mesopelagic fishes and squids were captured using a large midwater otter trawl. The trawl had an operational rectangular 2 mouth opening of 200 m . A depth telemeter was fixed to the top edge of the mouth of the net, in the center. One or two tows at 800 m (an hour at depth for each tow) was made at every station during the daytime, weather permitting. Seas of over three m were considered too large to tow in due to the stress on the towing cable and winch. During years 2000-2001 the large trawl was replaced with a smaller frame trawl. This trawl consisted of a square steel frame roughly three meters on each side and trailing a net with approximately 10 mm stretch mesh. The depth telemeter used on the large trawl was also used on the frame trawl. The frame trawl was towed to a depth of 900 meters and required less tum-around time than the large trawl, consequently it was frequently used twice at each station. Three tows were taken during the November cruise aboard the Townsend 2 Cromwell using a'" five m Isaac's Kidd midwater trawl with a 202 j.UJl mesh net. These tows were taken at night and ranged in depth from 160 to 300 m.

Sample manipulation and preparation Squids

Mantle length (measured from the dorsal anterior free mantle margin to the posterior end of the fins), sex, and species were determined onboard for all jig caught squids during all cruises. Each stomach (with cecum attached) was removed from each squid and placed in a small plastic bag, labeled, and then frozen at _200 C for later analysis in the lab. In the lab stomachs were thawed, excess water was blotted, and then the stomachs were weighed. The caecum, and any amount of esophagus or intestine, was trimmed off and the remaining stomach and contents were weighed again. The contents were removed and any excessive water was blotted and the contents were weighed. Using these three measurements the weights of the material in the stomach, the stomach itself, and the cecum were obtained for all squids captured. After weighing, the contents were placed into a 333 urn Nitex strainer and rinsed with running water. The contents were

12 then examined under a dissection microscope for identifiable parts. Squid food items are thoroughly chopped by the action of their beaks, making identification of prey difficult. Because the squids rely primarily on enzymatic digestion (Hochachka, 1994), calcium carbonate parts such as otoliths survive the feeding process in exceptional condition. Beaks of prey squids also survive well although often broken into pieces. To identify prey items within the stomachs, fish otoliths and cephalopod beaks were removed and stored separately from each other, in 100% ethanol or 50% isopropanol, respectively for later identification. To avoid the loss of small particles stuck in the folds of the stomach, the stomach lining was removed after the contents had been fully examined. The stomach lining was examined under the dissection microscope using transmitted light that shone through the lining and made any particles caught in the folds highly visible, many small otoliths were found in this manner. Sagittal otoliths recovered from the stomachs ofthe squids were compared with sagitta in an otolith library formed from fish caught during the cruises. Henceforth, any mention of otloliths will be referring to sagitta. Cephalopod beaks recovered from the stomachs of the squids were also identified by comparisons to a library of beaks. Normalized stomach fullness was calculated using a regression derived by Tung (1978) from values of stomach weights for given mantle lengths of S. oualaniensis. This regression was assumed to approximate a value of fullness that the stomach at a certain size of squid can hold. Fullness of each stomach was normalized as a percentage of the regression value for 100% fullness. These values were regressed against mantle length for male and female O. bartramii and S. oulaniensis to estimate an average percentage of fullness. Levins' measure of niche breadth (Levins, 1968) was calculated separately for otoliths and beaks recovered in the stomachs ofboth squid species. Fish and squid represented the two distinct and major sources of food for these squids and were thus considered separately to examine potential differences. This value was then standardized using the method from Hurlbert (1978). MacArthur and Levins' measure of niche overlap (1967) was also calculated separately for both species using the otolith and beak data. When calculating niche breadth only otoliths and beaks that could be identified were used, and categories were assigned to the lowest identifiable taxon.

13 The number of resource categories was determined as the total number of resources (fishes or squids) that were represented in all the stomachs ofboth squid species during the entire study. This was assumed to be the number of resource categories even though other resource categories may exist for both fishes and squids that were not represented in the stomachs sampled.

Potential Prey of O. bartramii and S. oualaniensis

Potential prey collected in the trawls were placed into gallon-size ziplock bags, labeled and frozen at _20 0 C. Only specimens that were in a condition that would allow for identification (photophores mostly intact, body mostly intact, etc..) were kept. As a result the identification library was incomplete. Not all otoliths and beaks could be identified using the libraries constructed for this study. Fishes caught in the trawls were taken back to the lab and identified to lowest possible taxon. Fishes in the family Myctophidae were generally classified to species level when possible, other fishes were usually only categorized to the family level. All fish caught in good condition during the years of 1996-1998 were used as the basis for the otolith library. The otoliths (sagitta, asteriscus, and lapillus) of all fish from these years were removed using a dissection microscope, scissors, and micro-forceps. The otoliths were cleaned of debris and placed in cardboard microslide containers and covered with a glass slide. Otoliths were collected from a total of 250 fish comprising 20 separate fish families including 17 species of myctophids.

Results

General information Over the six years of this study, 323 O. bartramii were captured, 267 males (82.3%) and 56 females (17.3%). The male 0. bartramii had a mean mantle length of 329 mm (SD =25 mm), while the females had a mean mantle length of 535 mm (SD =41 mm). There were 302 S. oualaniensis caught, 40 males (13.2%) and 262 females (86.7%). The male S. oualaniensis had a mean mantle length of 155 mm (SD = 15 mm),

14 while the females had a mean mantle length of 198 mm (SD = 40 mm). The total number of squids caught varied per year (Table 1.).

Table 1. The number of squids ofhoth species caught during each year (the number ofS. oualaniensis caught during 1999 includes specimens taken aboard the RIV Townsend Cromwell).

Ommastrephes Sthenoteuthis Year bartramii oualaniensis 1996 65 5 1997 46 19 1998 53 73 1999 71 48 2000 9 31 2001 16 31

A stomach was considered empty if there was no weighable material (Le., <0.001 g). Of the male O. bartramii stomachs examined, 19 (7.1 %) were empty, while nine (16%) of the female stomachs were empty. Of the male S. oualaniensis stomachs examined 12 (30%) were empty, while 44 (17%) of the female stomachs were empty. The stomach content weights for O. bartramii ranged from < 1 to 95.9 g in males (x = 9.7 g, SD = 13.7), and from < 1 to 435 g in females (x = 32.3 g, SD= 66.7). The stomach content weights for S. oualaniensis ranged from < 1 to 10.8 g in males (x = 2.18 g, SD = 2.95), and from < I to 31.0 g in females (x = 4.21, SD = 5.32). Stomach fullness did not increase appreciably with mantle length for either species (Fig. 5,6). S. oualeniensis showed an increase in maximum fullness between 125 and 200 mm ML (Fig. 6). Average percentages offullness for S. oualaniensis males and females were 24 and 13%, and for O. bartramii males and females averages were 8 and 5% (Fig. 5,6). The weight of the stomach organ increased with increasing size of both species of squids (Fig. 7). The weight of the stomach organ increased exponentially with mantle length with similar trends in both O. bartramii and S. oualaniensis (Fig. 7). Stomach

15 organ weight increased with mantle length for each squid according to the following regressions.

Stomach weighto bart,"mii = -3.0008 + [1.6922*1.0051(mantlelength)]

Stomach weights. oaalaniensis= -8.9057 + [6.8234*1.0024(mantle length)]

Fish and cephalopod remains dominated the stomach contents; rarely some decapod crustacean material was found but consisted of such small and unidentifiable portions that it was ignored. Other items such as thecate gastropods, amphipods, copepods, and unidentifiable pieces ofplastic were occasionally found. These smaller items (excluding the plastics) were assumed to be transit items (i.e., from the stomach of prey) as they were probably too small for the squids to feed on (Shchetinnikov, 1992). In 15 of the O. bartramii stomachs mesh-like masses of gelatinous tissue were found which were possibly the remains of pyrosomes, but could not be identified conclusively. Similar mesh-like prey were not found in S. oualaniensis. One O. bartramii stomach was filled with an egg mass from a pelagic octopus, probably Argonauta sp. A total of 2561 otoliths and 424 beaks were recovered during this study. For O. bartramii, otoliths were found in 264 (81.5%) of the stomachs and 169 (52.2%) stomachs contained cephalopod beaks. For S. oualaniensis, otoliths were found in 211 (69.6%) stomachs while beaks were found in 96 (31.7%) of the stomachs. The fish otoliths recovered from the stomachs of both squid species were categorized into 17 families. Of these 17 families only otoliths from the members of Myctophidae, Anoplogasteridae, and Omosudidae were identified to species level. Within the Myctophidae 16 species were identified. Four species of myctophids could only be identified to genus, they where Hygophum proximum or H. rheinhardti, and Myctophum lychnobium or M. spinosum. These four species were placed within two species groups. The beaks found in the stomachs of both species of squids represented 18 families of cephalopods from which 29 genera were identified, of these, 11 were identified to the species level. Otoliths or beaks occurring at < 1% frequencies in both species were ignored. Otoliths or beaks which made up < 1% of the total abundance in both species were ignored.

16 Otolith data (0. bartramii)-

Frequency: Otoliths belonging to the family Myctophidae were the most frequently occurring otoliths in both squid species. Myctophid otoliths were found in 63% of O. bartramii stomachs while unidentified otoliths occurred next most frequently (37%), followed by otoliths from Gonostomatidae (16%) (Fig. 8). Otoliths of seven families of fishes occurred in O. bartramii stomachs at frequencies of greater than or close to 5% (Fig. 8). Otoliths from eight species groups within the family Myctophidae were found at high (> 5%) frequencies in O. bartramii, while six other myctophid species groups were found at > 1% frequencies (Fig. 9).

Abundance: There were 1624 otoliths recovered from all the O. bartramii stomachs. The myctophid and unidentified categories comprised 63% of all otoliths in O. bartrami (Fig. 10). The three next most abundant family categories of otoliths found in O. bartramii, belonged to Gonostomatidae (11 %), Melamphaeidae (7%), and Stemoptychidae (5%) (Fig. 10). The remaining 12% of the otoliths found in 0. bartramii stomachs belonged to the families Photichthyidae, Everrnanellidae, Chauliodontidae, Gempylidae, and Scopelarchidae. Of the otoliths from the Myctophidae, Symbolophorus evermanni was recovered at the highest percentage (7.5%) in O. bartramii, while M. lychnobium or spinosum, Lobianchia gemellerii, and Myctophum selenoides were all recovered at similar proportions (;::; 5%) (Fig.11).

Fish eye lenses (0. bartramii)-

17 Fish eye lenses were collected from O. bartramii stomachs during the cruises in 1996-1999. There were a total of 925 fish eye lenses collected during those years, compared with 1511 otoliths and 255 beaks over the same period.

Otolith data (S. oualaniensis)-

Frequency: In S. oualaniensis 62% of stomachs contained myctophid otoliths, 19% contained unidentified otoliths, and 4% contained gonostomatid otoliths (Fig.8). Otoliths from three species groups within myctophidae (S. evennanni, C. wanningii, and H. proximumlrheinhardtz) were found at greater than or close to 20% occurrence, while seven other species within Myctophidae occurred at > 1% (Fig. 9). The only other otoliths that occurred at a frequency greater than 1% in S. oualaniensis stomachs were from Evermanellidae (1.65%).

Abundance: There were 937 otoliths recovered from the S. oualaniensis stomachs. The myctophid and unidentified categories comprised 97% of the otoliths in S. oualaniensis (Fig.l0). Gonostomatidae and Evermanellidae made up the remaining 3% ofotoliths in S. oualaniensis. Ofthe otoliths from Myctophidae found in S. oualaniensis stomachs, S. evennanni was the most abundant (37%), followed by C. wanningii and H. proximumlrheinhardti (both z 15%), and M. lychnobium (5%) (Fig.ll).

Beak data (0. bartramii)-

Frequency: At the family level, beaks from Onychoteuthidae occurred most frequently (14%) in O. bartramii, while Histioteuthidae, Enoploteuthidae, and unidentified beaks all occurred at similar frequencies (10-12%). Ommatrephidae and Pyroteuthidae occurred next most frequently (5%) while all other categories of beaks found in O. bartramii occurred at < 5% (Fig. 12). When the beaks were categorized to the lowest possible taxa, Onychoteuthis sp. occurred most frequently in O. bartramii (14%), while Histioteuthis

18 sp., and unidentified beaks occurred next most frequently (10%). Abraliopsis sp. A and Pyroteuthis addolux occurred in O. bartramii with similar frequencies (5%), while all the other categories had frequencies below 3% (Fig. 13).

Abundance: There were 288 cephalopod beaks recovered from the O. bartramii stomachs. At the family level, beaks of Onychoteuthidae were most abundant (24%). Beaks of Histioteuthidae, Ommastrephidae, Enoploteuthidae, were all recovered at similar abundances (10-12%) in O. bartramii stomachs (Fig.14). When beaks were categorized to the lowest possible taxa, Onychoteuthis sp. accounted for 24% of beaks in O. bartramii followed by Histioteuthis sp.(12%). Pyroteuthis addolux and Abraliopsis sp. A both accounted for similar percentages (6%) of the beaks in O. bartramii. Nine more categories of beaks were recovered in O. bartramii at varying frequencies between 2-3% (Fig. 15).

Beak data (S. oualaniensis)-

Frequency: At the family level in S. oualaniensis, Enoploteuthidae occurred most frequently (17%) followed by Onychoteuthidae (10%), while Ommastrephidae and unidentified beaks occurred in similar frequencies (;::; 5%) (Fig. 12). When beaks were categorized to the lowest possible taxa in S. oualaniensis stomachs, Abraliopsis sp A. occurred most frequently (12%) followed by Onychoteuthis sp. (10%), and Enoploteuthis sp. and Abraliopsis sp. (;::; 2%) (Fig. 13).

Abundance: There were 136 cephalopod beaks recovered from the S. oualaniensis stomachs. At the family level beaks of Enoploteuthidae were most abundant (46%) in S. oualaniensis stomachs, followed by Onychoteuthidae (22%), unidentified beaks (12%) and Ommastrephidae (11%) (Fig.14). When beaks recovered in S. oualaniensis stomachs were categorized to the lowest possible taxa Abraliopsis sp. A, and Onychoteuthis sp. accounted for the greatest percentages of beaks, 33 and 22% respectively (Fig. 13).

19 Abraliopsis sp. accounted for 7% ofbeaks in S. oualaniensis while three other genera occurred at < 3% for each category (Fig. 15).

Ommastrephidprey remains (0. bartramii and S. oualaniensis)-

O. bartramii: The remains of three O. bartramii were identified in O. bartramii stomachs, all three consisted of freshly consumed material. No identifiable remains of S. oualaniensis were found in O. bartramii stomachs. The remains of 16 unidentified ommastrephid squids were found in O. bartramii stomachs, 15 of these consisted offreshly consumed material such as arms and tentacles. The freshly consumed materials were considered to be artifacts because O. bartramii were frequently seen attacking others that had been hooked, and there were numerous pieces of caught animals (arms, tentacles) being lost overboard during fishing.

S. oualaniensis: The remains of two S. oualaniensis were identified in S. oualaniensis stomachs, one was fresh material while the other appeared to have been consumed at a significant time prior to capture. No identifiable remains of O. bartramii were found in S. oualaniensis stomachs. The remains of 16 unidentified ommastrephid squids were found in S. oualaniensis stomachs, 13 of these consisted of freshly consumed material such as arms and tentacles. The freshly consumed materials were considered to be artifacts because there were numerous pieces ofcaught animals (arms, tentacles) being lost overboard during fishing.

Gender comparisons (0. bartramii)-

20 Otoliths: Otoliths occurred in 230 (86.1 %) of the male O. bartramii stomachs, and 33 (58.9%) of the female stomachs. Otoliths occurred at higher frequencies in male than female O. bartramii stomachs for all families of fishes except Chauliodontidae, Apogonidae, Giganturidae, and Melanostomiatidae (Fig. 16). A similar trend was seen in myctophid otoliths that occurred in higher frequencies in male O. bartramii in all species groups except, C. warmingii, D. perspiculatus, and B. longipes (Fig. 17). Of the total otoliths recovered from O. bartramii, 1460 (90%) were from male stomachs while 164 (10%) were from females. Myctophids made up a greater relative abundance of the otoliths recovered from female O. bartramii (66%) than those recovered from males (48%) (Fig. 18). The abundance of unidentified otoliths was also higher in otoliths from female 0. bartramii (18%) than in males (13%) (Fig.18). The remaining 52% of the otoliths in the male O. bartramii stomachs showed similar values for the categories reported for both sexes combined (Fig. 10). The relative abundance ofotoliths recovered from female O. bartramii showed greater proportions than those recovered from males for Chauliodontidae, C. warmingii, L. tenuiformis, D. perspiculatus, M. nitidulum, and L. luminosa (Fig.19), while showing a general decrease in the proportions of otoliths recovered relative to males in all other categories (Fig. 18,19).

Beaks: For O. bartramii, beaks occurred in 143 (55.8%) males, and 26 (46.4%) females. Beaks of Onychoteuthidae occurred more frequently in male (17%) than in female (2%) O. bartramii stomachs (Fig. 20). In categories where both male and female 0. bartramii showed high relative frequencies ofbeaks, the males usually showed greater frequencies of occurrence for beaks than females (Fig. 20). Female O. bartramii did show higher frequencies of occurrence in categories ofbeaks that did not occur frequently in males such as Ommastrephidae (family level comparison), Nototodarus hawaiiensis, Octopoteuthis nielsenii, and Mastigoteuthis sp. (lower taxon-level comparisons) (Fig. 21). Of the total beaks recovered from O. bartramii, 245 (85.0%) were found in males while 43 (14.9%) were found in females. In categories where both male and female O. bartramii showed high relative abundances of beaks, the males usually showed greater

21 abundances than females (Fig. 22,23). Female O. bartramii did show greater abundances of beaks in categories of beaks that did not show high proportions in males (Fig. 22,23).

Gender comparisons (S. oualaniensis)-

Otoliths: For S. oualaniensis, otoliths occurred in 191 (72.9%) females, and 40 (71.4%) males captured. At the family level, otoliths of Myctophidae and unidentified otoliths were the only categories that occurred more frequently in S. oualaniensis females than in males (Fig. 24). Myctophid otoliths occurred more frequently in S. oualaniensis females than in males for all species groups except M. lychnobium or spinosum, L. luminosa, and L. tenuiformis (Fig 25). Of the total otoliths recovered from S. oualaniensis, 850 (91%) were from females and 87 (9%) were from males. Myctophid and unidentified otoliths made up similar percentages ofotoliths found in both female and male S. oualaniensis stomachs (97 and 95% respectively) with members of Evermanellidae, Gonostomatidae, and Nomeidae making up the remaining percentages (Fig. 26). Within Myctophidae both genders of squid contained similar percentages of S. evermanni and the species group M. lychnobium or spinosum (Fig. 27). Female S. oualaniensis contained larger percentages of C. warmingii and had more unidentified otoliths, while males contained more H. proximum or H. rheinhardti (Fig. 27).

Beaks: For S. oualaniensis, beaks occurred in 89 (34.0%) females, and 7 (17.5%) males. In S oualaniensis stomachs, beaks were found in higher frequencies in females than in males for all categories except Ommastrephidae and unidentified beaks (Fig. 28). In female S. oualaniensis stomachs beaks ofAbraliopsis sp. A were found at roughly twice the frequency found in males (Fig. 29). Ofthe total beaks recovered from S. oualaniensis, 124 (91.1%) were found in females while 12 (8.8%) were found in males. In the family level categories where both male and female S. oualaniensis showed high relative abundances of beaks, the males usually showed higher abundances for beaks than females, except for Enoploteuthidae prey (Fig. 30). Female S. oualaniensis showed

22 higher abundances of all lower taxon categories than males except for P. addolux (Fig. 31).

Niche breadth and overlap: Levins' standardized measure of niche breadth in O. bartramii was greater than niche breadth in S. oualaniensis using both otoliths and beaks as indices (Table. 2). Niche breadth for O. bartramii males was greater than females using otoliths, but niche breadth was greater in female O. bartramii than in males using beaks (Table. 2). Niche breadth for S. oualaniensis females and males was approximately equal using otoliths or beaks (Table 2.). The niche breadth for the different size classes of O. bartramii males showed no consistent trend, although the largest size class showed the lowest niche breadth, while niche breadth increased with size class in S. oualaniensis (Table. 2). The value of niche breadth for the large size class of O. bartramii was greatly affected by one stomach that contained 24 otoliths of S. evennanni (7% of all the otoliths in this category). Excluding this individual resulted in a niche breadth value of 0.33 for the large size class of 0. bartramii. The niche breadth of the largest size class of the S. oulaniensis was greater than the breadth calculated for the species as a whole (Table 2.).

23 Table 2. Comparisons ofLevins' modified measure of niche breadth (B) for different categories of squids using ditTerent prey iudices (unidentified items excluded).

Prey indices Ommastrephes bartramii Sthenoteuthis oualaniensis Total (otoliths) 0.56 0.086 Male (otoliths) 0.40 0.085 Female (otoliths) 0.23 0.083 Total (beaks) 0.22 0.097 Male (beaks) 0.19 0.096 Female (beaks) 0.39 0.096 Large size class (otoliths) 0.23 (396-336 mm)[males] 0.12 (300-201 mm) Medium size class (otoliths) 0.41 (335-320 mm)[males] 0.08 (198-161 mm) Small size class (otoliths) 0.34 (319-228 mm)[males] 0.06 (160-120 mm)

The overlap of S. oualaniensis on O. bartramii was greater than the overlap of O. bartramii on S. oualaniensis using both fish otoliths and squid beaks (Table. 3). The overlap of 0. bartramii males on females was greater using cephalopod beaks but less using fish otoliths. Overlap was virtually the same for both male and female S. oualaniensis using both otoliths and beaks (Table. 3). Niche overlap ofthe small O. bartramii size class increased with size of S. oualaniensis (Table 3.). Niche overlap of the different size classes of S. oulaniensis on the small O. bartramii increased with increasing size of S. oulaniensis (Table. 3).

24 Table 3. Comparisons ofMacArthur and Levins measure ofniche overlap for both species ofsquids using different prey indices and different size categories ofsquids. O. barlramii, small = 319-228 mm ML; S. oualaniensis, small = 160-120 nun, medium = 198-161 nun, large = 3()().201 mm. sm= small, md = medium, 19= large

Overlap of one group on another Fish Otoliths Cephalopod beaks O. bartramii on S. oualaniensis 0.22 0.46 S. oualaniensis on O. bartramii 1.16 0.93 O. bartramii male on female 0.31 0.85 O. bartramii female on male 0.52 0.45 S. oualaniensis male on female 0.95 0.80 S. oualaniensis female on male 0.97 0.80 sm O. bartramii on sm S. oualaniensis 0.08 --- sm O. bartramii on md S. oualaniensis 0.11 --- sm O. bartramii on Ig S. oualaniensis 0.15 --- sm S. oualaniensis sm O. bartrumii 0.32 --- md S. oualaniensis sm O. bartramii 0.35 --- Ig S. oualaniensis sm O. bartramii 0.36 ---

Otoliths from myctophid species that wereJound in substantial numbers in both O. bartramii and S. oualaniensis stomachs were significantly larger (otolith length along the long axes of sagitta) in O. bartramii stomachs (Table. 4).

25 Table 4. List of the mean length (long axis of sagitta) and nnmber of otoliths of ditTerent species of myctophids found in O. bartramii and S. oualaniensis stomachs along with the P-vaJues obtained for a two sample t-test against the hypothesis that the mean of otoliths in O. bartramii > otoliths in S. oualaniensis.

Fish species Mean length (mm) and Mean length (mm) and p. numbers for O. bartramii numbers for S, oualaniensis value S. evermonni 4.42 (25) 4.04 (107) 0.001 C. warmingii 3.60 (36) 2.87 (58) <0.001 H. proximum or rheinardti 1.61 (20) 1.47 (63) 0.006 M. lychnobium or spinosum 3.64 (42) 2.70 (29) <0.001 M. selenoides 3.10 (23) 2.53 (9) 0.003

Otolith size comparisons between male and female O. bartramii, conducted for the two categories ofmyctophids (L tenuiformis and C. warmingii) that had enough numbers for comparison, did not yield significant differences at the 95% confidence level. There were, however, differences in otolith sizes recovered from the stomachs of female and male S. oualaniensis for two of the four groups that contained enough otoliths for comparison (Table 5).

Table 5. List of the mean length (long axis of sagitta) and number ofotoliths of ditTerent species of myctopbids found in female and male S. ollalaniensis stomachs along with the P-valnes obtained for a two sample t-test against the hypothesis that the mean size ofotoliths in females> otoliths in males.

Fish species Mean length (mm) and Mean length (mm) and p- value numbers for females numbers for males C. warmingii 2.98 (52) 1.40 (6) <0.001 S. evermanni 4.11 (87) 3.69 (17) 0.03

H.proximum 1.49 (52) 1.32 (9) 0.056 or rheinhardti M. lychnobium 2.64 (22) 2.89 (6) 0.365 or spinosum

26 Discussion

General discussion In this study Teleostei and Cephalopoda comprised the main dietary components of both o. bartramii and S. oualaniensis by abundances and frequencies. These items are typical ofommastrephid prey items in general, and prey of O. bartramii and S. oualaniensis in particular (Young, 1975; Arraya, 1983; Nixon, 1987; Seki, 1993; Schetinnikov,1992). Crustacean material was virtually absent in the stomachs of both species, which has been observed for O. bartramii (Seki, 1993; Naito et al., 1977; Arraya, 1983; Dunning and Brandt, 1985) and for adult S. oualaniensis (Shchetinnikov, 1992; Tung, 1976). Shchetinnikov (1992) found that larger S. oualaniensis (> 155mm ML) switched from a diet of mostly crustaceans to that of fishes. Adult S. oualaniensis, however, have been observed feeding on crustaceans in certain specific areas on the windward sides of the main Hawaiian Islands and off Okinawa (Taguchi et aI., 1985; Tung, 1976). Murata (1990) found crustaceans in the stomachs of o. bartramii, but this was because his analysis included smaller immature squids. Attempts were made to sample individual S. oualaniensis that might show crustacean feeding habits in adult, but this behavior was not observed. Cannibalism has been reported frequently in schooling squids (Nixon, 1987) and has been reported commonly in O. bartramii and S. oualaniensis (Seki, 1993; Araya, 1983; Nigmatullin et al., 1983). Cannibalism was virtually absent in this study for either species. When material was found in the stomachs that suggested cannibalism, the material was almost always fresh and usually consisted of arms and tentacles. Arms and tentacles of the squids were frequently pulled off during attempted landings and thrown back into the water by the fisherman, or attached to the jigs and used as bait. Squids on deck waiting to be processed were also frequently seen chewing on themselves or a nearby squid, although attempts were made to ruinimize this occurrence. Squids, especially O. bartramii females, were often observed attacking other squids caught on jigs. Apparent cannibalism in this study was likely an artifact of sampling squids by jig fishing; Breiby and Jobling (1985) reported similar observations. Seki reported extensive cannabilism in gillnet caught O. bartramii, although this could have resulted from squid

27 feeding on "gilled" squid prior to their own entanglement. Dunning and Brandt (1985) found little or no evidence of cannibalism under similar fishing conditions. The stomach organ weights scaled exponentially with body length in both species. As expected this indicates that as the squid grows its ability to consume larger meals increases. Attraction based methods generally catch animals that are hungry. The largest amount of material was found in one female that consumed 435 g of material (almost entirely fresh squid material from other O. bartramii), which is only about 68% of total calculated fullness using the equation from Okutani and Tung, 1978. The percentages of fullness were greater for males than females in each species. Assuming that the categories with the greater numbers of individuals represent more robust data sets, the O. bartramii males and S. oualaniensis females show that the squids are between 5-13% full during the times when they were sampled.

Otolith data Otolith frequency data suggest that O. bartramii is a more generalized predator than S. oualaniensis. The data show that many groups of fish prey show up at moderate frequencies in O. bartramii stomachs while few fish prey groups show up at high frequencies in S. oualaniensis. The data on frequency of occurrence match the abundance data in both pattern and amount fairly well indicating that large bouts of feeding by certain individuals on select species probably did not bias the abundance data to a large degree. Otoliths had a much greater frequency ofoccurrence and were present in greater numbers overall than cephalopod remains in both O. bartramii and S. oualaniensis. These squids are feeding predominantly on fishes. The fish prey of S. oualaniensis was almost exclusively made up of myctophids while O. bartramii fed on myctophids and several other groups of fishes. Gonostomatids occurred fairly frequently as did several other groups ofmesopelagic fishes that were not fed upon by S. oualaniensis. Disparate feeding also occurred between O. bartramii and S. oualaniensis on prey items within the Myctophidae. Symbolophorus evermanni was a mainstay in the diet of S. oualaniensis ( > 35% of all otoliths found), as were C. warmingii and the species group ofH. proximum or rheinhardti, while the diet of O. bartramii is spread over a wide variety of species with

28 none greatly dominating. No single species of fish in O. bartramii accounted for greater than about 10% of all the otoliths found. In S. oualaniensis, three categories of myctophids accounted for almost 70% of all the otoliths found, while in O. bartramii the three most abundant myctophid categories only accounted for 17% of the total. This prevalence of myctophid prey in S. oualaniensis is supported by other studies from Hawaiian waters and Taiwan (Young, 1975; Tung, 1976; Wormuth, 1976; Okutani and Tung, 1978) and from areas of the Indian Ocean and Red Sea (Zuev, 1971; Filipova, 1974; Nigmatullin et al., 1983). The substantial amount of S. evermanni in the diet of S. oualaniensis is also confirmed by Shchetinnikov (1992) in the Eastern Pacific.

Beak data Similar trends to those seen in the stomach content data based on otoliths were also seen when comparing the two squid predators using beaks. The beak data indicate that the feeding spectra of S. oualaniensis and O. bartramii differ in both the type of species fed upon, and the frequency and proportions of feeding. As was seen with otoliths, O. bartramii fed on more species of cephalopods usually at lower percentages than S. oualaniensis, which fed on fewer species but with greater percentages in the categories it did feed on. Nine of the categories that O. bartramii fed on were not even present in S. oualaniensis stomachs. Both squids fed with high percentages on Onychoteuthis spp. O. bartramii fed with a high percentage on Histioteuthis spp. which S. oualaniensis did not feed on at all while S. oualaniensis fed with higher percentages than O. bartramii on Abraliopsis sp A. and on Abraliopsis spp. which 0. bartramii did not feed on. These results agree to some extent with previous studies that found members of Onychoteuthidae and Enoploteuthidae to occur in O. bartramii and S. oualaniensis stomachs (Murata, 1990; Shchetinnikov, 1992; Seki, 1993).

29 Gender differences O. bartramii- Results from the otolith data indicate that some differential feeding between adult males and the larger adult females of O. bartramii occurs. Females showed a higher abundance of myctophids (66%) than did males (48%). Males had a higher abundance for all other major family level categories except unidentified and Chauliodontidae. Within the Myctophidae the top three categories in males were not found in females. The three categories with the highest abundance percentages in females were found in lower percentages in males. The frequency of occurrence data also show a difference between males and females, prey categories in males occurred more frequently in all categories at the family level or within Myctophidae except for, Chauliodontidae, Apogonidae, Giganturidae, Melanostomiatidae, D. perspiculatus, C. warmingii, and B. longipes. The frequency of occurrence data have to be viewed with some skepticism because ofthe small sample size of females, 56 captured with only 33 stomachs containing otoliths. The abundance data may be more robust, with 164 total otoliths from females. The niche breadth for male O. bartramii was greater than for females but both were lower than that for the species as a whole. Dietary overlap between the sexes was also low indicating that the high values of niche breadth obtained with males and females combined is partly caused by differential feeding between the sexes, with the females feeding broadly enough to raise the overall niche breadth ofthe species. This result contrasts with a previous study done at a more general level farther north that did not find a difference in the feeding between sexes (Seki, 1993). Differential feeding by male and female O. bartramii was also found in the stomach content data using beaks. Males had higher abundance in a number of categories where both sexes fed, such as Onychoteuthis spp. Female O. bartramii had substantial abundances for categories where males did not feed at all such as Octopoteuthis nielseni and Nototodarus hawaiiensis. The trend of greater niche breadths in male relative to female O. bartramii found using otoliths was opposite to that found using beaks. Using the beak data, niche breadth for male O. bartramii was almost half of that for females. Worded another way, the niche breadth for males (using beak data) was less than the value for the whole species while niche breadth in females was greater than the species value. The trend in dietary overlap for male and female O.

30 banramii using beaks was also different than that found using otoliths. Overlap of females on males was low while overlap of males on female O. bartramii was high (> 0.8). These results indicate that, for cephalopod prey, O. bartramii females have a more generalized pattern offeeding than do O. bartramii males, which seem to concentrate more heavily on fewer categories. Caution should be taken when viewing the results of male/female comparisons using the beak data because of the small numbers ofbeaks available from females (43) that were used for these comparisons.

S. oualaniensis- Results from the otolith data on diet of S. oualaniensis did not indicate a substantial difference in prey items between the sexes. The frequencies of occurrence and abundances of otoliths in the diet both showed similar trends between males and females, with small differences in select categories, C. warmingii, M. lychnobium or spinosum, D. jragilis, and L luminosa. The niche breadth of each sex was small and virtually identical both to each other and to the value calculated for the species as a whole. Dietary overlap of males on females and females on males was high (> 0.9) in both cases indicating that the diets of both sexes are similar. The frequency of occurrence data may be suspect because, of the 40 male S. oualaniensis caught, only 19 contained otoliths. The abundance data may be more robust, with 87 otoliths in males. There were little or no differences in feeding by female and male S. oualaniensis based on the beak data. The proportions of several of the cephalopod categories were higher in males than females except for Abraliopsis sp. A, and the categories where males did not feed at all. The frequencies of occurrence were similar between males and females except for Abraliopsis sp. A, Onychoteuthis sp., and the categories where occurrence was zero in males. The niche breadths for each gender were practically identical indicating that their range of food items is the same. The overlap of males on females was also nearly identical to the overlap of females on males. Both of these measures indicate that there are no gender differences in diet between female and male S. oualaniensis. The results should be viewed with some caution because the number of beaks used to calculate these relationships is small (12 beaks occurring in 7 male

31 stomachs out of 40 males total). However, this trend in the beak data is in agreement with that found from the more robust otolith data.

Niche breadth and overlap- The dietary measurements indicate that in Hawaiian waters S. oualaniensis is a more specialized predator than O. bartramii.l..evins' measure relies only on the proportions of prey categories found in the stomachs and is not affected by the number of categories assigned by the researcher (although theoretically B would approach the number of categories evaluated if all feeding was in equal proportions across all categories), and these values reveal a large margin between the niche breadths of each squid, 20.1 for O. bartramii and 3.85 for S. oualaniensis. The standardized values reveal an even larger margin between the two, with O. bartramii having a niche breadth of 0.56 and S. oualaniensis at 0.086. Niche breadth increased slightly with size class for S. oualaniensis indicating that as the animal grows larger its diet expands. Schetinnikov (1992) also reported dietary changes with size for S. oulaniensis. The niche breadth values for small and large O. bartramii were nearly identical when the individual with the large numbers ofS. evermanni was removed. Although the difference between the smallest and largest breadths was similar for both species size classes (roughly 0.07 for each) the difference as a percent of the whole was much greater in S. oualaniensis. These data use the same categories of prey to calculate the standardized niche breadth for feeding in each species; this process does not take into account the fact that there may be other potential prey categories that were not represented in the individuals sampled. Oiven the sample sizes caught for each squid, most of the prey categories are probably represented. The unidentified otoliths were not included in the calculations. Because the unidentified otoliths only represented around 13% of the total for both species, and encompassed at least 10 easily differentiable categories, the effect of each category on niche breadth values would be minor. Many families of fishes were present only in O. bartramii and since several species appear to be encompassed in these family categories, niche breadth in O. bartramii is likely broader, and the actual difference in niche breadth between these squids probably greater than reported here.

32 Measures of dietary overlap calculated from the proportions of all the otolith categories (excluding the unidentified otoliths) revealed unequal overlap between the two squid species. Values of overlap> 0.7 (calculated with MacArthur and Levins' index) have been considered substantial in previous studies of other animals (Macpherson, 1981; Ellis et al., 1996; Heithaus, 2001). Some values of overlap from other studies ranged up to 1.8 but values above 1 were considered high (Heithaus, 2001). The overlap of O. bartramii diet onto that of S. oualaniensis was low (0.22), indicating that species fed upon by O. bartramii were not fed upon to the same degree, or at all, by S. oualaniensis. The overlap of S. oualaniensis diet on that of O. bartramii was high (1.16) indicating that species fed upon by S. oualaniensis were present in the diet of O. bartramii to a substantial degree. The results of these calculations indicate that diet composition of S. oualaniensis is a subset of the diet of 0. bartramii. This result is confirmed by finding only one category of fish, Nomeidae, that was present in S. oualaniensis and not in O. bartramii. Nomeidae only occurred once in S. oualaniensis with two otoliths being present. On the other hand, thirteen categories of fishes occurred exclusively in O. bartramii but not in S. oualaniensis, many of these categories occurred frequently and with high percentages, e.g. Stemoptychidae, Melamphaeidae, etc.. Dietary overlap increased for both species as the larger size classes of S. oualaniensis approached the smallest size class of O. bartramii. The increases in overlap were small but consistent throughout the size classes. The niche breadth of O. bartramii was broader than that of S. oualaniensis in the calculations based on cephalopod beaks. The difference in niche breadths was not as large as that found in the otoliths but breadth in O. bartramii was still more than twice that of S. oualaniensis. The calculations of niche overlap also showed the same trends with the beaks as they did with the otoliths. Overlap of O. bartramii diet on S. oualaniensis diet was low (0.46), while overlap of S. oualaniensis on O. bartramii was high (0.93). The trends in differences of niche breadth and dietary overlap between O. bartramii and S. oualaniensis found in the beak data were the same as those found in the otolith data but the magnitude of the differences was less extreme. The beak data indicate (as did the otolith data) that O. bartramii feeds moderately on many categories in a broad

33 range ofprey items while S. oualaniensis feeds more heavily on a few categories in a relatively narrow range of prey items. Statistically significant differences occurred in the sizes ofotoliths found in 0. bartramii versus S. oualaniensis. In the fi ve species groups of myctophids where there were substantial numbers of otoliths for comparison, larger mean lengths of otoliths were found in every group in O. bartramii stomachs. Shchetinnikov (1992) found three feeding stages in S. oualaniensis with regard to prey size: when the squids were small (40-100 mm ML) prey size remained constant, in medium sized squids (100-150 mm ML) the size ofprey increased proportionally with squid predator, and in larger squids (150-365 mm ML) the prey size increased more rapidly than did predator size. A trend of prey size increase with predator size within S. oualaniensis and O. bartramii was not evident in otolith length or beak length. The length of myctophid prey (calculated from regressions of otolith length versus fish length) for C. warmingii, H. proximumlrheinhardti, M. selenoides, and S. evermanni, ranged from 12 to 27% ofthe average length of S. oualaniensis. These values agree with the range of prey sizes relative to the group II sizes of S. oualaniensis (ML= 100-150 mm) found by Shchetinnikov (1992) of 16 to 21% for myctophids. S. oualaniensis shows a typical size selection ofprey in the presence of O. bartramii feeding on the same groups. Given the results of differential size selection of shared prey between O. bartramii and S. oualaniensis, and the differential feeding by male and female O. bartramii, a difference in size of prey within categories shared by O. bartramii genders might be expected. There were only two categories that contained sufficient numbers of otoliths to compare males and females of O. bartramii and neither prey category demonstrated a significant size difference between the genders. S. oualaniensis males and females did not exhibit differences in niche breadth and their dietary overlap was nearly identical, however, significant differences in the otolith sizes of two prey categories, C. warmingii and S. evermanni did occur. Feeding by S. oualaniensis on H. proximum or rheinhardti was not significantly different between genders at the 95% confidence level, although with a p value of 0.056 the differences were close to significant. In both cases of Significant differences of otolith size between gender, females preyed upon larger sizes of myctophids. Given the lack of strong trends between overall S. oualaniensis size and

34 otolith sizes of C. wanningii and H. proximumlrheinhardti these results apparently depict a gender specific feeding preference.

Conclusions

Understanding competition between wide ranging open ocean species is difficult given present limitations of the data. In addition, the habitat virtually excludes experimental manipulation, such as removal or exclusion experiments. Zuev et aI., (1975; quoted in Shchetinnikov, 1992) states "the systematic and ecological similarity of [So oualaniensis, O. bartramii, and other ommastrephids] has led to division of the open ocean...through interspecific competition." If so, one might expect these species to show strong competitive interactions in areas of syrnpatry. The proportions and frequencies of items in the diets of O. bartramii and S. oualaniensis, as well as the dietary measurements calculated from these values, indicate that, although S. oualaniensis feeds on a subset of the O. bartramii diet, their feeding patterns are basically different. The stomach contents data show that O. bartramii feeds on a broader selection of species than S. oualaniensis and feeds more evenly across these categories. Not only does O. bartramii feed more broadly than S. oualaniensis on fishes and squids at the family level, it also feeds more generally at the species level. When the prey items are classified to genus and species levels, substantial fine scale differences in diet emerge. Even when feeding on the same species of fishes, dietary differences are evident in the size of fishes eaten, with larger fish being eaten by O. bartramii relative to those eaten by S. oualaniensis. The diets of both squids give good agreement with the diets of other studies that have looked at these species separately, arguing against a possible niche shift when they are located together geographically. Whether these dietary differences are an effect of resolved competition, or simply a result of the size difference between these two species, presumably unrelated to competition, cannot be determined. However, the niche breadth of S. oualaniensis increased with size, and the dietary overlap increased between the smaller size O. bartramii and the increasing sizes of S. oulaniensis, indicating that size may be a major factor in controlling the feeding habits of these animals relative to each other. These

35 results raise the question of whether or not juveniles of similar size for each species represent the stage where competition between them occurs. Because of the difficulties in capturing the juvenile forms, little is known about the juvenile stage of either species, but the peak spawning periods differ between these squids and migration causes O. bartramii to be absent from S. oualaniensis habitats for a long period when these species would directly overlap in size. In the size ranges of squids sampled direct interactions between these species, such as predation, were not observed in the areas sampled. This finding is contrary to studies in other areas that have shown substantial feeding on closely related squids in this family with comparable size differential, and contrary to expectations. Apparently competition between o. bartramii and S. oualaniensis for food resources is low. Direct predation by one species on another was also not substantial during this study. These results call into question the underlying reasons for the geographical separation of these species, and the separation ofknown peaks in breeding seasons. Given the lack of evidence for direct interaction and/or competition for food resources between O. bartramii and S. oualaniensis, competitive exclusion of one species by another does not seem likely. The present feeding patterns displayed by each species could be the result of past competition. Perhaps intense competition between the juvenile forms of these two species accounts for the evolution of their partial geographical and temporal separation. Female O. bartramii are rarely caught or seen at the southern end of the migration, suggesting that they are occupying a deeper nighttime habitat in their southern range. If so, this deeper nighttime habitat would spatially segregate them from male O. bartramii and from S. oualaniensis, and lessen the probability of direct predation of large female O. bartramii on S. oualaniensis.

36 References:

Abrams, P. (1980). "Some comments on measuring niche overlap." Ecology 61: 44-49.

Andrews, P. L. R., and E. M. Tansey (1983). "The digestive tract of Octopus vulgaris: The anatomy, physiology and pharmacology of the upper tract." Journal of the Marine Biological Association of the United Kingdom 63: 109-134.

Araya, H. (1983). "Fishery biology and stock assessment of Ommastrephes bartramii in the North Pacific Ocean." Memorandum of the National Museum of Victoria 44: 269­ 283.

Ashmole, N., and M. Ashmole (1967). "Comparative feeding ecology of sea birds of a tropical oceanic island." Bulletin of the Peabody Museurn of Natural History 24: 1-131.

Bidder, A. M. (1966). Feeding and digestion in cephalopods. Physiology of Mollusca. K. M. Wilbur, and C. M. Yonge. New York, Academic Press. II: 97-124.

Bower, J. R. (1996). "Estimated paralarval drift and inferred hatching sites for Ommastrephes bartramii (Cephalopoda: Ommastrephidae) near the Hawaiian Archipelago." Fishery Bulletin 79(2): 277-292.

Bradbury, H. E., and F. A. Aldrich (1969). "Observation on feeding of the squid Illex illecebrosus (LeSueur, 1821) in captivity." Canadian Journal of Zoology 47: 913-915.

Brieby, A., and M. Jobling (1985). "Predatory role of the flying squid (Todarodes sagitattus) in north Norwegian waters." North American Fisheries Organization (NAPO) Science Council Studies 9: 125-132.

Brock, R. E. (1985). "Preliminary study of the feeding habits of pelagic fish around Hawaiian fish aggregation devices of can fish aggregation devices enhance local fisheries productivity?" Bulletin of Marine Science 37(1): 40-49.

Casteel, R. W. (1974). "Identification of the species of Pacific salmon (genus Oncorhynchus) native to North America based upon otoliths." Copeia 74: 305-311.

Clarke, A., P. G. Rodhouse, and D. J. Gore (1994). "Biochemical composition in relation to the energetics of growth and sexual maturation in the ommstrephid squid Illex argentinus. " Philosophical Transactions of the Royal Society ofLondon (B) 344: 201­ 212.

Clarke, M. R. (1966). "A review ofthe systematics and ecology of oceanic squids." Advances in Marine Biology 4: 91-300.

37 Clarke, M. R. (1986). A handbook for the identification of cephalopod beaks. Oxford, Clarendon Press.

Cortes, E., C. A. Manire, and R. E. Hueter (1996). "Diet, feeding habits, and diel feeding chronology of the bonnethead shark Sphyrna tiburo, in southwest Florida." Bulletin of Marine Science 58(2): 353-367.

Crisp, D. T., R. H. K. Mann, and J. C. McCormack (1978). "The effects of impoundment and regulation upon the stomach contents of fish at Cow Green, Upper Teesdale." Journal ofFish Biology 12: 287-301.

Dunning, M., and S. B. Brandt (1985). "Distribution and life history of deep-water squid of commercial interest from Australia." Australian Journal of Marine and Freshwater Research 36: 343-359.

Ellis, J. R., M. G. Pawson, and S. E. Shackley (1996). "The comparative feeding ecology of six species of shark and four species ofray (Elasmobranchii) in the north-east Atlantic." Journal of the Marine Biological Association of the United Kingdom 76: 89­ 106.

Fillipova, Y. A. (1974). "On the feeding habits of oceanic squid of the family Ommastrephidae." Trudy VNlRO 99: 123-132.

Fitch, J. E., and R. L. Brownell (1968). "Fish otoliths in cetacean stomachs and their importance in interpreting feeding habits." Journal of the Fisheries Research Board of Canada 25: 2561-2574.

Frost, W. E. (1954). "Thefood ofthe pike, Esox lucius, L., in Windermere." Journal of Animal Ecology 23: 339-360.

Harman, R. F., and R. E. Young (1985). "The larvae of Ommastrephid squids (Cephalopoda, Teuthoidea) from Hawaiian waters." Vie Milieu 35: 211-222.

Heithaus, M. R. (2001). "Predator-prey and competitive interactions between sharks (order Selachii) and dolphins (suborder Odontoceti): a review." Journal of Zoology, London 253: 53-68.

Holt, R. D. (1987). "On the relation between niche overlap and competition: the effect of incomensurable niche dimensions." Oikos 48: 110-114.

Hunt, 8. P., and N. F. Carbine (1951). "Food of young pike and associated fishes in Peterson's Ditches, Houghton Lake, Michigan." Transactions of the American Fish Society 80: 67-73.

38 Hunt, P. C., and J. W.Jones (1972). "The food ofbrown trout in Llyn Alaw, Anglesey, North Wales." Journal ofFish Biology 4: 333-352.

Hunter, S. (1983). "The food and feeding ecology of the giant petrels Macronectes halli and M. giganteus at South Georgia." Journal of Zoology 200: 521-538.

Hurlbert, S. H. (1978). "The measurement of niche overlap and some relatives." Ecology 59: 67-77.

Hyslop, E. J. (1980). "Stomach contents analysis- a review of methods and their application." Journal ofFish Biology 17: 411-429.

Ikusemiju, K., and C. 1. O.OIaniyan (1977). "The food and feeding habits of the catfishes Chrysichthys walkeri (Gunther), Chrysichthys filamentosus (Boulenger) and Chrysicthys nigrodigitatus (Lacepede) in the Lekki Lagoon, Nigeria." Journal ofFish Biology 10: 105-112.

Kajimura, H., and T. R. Loughlin (1988). "Marine mammals in the oceanic food web of the eastern subarctic Pacific." Bulletin of the Oceanic Research Institute, University of Tokyo 26(2): 187-223.

Kelly, K. A. (1980). "Gastric emptying of liquids and solids: Roles of proximal and distal stomach." American Journal of Physiology 239: G71-G76.

King, J. E" and J. J. Ikehara (1956). "Comparative study of food ofbigeye and yellowfin tuna in the Central Pacific," Fishery Bulletin 57: 61-85.

Kornilova, G. N. (1981). "Feeding of yellowfin tuna, Thunnus albacares, and bigeye tuna, Thunnus obesus, in the Equatorial Zone of the Indian Ocean." Journal of Ichthyology 20(6): 111-119.

Krebs, C. J. (1989). Ecological Methodology. New York, Harper Collins.

LeDrew, B. R., and J. M.Green (1975). "Biology of the radiated shanny Ulvaria subbifurcata Storer in Newfoundland (piscesL Stichaeidae)." Journal ofFish Biology 7: 485-495.

Levins, R. (1968). Evolution in Changing Environments. Princeton, New Jersey, Princeton University Press.

Linton, L. R" R. W. Davies, and F. J. Wrona (1981). "Resource utilization indices: an assessment." Journal of Animal Ecology 50: 283-292.

MacArthur, R. H., and R. Levins (1967). "The limiting similarity, convergence, and divergence of coexisting species." American Naturalist 101: 377-385.

39 Macpherson, E. (1981). "Resource partitioning in a Mediterranean demersal fish community." Marine Ecological Progress Series 4: 183-193.

Matsumoto, W. M. (1984). "Synopsis of biological data on skipjack tuna, Katsuwonus pelamis." NOAA Technical Report 451: 1-92.

Meyer, J. H. (1980). "Gastric emptying of ordinary food: Effect of antrum on particle size." American Journal of Physiology 239: G133-G135.

Moiseev, S. I. (1991). "Observation of the vertical distribution and behavior of nektonic squids using manned submersibles." Bulletin of Marine Science 49: 446-456.

Morrow, J. E. (1979). "Preliminary keys to otoliths of some adult fishes of the Gulf of Alaska, Bering Sea and Beaufort Sea." NOAA Technical Reports, National Marine Fisheries Service Circulations 42: 1-32.

Murata, M. (1990). "Oceanic resources of squids." Marine Behavioral Physiology 18: 19­ 71.

Murata, M., and Y. Nakamura (1998). Seasonal migration and diel vertical migration of the neon flying squid, Ommastrephes bartramii, in the North Pacific. International symposium on large pelagic squids, Japan Marine Fishery Resources Research Center, Tokyo.

Naito, M., K. Murakami, and T. Kobayashi (1977). "Growth and food habits of oceanic squids (Ommastrephes bartramii, Onychoteuthis borealijaponicus, Berryteuthis magister, and Gonatopsis borealis) in the western subarctic Pacific region." Research Institute of North Pacific Fisheries Special Volume: 339-351.

Nakamura, Y. (1985). "Billfishes of the world. An annotated and illustrated catalogue of marlins, sailfishes, spearfishes, and swordfishes known to date." FAO species catalogue 125(5): 1-65.

Nakamura, Y. (1993). Vertical and horizontal movements ofmature females of Ommastrephes bartramii observed by ultrasonic telemetry. Recent advances in CephalOPOd fisheries. T. Okutani, R. K. O'Dor, and T. Kubodera. Tokyo, Tokai University Press: 31-36.

Nesis, K. N. (1993). Population structure of oceanic ommastrephids with particular reference to Sthenoteuthis oualaniensis: A review. Recent advances in cephalopod fisheries biology. T. Okutani, R. K. O'Dor, and T. Kubodera. Tokyo, Tokai University Press: 375-383.

NigmatuUin, C. M., A. S. Shchetinnikov, S. I. Bazanov, and A. M. Pinchukov (1983). Food spectrum ofthe squid Sthenoteuthis oualaniensis in the Indian Ocean and Red Sea.

40 Taxonomy and ecology of Cephalopoda. Y. I. a. K. N. N. Starobogatov. Leningrad, Russia, Zoological Institute of the USSR: 94-96.

Nixon, M. (1987). Cephalopod diets. Cephalopod Life Cycles. P. R. Boyle. London, Academic Press Inc. (London) Ltd. II: 201-219.

Norden, C. R. (1961). "Comparitive osteology of representative salmonid fishes, with particular reference to the grayling (Thymallus areticus) and its phylogeny." Journal of the Fisheries Research Board of Canada 18: 679-791.

Okutani, T., and I. H. Tung (1978). "Reviews of biology of commercially important squids in Japanese and adjacent waters. I. Symplectoteulhis oualaniensis (Lesson)." Veliger 21: 87-94.

Okutani, T., and Y. Satake (1978). "Squids in the diet of 38 sperm whales caught in the Pacific waters off northeastern Honshu, Japan, February 1977." Bulletin of the Tokai Regional Fisheries Research Laboratory 93: 13-27.

Okutani, T., Y. Satake, S. Ohsumi, and T. Kawakami (1976). "Squids eaten by sperm whales caught offJoban district, Japan during January-February 1976." Bulletin of the Tokai Regional Fisheries Research Laboratory 87: 67-113.

Parker, R. R. (1963). "Effects of formalin on length and weight of fishes." Journal ofFish Research of Canada(20): 144-155.

Pierce, J. P., P. R. Boyle, L. C. Hastie, and M. B. Santos (1994). "Diets of squid Loligo forbesi and £Oligo vulgaris in the northeast Atlantic." Fisheries Research 21: 149-163.

Pillay, T. V. R. (1952). "A critique of the methods of study of food in fishes." Journal of the Zoological Society ofIndia 4: 185-200.

Pinkas, L., M. S. Oliphant, and I. L. K. Iverson (1971). "Food habits of albacore, bluefin tuna, and bonito in California waters." Fishery Bulletin 152: 1-105.

Pitcher, K. W. (1980). "Stomach contents and feces as indicators of harbor seal Phoca vitulina foods in the Gulf of Alaska." Fishery Bulletin 78: 797-798.

Porteiro, F., H. Martins, and R. T. Hanlon (1990). "Some observations on the behavior of adult squids, £oligoforbesi, in captivity." Journal of the Marine Biological Association of the United Kingdom 70: 459-472.

Reintjes, J. W., and J. E. King (1953). "Food of yellowfin tuna in the central Pacific." Fishery Bulletin 54: 91-110.

41 Roper, C. F. E., M. J. Sweeney, and C. E. Nauen (1984). "Cephalopods of the world. An annotated and illustrated catalogue of species of interest to fisheries." FAO species catalogue 125(3): 1-277.

Saito, H., and T. Kubodera (1998). Distribution ofommastrephid rynchoteuthion paralarvae (Mollusca:Cephalopoda) in the Kuroshio Region. Recent advances in cephalopod fisheries biology. T. Okutani, R. K. ODor, and T. Kubodera. Tokyo, Tokai University Press: 457-466.

Schmidt, W. (1968). "Vergleichend morphologische Studie uber die Otolithen mariner Knochenfische." Arkiv fur Fishereiwissenschaft 19: 1-96.

Scott, T. (1906). "Observations on the otoliths of some teleostean fishes." Reports of Fisheries Research of Scotland 3: 48-82.

Seki, M. P. (1993). Trophic relationships of Ommastrephes bartramii during winter migrations to subtropical waters north of the Hawaiian islands. Recent Advances in Fisheries Biology. T. Okutani, O'Dor, R. K. and Kubodera, T. Tokyo, Tokai University Press: 523-529.

Seki, M. P., and C. S. Harrison (1989). "Feeding ecology of two subtropical seabird species at French Frigate Shoals, Hawaii." Bulletin of Marine Science 45: 52-67.

Shchetinnikov, A. S. (1992). "Feeding spectrum of squid Sthenoteuthis oualaniensis (Oegopsida) in the Eastern Pacific." Joumal ofthe Marine Biological Association of the United Kingdom 72: 849-860.

Suzuki, T. (1990). "Japanese common squid- Todarodes pacificus Steenstrup." Marine Behavioral Physiology 18: 73-109.

Swynnerton, G. H., and E. B. Worthington (1940). "Notes on the food of fish in Haweswater (Westmorland)." Journal of Animal Ecology 9: 183-187.

Tung, 1. H. (1976). "On the food habit ofcommon squid Symplectoteuthis oualaniensis (Lesson)." Report of the Institute ofFisheries Biology, Taipei 3(2): 49-66.

Wallace, 1. C., R. K. O'Oor, and T. Amaratunga (1981). "Sequential observations on the digestive process of the squid Illex illecebrosus." North American Fisheries Organization (NAPO) Scientific Council Studies 1: 65-69.

Wormuth, J. H. (1976). "The biogeography and numerical taxonomy of the oegopsid squid family Ommastrephidae in the Pacific Ocean." Bulletin of the Scripps Institution of Oceanography 23: 1-90.

Young, J. W., T. D. Lamb, D. Le, R. W. Bradford, and A. W. Whitelaw (1997). "Feeding ecology and interannual variations in diet of southern bluefin tuna, Thunnus maccoyii, in

42 relation to coastal and oceanic waters off eastern Tasmania, Australia." Environmental Biology of Fishes 50(275-291).

Young, R. E. (1975). "A brief review of the biology of the oceanic squid, Symplectoteuths oualaniensis (Lesson)." Comparitive Biochemistry and Physiology 52B: 141-143.

Young, R. E., and J. Hirota (1990). "Description of Ommastrephes bartramii (Cephalopoda: Ommastrephidae) paralarvae with evidence for spawning in Hawaiian waters." Pacific Science 44: 71-80.

Young, R. E., and J. Hirota (1998). Review of the ecology of Sthenoteuthis oualaniensis near the Hawaiian Archipelago. International symposium on large pelagic squids, Japan Marine Fishery Resources Research Center, Tokyo.

Yuen, H. S. H. (1979). "A night handline fishery for tunas in Hawaii." Marine Fisheries Review 41(8): 7-14.

Zuev, G. V. (1971). Cephalopods from the north-western part of the Indian Ocean. Kiev, Naukova Dumka.

43 List of Tables

TABLE 1. THE NUMBER OF SQUIDS OF BOTH SPECIES CAUGHT DURING EACH YEAR (THE

NUMBER OFS. OUALANlENSlS CAUGHT DURING 1999 INCLUDES SPECIMENS TAKEN

ABOARD THE RIV TOWNSEND CROMWELL) 15

TABLE 2. COMPARISONS OF LEVINS' MODIFIED MEASURE OF NICHE BREADTH (B) FOR

DIFFERENT CATEGORIES OF SQUIDS USING DIFFERENT PREY INDICES (UNIDENTIFIED

ITEMS EXCLUDED) 24

TABLE 3. COMPARISONS OF MACARTHUR AND LEVINS MEASURE OF NICHE OVERLAF FOR

BOTH SPECIES OF SQUIDS USING DIFFERENT PREY INDICES AND DIFFERENT SIZE

CATEGORIES OF SQUIDS. O. BARTRAMll, SMALL = 319-228 MM ML; S. OUALANlENSlS,

SMALL = 160-120 MM, MEDIUM = 198-161 MM, LARGE = 300-201 MM. SM= SMALL,

MD = MEDIUM, LG= LARGE 25

TABLE 4. LIST OF THE MEAN LENGTH (LONG AXIS OF SAGillA) AND NUMBER OF OTOLITHS

OF DIFFERENT SPECIES OF MYCTOPHIDS FOUND IN O. BARTRAMll AND S. OUALANlENSlS

STOMACHS ALONG WITH THEP-VALUES OBTAINED FORA TWO SAMPLE T-TEST

AGAINST THE HYPOTHESIS THAT THE MEAN OF OTOLITHS IN O. BARTRAMIl> OTOLITHS

IN S. OUALANIENSlS 26

TABLE 5. LIST OF THE MEAN LENGTH (LONG AXIS OF SAGITTA) AND NUMBER OF OTOLITHS

OF DIFFERENT SPECIES OF MYCTOPHIDS FOUND IN FEMALE AND MALE S. OUALANIENSIS

STOMACHS ALONG WITH THE P-VALUES OBTAINED FOR A TWO SAMPLE T-TEST

AGAINST THE HYPOTHESIS THAT THE MEAN SlZE OF OTOLITHS INFEMALES> OTOLITHS

IN MALES 26

44 List of Figures

Figure 1. Map showing the ship tracks for the cruises from 1996-2001 aboard the FrS Hokusei Maru and the stations for the two cruises in 1999 aboard the RJV Townsend Cromwell...... 48 Figure 2. A picture of typical jig lures used in catching squids, jigs shown here are barbless type (modified from Suzuki, 1990) 49 Figure 3. An automated jig machine similar to the machines used on the Hokusei Maru. Picture taken from Suzuki, 1990 50 Figure 4. Drawing of the driftline used during cruises aboard the Hokusei Maru. Drawing depicts a set of ten jigs with lights place at 5 jig intervals. Driftline constructed by Jed Hirota. Jigs and lights modified from Suzuki, 1990 51 Figure 5. Percent fullness versus mantle length for 0. bartramii. Two clusters are apparent in the mantle lengths representing males (smaller) and females (larger). Percent fullness was calculated from Tung (1978) and represents the relationship between stomach content weight of "full" stomachs and mantle length for a confamilial species, S. oualaniensis 52 Figure 6. Percent fullness versus mantle length for S. oualaniensis. Percent fullness is calculated from Tung (1978) that represents the relationship between stomach content weight of "full" stomachs and mantle length for S. oualaniensis 53 Figure 7. Weight of stomach organ versus mantle length for the squids, O. bartramii and S. oualaniensis caught between 1999-2001. 54 Figure 8. The frequency of occurrence of otoliths ofdifferent families ofprey in the stomachs of O. bartramii and S. oualaniensis. Frequency ofoccurrence is the proportion of stomachs that a specific category is found in 55 Figure 9. The frequency of occurrence of otoliths ofdifferent myctophid prey in the stomachs of O. bartramii and S. oualaniensis. Frequency of occurrence is the proportion of stomachs that a specific category is found in 56 Figure 10. The relative abundance ofotoliths ofdifferent families of prey in the stomachs of O. bartramii and S. oualaniensis. Relative abundance is the proportion of otoliths of that category out of the total from all stomachs 57 Figure 11. The relative abundance of otoliths ofdifferent myctophid prey in the stomachs of O. bartramii and S. oualaniensis. Relative abundance is the proportion of otoliths of that category out of the total from all stomachs 58

4S Figure 12. The frequency of occurrence of beaks ofdifferent prey, identified to the family level, in the stomachs of O. bartramii and S. oualaniensis. Frequency ofoccurrence is the proportion of stomachs out ofall stomachs that contain a specific category 59 Figure 13. The frequency ofoccurrence of beaks of different prey, identified to the lowest possible taxa, in the stomachs of O. bartramii and S. oualaniensis. Frequency of occurrence is the percentage of stomachs out of all stomachs that contain a specific category 60 Figure 14. The relative abundance of beaks ofdifferent prey, identified to the family level, in the stomachs of O. bartramii and S. oualaniensis. Relative abundance is the proportion of beaks in that category out of the total from all stomachs 61 Figure 15. The relative abundance of beaks of different prey items, identified to the lowest possible taxa, in the stomachs of O. bartramii and S. oualaniensis. Relative abundance is the proportion of beaks in that category out ofthe total collected from all stomachs 62 Figure 16. The frequency of occurrence of otoliths of different families of prey in the stomachs of both O. bartramii genders. Frequency of occurrence was calculated for the number of stomachs in each category 63 Figure 17. The frequency of occurrence of otoliths ofdifferent myctophid prey in the stomachs of both 0. bartramii genders. Frequency ofoccurrence was calculated for the number of stomachs in each category 64 Figure 18. The relative abundance ofotoliths ofdifferent families ofprey in the stomachs of both O. bartramii genders. Relative abundance is the proportion ofotoliths of that category out of the total otoliths from the stomachs 65 Figure 19. The relative abundance ofotoliths ofdifferent myctophid prey in the stomachs ofboth 0. bartramii genders. Percent abundance is the proportion of otoliths of that category out of the total in the stomachs 66 Figure 20. The frequency ofoccurrence of beaks ofdifferent prey, identified to the family level, in the stomachs of both genders of O. bartramii. Frequency of occurrence is the proportion ofstomachs out stomachs in each gender that contain a specific category 67 Figure 21. The frequency of occurrence of beaks ofdifferent prey, identified to the lowest possible taxa, in the stomachs of both 0. bartramii genders. Frequency ofoccurrence was calculated as a percentage of male and female stomachs that contain a certain category...... 68 Figure 22. The relative abundance of beaks ofdifferent prey, identified to the family level, in the stomachs of both genders of O. bartramii 69 Figure 23. The relative abundance of beaks ofdifferent prey, identified to the lowest possible taxa, in the stomachs of both 0. bartramii genders 70

46 Figure 24. The frequency ofoccurrence ofotoliths ofdifferent families ofprey in the stomachs of both S. oualaniensis genders 7l Figure 25. The frequency of occurrence of otoliths of different myctophid prey in the stomachs of both S. oualaniensis genders 72 Figure 26. The relative abundance ofotoliths of different families of prey in the stomachs of both S. oualaniensis genders 73 Figure 27. The relative abundance of otoliths ofdifferent myctophid prey items in the stomachs of both S. oualaniensis genders 74 Figure 28. The frequency of occurrence of beaks of different prey, identified to the family level, in the stomachs ofboth genders of S. oualaniensis 75 Figure 29. The frequency ofoccurrence of beaks of different prey, identified to the lowest possible taxa, in the stomachs of both S. oualaniensis genders 76 Figure 30. Relative abundance of beaks of different prey, identified to the family level, in the stomachs of both genders of S. oualaniensis 77 Figure 31. The relative abundance ofbeaks of different prey items, identified to the lowest possible taxa, in the stomachs of both S. oualaniensis genders 78

47 S hip Tracks • 1996 (' 1998 e 2000 o 1997 .1999 02001 19 C -==-=_ 30 N eTC99 19C K"C • J.~

o 25 N o • .. ' .. \ ' .. -..'

20 N

r--J 100 km

170 W 165 W 160 W 155 W 150 W

Figure 1.- Map showing the ship tracks for the cruises from 1996-2001 aboard the FTS Hokusei Maru and the stations for the two cruises in 1999 aboard the RNTownsend Cromwell.

48 Figure 2. -A picture oftypical jig lures used in catching squids, jigs shown here are barbless type (modified from Suzuki, 1990).

49 Figure 3.- An automated jig machine similar to the machines used on the Hokusei Maru. Picture taken from Suzuki, 1990.

50 Occan surface

Weight

Jig Light

Figure 4. - Drawing ofthe driftline used during cruises aboard the Hokusei Maru. Drawing depicts a set often jigs with lights placc at 5 jig intervals. Driftline constructed by Jed Hirota. Jigs and lights modified from Suzuki, 1990.

51 100 0 Male QI) Female 80 ,....~co 0 ....OJ 0 Ol 60 c :3 l- 8 0 E 0 40 oJ> ~ ~ 8'@ 0 -rn 0 rn o~o (]) c ~O 0 20

200 300 400 500 600 700 Mantle length (mm) O. bartramii Figure 5. Perceut fulluess versus mantle length for O. bartramii. Two clusters are apparent in the mantle lengths representing males (smaller) and females (larger). Percent fullness was calculated from Tung (1978) and represents the relationship between stomach content weight of"full" stomachs and mantle length for a confamilial species, S. oualaniensis.

52 100 -,------;======::;l a a a Female II Male 80 a o Ol a a C ::l 60 l- a E e II 't:- a Ul 40 o Ul (]) c "5 o o 0 -~ o a o 20 o o o o a o +-----€~1lIlH1lH 100 150 200 250 300 350 Mantle length (mm) of S. oualaniensis Figure 6. Percent fullness versus mantle length for S. ouaJaniensis. Perceut fulluess is calculated from Tung (1978) that represents the relationship between stomach content weight of "full" stomachs and mantle length for S. oualaniensis

53 70 CIt 0. bartramii 60 A S. Qua/aniensis

~ Ul E 50 ~ IIll / '"Cl ~ .s:::.- / Cl 40 'iii lit ~ / r::: 30 Cl ~ '"0 .s:::. u 20 E 0'" en- 10

0

0 200 400 600 Mantle length (mm)

Figure 7. Weight of stomach organ versus mantle length for the sqnids, O. barlramii and S. ouaillnwnsis canght between 1999·2001.

54 70 E .0. bartramii .~ 60 >- ms. oualaniensis ~ C- Ol 50 cc ro E 0 40 (.) en ..c (.)

Fish family

Figure 8. The frequency of occurrence ofotoliths of different families of prey in the stomachs of O. bartramii and S. oualaniensis. Frequency ofoccurrence is the proportion of stomachs that a specific category is found in.

55 40 • O. bartramii ~ 35 0. • S. oualaniensis g' 30 '2:

Sc: 25 o (,) l/) .s::. 20 ~ E 15 ~ "C 'S CT l/) '0 "#

~tophid species

Figure 9. The frequency ofoccurrence ofotoliths of different myctophid prey in the stomachs ofO. bartramii and S. oualaniensis. Frequency ofoccurrence is the proportion ofstomachs that a specific category is found in.

56 90 ,------, 80 .0. bartramii 70 • S. oualaniensis ~ :c: 60 ~ 50 ~ 40 '0 *- 30 20 10 ­ o

Figure 10. The relative abundance ofotoliths ofdifferent families of prey in the stomachs of O. bartramii and S. oualaniensis. Relative abundance is the proportion ofotoliths of that category out of the total from all stomachs.

57 40 35 • O. bartramii • S. oualaniensis 30

III E 25 ~ 20 iii:g '0 15 ?fl. 10 5 0

Figure 11. The relative abundance ofotoliths ofdifferent myctophid prey in the stomachs ofO. bartramii and S. oualaniensis. Relative abundance is the proportion ofotoliths of that category out of the total from all stomachs.

58 18 -,------., E 16 ~ • 0. bartramii ~ 14 C s. oualaniensis a. 12 Ol c: C 10 (ij C 8 o o 6 ­ .ctil o ell 4 E o ~ 2 '0 o

Cephalopod family

Figure 12. The frequency ofoccurrence of beaks of different prey, identified to the family level, in the stomachs ofO. bartramii and S. oualaniensis. Frequency of occurrence is the proportion ofstomachs out ofall stomachs that contain a specific category.

59 16 ~ 14 • O. bartramii a. • S. oualaniensis en c 12 "c "(ij 10 c -o o 8 ..cIII 6 ~ E 4 ~ 2 :Q ::J CT III o 15 o~

Cephalopod taxa

Figure 13. The frequency ofoccurrence ofbeaks of different prey, identified to the lowest possible taxa, in the stomachs ofO. bartramii and S. oualaniensis. Frequency ofoccurrence is the percentage ofstomachs out of aU stomachs that contain a specific category.

60 50 -.------, 45 .0. bartramii 40 • S. oualaniensis ~ 35 .c:ll 30 ~ 25 '0 20 ~ 15 10 5 o

Figure 14. The relative abundance of beaks ofdifferent prey, identified to the family level, in the stomachs ofO. bartramii and S. oualaniensis. Relative abundance is the proportion of beaks in that category out ofthe total from all stomachs.

61 35

30 bartramii en .0. ~ I!I S. oualaniensis Q) 25 .0 (ij 20 (5 '0- 15 '#. 10- 5 o

Figure 15. The relative abundance of beaks ofdifferent prey items, identified to the lowest possible taxa, in the stomachs ofO. barlramii and S. oualaniensis. Relative abundance is the proportion of beaks in that category out ofthe total collected from all stomachs.

62 70 • Male I!lFemale 50

Ul E ~ 40 >- ~ .c 30 o Co CIl Ql E 20 ,g .cUl 10 o CIl E o 0 Cii "0 eft

Rsh family

Figure 16. The frequency ofoccurrence ofotoliths ofdifferent families ofprey in the stomachs of both O. bartramii genders. Frequency of occurrence was calculated for the number ofstomachs in each category.

63 OJ .!: r::: 'iii 'E o o Q; "0 16 r::: • Male OJ 14 12 rn Female 10 - 8 6 4 2 o

~tophid species

Figure 17. The frequency of occurrence of otoliths of different myctophid prey in the stomachs of both O. bartramii genders. Frequency of occurrence was calculated for the number ofstomachs in each category.

64 70 • Male 60 - o Female 50

40

30

20

10

0

Fish family

Figure 18. The relative abundance ofotoliths ofdifferent families of prey in the stomachs of both O. bartramii genders. Relative abundance is the proportion ofotoliths of that category out of the total otoliths from the stomachs.

65 12 • Male 10 - IIFemale 8

6 4

2 0 ~~ ·r:i;~ ~ #' ~o ~\ '~r:, ~r:, ~vr:, F:J~ ~~ If~ ~"f o~ ~0 'S'~ ~\r:, #'~ .o(f' .~~ .s.~ .~~o .oJ. o .~o ~0~ 0~ 0~ ~. ~ ~ .~c; ~r:,Q~ ~ ,0~J \V~ ~ 0":> ",0 ~ i~ -<,. tJ<:l ~. <:». '1S ~. V- Cr V- V- ~~ ~o ~~ Q.40 ~ (l fv'¥:tophid species ~.\'f' -(-..

Figure 19. The relative abundance ofotoliths of different myctophid prey in the stomachs ofboth O. bartramii genders. Percent abundance is the proportion ofotoliths ofthat category out of the total in the stomachs

66 18 16 • Male 14 - IIFemale 12 10 8 6 4 2 o

Figure 20. The frequency ofoccurrence of beaks ofdifferent prey, identified to the family level, in the stomachs of both genders ofO. bartramii. Frequency ofoccurrence is the proportion ofstomachs out stomachs in each gender that contain a specific category.

67 Qj "'C c: Q) 01 18 16 • Male E 14 o Female ~ >. 12 ~ Q. 10 01 c: '2 8 'iii "E 6 0 0 4 2 0

Figure 21. The frequency ofoccurrence of beaks of different prey, identified to the lowest possible taxa, in the stomachs of both O. barlramii genders. Frequency ofoccurrence was calculated as a percentage of male and female stomachs that contain a certain category.

68 30 tv1ale '0 25 CD DFemale "0 C Ql Cl 20 - ~ U CIl Ql 15 E ,g VI 10 ~ CIl Ql .0 5 t\1 B '0

Figure 22. The relative abundance of beaks ofdifferent prey, identified to the family level, in the stomachs ofboth genders of O. bartramii. Relative abundance is the proportion ofbeaks in that category out of the total for that category.

69 ... Ql "tJc:: 30 Ql C) .rv1ale 25 - o Female 20 -

15 -

10

5

Cephalopod taxa

Figure 23. The relative abundance of beaks ofdifferent preyt identified to the lowest possible taxat in the stomachs of both O. bartramii genders. Relative abundance was calculated for beaks found in male or female stomachs only.

70 ... CD "t:l I: CD 70 Cl .~ ofemale ~ 60 .S:! .male c: III oS! E 50 - ~~ 0'- (Jj~ 40 - Q. ~Cl ~ .5 CD I: 30 ES o I: 20 -=III 8 ~ as0 10 E 0 iii '0 0 ~ ~0 0 ~?Y o°<:l ~"

Fish family

Figure 24. The frequency ofoccurrence ofotoliths ofdifferent families ofprey in the stomachs of both S. oualaniensis genders. Frequency ofoccurrence was calculated for the number ofstomachs in each category.

71 "- CD "'0 c: CD Cl III 40 'iii c: CD 35 o female '2 III ~ E 30 - .male ~ CD ~.:!:: o >. 25 u:i !!! .r:. a. 20 c.> ~ ~ :5 15 ES e § 10 -III c.> .r:. ~ 5 - E o o iii '0 '#.

~tophid species

Figure 25. The frequency ofoccurrence ofotoliths ofdifferent myctophid prey in the stomachs of both S. oualaniensis genders. Frequency ofoccurrence was calculated for the number ofstomachs in each category.

72 ·le ~ .~ 100 t: .!l! • Female § 90 0 • Male (Jj 80 '0... 70 Q) "Cc: Q) 60 Cl ~ U 50 - ltI Q) E 40 ~ 30 III ~ "0 20 '0 ~ 10 '0 0 ~ 0 Mjctophidae unidentified EVerrmnellidae Gonostorratidae ttlrneidae Fish family

Figure 26. The relative abundance ofotoliths ofdifferent families of prey in the stomachs of both S. oualaniensis genders. Relative abundance is the proportion ofotoliths of that category out ofthe total otoliths in that category.

73 45 (jj "0 c 40 oFemale Q) Cl .s:::. 35 • Male 0 til Q) 30 E ,g 25 til € 20 - (5 '0 15 lU § 10 '0 5 ~ 0 0

~tophid species

Figure 27. The relative abundance ofotoliths of different myctophid prey items in the stomachs of both S. oualaniensis genders. Relative abundance is the proportion ofotoliths of that category out of the total otoliths in that category.

74 20 ,------, 18 -1'--, oFemale 16 • Male 14 12 10 8 6 4 2 o -jL-----

Cephalopod family

Figure 28. The frequency ofoccurrence of beaks ofdifferent prey, identified to the family level, in the stomachs ofboth genders ofS. oualaniensis. Frequency ofoccurrence is the proportion of stomachs out ofall stomachs in each gender that contain a specific category.

75 ~ ~ 14 ,------, .~ ~ .~ 12 c: o Female ~ j 10 • Male ;:) .­ o >- .Q) CI) a. 8 or; Cl g .5 Q).5 6 I ES 8 : 1 i~ 0 --1'1__ '0 Abraliopsis sp. Onychoteuthis Abraliaopsis 61oploteuthis Eluninosa P. addolux #. A sp. sp. Cephalopod taxa

Figure 29. The frequency ofoccurrence ofbeaks ofdifferent prey, identified to the lowest possible taxa, in the stomachs ofboth S. oUillaniensis genders. Frequency ofoccurrence was calculated as a percentage ofmale and female stomachs that contain a specific category.

76 .!a:g .~ ~60,------, ~ o oFemale (Jj 50 • Male '0 Q) 40 "0c: Q) ~30 oco ~ 20 ­ f:? -~ 10- co Q) .c ~ 0-/'---- '0

Cephalopod tam i1y

Figure 30. Relative abundance of beaks ofdifferent prey, identified to the family level, in the stomachs of both genders ofS. oualaniensis. Relative abundance is the proportion of beaks in that category out ofthe total in that category.

77 ... Ql "t:l 5i 40-r------, Cl OFemale .~ 35 ­ ~ • Male .~ 30 ~ ~ 25 o (fj 20 ~ m 15- ~ 10 -~ 5 III Ql .Q o '0 o~

Cephalopod ta>ea

Figure 31. The relative abundance of beaks ofdifferent prey items, identified to the lowest possible taxa, in the stomachs ofboth S. oualaniensis genders. Relative abundance was calculated for beaks found in male and female stomachs.

78 CHAPTER II: STABLE ISOTOPE ANALYSIS (SIA) 80

INTRODUCTION 80 A briefhistory ofSIA. 80 Limitations ofusing SIA 82 Benefits ofSIA 84

MATERIALS AND METHODS 85 Notation. 85

SAMPLE COLLECTION 86 Squids (Ommastrephes bartramii and Sthenoteuthis oualaniensis) 86 Potential prey items of O. bartramii and S. oualaniensis 86 Large fishes 86 Plankton 86 Particulate organic matter (POM) 87

SAMPLE MANIPULATION AND PREPARATION 87 Squids 87 Potential prey items of O. bartramii and S. oualaniensis 89 Large fishes 90 Plankton 90 Particulate organic matter 91 Elemental analyses and mass spectrometry 91

RESULTS 91 Filters 91 Plankton 93 Potential prey items of(O. bartramii and S. oualaniensis) 94 Large fishes 96 Tuna (combined) 96 Thunnus albacares 96 Squids (0. bartramii and S. oualaniensis) 96 O. bartramii 96 S. oualaniensis 99 Species comparisons 103

79 DISCUSSION 104 Filters 104 Plankton 104 Potential prey 104 Largefishes 105 Squids (D. bartramii and S. oualaniensis) 106 O. bartramii 106 S. oualaniensis 108 Species comparisons 110 Isotopes at indicators oftrophic level 114

CONCLUSIONS 114

REFERENCES 117

LIST OF TABLES 125

LIST OF FIGURES 127

Chapter II: Stable isotope analysis (SIA)

Introduction

A briefhistory ofSIA

The term "isotope" was first introduced in 1914 by Frederick Soddy to explain certain ambiguities in atomic research such as the observation that the same elements from different geographical areas had different atomic weights, and that atomic weight did not always increase with atomic number. Isotopes were defined as distinctive atoms of a certain element that had different atomic weights. There are actually three types of isotopes, radioactive isotopes (isotopes which are unstable and decay), radiogenic stable isotopes (those that are stable and formed from radioactive decay), and non-radiogenic

80 stable isotopes (Faure, 1986). Radioactive isotopes have been used to measure long-term processes because oftheir long half-lives and predictable rates of decay. Unlike radiogenic stable isotopes, non-radiogenic stable isotopes (which will henceforth be referred to only as stable isotopes) were all formed during the creation of the Universe. Because these isotopes have all been present since the beginning of time and do not decay, they have been used extensively as tracers of past geological and paleological conditions (Isaacs et at., 1963; MacNeish, 1967; Weiner et at., 1976). Observations that the ratios of the stable isotopes of some compounds varied in natural substances were made fairly early on (Hoering, 1955). The utility of nitrogen stable isotopes in biological and ecological processes became evident when it was 5 determined that the stable isotope ratio of nitrogen ct N /4N) in an animal is affected by its diet (DeNiro and Epstein, 1981). This process. was at first looked upon as a complication to using stable isotopes as tracers of paleological processes (Steele and Daniel, 1978). An enrichment of roughly 3.4%0 was determined for the &15N at each trophic level that the food material passes through (Mingawa and Wada, 1984, Peterson and Fry, 1987). Mingawa and Wada (1984) proposed that this enrichment of nitrogen was caused by the preferential excretion of 14N in urea and ammonia, and the retention of 15N in the tissues. These discoveries have led to the use of stable isotopes in a plethora of studies on food webs and trophic relationships (Hobson and Welch, 1992; Kling et aI., 1992; Hobson et al., 1994), animal migrations (Fry, 1988; Kline et al., 1998; Marra et al., 1998), and nutrient biogeochemical cycling (Hoch et aI., 1996; Karl et aI., 1997). There has been a substantial effort devoted to investigating the variation of isotopic composition at the phytoplanktonlbacteriallevel in the ocean (Hoch et aI., 1992; Waser et al., 1998a, 1998b) Much of the higher trophic level work in the marine realm has been relegated to coastal and nearshore environments (Thomas and Cahoon, 1993; Fantle et al., 1999; Montcreif and Sullivan, 2001). Less work has been done on higher trophic levels in open ocean ecosystems (Rau, 1983; Sholto-Douglas et aI., 1991; Rau et al., 1992; Kline et aI., 1998; Takai et al., 2000). Stable isotopes of nitrogen represent a potentially powerful ecological tool to determine trophic and foodweb relationships in the open ocean, however interpretation of isotopic data can be difficult.

81 Limitations ofusing SIA

One of the most obvious drawbacks ofusing stable isotope analysis to study trophic relationships is that it only reflects the metabolic processes that resulted in the signal of the isotope you are using. Stable isotope analysis does not identify prey types or feeding patterns, which can be critical in understanding material flow in an ecosystem. Because stable isotope signatures are an integrated value of feeding many potential prey of a given predator cannot be ruled out. Using stable isotopes alone there is no way to determine which species preyed upon by a predator and which aren't if potential prey items have similar signatures. To use stable isotopes effectively in evaluating trophic positions the "baseline" signal within the environment must be determined, that is, if every trophic level causes a fixed enrichment in an isotope, one must first determine the starting value of the isotope was in the environment. However, this starting value can be difficult to determine. The cycling of basal resources in an ecosystem (carbon, nitrogen, phosphorus) will affect the isotopic signature of the nutrient pools, in tum affecting the isotopic signature ofthe primary producers that make use of these nutrients (Michener and Schell, 1994). The different metabolic pathways used to fix nitrogen in plants have been shown to effect isotopic fractionation and therefore the baseline Sl5N value for the ecosystem. Fractionation of nitrogen during assimilation depends on the enzymatic processes 3 2 3 involved in assimilating a particular nitrogen source molecule (N0 ., N0 ., NH +, and

N2). Each of these species of nitrogen undergoes different active/diffusive pathways into the cell for assimilation, and then follows different metabolic pathways to become fixed as cellular nitrogen (Fogel and Cifuentes, 1993). These pathways, both physical and enzymatic, can fractionate nitrogen to different extents depending on the ambient concentrations of each nitrogen source molecule (Miyake and Wada, 1971; Wada and Hattori, 1978; Hoch et al., 1992; Waser et al., 1998; and others). A range of fractionation values for S15N has been reported under these conditions varying from -4 to 27 %. (summarized in Fogel and Cifuentes, 1993).

82 Other difficulties arise from processes that occur as material is passed up to higher trophic levels. One problem is that animals have different assimilation efficiencies for different types of compounds (Karasov, 1990) and that some compounds such as certain amino acids can be either synthesized or incorporated directly into the animals tissues. Synthesized material can undergo different fractionation from deposited material (Deniro and Epstein, 1977; Monson and Hayes, 1982). Different chemical compositions within tissues of an organism also can result in different fractionations (DeNiro and Epstein, 1981; Hesslein et ai., 1993; PfeiIer et ai., 1998). Different metabolic pathways of amino acid synthesis can show different amounts of fractionation of material (Macko et ai., 1987). "Isotopic routing" is another phenomenon that can cause difficulties in stable isotopic studies of trophic interactions (Schwarz, 1991). The biochemical components of the diet do not constitute a uniform pool isotopically. Each component within the diet may undergo different routes to different tissue compartments within the predators tissues (Tieszen and Fagre, 1993) therefore, the isotopic value obtained from a certain tissue of the predator may not be directly related to the isotopic value of the bulk diet. A subset of isotopic routing is the phenomenon of protein sparing where an animal with a varying diet will use some portions ofthe diet for tissue synthesis and other portions for metabolism (Castellani and Rea, 1992). This would cause the isotopic value of the predator to be reflected to a greater extent in the high protein portions of the diet rather than the bulk diet itself. The condition of the predator has also been shown to have an effect on the stable isotope signal. Hobson et al. (1993) showed that starving birds showed in increase in

1115N with loss of muscle mass. They hypothesized that the birds catabolized their own tissues for metabolism, becoming progressively enriched through deamination and excretion of lighter nitrogen. A related problem with the use of stable isotopes in ecology is that metabolic turnover rates can effect the stable isotope signal of animals which have changed their diets either through dietary shifts, migration, or seasonal changes (Ben­ David et ai., 1997; Minami and Ogi, 1997; Kline et ai., 1998; Marra et ai., 1998). Different tissues within an organism show different turnover times with which they incorporate a new isotopic signal from the diet (Hobson and Clarke, 1992; Hesslein et al., 1993). Some tissues such as those in the blood and liver have fast turnover times, e.g.

83 they incorporate a new and different signal in the diet quite rapidly (Tiezsen et al., 1983; Hobson and Clarke, 1992) while tissues such as muscle take longer to incorporate a dietary signal (Hobson and Clarke, 1992; Hesslein et at., 1993). Turnover times can also affect the isotopic signal being measured if the signal was incorporated during growth of the organism rather than during tissue maintenance (Frazer et al., 1997). Preparation and storage of samples is another source of potential problems when using SIA. Tissues used in SIA need to be protected from bacterial degradation and the concomitant fractionation that this process would impart to these tissues. Traditional methods of preservation and storage of tissues can effect their isotopic composition. Formalin, mercuric chloride, and di-methyl sulfoxide (DMSO), and certain lysis buffers have all been shown to significantly alter the 15N composition of muscle and blood (Hobson et at., 1997; Bosley and Wainright, 1999). Freezing, drying, storage in 70 % ethanol and acidification did not have any effect on isotopic composition of samples (Hobson et al., 1997; Bosley and Wainright, 1999). Finally, the behavior of the organisms being studied can affect isotopic signatures. Animals with widely varying or specialized feeding habits, both within and between populations of a single species, can affect the outcome of trophic level results using SIA to look at species level comparisons (Vander-Zanden et at., 2000; Beaudoin et at., 1999). SLA used alone without observational or stomach content data would be unable to explain such individual variation.

Benefits ofSIA

Even with all the difficulties involved in using SLA to look at food webs and trophic relationships SLA remains a powerful tool in ecology. This technique can assign an objective value to an animals' trophic position in an ecosystem. Assigning trophic relationships based on stomach contents and/or observations of feeding is difficult, time consuming, has been highly criticized do to the subjective nature of these techniques (Paine, 1988). SLA has shown to be an effective method of assigning trophic position in simple lake ecosystems (Estep and Vigg, 1985; Kling et al., 1992). Because SIA is an integrated signal over a certain time period (depending on the tissue sampled) it provides a view of feeding which is unlikely to be biased by collection methods. The uncertainties

84 involved with fractionation in different tissues and their associated turnover times have been the subject of several studies and the ambiguities are likely to become better understood in the near future (Hobson and Clark, 1992; Hobson and Clark, 1993; Hesslein et aI., 1993). Recently, studies have taken advantage of the fact that isotopic signatures that are laid down in a non-metabolic material can give insight into the trophic history of an organism. A material within an organism that is non-metabolic will not have a turnover time due to maintenance and will only possess an isotopic signature that is effectively locked in place during fonnation and will remain unchanged within that portion of the tissue. Researchers have begun to use this retrospective method to look at historical signals ofmigration and seasonal changes in certain animals such as birds and whales which can lock isotopic signals within their non-metabolic structures such as feathers and baleen respectively (Hobson and Schell, 1998; Best and Schell, 1996; Mizutani et aI., 1992). There have even been attempts to look at long-tenn trophic changes within ecosystems by using the isotopic signals of archived non-metabolic hard parts such as fish scales (Wainright et al., 1993), which opens up the possibility of studying decadal scale or longer ecological processes in history. The amount and complexity of problems associated with SIA should not be viewed as reasons to curb the usage of this technique. These problems should be cited and acknowledged as infonnation that a prudent researcher will take into account when designing and conducting research involving the use of SIA.

Materials and Methods

Notation

The standardized 0 notation was used to report all values for stable isotope ratios. The 0 notation is expressed in the following relationship: oX = [(RsamplJRstandard)-l] x 1000

X is 15N and Rsamp1e is the ratio of the heavy to light isotope 15N/4N in the sample.

Rstandard is theratio of heavy to light isotope or 15N/4N in atmospheric nitrogen.

85 Sample Collection

Squids (Ommastrephes bartramii and Sthenoteuthis oualaniensis) Squids were collected as noted in the section on SCA.

Potential prey items ofO. bartramii and S. oualaniensis Potential prey items were collected as noted in the section on SCA.

Largefishes

Tuna (Thunnus obesus, Thunnus albacares, Thunnus alalunga and Katsuwonus pelamis) along with Mahi Mahi (Coryphaenus hippurus) were caught during the two cruises aboard the Townsend Cromwell in 1999. Tuna were caught using both a longline and rod and reel. Mahi were caught using a rod and reel while chumming the waters at station, and by trolling as the ship was underway.

Plankton

Three plankton tows were conducted during the day at each station, weather permitting. Tows were conducted in seas up to four or five meters before conditions were deemed inappropriate for towing operations. A one m2 ring plankton net with a 333 pm mesh was suspended from a custom modified bridle system used to minimize the effect of equipment on the avoidance of maneuverable plankton such as squid paralarvae. Two tow points were welded to the outside of the one m2 ring on opposite sides. Wire bridles ran from the two points on the ring to a spreader bar about one meter in length. From the spreader bar a pair of short bridles attached to the main towing cable. A pair of five meter lengths ofrope ran from the tow points on the ring to a 20 kg weight. In this system, therefore, the side attachment points of the bridle and weight allowed the net to tow without the attachment mechanism directly in front of the netopening. A plastic cod end was attached to the tail end of the net. Each plankton tow was approximately 25 minutes long. The plankton net was lowered to roughly 40 m (estimated from wire angle and amount ofcable) depth and was towed for 5 min. The net was then raised in 10m

86 increments every 5 minutes until it was just below the surface at which point it towed for another 5 min, and then taken aboard.

Particulate organic matter (POM)

A positive pressure filtration system was designed to filter approximately 12 L of water for each sample. This quantity of water was needed to obtain enough nitrogen for stable isotope analysis. A 64 cu. in. nitrogen cylinder was used as the pressure source. The nitrogen cylinder was connected to three 15 L polycarbonate carboys. Holes were drilled in the carboy caps and a polycarbonate barbed fitting was sealed into each one with silicone sealant. Each carboy was connected in series to its neighbor with tygon tubing while the third carboy was connected to the nitrogen cylinder using a high­ pressure two-stage regulator. A 25 mm inline filter was connected to the nozzle of each carboy and 25 mm Whatmann GFIF's « 0.7 ~ pore size) were used to filter the water. The whole system was pressurized to 15 psi; sufficient pressure to filter the water in a reasonable amount of time (2-3 hrs) while maintaining the integrity of the filters. Surface water was collected by fixing a 5 L bucket to 20 m length of rope that was thrown from the bow of the ship into the water. The bucket was thrown away from the ship as far as possible to avoid possible contaminants from the ship such as oil or particles. Once retrieved the sample of water was visually inspected to make certain that there were no large particles, and the water was used to fill the carboys. When filtration was competed carboys, filter holders, and their respective tubing, were rinsed with aIM Hel solution and then rinsed again with distilled water to insure that growth did not take place within the system. The system was designed to collect a maximum ofthree samples of particulate organic matter per station. Sometimes certain areas of the system developed leaks that prevented the full compliment of samples from being taken at each station.

Sample manipulation andpreparation

Squids

During the cruises from 1998-2001 samples were also taken from captured squids for stable isotope analyses. A total of 143 0. bartramii and 160 S. oualaniensis were

87 sampled. After a squid was caught, weighed, and had its stomach removed, a 2 cm2 segment of muscle was cut away from the anterior mantle, in an area between the gladius and the mantle locking-cartilage, using metal forceps and dissection scissors. Forceps and scissors were cleaned with 100% methanol and rinsed with distilled water between samples. Care was taken to only sample muscle without cartilage or excessive photophore material (the dorsal photophore in S. oualaniensis was specifically avoided). In 1998 the muscle segments were rinsed lightly with distilled water and wrapped in foil that was then placed in a labeled bag. The samples were then frozen at _200 C for later manipulation and analysis in the lab. This procedure was time consuming and resulted in corrosion of the foil around the sample that complicated further manipulations. For the years 1999-2001 the muscle samples were placed in 20 ml glass scintillation vials that were capped, labeled and frozen as before. For analysis the muscle sample was thawed and removed from the vial. A thin (3-4 mm thick) section of the muscle sample was cut away using dissection scissors and the edges were trimmed, leaving a section of muscle which had been unexposed to any outside materials. The remaining sample was placed back into the vial and refrozen. The sectioned muscle sample was then cut into smaller pieces to increase surface area, placed into a clean 20 mI glass scintillation vial and dried at 600 C overnight. These smaller pieces were then ground into a fine powder using a Wig-L-Bugdental amalgamater (Crescent Dental MFG. Co, Model 5AR). The resulting powder was then weighed out to between 300 and 1000 flg on a microbalance and placed into foil boats that were crushed into cubes to seal the material inside. This weight range was calculated so that sufficient N was obtained considering the C:N ratios of the samples used. In addition to muscle, various other tissues were also sampled from select squids in order to compare the isotopic fractionation and turnover times between tissues and to investigate possible ontogenetic signals in secreted hard structures. The eyes of many of the squids were removed and frozen for later analysis. In the lab select eyes were thawed and the lenses removed and any residual tissue or fluid was cleaned. The distal lens was removed and discarded leaving the larger proximal lens. Using micro-forceps, separate distinct layers or "shells" of a lens were separated from each other. Starting from the innermost shell (smallest) of a lens, the shells were sequentially separated, ending with

88 the outermost (largest). The shells were placed in foil lined plastic scintillation caps and the caps were labeled. After separation the thickness of each shell was measured using an ocular micrometer fixed in a dissection microscope. Each shell was then dried, ground, and weighed as stated above. During growth the eye lens must be constantly re-adjusted to maintain the proper gradient of refractive index. This presumably occurs by shrinkage from loss of water. Determination of an accurate correction factor for this effect was not possible. The shrinkage effect would be a slight overestimate of ML in the retrospective values. Samples of blood were also taken from large squids. Blood samples were taken before all other measurements in order to minimize the amount of blood lost to bleeding from the incision into the mantle. Blood was collected from two areas in the squids, the efferent branchial vessels coming from the gills, and the cephalic vein before it divides into the vena cavae. Ten cc syringes fixed with 18 gauge needles were used to collect the blood. The syringes were rinsed with 100% methanol and then twice with distilled water before sampling a new animal. Squids of around 200mm ML or larger were usually required to obtain enough blood, however squids of smaller sizes (180-190mm ML) would occasionally yield a sufficient quantity. Blood sample sizes ranged from less than one m] to over 20 ml from the largest squids caught. Blood samples were dried and ground as above.

Potential prey items ofO. bartramii and S. oualaniensis

Fishes caught by trawl during cruises in 1999-2001 were used for SIA and to determine otolith size versus fish weight relationships. Only species that could be identified were used for these analyses. Identified fish were measured for their standard lengths (measurement starts at foremost point on fish and ends at the fork of the tail). The fish were also weighed, their otoliths were removed and measured using an ocular micrometer and dissection microscope, and a muscle sample was excised. When taking the muscle sample, the skin and scales were removed and a fillet of meat was taken from the dorsal musculature near the dorsal fin, care was taken to avoid bone tissue. The muscle sample was then dried, ground and weighed as above.

89 Large fishes

Muscle samples were taken from all tuna (Thunnus obesus, T. albacares, T. alalunga, and Katsuwonus pelamis, ) and mahi (Coryphaenus hippurus) caught which were not being utilized for other research. Samples of white muscle were excised from just behind the pectoral fin of the tuna and mahi mahi. AII samples were placed in labeled 20 ml glass scintillation vials and frozen at _20° C. In the laboratory samples of white muscle ofboth species were thawed, trimmed, dried, ground and weighed following the same protocol as previous muscle samples.

Plankton

Plankton samples were kept in 10 L buckets of seawater chilled with bags of ice to keep them from overheating during sorting. Samples were examined for paralarvae of squids and any paralarvae were identified. O. bartramii and S. oualaniensis paralarvae were placed in 20 ml glass scintillation vials, labeled, and frozen at _20° C for later analysis. The remaining plankton sample was filtered through a 333 urn Nitex mesh and the resulting bolus of material was placed in a labeled plastic ziplock® bag and frozen at­ 20°C. In the laboratory paralarval mantle lengths of O. bartramii and S. oualaniensis were measured using an ocular micrometer on a dissection microscope. Because of the small size of the paralarvae the specimens were not subjected to the grinding process and whole animals were dried and weighed for isotopic analysis as noted above. The mantle and head/viscera were separated from a single large O. bartramii paralarva. The plankton samples were thawed and a "representative" sub-sample was taken from the mass of zooplankton. A small amount of the whole bolus ofplankton was taken, dried, ground and weighed as above. Typically, no attempt was made to categorize the zooplankton or to estimate the contribution of one group or another to the sample as a whole. From one tow in 2000 four categories of organisms were separated, euphausids, hyperiid amphipods, chaetognaths, and radiolaria. Multiple individuals of each group were combined and treated as above.

90 Particulate organic matter

The filters were dried in 20 ml scintillation vials as above. The filters were then placed in flattened foil boats and rolled and then placed into a small glass tube. The foil and filter were then tamped down on both ends by small glass rods that fit inside the tube, this resulted in a compact foil covered filter which was then analyzed.

Elemental analyses and mass spectrometry

All samples which were investigated for the isotopes 15N and l3C were analyzed using the same system located in the Isotope Biogeochemistry Laboratory of the Marine Geology & Geophysics Division of the University of Hawaii at Manoa. This system consisted of a ThennoquestlCE Instruments Automated Elemental Analyzer (model 1110 NC 2500) interfaced to a Finnigan Mat Delta-S stable isotope ratio mass spectrometer via a Finnigan MAT ConFlo n Elemental Analyzer (model 1110 NC 2500) using a Finnigan MAT ConFlo n (Continuos Flow) interface. The ConFlo n allowed for fully automated dual element analysis (C and N) from each combustion including determination of C and N elemental concentrations. Ciclosanone was used as a reference standard and a total of 229 analyses of the standard were perfonned. Standards were analyzed at the beginning of each sampling session, and internally (for every 5-7 samples a standard was analyzed). For 015N the mean and standard deviation were -4.8%0 and 0.2%0. The minimum and maximum 015N values obtained for ciclosanone in this study were -4.0 and -5.4%0.

Results

Filters

91 The Ol5N values for filtered POM samples collected between 1999 and

2001 ranged from -0.7 to 6.8%0. The mean Ol5N values for filtered material varied by up to 3.04%0 between years (Table 2-1). There was significant difference (one-way

ANOVA, p < 0.001) between the Ol5N of POM collected during 1999 and other years. A two sample t-test showed that the Ol5N ofPOM collected during the 1999 Hokusei Maru cruise was not significantly different from the Ol5N of POM collected during the cruises aboard the Townsend Cromwell (p = 0.057). There were no clear trends in ol5N with latitude or longitude, although variation between years was evident (Figs. 1, 2).

92 Table 1. The mean and standard deviations of6lSN valnes for all POM collected on separate ships dnring different years. HM= Hoknsei Maru, TC= Townsend Cromwell, 1999 total is the combined valnes for the HM and TC cruises. The Townsend Cromwell cruises took place in November and January; all other samples were taken during February.

Year (Ship) Mean /}"N (SD) Min-max # of filters 1999 (TC) 2.0 (0.8) 0.5-4.0 14 1999 (HM) 1.0 (1.3) -0.7-3.0 8 1999 (total) 1.7 (1.1) -0.7-4.0 22 2000 (HM) 3.5 (1.5) 0.8-5.7 16 2001 (HM) 4.0 (1.9) 2.1-6.8 13 All years combined 2.8 (1.6) -0.7-6.8 51

Plankton

The /}lsN values for plankton samples collected during 1999 and 2000 ranged from 0.1 to 6.5%0. The mean /}lsN values for bulk plankton material varied by 0.7%0 between 1999 and 2000 (Table 2-2). The /}lsNvalues for the bulk plankton did not show significant differences between 1999 and 2000 using two sample t-tests at p '" 0.294.

93 Table 2. The mean (i"N values for plankton collected during 1999 and 2000 aboard the Hokusei Maru. The plankton category for 2000 includes the various subgroups listed separately. The number ofsamples in the categories with specific taxa represent number ofcombined individuals.

Year (category) Mean ll"N (SD) Min-max # of samples/individuals 1999 (bulk) 3.9 (1.1) 2.3-6.4 10 2000 (bulk) 4.6 (1.8) 0.1-6.5 15 2000 (Euphausidaceae) 5.4(-) (-) 5 2000 (Hyperiidae) 5.1 (-) (-) 5 2000 (Radiolaria) 0.1 (-) (-) 10 2000 (Chaetognatha) 5.6 (-) (-) 5

Potential prey items of(O. bartramii and S. oualaniensis)

SIA was conducted on 113 individuals comprising 7 families of fishes and 2 families of squids that are known or potential prey items for O. bartramii and/or S. oulaniensis. Thirteen species of myctophids were analyzed.

94 Table 3. Tbe mean lj15N values for different categories of known and potential prey items ofO. hartramii and/or S. oualaniensis. Standard deviations are listed in parentbeses. All species of myctophid listed are included in tbe category Myctophidae. Heavily oatlined cells represent myctophid species.

Category o"N (SD) Min-max # ofindividuals Myctophidae 8.4 (1.8) 5.1-13.1 81 Gonostomatidae ILl (1.7) 8.5-12.6 11 Melamphaeidae 12.6 (1.2) 11.2-14.1 6 Stemoptychidae 9.5 (1.0) 7.8-10.8 7 Malacosteidae 9.9 (0.7) 9.5-10.4 2 Astronesthidae 10.6 (-) (-) 1 Orrwsudis onwsudis 10.0 (-) (-) 1 Abraliopsis sp. 8.3 (0.7) 7.8-8.8 2 Histioteuthis hoylei 11.0 (-) (-) 1 Abralia trigonura 10.1 (-) (-) 1 Bolinichthys longipes 7.3 (0.7) 6.6-8.4 5 Ceratoscopelus 8.9 (2.2) 6.5-13.1 27 warmingii Diaphus trachops 8.6 (0.1) 8.5-8.7 2 Electrona risso 10.3 (-) (-) 1 Hygophum proximum 7.5 (0.8) 6.3-9.4 13 Hygophum rheinhordti 7.5 (0.7) 6.9-8.4 4 Lampadena luminosa 9.6 (0.5) 9.3-10.0 2 Lobianchia gemellerii 8.3 (Ll) 6.4-10.2 13 Lampanyctus nobilis 7.0 (-) (-) 1 Lampanyctus 10.6 (1.4) 9.0-12.1 5 tenu~formis Myctophum lychnobium 6.7 (1.3) 5.2-8.2 4 Myctophum selenoides 10.8 (-) (-) 1 Symbolophorus 7.1 (0.6) 6.4-7.5 3 evermanni

Individuals of Stemoptchyidae, Gonostomatidae, L. tenuiformis, M. lychnobium, and Melamphaeidae (minus one outlier) all showed linear increases in 015N with length (fork length) (Fig. 3). The remaining groups of myctophids did not show clear increases in 015N with length (Fig. 4).

95 Largefishes

SIA was conducted on 53 individuals from two families of fishes comprising five species of large epipelagic fishes (Table 2-4).

Table 4. The mean 01~ values for different categories ofepipelagic fishes, standard deviations are listed in parentheses. The tona category is comprised of the four species listed in the heavily outlined cells.

Fish Category /)ION (SD) Min-max # of individuals Coryphaena hippurus 9.2 (1.2)· 7.9-12.2 23 Tuna (combined) 10.6 (1.2) 8.2-13.3 30

Thunnus obesus 10.6 (1.0) 8.7-13.3 21

Thunnus albacares 10.0 (1.6) 8.2-12.7 7 Thunnus alalunga 11.8 (-) (-) 1

Katsuwonus pelamis 12.4 (-) (-) 1

The values of /)15N were significantly different (two sample t-test, p < 0.001) between the combined tuna values and mahi mahi (Coryphaena hippurus). There were no significant differences in between species comparisons of the mean values of /)15N for Thunnus obesus and Thunnus albacares (two sample t-test, p = 0.391). There was no relationship of /)15N with fork length for all tuna combined or for the mahi mahi (Fig. 5).

Squids (0. bartramii and S. oualaniensis)

O. bartramii

SIA was conducted on the mantle muscle of 174 separate individuals of O. bartramii that were divided into five categories based on mantle length (Table 2-5). The 1-7 mm category is composed of 27 paralarvae where the whole animals were analyzed.

96 The mantle and head/viscera were separated for one paralarvae (ML = 5mm) and their results were comparable, mantle: 015N= 5.23%0; head/viscera: 015N= 4.92%0.

Table S. The mean /j15N values of mantle muscle for different categories of O. bartramii, standard deviations are listed in parentheses. Whole animals were analyzed in the category of 0-7 mm ML.

Mantle length (mm) Mean o"N (SD) Min-max # of individuals 1-7 6.4 (1.2) 4.0-8.5 27 75-200 6.9 (0.7) 6.0-8.0 8 200-300 11.1 (1.5) 9.7-14.4 40 300-400 13.3 (1.0) 10.4-15.4 107 400-570 12.8 (1.4) 10.0-16.8 19

The 30 squid taken from northern latitudes exhibited a narrow range of 015N values (std dev =0.3) over a mantle length range of 193 to 292 mm. The 015N values of mantle muscle tissue showed a sigmoidal relationship (4-parameter logistic curve, ~ = 0.8) with mantle length for all O. bartramii samples (Fig. 7). No statistically significant differences were found between Ol5N values and latitude or longitude of collection stations within individual years for mantle muscle tissue taken from male O. bartramii with mantle lengths of> 300 mm (one-way ANOVA, 1998- P= 0.206, 1999- P = 0.057, 2001- P = 0.82). Between-year samples, however, did show some differences. The &15N values were significantly higher for mantle muscle taken from males> 300 mm ML captured in 1998 and 2000 (n = 44, !! = 14.0%0, SD = 0.75) than for males> 300 mm ML captured in 1999 and 2001 (n = 62, !! = 12.8%0, SD = 0.85) (2-sample t-test, p < 0.001). There were no significant differences in male mantle muscle 015N values taken from males> 300 mm ML between years 1998 and 2000 or 1999 and 2001 (2-sample t-test, p =0.980 and p =0.315 respectively). There was no significant difference between the mantle muscle 015N values of male and female 0. bartramii captured in 1999 (2-sample t-test, p = 0.178) or 2001 (2-sample t-test, p = 0.46). Only one female sample was obtained during 1998 and another in 2000. SIA was conducted on 181 segments of eye lenses from 14 different O. bartramii individuals (six males, five females and three juveniles). The 015N values of eye lens

97 samples ranged from a minimum of 3.6%0 to a maximum of 16.8%0 for all individuals. The lilsN values of eye lens samples showed a sigmoidal relationship (4-parameter logistic curve) with estimated mantle length (based on eye lens radius) for each male and female sample, and for all samples combined (Table 2-5, Fig. 8). The r for the logistic curve fitted to the li 15N of the eye lens segments versus estimated mantle length from all individuals combined was 0.8. However, the mean r2 for each specific eye lens segment curves was 0.9 indicating considerable difference between individuals.

Table 6. The r' values of the sigmoidal relationship for &15N of eye lens segments and estimated mantle length (hased on eye lens radins), and nnmber ofsamples for different O. bartramii individuals.

Sex, date caught, r"-value Min-max # of samples (eye ML segments) Male, 2-819-00, 296 0.99 4.8-10.9 12 mm Male, 2-8/9-00, 318 0.99 4.6-14.5 11 mm Male, 2-7/8-00, 333 0.99 4.2-12.7 8 mm Male, 2-8/9-00, 354 0.98 6.5-11.7 9 mm Male, 2-9/10-00, 360 0.88 6.1-12.6 14 mm Male, 2-11112-01, 370 0.98 5.1-13.7 14 mm Female, 2-16/17-99, 0.97 3.7-16.7 20 446mm Female, 2-11112-01, 0.94 4.2-15.8 21 534mm Female, 2-9/10-00, 0.97 4.2-14.5 19 540mm Female, 2-11/12-01, 0.81 7.4-10.3 20 557mm Female, 2-11/12-01, 0.98 3.9-16.8 17 570mm

The li 15N values of eye lens segments for the three juvenile O. bartramii did not show a sigmoidal relationship with estimated mantle length, but a nearly straight and horizontal line. This shape is similar to the corresponding portion of the lilsN versus

98 mantle length curve for mantle muscle and therefore supports the initial shape of the sigmoidal curve (Fig. 9). SIA was conducted on 61 samples of O. bartramii blood each from a separate individual (Fig. 10). The /)t5N values ranged from 9.7 to 14.7%0 with a mean value of 12.2%0, SD = 0.9. A separate logistic curve could not be fit to the blood samples because the smaller animals caught did not have enough blood for analysis. There was a significant difference between the mean /)t5N values of blood taken from individuals in 1998 and 2000 (n =10, ~ =13.4%0, SD =0.6) and 1999 and 2001 (n =51, ~ =12.0%0, SD =0.8) (2-way t-test, p < 0.001).

S. oualaniensis

SIA was conducted on the 167 separate S. oualaniensis individuals (153 subadults/adults and 14 paralarvae). The mean /)15N value for sub-adult and adult mantle muscle (ML range, 128 to 324 mm) was 8.2 %0, SD = 1.12 with a range of6.2-11.2%0. The mean /)t5N value for paralarvae (whole body sample) was 6.2%0, SD =1.44 with a range of 4.2 to 8.0%0. The /) 15N values for all S. oualaniensis mantle muscle samples showed an exponential increase (single exponent, 3-parameter) with mantle length (Fig. 11, r2 = 0.56). Samples taken in 1999-2001 analyzed separately showed a stronger exponential relationship between /)15N and mantle length, and samples in 1998 showed a weaker relationship (Table 2-6.).

Table 7. The'" values, and number of samples of (mantle muscle) from all sizes(whole body) taken for each year. The'" values are for single exponent 3-parameter exponential relationships between mantle length and I;t'N.

Year r' Min-max # of samples 1998 0.38 6.1-9.2 32 1999 0.75 4.2-11.2 50 2000 0.82 4.2-9.0 38 2001 0.79 5.4-10.6 47

99 A subgroup of samples from different years showed high 1)15N values for the given mantle lengths. Station 2-12113-00 (2 l.l7°N, 152.19°W) had two (out of six caught at that station) female samples as well as three paralarvae (all from that station) that were isotopically heavy for their size (Fig. 11). Samples taken from a nearby station the following year 2-10/11-01 (21.20oN, 152.300 W) did not show any substantially higher 1)15N values from average. Squids from two northwestern stations, 2-9/10-99 (28.59°N, 164.59°W) and 2-14/15-99 (28.28°N, 163. 12°W) also showed an isotopically heavy mean 1)15N for their size (Fig. 12). There were statistically significant increases in adult mantle length and 1)15N of muscle for S. oualaniensis caught during successive years (one-way ANOVA, p < 0.001 for ML and 1)15N against year) (Table. 2-7).

Table 8. The mean and standard deviation of mantIe length values (mm) of S. oualaniensis individuals used for SIA during years 1998·2001. 'The mean lllSN values for 1999 were calculated without the two isotopically heavy stations, n=32.

Year # of individuals Mean ML (SD) Mean 1)"N%o (SD) 1998 32 168 (17) 7.1 (1.3) 1999 46 187 (41) 7.7 (0.9)' 2000 44 204 (32) 8.0 (1.2) 2001 43 234 (56) 8.5 (1.2)

Because of the relationship of 1)15N with mantle length, residuals of the S15N data against mantle length (calculated using the exponential relationship noted above) were used to explore geographical and temporal relationships of the S15N values for S. oualaniensis mantle muscle. There were no statistical differences in the residuals between years when the two northwestern stations in 1999 were excluded (one-way ANOVA, p =0.242) (Table. 2-8). There were also no statistical differences in residuals of all years combined against binned latitude (one-way ANOVA, p = 0.247) or binned longitude (one-way ANOVA, p =0.142) (Table 2-8).

100 Table 9. The mean and standard deviation of residual valnes of l)lSN(mantie muscle) for S. ouafaniensis mantle samples for year, latitude, and longitude. *The mean residual values for 1999 were calculated witbout the two isotopically heavy stations.

YearlLatitude/Longitude number of samples mean of residuals (SD) 1998 32 -0.351 (0.717) 1999* 32 -0.004 (0.624) 2000 44 -0.108 (0.869) 2001 43 -0.189 (0.547) 18-20oN 17 0.061 (0.636) 20-22 Q N 80 -0.283 (0.691) 22-24°N 16 -0.077 (0.629) 25-28°N 28 -0.033 (0.851) 28-30oN 10 -0.047 (0.545) 152-156°W 79 -0.129 (0.672) 156-158°W 20 -0.326 (0.714) 158-161°W 31 -0.309 (0.831) 161-165°W 21 0.099 (0.580)

There was a statistical difference (two-sample t-test, p

101 segments and estimated mantle length (Fig. 12). The exponential curves for eye lens segments in the typical and atypical classes were different in curvature and elevation from the combined mantle muscle 015N versus mantle length curve.

Table 10. The number ofsamples, minimum and maximum ,,1'N values of eye lens segments for S. oualaniensis individuals from the three different classes and the r' values for their exponential relationship with estimated mantle length (based on eye lens radius).

Sex, date caught, ML class Min-max o"N (#) r" Male, 2-12/13-00, 158 nun typical 3.7-5.3 (9) 0.92 Female, 2-14115-99, 200 nun typical 4.2-9.0 (11) 0.93 Female, 2-12/13-00, 209 mm typical 4.3-6.7 (13) 0.94 Female, 2-7/8-00, 210 mm typical 3.6-5.8 (10) 0.78 Female, 2·14/15-99, 223 mm typical 5.4-9.8 (12) 0.96 Female, 2-13/14-00, 251 nun typical 4.2-6.2 (12) 0.88 Female, 2-12/13-01, 255 mm typical 2.8-9.4 (12) 0.97 Female, 2-12/13-01, 302 nun typical 3.8-8,8 (II) 0.96 Female, 2-12/13-01, 304 nun typical 3.3-9.6 (10) 0.98 Female, 2-12/13-01, 308 nun typical 3.1-9.8 (12) 0.96 Female, 2-14/15-99, 324 mm typical 4.2-10.0 (10) 0.96 Female, 2-12/13-00,160 nun atypical 4.2-8,9 (11) 0.95 Female, 2-12/13-00,173 nun atypical 4.6-8.5 (10) 0.94 Female, 2-12/13-01, 181 mm anomalous 11.3-12.6 (10) N/A

SIA was conducted on 21 blood samples, each from a separate S. oualaniensis individual, only two of these individuals were under 200 mm ML. The 015N values from the blood samples had a mean of 8.0%0, SD= 1.2 and ranged from 6.4 to 10.1%0 (11.3%0 for the anomalous individual). The blood samples from 2000 and 2001 were significantly different (two-sample t-test, p=0.025) with blood from 2001 being heavier (2000: n = 14, 1l015N = 7.8%0, SD = 1.1; 2001: n = 6, 1l015N = 8.7%0, SD = 0.736). There was an exponential increase in blood 015N with mantle length (r= 0.70, Fig. 13) that differed only slightly from the mantle muscle Ol5N versus ML curve.

102 Species comparisons

The mean /)15N values for all animals sampled increased with certain taxa (Fig.

14). The /)15N values for mantle muscle of S. oualaniensis and 0. bartramii were significantly different (two-sample t-test, p < 0.001) with O. bartramii being 4.5%0 greater on average than S. oualaniensis. Unless otherwise stated all comparisons are based on adult and sub-adults squid. The /)15N value (mean of means) from the muscle of the top five most abundant fish prey items found in O. bartramii stomachs was 9.7%0, which is 3.55%0 less than the mean value for all muscle taken from O. bartramii. The

/)15N value (mean of means) of the top five most abundant fish prey items found in S. oualaniensis was 7.5%0 which is 0.6%0 less than the mean muscle value for all S. oualaniensis.

The /)15N values for blood of S. oualaniensis and O. bartramii were significantly different (two-sample t-test, p < 0.001) with O. bartramii being 4.2%0 greater on average than S. oualaniensis. O. bartramii and S. oualaniensis each had significantly higher /)15N values for the mantle than the blood. In O. bartramii the mantle muscle was 0.6%0 heavier than the blood (paired t-test, n =21, P < 0.001) and in S. oualaniensis the mantle was 0.59%0 heavier than the blood (paired t-test, n = 20, P < 0.001) (Figs. 15, 16).

The /)15N values, within single individuals, for outer eye lens segments of O. bartramii were significantly different from mantle muscle values (n = 13, P < 0.001). For O. bartramii there a mean 1.0%0 difference between mantle and outer eye lens segments.

The /)15N values, within single individuals, for the outer eye lens segments in S. oualaniensis were significantly different from the mantle muscle values (paired T-tests, n =14, P =0.004). For S. oualaniensis there was a mean difference of 1.1%0 between mantle and outer eye segments.

103 Discussion

Filters

The /)15N values collected for the POM are within the range of other published values for PON in the North Pacific (Wada et al., 1975; Wada and Hattori, 1976; Wada and Hattori, 1991; Karl et al., 1997). The mean of all years (2.8%0) is similar to annual means (3.1%0) for /)15N of sinking particles reported by Karl et al. (1997) in the same waters. The /) 15N values of POM collected in 1999 were significantly lighter from other years indicating a possible increased contribution of N2 fixation during that year. The total range of values was 7.6%0, but because of the wide range of processes that can contribute to /) 15N signatures of POM (see Michener and Schell, 1994) this range does not necessarily translate into differences in trophic levels. The /)15N values reported here are lighter than those reported for bulk phytoplankton in the North Pacific Ocean reported by Miyake and Wada (1967) which had a mean of 5.6%0. The samples of filtered material were designed to represent an approximate baseline signal for the trophic calculations in this ecosystem and were assigned a trophic level of one.

Plankton

The mean /)15N value of bulk plankton was 1.1%0 heavier than the mean value for POM; much less than the 3.4%0 indicative of a trophic level enrichment. Because of the complexity of the planktonic assembledge, bulk plankton values may not be a wholly applicable value. This variability is evident in the values obtained for separate groups of plankton. Values for colonial radiolaria were among the lightest found in this study. The combined value for euphausids, chaetognaths, and hyperiids (5.4%0) agrees with other published results (Kaehler et al., 2000) and was 2.6%0 greater than the estimated trophic level one value, however these organisms cannot all be considered herbivores.

Potential prey

104 In general, the myctophid categories were lighter than the other mesopelagic categories. The myctophid fishes had a mean I)15N value that was 4.5%0 heavier than the 15 mean plankton value, and 3.0%0 heavier than the combined 1) N values for euphausids, chaetognaths, and hyperiids. The range within the myctophids was 8.0%0, and much of this variation was due to C. warmingii, L. tenuiformis and L. gemellerii indicating substantial variation between related species and between individuals within a species. Three species of myctophids exhibited increases in (\ 15N with length that could account for a certain amount of variation in some species such as L. tenuiformis. Very little published information exists on 1)15N values of myctophids, Kline (1997) lists "Northern lanternfish" as having values of approximately 13.5%0 with a variation of about 0.5%0 and an assigned trophic level of close to 4. Gould et at. (1997) found "Ianternfish" in the transition zone of the North Pacific to have 1)15N values of 10.1 to 11.2%0, but did not assign a trophic level. In this study the mean value for myctophids indicate a trophic position of 2.6. Clarke (1972) estimated that the Myctophidae occupy the third trophic level based on his biomass estimations. Other mesopelagic fishes (gonostomatids, melamphaeids, and sternoptychids) exhibited stronger relationships of 1)15N with length for muscle tissue. The 1)15N values of the gonostomatids and melamphaeids were 2.7 and 4.2%0 greater than the average myctophid value and indicated trophic levels of 3.4 and 3.8 respectively. Gonostoma spp. have been shown to feed on euphausids in the Gulf of Mexico (Lancraft et al., 1988) however the 1)15N values of the gonostomatids (a mixture of Gonostoma species) measured in this study were much higher than one would anticipate if they had substantial feeding on euphausids.

Large fishes

15 The tuna and mahi mahi showed no increase of 1) N with size and the data were highly scattered. The difference in diving profiles of bigeye and yellowfin, and the greater abundance. of mesopelagic prey species in the stomachs of deeper diving bigeye tunas (Tsuchiya, 1998), indicates fundamental foraging differences between these species. These foraging differences did not manifest as different stable isotopic signatures obtained for these two species.

105 Squids (0. bartramii andS. oualaniensis) O. bartramii

The overall l) lsN values for O. bartramii are in general agreement with other published results. Takai et al. (2000) report l)lsN values ranging from 11.8 to12. 1%0 for five female O. bartramii with mean mantle lengths of 360 mm (SD = 19 mm) captured off Japan. Gould et al. (1997) report l)lsN values of between 10.0-15.1%0 for 44 O. bartramii (170-446 mm) taken from albatross stomachs in the North Pacific transition zone (35-46°N, 170oE-148°W). The most distinctive feature of O. bartramii, besides the absolute values, was the increase in l)lSN (mantle muscle and eye segments) with size and the sigmoidal shape of the curve. Body size is one of the more important biological factors in structuring aquatic food webs and directing the flow of energy from the smallest organisms to the largest (Borgmann, 1987; Peters et aI., 1996; France et al., 1998). Ecosystem wide trophic enrichments of l)lsN with increasing size across taxa have been documented in several marine ecosystems (Sholto-Douglas et al., 1991; France et al., 1998; Fry and Quinones, 1994; Kline, 1997). Sholto-Douglas et al. (1991), however, found that l)ISN values for anchovy (Engraulis capensis) and round herring (Etrumeus whiteheadi) decreased with increasing fish length. Two hypotheses were put forward for this reverse trend, a physiological basis (i.e., changes in growth or assimilation effeciencies), or a trophic basis, (i.e., reflecting a shift from zoophagy to phytophagy). The range of l)ISN in the lengths of fishes they tested was only 1.2%0, much less than was observed in this study. Kline et al. (1998) observed species-specific increases in l)lSN with increasing fish length ('" 60-480 rom) for four species of salmonids in waters off Northern Alaska. The l)ISN versus length data for all four species were fitted with second-order polynomial regressions that detailed rapid increase in l)lsN with length followed by a flattening of the curve and a plateau at the greatest fish lengths. The greatest range of 1l 1sN values in one species was roughly 10%0, comparable to what was found over the whole size range of O. bartramii in this study. In a separate study Kline (1997) observed a sigmoidal relationship between l)lsN and. length for combined nekton classes (fishes and squids) in Prince William Sound that ranged from 2 to 4 trophic levels ('" 8-16%0). This range

106 coincides closely with the overall range of O. bartramii as it makes its transition from paralarvae ('" 7%0, Trophic level (TL) '" 2) to adult ('" 13%0, TL '" 4).

The I)lsN mantle data for O. bartramii shows an initial plateau followed by a rapid increase in trophic position followed by a trophic plateau beginning around 250-300 mm mantle length. The latter plateau corresponds to a trophic position of4.1. Size appears to be a major factor in determining TL in this species. However, the logistic relationship seen in O. bartramii mantle data (rather than an exponential relationship seen in S. oualaniensis) indicates that the causes are more complex. The shapes of the curves will be discussed in more detail under the Species Comparisons section. Mean mantle length of O. bartrtnaii males increased in every year sampled for 1)15N but the difference was not always significant. There was no significant difference between mean mantle lengths in 1998-2000 (one-way ANOVA, p =0.055; mean ML: 1998 = 318mm, 1999 = 321 mm, 2000 = 329 mm). There was a significant difference in mean mantle length of O. bartramii males caught in 2001 (mean ML = 366 mm) versus other years (one-way ANOVA, p < 0.001). Overall O. bartramii did not show a change in 1)15N with size above 300 mm ML using mantle muscle, however overall differences in mean mantle length could indicate differential yearly growth regimes. The differences in mean 1)15N of mantle muscle samples for O. bartramii did not, however, coincide with the mean mantle length differences between years as 1)15N for 2001 was more similar to 1999 while 1998 was more similar to 2000.

Aside from mantle length, the main source of variation in I)ISN values of mantle muscle was variation between years, in spite of the fact that within some years samples were taken over a wide geographical range. Uncertainty exists as to the nature of the observed yearly differences in O. bartramii. The 1999 paM was significantly lighter than 2000 and 2001, and the mantle values for 1999 O. bartramii males were lighter than those sampled in 2000 but not different from mantles sampled in 2001. Thus, heavier mantle values in 2000 correspond with heavier paM in 2000 but the trend is opposite in 2001. The results, therefore, are equivocal as to whether or not yearly changes in mantle values represent a yearly change in the baseline trophic signal of the environment. Changes in food chain length may occur with differences in species assemblages above the baseline signal that could account for yearly variation.

107 The retrospective <')15N values of the eye lenses reveal an almost identical logistic relationship to estimated mantle length as seen in the <')15 N values of muscle versus mantle length. The fit of the logistic curve, however, was better for each individual than for all individuals combined, indicating that variation between individuals has a greater effect on this relationship than variation within individuals. Individual curves are considered in the next chapter. The interpretation of the sigmoidal curve is discussed under Species Comparisons. While offset somewhat from mantle values the blood samples followed the mantle patterns during these years. Blood has been found to have a fast turnover rate in birds when incorporating isotopic signals (1i 13C) (Hobson and Clarke, 1992, 1993). Within the blood Hobson and Clarke (1992, 1993) found that the half-life the blood plasma ofcrows was 2.9 days. Whole cellular fraction had a tenfold longer turnover time than plasma (29 days). Muscle tissue in avians was found to have a much longer half-life than blood plasma, and Hesslein et al. (1993) found that it took over 100 days for a new <')15 N signal to be wholly incorporated after a switch in diet in broad whitefish. While cephalopod blood is very different (it does not contain cells for oxygen transport) the blood signal could be incorporated in a relatively short time frame relative to the muscle. If so, the consistent relationship in <') 15N between blood and muscle indicates that either there is little difference in isotopic signal in waters that O. bartramii passes through on its southerly migration, or that at the time of capture the O. bartramii individuals had been residing in local waters long enough for the muscle tissue to incorporate a local signal.

s. oualaniensis

15 The <') N values reported here are generally in agreement with published values.

Takai et al. (2000) reported mean <')15 N of 10.0%0 (5 individuals; mean ML '" 217mm) for S. oualaniensis caught off Japan (26.30o N, 144.000 E). This value is higher than the mean value obtained for adults and sub-adults for this study, but well within the range of variation for adult and sub-adult Hawaiian squids. Squid from the same family (Todarodes pacificus) with a similar size range (207 mm, SD '" 7) taken from the Japan

Sea also had Ol5N values (10.5%0) comparable to S. oualaniensis of that size. Takai et al.

108 (2000) found large regional differences in 615N values for S. oualaniensis throughout both the Pacific and other oceans. The greatest 615N values Takai et al. found for S. oulananiensis occurred off ofPeru (14.00oS, 85.000 W) and were exceptionally high (16.3 %0), a fact they attributed to denitrification of upwelled nitrate in the area (Cline and Kaplan, 1975). The exponential increase in 615N with ML is different from O. bartramii and will be discussed under Species Comparisons. There were no obvious differences in 615N with geography in this study, with the exception of the two stations in 1999 that showed enrichment of 615N relative to all other stations. Two squids from a single station in 2000 also showed high 1)15N values but nothing unusual was found in the stomachs of these squids. A distinction between geographically induced differences and intrinsic individual variation is difficult to determine without a better understanding of individual feeding history. Individual variation due to feeding specialization has been shown to cause I)15N to range by >6%0 in northern pike within the same lake (Beaudoin et al., 1999). The specialization found by Beaudoin et aI. was seen not only in the isotopic values but also in feeding determined from stomach contents. The increase of 615N with size for mantle niuscle samples in S. oualaniensis did not plateau but increased exponentially over the entire size range of squid caught during this study. This indicates that S. oualaniensis does not reach a trophic plateau over the size ranges sampled, as occurred in O. bartramii. The exponential relationship of 1)15N with mantle length was stronger in individual years than for all the data combined, except in 1998, which had a weak relationship. The weak exponential relationship between 1)15N and mantle length for 1998 is probably due to the lack of samples on each end of the size spectrum. In 1998 there were no paralarval samples taken, and few S. oulaniensis adults above 200 rom ML were captured. The 615N values taken for the blood samples of S. oualaniensis support the general trend ofexponential increase in 1)15N with increased size. The blood samples were statistically different from the mantle samples but there was a unifonn difference over the mantle sizes tested. This would indicate that isotopic signals that are presumably incorporated over a short time period in the blood (weeks?) are similar to those presumably incorporated over a longer time period in the muscle (months?).

109 The li15N values taken from retrospective analyses ofthe eye lens segments ofthe "typical" group show an exponential increase with mantle size (estimated from eye lens radius) similar to that seen for mantle muscle, although falling well below the curve for mantle muscle at smaller sizes. The reasons for this discrepancy at smaller sizes are unknown. All of the individual retrospective relationships between li15N and estimated mantle length were stronger than for the combination of all individuals. The differences between the two steeply increasing "atypical" individuals and the more gradually increasing "typical" individuals which came from the same station could be a result of individual feeding specialization, beyond where the curves join near 75 mm ML, by the heavier group of animals. Many of the values from estimated sizes of ca. 110-150 mm

ML fell in the same 1)15N range as the "typical" group. Another explanation for the "atypical" values seen in these individuals is different growth rates between the two groups. If fractionation is less during periods of rapid growth (Fantle et al., 1999) the animals with slower growth rates would have higher li15N values. The "anomalous" individual is problematic and will be considered in more detail in the next chapter. For other squid the li15N values observed for the eye lenses represent an increase in trophic position with size. More detailed individual analyses of these groups is conducted in Chapter ill.

Species comparisons

The li15N values seem to describe trophic positions fairly well in this study. The increase in li15N from POM up to large animals such as squids and tunas agrees in general with existing data on the trophic position of these organisms. The fact that O. bartramii occupy a higher trophic level than tunas, and that the largest sized S. oualaniensis approach the trophic level of the tunas sampled may be counterintuitive when considering the size differences. However, size alone, is insufficient to determine trophic levels especially when the animals in question belong to unrelated taxa. In general, adult O. bartramii is more then a trophic level above S. oualaniensis (4.6%0, 1.3 TLs), which is compatible with S. oualaniensis as prey for O. bartramii (a situation which was presumed at the beginning of this study). From the stomach analyses,

110 however, S. oualaniensis is rarely preyed upon by O. bartramii. This situation re-affinns the need to couple SIA with known feeding data. The higher trophic position of O. bartramii, relative to S. oualaniensis, can be partially explained by prey items found during the SCA. The five most abundant prey species found in O. bartarmii had a mean /)uN value that was almost exactly a trophic level below the mean value of adult O. bartramii. This is somewhat surprising given the result that O. bartramii was a generalist feeder and did not specialize heavily on anyone prey item. The five most abundant prey items in S. oualaniensis, however, were only about one fifth of a trophic level below the mean value of S. oualaniensis, in spite of the fact that S. oualaniensis is a more specialized predator; only five prey species make up 75% of the otoliths recovered from S. oualaniensis stomachs. The prey fishes sampled for SIA, however, were too large to adequately represent S. oulaniensis and O. bartramii prey sizes. The otoliths for the five most abundant species collected from .the stomachs of adultlsub-adult S. oualaniensis were on average only 58% of the length of otoliths from the same species used for SIA, compared to otoliths from adultlsub-adult O. bartramii which were 74% of the size sampled for SIA. Using the genus group of M. lychnobium and M. spinosum this effect would cause a 1.0%0 difference in S. oualaniensis and a 0.8%0 difference in O. bartramii. The fishes sampled for SIA, therefore, were most likely heavier (isotopically) than actual prey fed upon by both squids, which still places S. oualaniensis only about half a trophic level above, and O. bartramii about one and a half trophic levels, above their primary prey (using 3.4%0 per trophic level). Fishes sampled for SIA were closer in size to those fed upon by O. bartramii and therefore more accurately reflect the dietary /)15N signal of O. bartramii than of S. oualaniensis. There is also the possibility that a substantial component of the diet ofS. oualaniensis was missing from the stomach content study.

The patterns of /) 15 N change with size for the mantle samples of O. bartramii and S. oualaniensis were different, a sigmoidal curve versus an exponential curve. Among the most likely factors that would contribute to these differences are: differences in feeding, growth rate, and migration. S. oualaniensis and O. bartramii clearly show differential feeding patterns according to SCA but data are few for O. bartramii in the size range where increases in /) 15N versus ML are most rapid.

111 The migrational pattern of O. bartramii may lead to the animal being subjected to isotopically different ecosystems. When O. bartramii migrates from near Hawaiian waters, north to the feeding grounds it may enter into an ecosystem that has a totally different baseline signal for 015N. Nitrogen fixation by organisms such as Trichodesmium sp. can introduce low (close to zero) 015N nitrogen into the ecosystem (Mingawa and Wada, 1986) and may be important in oligotrophic waters around Hawaii. Upwelled nitrate, on the other hand, has more enriched 015N values (Altabet, 1989; Liu and Kaplan, 1989) and in the transition zone (the northern end of O. bartramii migration) may playa more important role in supplying the nitrogen demands of the phytoplankton resulting in a higher baseline trophic signal. Wada and Hattori (1976) found a range of PON in the North Pacfic from -1.7 to 9.7%0 while Miyake and Wada (1967) found a range of 3.4 to 7.2%0 for PON. Two studies have shown possible migratory patterns of 015N in whale baleen (Best and Schell, 1996; Hobson and Schell, 1998) on the order of 2-4%0. Hypothetically the migration of these squids from a lighter ecosystem to a heavier one (see Wada and Hattori, 1991) and then back again should result in a more bell shaped curve of the eye lens segments. Migration south shows no obvious effect on 015N in spite of seemingly mandatory changes in diet with region. The greatest range of 015N for the eye lenses of O. bartramii was 13%0, which is a large change to explain using migration alone. Values for O. bartramii caught offJapan (Takai, et ai., 2000) (39°30'N, 155"E) and in the North Pacfic transition zone (Gould et al., 1997) (35-46"N, 170"E-148°W), also did not show substantially different values, 12.1 and 11.7%0 respectively. However, if growth rates are taken into consideration as well, the situation changes. If O. bartramii individuals are putting on the majority of their mass over a relatively short time period at the feeding grounds, then the northern signal will be incorporated and the muscle signal will remain relatively high even though the animals are moving south again because the squid are presumably not growing as rapidly (much of the growth would be in reproductive tissue) and hence incorporating little of the lighter southern signal into muscle. The signal of the tissues, therefore, will be a combination ofrapid growth ofnew tissue then maintenance turnover of old tissue. If turnover times of the tissues in question are too long then the southern signal will not be observed from either growth or turnover. The time periods that these animals remain in each area would have a large effect on the

112 shape of the eye lens segments 1)15N curve with length. In conjunction with these hypotheses, feeding on different prey with heavier signals during rapid growth of the squid predator could playa role in shaping the sigmoidal curve, regardless of the baseline ecosystem signal. Little is known about the size at which O. bartramii begins its migration northward. At 175 mm ML S. oualaniensis begins to exhibit a variety of increasing 1)15N trajectories based on retrospective eye lens data (see next chapter) presumably due to growth or diet differences. The steep trajectory of O. bartramii appears to start around the same size and could have the same cause since, by this size, they should be well into their northern migration and encountering different feeding regimes. Around this size growth (in length) should also be at a maximum, and much faster than in S. oualaniensis, which could contribute as well.

The northern group of 0. bartramii showed no increase in 1)15N with ML suggesting that, at this locality, prey type was independent of size and that squid had sufficient time to accommodate prey values with their tissues (this will be examined more closely in Chapter III). Presumably 1)15N increase with size resumed as migration continued. In general O. bartramii shows a final plateau in trophic level that begins at sizes that are near S. oualaniensis' maximum size in Hawaiian waters. S. oualaniensis does not show a plateau. If absolute size is the primary factor determining the presence of a plateau in these squids, then S. oualaniensis doesn't get quite large enough to reach a trophic plateau.

The 1)15N values for blood samples showed similar patterns for S. oualaniensis and 0. bartramii to those found using muscle and eye lens segments. The mean difference between O. bartramii and S. oualaniensis using blood was 4.2%0 (1.2 TLs) compared to 4.6%0 (1.3 TLs) obtained for muscle of adults/sub-adults. The difference between muscle and blood within O. bartramii and within S. oualaniensis averaged 0.6%0 and 0.6%0 respectively and was consistent between most individuals. This indicates that a distinct difference in fractionation occurs between these tissues. Correcting for this tissue bias, no differences were apparent that might reflect different turnover times between the supposedly slow muscle and fast blood tissues. The implications are that O. bartramii

113 individuals sampled may have been present in the area on long enough time scales to wholly incorporate this signal in both tissues, or that there is no signal difference between waters near Hawaii and the northern range of O. bartramii's migration.

Isotopes at indicators oftrophic level

This study brings into question the efficacy of using the value of 3.4%0 as the standard increase in 1i 15N per trophic level. The bulk plankton values obtained from the net tows were fairly close in value to the mean 1i15N values obtained for the paM. Even the range of values for the bulk plankton and the paM were similar in most cases. Certain organisms such as chaetognaths had 1i 15N values which, using the 3.4%0 per TL, would identify them incorrectly as herbivores. Other problems with using this standard increase occurred with the O. bartramii paralarvae showing very similar values to juveniles, and the 0.5 TL calculated for S. oualaniensis and the 1.5 TL for O. batramii over their prey. These problems point to greater inherent isotopic variability between trophic positions at lower trophic levels. The greater isotopic variability between trophic positions at lower trophic levels probably reflects the complexity of the planktonic ecosystem. The effects of nutrient recycling, new production, and mixotrophy, and increased assimilation efficiencies should be most notable at lower trophic levels as well. Presumably, the varied diet of larger animals and the long integration times of their tissues, produce long-term averages that progressively damp out many of the initial factors that affect Ii15N at the higher trophic levels. This, however, is confounded by individual trophic variability and taxa/tissue specific biases at higher trophic levels. Turnover times probably change with height in the trophic pyramid, for example, the large variability of the myctophids (- 8%0) suggests that there may be factors other than diet that are involved in determining the isotopic signal and would therefore compromise the canonical 3.4%0 value.

Conclusions

Several limitations with using stable isotope analyses as a method to elucidate the trophic ecology of organisms were encountered during this study. If 1i 15N is reflective of trophic level, and the trophic level of an animal is dictated by its size, then the 1i15N signal

114 will, to some extent, be dependent on a number of variables affecting the animal's size such as growth rate, food composition, and condition of the animal. This study has also shown that using different tissues will not necessarily yield similar results. While some tissues such as the mantle, blood, and eye lenses all supported the general trends of change in olsN with size for both S. oualaniensis and O. bartramii, there were substantial differences in the increase determined for S. oualaniensis mantle and eye lens tissue. These discrepancies between and within some tissues require that either specific tissues be carefully chosen when addressing certain questions, or that whole organisms be used throughout a study. The variability in using SIA requires that its use be coupled with other techniques that will provide a context or background within which the results can be analyzed. Stomach contents analysis or behavioral feeding studies would provide the best framework to base SIA results on. The differences in feeding by S. oualaniensis and O. bartramii revealed in the stomach contents clearly indicates different trophic ecology for these squids. However, the wide difference in trophic level determined by olsN was unexpected. Both squids fed heavily on Myctophidae, although O. bartramii also clearly feeds on certain piscivorous fishes. The SIA results indicate that these animals are substantially more than a trophic level apart based on the mantle, blood, and terminal eye segments, although individual variability was great. Obviously, placing undue confidence in one or a few measurements is risky. One must also be cautious, therefore, and rely on average values until the causes of variability are better understood. Future studies of SIA should include rigorous controlled experimental studies to determine the exact sources of 015N variability. Unfortunately this approach is not always possible. This study, for example, examined animals that do not yield easily, if at all, to experimental manipulation. Nevertheless, SIA remains one of the few techniques to assess trophic interactions within ecosystems that are difficult at best to study. In such cases, constraining the results of SIA by using different tissues and supporting techniques is crucial for obtaining useful results.

The SIA results of this study indicate that the 015N values can show consistent general trends that are indicative of trophic interactions, in spite of the many factors that affect the results. Not surprisingly, substantial variability within these general patterns

115 exists. A fundamental question of stable isotope analyses still exists; how much of the variability represents true trophic variability and how much represents other sources of bias? If this variability represents true trophic variability, then it suggests that trophic structure is.much more complex than previously thought. To better determine how these general values (and their variability) reflect the feeding ecology of an organism it is also necessary to look at specific individual variation to see how it compares to the larger general trends. This is the subject of the next chapter.

116 References

Beaudoin, C. P., W. M. Tonn, E. E. Prepas, and L. I. Wassenaar (1999). "Individual specialization and trophic adaptability of northern pike (Esox lucius): an isotope and dietary analysis." Oecologia 120: 386-396.

Ben-David, M., R. W. Flynn, and D. M. Schell (1997). "Annual and seasonal changes in diets of martens: evidence from stable isotope analysis." Oecologia 111: 280-291.

Best, P. B., and D. M. Schell (1996). "Stable isotopes in southern right whale (Eubalaena australis) baleen as indicators of seasonal movements, feeding and growth." Marine Biologv 124: 483-494.

Borgmann, U. (1987). "Models on the slope of, and biomass flow up, the biomass size spectrum." Canadian Journal ofFisheries and Aquatic Sciences 44(Supplement 2): 136­ 140.

Bosley, K. L., and S. C. Wainright (1999). "Effects of preservatives and acidification on the stable isotope ratios eSN: 14N, l3C:12C) of two species of marine mammals." Canadian Journal ofFisheries and Aquatic Sciences 56: 2181-2185.

Castellani, M. A., and L. D. Rea (1992). "The biochemistry of natural fasting at its limits." Experientia 48: 575-582.

Cline, J. D., and I. R. Kaplan (1975). "Isotopic fractionation of dissolved nitrate during denitrification in the eastern tropical North Pacific Ocean." Marine Chemistry 3: 271­ 299.

117 DeNiro, M. J., and S. Epstein (1981). "Influence of diet on the distribution of nitrogen isotopes in animals." Geochmica et Cosmochimica Acta 45: 341-351.

Estep, M. L. F., and S. Vigg (1985). "Stable carbon and nitrogen isotope tracers of trophic dynamics in natural populations and fisheries of the Lahontan Lake, system, Nevada." Canadian Journal of Fisheries and Aquatic Sciences 42: 1712-1719.

Fantle, M. S., A. I. Dittel, S. M. Schwalm, C. E. Epifanio, and M. L. Fogel (1999). "A food web analysis of the juvenile blue crab, Callinectes sapidus, using stable isotopes in whole animals and individual amino acids." Oecologia 120: 416-426.

Faure, G. (1986). Principles ofIsotope Geology. New York, John Wiley and Sons.

Fogel, M. L., and L. A. Cifuentes (1993). Isotope frationation during primary production. Organic Geochemistrv. a. S. A. M. M. H. Engel. New York, Plenum Press: 73-98.

France, R., M. Chandler, and R. Peters (1998). "Mapping trophic continua of benthic foodwebs: body size-li1sN relationships." Marine Ecology Progress Series 174: 301-306.

Frazer, T. K., R. M. Ross, L. B. Quetin, and J. P. Montoya (1997). "Turnover of carbon and nitrogen during growth of larval krill, Euphausia superba Dana: a stable isotope approach." Journal of Experimental Marine Biology and Ecology(212): 259-275.

Fry, B. (1988). "Food web structure on Georges Bank from stable C, N, and S isotopic composition." Limnology and Oceanography 33: 1182-1190.

Gould, P., P. Ostrom, and W. Walker (1997). "Trophic relationships of albatrosses associated with squid and large-mesh drift-net fisheries in the North Pacific Ocean." Canadian Journal of Zoology 75: 549-562.

118 Hesslein, R. H., K. A. Hallard, and P. Ramlal (1993). "Replacement of sulfur, carbon, and nitrogen in tissue of growing broad whitefish (Coregonus nasus) in response to a change in diet traced by 334S, 3l3C, and 315N." Canadian Journal ofFisheries and Aquatic Sciences 50: 1993.

Hobson, K. A., and D. M. Schell (1998). "Stable carbon and nitrogen isotope patterns in baleen from eastern Arctic bowhead whales (Balaena mysticetus)." Canadian Joumal of Fisheries and Aquatic Sciences 55: 2601-2607.

Hobson, K. A., and H. E. Welch (1992). "Determination of trophic relationships within a high Arctic marine food web using 3l3C and 315N analysis." Marine Ecology Progress Series 84(9-18).

Hobson, K. A., H. L. Gibbs, and M. L. Gloutney (1997). "Preservation of blood and tissue samples for stable-carbon and stable-nitrogen isotop analysis." Canadian Journal of Zoology 75: 1720-1723.

Hobson, K. A., J. F., Piatt, and J. Pitocchelli (1994). "Using stable isotopes to determine seabird trophic relationships." Journal of Animal Ecology 63: 786-798.

Hobson, K. A., R. T. Alisauskas, and R. G. Clark (1993). "Stable-nitrogen isotope enrichment in avian tissues due to fasting and nutritional stress: implications for isotopic analysis of diet." Condor 95: 388-394.

Hoch, M. P., M. L. Fogel, and D. L. Kirchman (1992). "Isotope fractionation associated with ammonium uptake by a marine bacterium." Limnology and Oceanography 37(7): 1447-1459.

Hoch, M. P., R. A. Snyder, L. A. Cifuentes, and R. D. Coffin (1996). "Stable isotope dynamics of nitrogen recycled during interactions among marine bacteria and protists." Marine Ecology Progress Series 132: 229-239.

119 Kaehler, S., E. A. Pakhomov, and C. D. McQuaid (2000). "Trophic structure of the marine food web at the Prince Edward Islands (Southern Ocean) determined by /i13C and /i15N analysis." Marine Ecology Progress Series 208: 13-20.

Karasov, W. H. (1990). Digestion in birds: chemical and physiological determinants. and ecological implications. Lawrence, Kansas.

Karl, D., R. Letelier, L. Tupas, J. Dore, J. Christian, and D. Hebel (1997). "The role of nitrogen fixation in biogeochemical cycling in the subtropical North Pacific Ocean." Nature 388: 523-538.

King, J. B., and J. J. Ikehara (1956). "Comparative study offood of bigeye and yellowfin tuna in the Central Pacfic." Fishery Bulletin 57: 61-85.

Kline, T. C. (1997). Confinning forage fish food web dependencies in Prince William Sound using natural stable isotope tracers. Forage Fishes in Marine Ecosystems, Alaska Sea Grant College Program.

Kline, T. C. (1999). "Temporal and spatial variability of 13CfI2C and 15N114N in pelagic biota of Prince William Sound, Alaska." Canadian Journal of Fisheries and Aquatic Sciences 56: 94-117.

Kline, T. C., W. J. Wilson, and J. J. Goering (1998). "Natural isotope indicators of fish migration at Prudhoe Bay, Alaska." Canadian Journal of Fisheries and Aquatic Sciences 55: 1494-1502.

Kling, G. W., B. Fry, and W. J. O'Brien (1992). "Stable isotopes and planktonic trophic structure in arctic lakes." Ecology 73(2): 561-566.

120 Lancraft, T. M., T. L. Hopkins, and J. J. Torres (1988). "Aspects of the ecology of the mesopelagic Gonostoma elongatum (Gonostomatidae, Stomiiforrnes) in the eastern Gulf of Mexico." Marine Ecology Progress Series 49(1-2): 27-40.

Liu, K, and 1. R. Kaplan (1989). "The eastern tropicl Pacific as a source of 15N-enriched nitrate in seawater off southern California." Limnology and Oceanography 34(5): 820­ 830.

Macko, S. A., M. L. Fogel-Estep, M. H. Engel, and P. E. Hare (1986). "Kinetic fractionation of nitrogen isotopes during amino acid transamination." Geochmica et Cosmochimica Acta 50: 2143-2146.

Macko, S. A., M. L. Fogel-Estep, M. H. Engel, and P. E. Hare (1987). "Isotopic fractionation of nitrogen and carbon in the synthesis of amino acids by microorganisms." Chemical Geology 65: 79-92.

Macko, S. A., W. Y. Lee, and P. L. Parker (1982). "Nitrogen and carbon fractionation by two species of marine amphipods: laboratory and field studies." Journal ofExperimental Marine Biology and Ecology 63: 145-149.

Marra, P. P., K A. Hobson, and R. T. Holmes (1998). "Linking winter and summer events in a migratory bird by using stable-carbon isotopes." Science 282: 1884-1886.

Massuti, E., S. Deudero, P. Sanchez, and B. Morales-Nin (1998). "Diet and feeding of dolphin (Coryphaena hippurus) in western Mediterranean waters." Bulletin of Marine Science 63(2): 329-341.

Michener, R. H., and D. M. Schell (1994). Stable isotope ratios as tracers in marine aquatic food webs. Stable isotopes in ecology and environmental science. R. H. Michener. Oxford, England, Blackwell: 138-157.

121 Minami, H., and H. Ogi (1997). "Detennination of migratory dynamics of the sooty shearwater in the Pacific using stable carbon and nitrogen isotope analysis." Marine Ecology ProgreSS Series 158: 249-256.

Mingawa, M., and E. Wada (1984). "Stepwise enrichment of 15N along food chains: further evidence and the relation between 015N and animal age." Geochmica et Cosmochimica Acta 48: 1135-1140.

Miyake, Y., and E. Wada (1971). "The isotope effect on the nitrogen in biochemical oxidation-reduction reactions." Recent Oceanographic Works from Japan 11: 1-6.

Mizutani, H., M. Fukuda, and Y. Kabaya (1992). ,,13C and 15N enrichment factors of feathers of II species of adult birds." Ecology 73(4): 1391-1395.

Moncreiff, C. A., and M. J. Sullivan (2001). "Trophic importance of epiphytic algae in subtropical seagrass beds: evidence from multiple stable isotope analyses." Marine Ecology Progress Series 215: 93-106.

Paine, R. T. (1988). "Food webs: road maps of interactions or grist for theoretical development." Ecology 69: 1648-1654.

Peters, R. H., G. Cabana, O. Choulik, T. Cohen, O. Griesbach, and S. J. McCanny (1996). "General models for trophic fluxes in animals based on their body size." Ecoscience 3: 365-377.

Petersen, B. J., and B. Fry (1987). "Stable isotopes in ecosystem studies." Annual Revue of Ecology and Systematics 18: 293-320.

Rau, G. H., D. G. Ainley, J. L. Bengston, J. J. Torres, and T. L. Hopkins (1992). ,,15N/14N and 13CPC in Weddell Sea birds, seals, and fish: Implications for diet and trophic structure." Marine Ecology Progress Series 84: 1-8.

122 Rientjets, J. W., and J. B. King (1953). "Food of yellowfin tuna in the Central Pacific." Fishery Bulletin 81(91-110).

Schwarcz, H. P. (1991). "Some theoretical aspects of isotope paleodiet studies." Journal of Archeological Science 18: 261-275.

Sholto-Douglas, A. D., J. G. Field, A. G. James, and N. J. van der Merwe (1991). 14 "nC/12C and 15N/ N isotope ratios in the Southern Benguela ecosystem: indicators of food web relationships among different size-classes of plankton and pelagic fish; differences between fish muscle and bone collagen tissues." Marine Ecology Progress Series 78: 23-31.

Steele, K. W., and R. M. Daniel (1978). "Fractionation of nitrogen isotopes by animals: A further complication of the use of variation in the natural abundance of 15 N for tracer studies." Journal of Agricultural Science 90: 7-9.

Takai, N., S. Onaka, Y. Ikeda, A. Yatsu, H. Kidokoro, and W. Sakamoto (2000). "Geographical variations in carbon and nitrogen stable isotope ratios in squid." Journal of the Marine Biological Association of the United Kingdom 80: 675-684.

Thomas, C. J., and L. B. Cahoon (1993). "Stable isotope analyses differentiate between different trophic pathways supporting rocky-reef fishes." Marine Ecology Progress Series 95: 19-24.

Tsuchiya, K., H. Okamoto, and Y. Uozumi (1998). "Cephalopods eaten by pelagic fishes in the tropical East Pacific, with special reference to the feeding habit of pelagic fish." Umilla mer 36(2): 57-66.

Vander-Zanden, M. J., B. J. Shuter, N. P. Lester, and J. B. Rasmussen (2000). "Within­ and among- population variation in the trophic position of a pelagic predator, lake trout

123 (Salvelinus namaycush)." Canadian Journal ofFisheries and Aquatic Sciences 57: 725­ 731.

Wada, E., and A. Hattori (1991). Nitrogen in the sea: founs, abundances, and rate processes. Boca Raton, CRC Press.

Wada, E. a. A. H. (1976). "Natural abundance of 15N in particulate organic matter in the North Pacific Ocean." Geochmica et Cosmochimica Acta 40: 249-251.

Wada, E. a. A. H. (1978). "Nitrogen isotope effects in the assimilation of inorganic nitrogenous compounds by marine diatoms." Geomicrobiological Journal 1: 85-101.

Wainright, S. C., M. J. Fogarty, R. C. Greenfield, and B. Fry (1993). "Long-term changes in the Georges Bank food web: trends in stable isotopic compositions of fish scales." Marine Biology 115: 481-493.

Waser, N. A., K. Yin, Z. Yu, K. Tada, P. J. Harrison, D. H. Turpin, S. E. Calvert (1998b). "Nitrogen isotope fractionation during nitrate, ammonium and urea uptake by marine diatoms and coccolithophores under various conditions ofN availability." Marine Ecology Progress Series 169: 29-41.

Waser, N. A. D., P. J. Harrison, B. Nielsen, and S. E. Calvert (1998a). "Nitrogen isotope fractionation during the uptake and assimilation of nitrate, nitrite, ammonium, and urea by a marine diatom." Limnology and Oceanography 43(2): 215-224.

124 List of Tables

TABLE 1. THE MEAN AND STANDARD DEVIATIONS OF '" 15N VALVES FOR ALL POM

COllECTED ON SEPARATE SHIPS DURING DIFFERENT YEARS. HM= HOKUSEI MARU,

TC= TOWNSEND CROMWELL, 1999 TOTAL IS THE COMBINED VALVES FOR THE HM

AND TC CRUISES. THE TOWNSEND CROMWELL CRUISES TOOK PLACE IN NOVEMBER

AND JANUARY; ALL OTHER SAMPLES WERE TAKEN DURING FEBRUARy 93

TABLE 2. THE MEAN '"15N vALVES FOR PLANKTON COllECTED DURING 1999 AND 2000

ABOARD THE HOKUSEI MARU. THE PLANKTON CATEGORY FOR 2000 INCLUDES THE

VARIOUS SUBGROUPS LISTED SEPARATELY. THE NUMBER OF SAMPLES IN THE

CATEGORIES WITH SPECIFIC TAXA REPRESENT NUMBER OF COMBINED INDIVIDUALS. 94

TABLE 3. THE MEAN '"15N vALVES FOR DIFFERENT CATEGORIES OF KNOWN AND POTENTIAL

PREY ITEMS OF O. BARTRAMll AND/OR S. OUALANIENSIS. STANDARD DEVIATIONS ARE

LISTED IN PARENTHESES. ALL SPECIES OF MYCTOPHID LISTED ARE INCLUDED IN THE

CATEGORY MYCTOPHIDAE. HEAVILY OUTLINED CELLS REPRESENT MYCTOPHlD

SPECIES 95

TABLE 4. THE MEAN '"15N vALVES FOR DIFFERENT CATEGORIES OFEPlPELAGIC flSHES,

STANDARD DEVIATIONS ARE LISTED INPARENIHESES. THE TUNA CATEGORY IS

COMPRISED OF THE FOUR SPECIES LISTED IN THE HEAVILY OUTLINED CELLS 96

TABLE 5. THE MEAN '"15N VALVES OF MANTLE MUSCLE FOR DIFFERENT CATEGORIES OF O.

BARTRAMll, STANDARD DEVIATIONS ARE LISTED IN PARENTHESES. WHOLE ANIMALS

WERE ANALYZED IN THE CATEGORY OF 0-7 MM ML 97 2 TABLE 6. THE R VALVES OF THE SIGMOIDAL RELATIONSHIP FOR'"15N OF EYE LENS

SEGMENTS AND ESTIMATED MANTLE LENGTH (BASED ONEYE LENS RADIUS), AND

NUMBER OF SAMPLES FOR DIFFERENT O. BARTRAMIIINDIVIDUALS 98 2 TABLE 7. THE R VALVES, AND NUMBER OF SAMPLES OF (MANTLE MUSCLE) FROM ALL 2 SIZES(WHOLE BODY) TAKEN FOR EACH YEAR. THE R VALVES ARE FOR SINGLE

EXPONENT 3-PARAMETER EXPONENTIAL RELATIONSHIPS BETWEEN MANTLE LENGTH

AND ",15N 99

TABLE 8. THE MEAN AND STANDARD DEVIATION OF MANTLE LENGTH VALVES (MM) OF S.

OUALANIENS1S INDIVIDUALS USED FOR SIA DURING YEARS 1998·2001. *THE MEAN

125 tJ.15 N vALVES FOR 1999 WERE CALCULATED WITIlOUT TIIE TWO ISOTOPICAlLY HEAVY

STATIONS, N'=32...... •••...•...... •...... •...... •...... 100

TABLE 9. THE MEAN AND STANDARD DEVIATION OF RESIDUAL VALVES OF tJ.15 N (MANnE

MUSCLE) FOR S. OUALAN1ENS1S MANTLE SAMPLES FOR YEAR, LATITUDE, AND

LONGITUDE. *THE MEAN RESIDUAL VALVES FOR 1999 WERE CALCULATED WITIlOUT

TIIE TWO ISOTOPICALLY HEAVY STATIONS...... •...... •...... 101

TABLE 10. THE NUMBER OF SAMPLES, MINIMUM AND MAXIMUM tJ.15N VALVES OF EYE LENS

SEGMENTS FOR S. OUALANIENSIS INDNIDUALS FROM TIIE IHREE DlFFERENT CLASSES 2 AND TIIE R VALVES FOR TIIEIR EXPONENTIAL RELATIONSHIP WITII ESTIMATED

MANnE LENGlH (BASED ON EYE LENS RADIUS)...... •...... 102

126 List of Figures

FIGURE 1. S15N OF FILTERED MATERIAL TAKEN DURING CRUISES ON THREE SUCCESSNE

YEARS...... •...... 130 15 FIGURE 2. THES N OF FILTERED MATERIAL TAKEN DURING CRUISES ON THREE

SUCCESSNE YEARS 131

FIGURE 3. THE S15N OF MUSCLE TISSUE FOR FNE GROUPS OF MESOPELAGIC FISHES

SHOWING THE LINEAR INCREASE IN S15N WITH LENGTH. THE SINGLE MELAMPHAEID

INDNUDUAL (CIRCLED VALUE) WAS EXCLUDED FROM THE REGRESSION FOR

MELAMPHAEIDAE 132

FIGURE 4. THE S15N OF MUSCLE TISSUE FROM FIVE SEPARATE MYCTOPHID GROUPS. THESE

GROUPS DID NOT SHOW OBVIOUS TRENDS WITH LENGTH 133

FIGURE 5. THE S15N OF WHITE MUSCLE FROM TWO SPECIES OF TUNA, YELLOWFlN

(THUNNUS ALBACARES) AND BIGEYE (THUNNUS 08ESUS) 134

FIGURE 6. THE S15N OF WHITE MUSCLE FROM MARl MARl (CORYPHAENA HlPPURUS) 135

FIGURE 7. THE S15N VALUES FOR MANTLE MUSCLE TAKEN FROM O. BARTRAMll DURING ALL

FOUR YEARS OF SAMPLING FOR SIA. THE CURVE IS A FOUR PARAMETER LOGISTIC

CURVE, COMPUTER FIT WITH SIGMAPLOT 2000. ALL SQUIDS WERE CAUGHT ON THE

HOKUSEI MARU EXCEPT THE NORTHERN 2000 SAMPLES. F=Yo+A*EXP(-EXP(-(X-

Xo)/B)); A= 6.8482, B= 38.0979, Xo= 183.2625, Yo= 6.4261. 136

FIGURE 8. THE S 15N VALUES FOR EYE LENS SEGMENTS FOR 14 SEPARATE O. BARTRAMll

INDNIDUALS. MANTLE LENGTH WAS ESTIMATED FROM REGRESSIONS OF WHOLE EYE

LENS RADII OF DIFFERENT SIZES OF INDNIDUALS. THE SOLID LINE IS THE REGRESSION

CALCULATED FOR EYE LENSES, THE DOTTED LINE IS THE REGRESSION CALCULATED

FOR MANTLE MUSCLE. F=YO+A*EXP(-EXP(-(X-XO)/B)); A= 7.8561, B= 38.0665, xo=

195.6045, Yo= 5.2977. EQUATION FOR MANTLE MUSCLE CURVE GNEN IN FIGURE 7...... 137

FIGURE 9. THE S15N VALUES FOR EYE LENS SEGMENTS OF FOUR JUVENILE O. BARTRAMll.

THE LINE SHOWS THE REGRESSION LINE CALCULATED FOR EYES OF ALL SQUIDS

COMBINED. MANTLE LENGTHS WERE CALCULATED USING EYE LENS RADIUS OF WHOLE

EYE LENS RADII OF DIFFERENT SIZED SQUIDS 138

127 FIGURE 10. THE 015N v ALVES FOR BLOOD FROM O. BARTRAMIl INDNIDUALS OF BOTII

GENDERS. THE DOTfED LINE IS mE REGRESSION LINE CALCULATED FOR mE MANTLE

S15N VALVES VERSUS MANTLE LENGTII...... •...... •...... •.....•...... 139

FIGURE 11. THE 015N VALVES FOR MANTLE MUSCLE TAKEN FROM S. OUALANIENSlS DURING

ALL FOUR YEARS OF SAMPLING FOR SIA. THE CURVE IS A SINGLE TIIREE PARAMETER

EXPONENTIAL CURVE, COMPUTER FIT TO mE DATA USING SIGMAPLOT 2000. THE

CIRCLED VALVES REPRESENT UNUSUALLY HEAVY SAMPLES OBTAINED AT STATIONS 2­ x 12/13-00,2-9/10-99, AND 2-14115-99. F=Yo+A*B , Yo= 4.2081, A= 1.8726, B= 1.0037 140 15 FIGURE 12. THEIl N VALVES FOR EYE LENS SEGMENTS FOR 14 SEPARATES. OUALANIENSlS

INDNIDUALS. MANTLE LENGTII WAS ESTIMATED FROM REGRESSIONS OF WHOLE EYE

LENS RADII OF DIFFERENT SIZES OF INDNIDUALS. THE THICK LINE IS mE REGRESSION

CALCULATED FOR "TYPICAL" EYE LENSES, mE TIIIN LINE IT mE REGRESSION FOR mE

"ATYPICAL" EYE LENSES, AND mE DOTfED LINE IS mE REGRESSION CALCULATED FOR x MANTLE MUSCLE. F=YO+A*B . "TYPICAL" CURVE Yo= 2.9207, A= 0.3376, B= 1.0096;

"ATYPICAL" CURVE, Yo= 2.7582, A= 0.3712, B= 1.0125 141

FIGURE 13. THE 015N VALVES FOR BLOOD FROM S. OUALANIENSlS INDNIDUALS. THE SOLID

LINE IS mE REGRESSION CALCULATED FOR THE BLOOD SAMPLES, THE DOTTED LINE IS

mE REGRESSION CALCULATED FOR mE MANTLE MUSCLE. THE BLOOD REGRESSION x WAS CALCULATED WITHOUT mE "ANOMALOUS" VALVE. F=Yo+A*B ; Yo= 5.2274, A=

0.4745, B= 1.0071 142 15 FIGURE 14. THE 1l N VALVES FOR DIFFERENT CATEGORIES OFORGANISMS SAMPLED.

VALVES SHOW AN INCREASING TREND WITH TAXON. POINTS REPRESENT MEAN

VALVES WHILE mE ERROR BARS REPRESENT ONE STANDARD DEVIATION FROM mE

MEAN...•.•...... •...... •...... •.....•.•...... •.. 143 15 FIGURE 15. THE 1l N OF BLOOD AND MANTLE SAMPLES FROM mE SAME O. BARTRAMll

INDNIDUALS. ALL FOUR YEARS FOR WHICH BLOOD SAMPLES WERE TAKEN ARE

SHOWN. THE * DENOTE mE VALVES FOR MANTLE AND BLOOD TIIAT CAME FROM mE

SAME ANIMAL WHERE SYMBOLS ARE OBSCURED 144 15 FIGURE 16. THE 1l N OF BLOOD AND MANTLE SAMPLES FROM mE SAME S. OUALANIENSIS

INDNIDUALS. ALL TIIREE YEARS FOR WHICH BLOOD SAMPLES WERE TAKEN ARE

128 SHOWN. 1HE * DENOTE TIffi VALUES FOR MANnE AND BLOOD lHAT CAME FROM TIffi

SAME ANIMAL WHERE SYMBOLS ARE OBSCURED 145

129 8

'" 0 1999 6 [J 2000 0 2001 '" B !l. '" "'EJ 0 0 ::;; 4 0 0 '" 0 a. '" A'" '" 0 o 0 -0 ~ ~A 0 Z 0 2 ~e.o'" Cb o tt 0 0 0 00 0 0 [J 0 0 0 0 0 0 0

-2 16 18 20 22 24 26 28 30 32 Latitude (degrees N)

Figure 1. 315N offiltered material takeu during cruises on three successive years.

130 8

b. 0 1999 6 0 2000 b. 2001 b. 0 B ~ Ill. b. OJ! ::i:4 0 0 0 b. b. 0- b. 0 0 0 0 b. - 0 0 0 Z 0 b. :'? 2 as b. <.0 0 gp 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

-2 166 164 162 160 158 156 154 152 150 Longitude (degrees W)

Figure 2. The l)1SN offiltered material taken during cruises on three successive years.

131 16 -,------,

14 ,. ,. ./ r"=O.52 (!) / 12 • ., Q) / "0 J...... III ...... ::l 0/0 •• ...... E r"=0.7V ...... 10 ...... 0 r"=0.7200/ ., ...... Z 0/"0 ...... , -00 '" III • 8 0 • • III M. fychnobium / ./. ., L. tenuiformis ./Ii Gonostomatidae 6 / ~. r"=0.84 • Melamphaeidae 0• Sternoptychidae 4 0 50 100 150 200 250 Length (mm)

Figure 3. The /j15N ofmuscle tissue for five groups ofmesopelagic fishes showing the linear increase in /jl'N with length. The single Melamphaeid indivudual (circled value) was excluded from the regression for Melamphaeidae.

132 Figure 4. The 1115N of muscle tissue from five separate myctophid groups. These groups did not show ohvious trends with length.

133 14 -.------,

o T. albacares o 13 o T.obesus o

12 o o (I) "0 o en 11 0 00 ::::l o E 00 0 o o 0 o '0 o 0 z 10 o o '"7-0 o 9 o o DO o 8

7+----,----,-----.------r------.---,---,------j o 200 400 600 800 1000 1200 1400 1600 Length (mm)

Figure 5. The 1115N of white muscle from two species of tuna, yellowliu (Thunnus albaeares) and Bigeye (Thunnus obesus).

134 13

0 12 0 C. hippurus I 0

11

0 Z 0 10 0 0 -00'" 0 0 0 0 9 0 0 0 0 0 0 0 0 0 o 0 0 8 0

7+.-----,-----,-----,-----,------1 800 900 1000 1100 1200 1300 Length (mm)

Figure Ii. The a1'N of white muscle from mahi mahi (Coryphaena hippurus).

135 18

16 •

14

12 •

10 Z • '"-00 8 • 1998 • 1999 6 • 2000 'VI'" Norlhern2000 4 2001 • Paralarvae1998 • Paralarvae1999 2 • • Paralarvae2000 0 0 200 400 600 Mantle length (mm)

Figure 7. The /l,sN values for mantle muscle taken from O. bartramii dnring all four years of sampling for SIA. The curve is a four parameter logistic curve, computer fit with Sigmaplot 2000. All squids were caught on the HOKUSEI MARU except the northern 2000 samples. f=yo+a*exp(.exp(-(x­ x.)Ib»; a= 6.8482, b= 38.0979, x.= 183.2625, Yo= 6.4261.

136 18 -,------,

.!!lc 16 o (]) o E g' 14 en o en c o ~ 12 ~=0.77 ~ (]) 10 oo :E o o 0 l!! '1:: 2 8 c:5 6 '5 z '"<.0 4 2+----,------r-----.-----..,-----,------I o 100 200 300 400 500 600 Estimated mantle length (mm) based on eye lens radius

Figure 8. The ll''N values for eye lens segments for 14 separate O. bartramii individuals. Mantle length was estimated from regressions of whole eye lens radii ofdifferent sizes of individuals. The solid line is the regression calculated for eye lenses, the dotted line is the regression calculated for mantle muscle. f=yo+a*exp(-exp(-(x-x,,)Ih)); a= 7.8561, h= 38.0665, x,,= 195.6045, Yo= 5.2977. Equation for mantle muscle curve given in Figure 7.

137 Estimated mante length (mm) from eye lens radius

Figure 9. The 1115N values for eye leus segmeuts offour juveuile O. bartramii. The line shows the regression line calculated for eyes ofall squids combined. Mantle lengths were calculated using eye lens radius ofwhole eye lens radii of different sized squids.

138 16 1j======::::::;------, • male 0. bal1ramii <:> female O. bal1ramii • 14 - .. <:> "8o /--...L~. ------~-- :c 12­ <:>g<:> :E / .+ft'.. Q~<:> l!! / • •• <:> 'l::: I • $ ~ <:> I • ~ o 10­ '0 / <:> z / / "'<.0 8 / I / e.-----/ 6 +------.----,------",----,,-----.-,----1 o 100 200 300 400 500 600 Mantle length (mm)

Figure 10. The l)''N values for blood from O. barlramii indIviduals ofboth genders. The dotted line Is the regression line calculated for the mantle l)''N values versus mantle length.

139 12

0 0 unusually heavy " qs> individuals 10 '"

~ 1 2 (]) r ",O.56 "0 8 Ul :J CD E ~ z •: 6 "'-00 • 1998 • 0 1999 2000 4 ".., 2001 • lliI 1999 paralarvae 2000 paralarvae • 2001 paralarvae 2 • 0 100 200 300 Mantle length (mm)

Figure 11. The 61'N values for mantIe muscle taken from S. oualllnknsis during all four years of sampling for SIA. The curve is a single three parameter exponential curve, computer fit to the data using Sigmaplot 2000. The circled values represent unusually heavy samples obtained at stations 2­ x 12/13-00,2-9/10·99, and 2-14/15-99. f=yo+a*b , Yo= 4.2081, a= 1.8726, b= 1.0037.

140 22 o typical '"c: 20 JIJ. atypical -(J) E • anomalous Ol 18 (J) 16 '"c: ~'" 14 (J) 1D' 12 .!!? •••• lJ) •••••• c: .!!! 10 c: ~ 8 § 0 6 ----- (Jj -0 4 Z "'<.0 2 0 0 50 100 150 200 250 300 350 Mantle length (mm) estimated from eye lens radii

Figure 12. The I)''N values for eye leus segments for 14 separate S. oualaniensis individuals. Mantle length was estilDllted from regressions of whole eye lens radii of differeut sizes of individuals. The thick line is the regression calculated for "typical" eye lenses, the thin line it the regression for the "atypiClll" eye lenses, and the dotted line is the regression calculated for IDllntle muscle. f=Yo+a*b'. ''typical'' curve Yo= 2.9207, a= 0.3376, b= 1.0096; "atypiClll" curve, yo= 2.7582, a= 0.3712, b= 1.0125.

141 Figure 13. The 1l''N values for blood from S. oualanumsis individuals. The solid line Is the regression calculated for the blood samples, the dotted Iiue is the regressiou calculated for the mantle muscle. x The blood regression was calculated without the "anomalous" value. f=Yo+a*b ; Yo= 5.2274, a= 0.4745, b= 1.0071.

142 18,------, 17 IT,.~:5------16 15 ~~ ~.4------i---

8 :_~2------1r-~-j------:-~~1'-~------1 O+-----,------,----.----,------r--,----,---,r--r--,---,---1, , .. :::' UJ rII Ql ~'" c: .~ CD :~ Q) :=: en .~ (/J rII <1>" E"" :< E.l!1 ",= -0 <: :g .c: .- ~ "'~e:~ ~~ m'- e:=> E '" o~ ~~ ~ c: -e~ \.: c: .S!~ a. Cl. c: .!!l .S! '" t::~ e:_ 0 .9-" t: Qi ~e lP~ -<:: => a. 21 a. -", .S1.g ~ 0 §a. .0'- (j r= Cia. §e E '" a 0 '" a 0 E

Figure 14. The 1115N values for different categories oforganisms sampled. Values show an increasing trend with taxon. Points represent mean values while the error bars represent one standard deviation from the mean.

143 0. bartramii mantle length (mm) O. battramii mantle length (mm)

529 530 531 532 280 300 320 340 360 380 15 15 <> Blood 1999 • Mantle 1999 no mantra". 14 14 - z z ! 13 '" '"<0 ~ <0 13 II I ! • 12 <> Blood 19981 12 I '------'- 11 15 15

! 14

14 - 13 z II 11JJ '"<0 I 1 12 13 1 I I 1 11 <> Blood 2000 <> Blood 2001 II • Mantle 2000 • Mantle 2001 12 10 310 320 330 340 360 360 370 320 330 340 350 360 370 380 390 400 O. bartramii mantle length (mm) O. battramii mantle length (mm)

Figure 15. The 1115N of blood and mantle samples from the same O. barlramii Individuals. All four years for which blood samples were taken are shown. The * denote the values for mantle and blood that came from the same animal where symhols are obscured.

144 S. Dualaniensis mantle length (mm)

279 140 160 160 200 220 240 260 260 -;------~----~---___j 12

• anomalous squid

<> 2000 Blood " 2000 Mantle • 10

9 ~ 11 8 rn! 7 I 6

!ll

<> 2001 Blood • 2001 ManUe 6+---.--,------'-.....- ...... ---'1 220 240 260 280 300 320 S. Dualaniensis mantle length (mm)

Figure 16. The alSN ofblood and mantle samples from Ibe same S. oualanumsis individuals. All Ibree years for which blood samples were taken are shown. The * denote the values for mantle and blood that came from the same animal where symbols are obscured.

145 CHAPTER III: INDIVIDUAL ISOTOPIC ANALYSES 146

INTRODUCTION 146 MATERIALS AND METHODS 148

RESULTS 148 Ommastrephes bartramii ,ysN(adults) 148 Ommastrephes bartramii ;PN (northern samples) 150 Ommastrephes bartramii l/SNjuveniles 151

Sthenoteuthis oualaniensis (515N (adults) 152

DISCUSSION 154 Ommastrephes bartramii ?pN (adults) 154

Ommastrephes bartramii (515N (Northern samples, sub-adults) 156

Ommastrephes bartramii (515N (juveniles) 156

Sthenoteuthis oualaniensis (515N 157

CONCLUSIONS 158

REFERENCES '" 160

LIST OF FIGURES 163

Chapter III: Individual isotopic analyses

Introduction

The differential fractionation of OI5N and Ol3C by different tissues has been recognized since the inception of stable isotopic techniques in ecology (DeNiro and Epstein, 1981). More recently, researchers have begun to experimentally quantify tissue dependent differences both in scope and relative timeframe of signal incorporation (turnover time) (Hesslein et at., 1993). Mizutani et at. (1991,1992) assessed the enrichment ofnon-metabolic structures (feathers) in a variety ofcaptive birds to determine the efficacy of using these structures as indicators of past feeding during the time periods of feather formation. They found that the feathers did not show an age dependent change, and that values had slight species dependent enrichments. Hobson

146 (1993) attempted to use liver, muscle and bone collagen to determine short, intermediate, and long-term trophic levels of seven species of seabirds. Each tissue was fairly effective in separating the birds by trophic level corresponding to known feeding habits. The isotopic signals from baleen of migratory whales have also been used in attempts to determine past feeding history (Best and Schell, 1996; Hobson and Schell, 1998). Both 13 studies noted substantial oscillations in 1) C and 1)15N with baleen length (age), one in soutbern right whales (Eubalaena australis) and the other in Arctic bowhead whales (Balaena mysticetus). Neither Best and Schell (1996) nor Hobson and Schell (1998) could isolate a single cause of the isotopic isolations and Hobson and Schell (1998) theorized that the shifts were due to either an annual movement between areas of different baseline 1)15N signal, an annual shift in one third of a trophic level, or that the animals underwent seasonal fasting which resulted in enrichment via protein catabolism. Retrospective isotopic analyses should reveal much about individual feeding history, and the extent to which individuals within the same speices may differ from each other. Individual variation in recent feeding history has been shown by isotopic studies. For example, Beaudoin et al. (1999) looked at the individual feeding specialization of the northern pike (Esox lucius) in five lakes in northern Alberta Canada. Both the stomach contents analysis (SCA) and the SIA revealed considerable within population variation, with some individuals differing by up to two TLs within a lake population. Beaudoin et al. (1999) found that specialization on fish versus benthic invertebrates between individual pike in certain lakes was not only a long-term trait (not ephemeral random feeding events) but depended on the size of the pike and the presence of other fish assemblages between lake systems; indicating size dependent trophic shifts in response to a competitive void. Vander Zanden et al. (2000) however found that size of lake trout (Salvelinus namaycush) and individual specialization did not account for the majority of isotopic variation in the species but that the majority of variation occurred between populations of different lakes. Clearly SIA based investigation of trophic interactions can sample tissues that hold historical records. These historical records can be used to determine historical feeding and to investigate the isotopic variation that occurs within and between individuals of a population. In chapter II some overall trends emerged for both species of

147 squids based on composites ofmany individuals. The general trends were most likely controlled by the animals' trophic position however substantial variation within the trends was observed. The focus of this chapter is to make a more detailed retrospective analysis of eye lens and gladius material to address the individual variation within the more general composite trends observed.

Materials and Methods

See chapter IT for details on the materials and methods.

Results

Ommastrephes bartramii 0/5N (adults)

2-14115-99 #14 F (570):

The (i15N of the lens segments of this squid ranged from 3.9 to 16.8%0; the mantle value was 12.8%0 and the blood was 12.2%0. The lens values increased from 159mm ML up to 372mm ML and then stayed relatively stable until after 458 mm ML where the last value decreased to 13.7%0 (Fig. 1).

2-16/17-99 # 5 F (446):

The (i15N of the lens segments of this squid ranged from 3.7 to 16.6%0; the mantle value was 12.2%0 and the blood was 11.6%0. The lens values decreased once after the first point to their lowest value at 128mm ML then increased fairly consistently (with one peak of 14.3%0 at 245mm ML) to 16.6%0 at 304mm.ML and then stayed relatively stable until after 363mm ML where the last value decreased to 13.3%0 (Fig. 2).

2-7/8-00 #1 M (333):

The (i15N of the lens segments of this squid ranged from 4.2 to 12.7%0; the mantle value was 12.1%0. The lens values increased immediately from the first point at 125mm

148 ML (4.2%0) to the 270mm ML (12.7%0) and then decreased to the last point at 333mm ML (12.1%0) (Fig. 3).

2-8/9·00 #1 M (318): The Sl5N ofthe lens segments ranged from 4.6 to 14.5%0; the mantle value was 13.0%0. The lens values decreased after the initial point at 107mm ML (4.9%0) and then increased steadily to 225mm ML (12.9%0), after that the lens values continued to increase at a slower rate culminating at 318mm ML (14.5%0)(Fig. 4).

2-819-00 #3 M (296): The 815N of the lens segments ranged from 4.8 to 10.9%0; the mantle value was 14.4%0. The lens values stayed relatively constant from 96mm ML (4.8%0) until 150mm ML (5.4%0), then increased to 221mm ML (10.9%0) then decreased to 296mm ML (10.0%0) (Fig. 5).

2-8/9-00 #4 M (354):

The 1)15N of the lens segments ranged from 6.5 tol1.7%0; the mantle value was 14.4%0 the blood value was 13.1%0. The lens values stayed relatively constant until 141 mm ML (6.5%0) and increased to 209mm ML (10.0%0) where the values continued to increase at a slower rate up to 354mm ML (11.7%0) (Fig. 6).

2-9/10-00 #4 (540): The S15N of the lens values ranged from 4.2 to 14.5%0; the mantle value was 13.7%0 and the blood was 12.5%0. The lens values showed a slight initial increase to 238 mm ML (5.3%0) followed by a rapid increase and plateau at 296mm ML (11.4%0) to 343mm ML with another increase and plateau at 430mm ML to 517mm ML (14.5%0) followed by a final decrease at 540mm ML (12.5%0) (Fig. 7).

2-9110-00 #5 (360):

149 The 015N of the lens values ranged from 6.2 to 12.5%0; the mantle value was 12.9%0 and the blood was 12.9%0. The lens values showed an initial decrease to 134mm ML (6.2%0) and then increased steadily, with a peak at 206mm ML (9.0%0), up to 256mm ML (12.5%0) followed by a general decrease to 360mm ML (11.1%0) (Fig. 8).

2-11112-01 #2 (370): The Ol5N of the lens values ranged from 5.1 to 13.7%0; the mantle value was 12.8%0 and the blood was 12.6%0. The lens showed an initial plateau to 157mm ML (5.3%0), a general increase to 273mm ML (13.7%0) followed by a final decrease to 370mm ML (12.2%0) (Fig. 9).

2-11112-01 #5 (534): The o15N values of the lens ranged from 4.2 to 15.8%0; the mantle value was 16.8%0. The lens showed an overall increase from 140mm ML (4.2%0) to 534mm ML (15.8%0) with a dip in the trend at 357mm ML (11.2%0) (Fig. 10).

2-11112-01#1 (557):

The 015N of the lens values ranged from 7.4 to 10.2%0; the mantle value was 14.4%0 and the blood was 10.8%0. The lens values showed an initial decrease to 172 (7.4%0) then generally increased to 274mm ML (10.2%0) followed by a general plateau (with fluctuations) to 557 mm ML (9.8%0) (Fig. 11).

Ommastrephes bartramii 015N (northern samples)

#21 (220):

The 015N ofthe lens values ranged from 4.76 to 7.30%0, and the mantle value was 10.20%0. The lens showed an initial decrease from 130mm ML (5.82%0) to 173mm ML (4.76%0) followed by an increase to 220mm ML (7.30%0) (Fig. 12).

#26 (201):

150 The l)15N of the lens values ranged from 5.03 to 7.11%0, and the mantle value was 10.22%0. The lens showed an increase from 117 mm ML (5.03%0) to 20lmm ML (7.11%0) (Fig. 13).

#28 (255):

The l)15N of the lens values ranged from 5.51 to 7.47%0 I and the mantle value was 10.69%0. The lens showed an initial plateau from 144mm ML (5.51%0) to 193mm ML (5.86%0) followed by an increase to 255mm ML (7.47%0) (Fig. 14).

15 Ommastrephes bartramii l) N juveniles

2-16117-98 #25 (142): The 815N of the lens values ranged from 5.14 to 5.44%0, and the mantle value was 6.33%0. The lens showed a slight increase from 94mm ML (5.14%0) to 149mm ML (5.44%0)(Fig. 15). 2-16/17-98 #26 (151):

The l)15N of the lens values ranged from 4.88 to 5.27%0, and the mantle value was 6.36%0. The lens showed fairly constant values from 96mm ML (5.27%0) to 151mm ML (5.05%0) (Fig. 16).

2-11112·99 #5 (161):

The l)15N of the lens values ranged from 4.84 to 7.07%0, and the mantle value was 8.01%0. The lens showed fairly constant values from 94mm ML (5.14%0) to 142mm ML (5.11%0}followed by a final increase to 161mm ML (7.07%0)(Fig. 17).

2-11112·99 #6 (150):

The l) 15N of the lens values ranged from 5.23 to 6.20%0, and the mantle value was 7.57%0. The lens showed the highest value from the initial sample at 82mm ML (6.2%0) and showed an overall oscillatory decline to 150mm ML (5.330/00) (Fig. 18).

151 Sthenoteuthis oualaniensis ,yi5N (adults)

2-14/15-99 #1 (324):

The 1)lsN values of the lens ranged from 4.24 to 10.00%0, and the mantle value was 11.14%0. The lens showed a slight increase from 129mm ML (4.24%0) to 172mm ML (4.7%0) followed by a rapid increase to 286mm ML (10.00%0) and a small decrease to 324mm ML (9.83%0) (Fig. 19).

2-14/15-99 #3 (200):

The 1)15N values of the lens ranged from 4.20 to 9.04%0, and the mantle value was 10.46%0. The lens showed a decrease from 7Imm ML (4.95%0) to a plateau between 95mm ML (4.20%0) and 109mm ML (4.22%0) followed by a rapid increase to 200mm ML (9.04%0) (Fig. 20).

2·14/15-99 #10 (223):

The 1)15N values of the lens ranged from 5.36to 9.75%0, and the mantle value was 11.24%0. The lens values remained relatively constant from 77mm ML (5.55%0) to 129mm ML (5.71%0), with one peak at 112mm ML (5.90), followed by an increase to 223mm ML (9.75%0) (Fig. 21).

2-7/8-00 #1 (210):

The 1)lsN values of the lens ranged from 3.56 to 5.85%0, and the mantle value was 8.95%0. The lens values decreased initially from 105mm ML (3.87%0) to 118mm ML (3.56%0) and then showed an overall increase to 210mm ML (5.85%0), with one peak at 166mm ML (5.22%0) (Fig. 22).

2-12/13-00 #1 (209):

152 The (j15N values of the lens ranged from 4.31 to 6.66%0, and the mantle value was 7.92%0. The lens values remained fairly constant from 90mm ML (4.44%0) to 115mm ML (4.44%0) and then showed an overall increase to 209mm ML (6.66%0) (Fig. 23).

2-12113-00 #2 (160):

The (j15N values of the lens ranged from 4.18 to 8.86%0, and the mantle value was 9.20%0. The lens values initially decreased from 78mm ML (4.62%0) to 84mm ML (4.18%0) then increased slightly to 111mm ML (4.87%0) followed by a rapid increase to 160mm ML (8.96%0) (Fig. 24).

2-12113-00 #3 (181):

The (j15N values of the lens ranged from 11.35 to 12.59%0, the mantle value was 11.26%0, and blood was 9.44%0. The lens values initially decreased from 94mm ML (11.99%0) to 115mm ML (11.35%0) then increased slightly to 168mm ML (12.59%0) and remained constant to 181mm ML (12.53%0) (Fig. 25).

2-12113-00 #4 (173):

The (j15N values of the lens ranged from 4.56 to 8.50%0, the mantle value was 10.15%0. The lens values initially decreased from 91mm ML (5.12%0) to 110mm ML (4.56%0) then increased to 173mm ML (8.50%0) (Fig. 26).

2-12/13-00 #6 (158): 15 The (j N values of the lens ranged from 3.72 to 5.32%0, the mantle value was 7.74%0. The lens values increased from 85mm ML (3.72%0) to 158mm ML (5.32%0) with a dip at 123mm ML (4.33%0) (Fig. 27).

2-13/14-00 #3 (251):

The (j15N values of the lens ranged from 4.20 to 6.21%0, the mantle value was 8.07%0 and blood was 7.68%0. The lens values increased from I11mm ML (4.20%0) to 251mm ML (6.20%0) with a peak at 166mm ML (4.80%0) and 221mm ML (6.21%0) (Fig. 28).

153 2-12/13-01 #1 (225):

The 1)15N values of the lens ranged from 2.79 to 9.40%0, the mantle value was 8.48%0 and blood was 7.71%0. The lens values increased from 108mm ML (2.79%0) to 225mm ML (9.40%0) (Fig. 29).

2-12/13-01 #11 (304):

The i)15N values of the lens ranged from 3.35 to 9.65%0, the mantle value was 10.08%0 and blood was 9.20%0. The lens values showed a slight general increase from 98mm ML (3.35%0) to 159mm ML (3.82%0), followed by a rapid increase to 304mm ML (9.65%0) (Fig. 30).

2-12/13-01 #12 (302):

The i)15N values of the lens ranged from 3.84 to 8.80%0, the mantle value was 9.41%0. The lens values remained relatively constant from 101mm ML (4.00%0) to 161mm ML (3.99%0), followed by a rapid increase to 302mm ML (8.80%0) (Fig. 31).

2-12/13-01 #16 (308):

The 1)15N values of the lens ranged from 3.08 to 9.78%0, the mantle value was 9.92%0 and blood was 9.23%0. The lens values decreased initially from 95rom ML (3.35%0) to 112mm ML (3.08%0), followed by an overall increase to 308mm ML (9.78%0) (Fig. 32).

Discussion

Ommastrephes bartramii ;/5N (adults)

The retrospective i)15N values of the eleven adult O. bartramii eyes sampled, ten showed a more or less sigmoidal relationship (with some variation) ofeye lens segment i)15N with estimated mantle length (Figs. 1-10). The earliest values attainable were in the eye region corresponding to 100mm ML and were in the vicinity of 4.0-6.0%0. The eye

154 lens values showed a rapid increase, mostly between ML of 175 and 300. While individual variation occurred in size at which ascent began, along with the steepness and smoothness of the ascent, the overall patterns were the same. This perhaps is not surprising since the curve placed through the combined eye lens data was very similar to the data taken for mantle muscle of different sized individuals as seen in the previous chapter.

Although most squids showed the rapid increase in eye lens segment /) 15N at around 175mm ML some squids deviated from this trend. Squids 2-8/9-00 #3, 2-9/10-00

#5 showed rapid increases in eye lens segment /) 15N earlier than the other squids, at or before 150mm ML (Figs. 5, 8). Squid 2-8/9-00 #3 had a typical sigmoidal pattern but it showed a plateau at an earlier size than typical O. bartramii, roughly 225mm ML and at a

lower /) 15N value (Fig. 5). Squid 2-9/10-00 #5 did not show a typical smooth sigmoidal pattern but had a number of deviations within the overall curve (Fig. 8). Squid 2-11/12-01 #5 also did not show a typical smooth sigmoidal pattern but increased rapidly and then showed a dip around 350mm ML before rising to follow its previous trajectory (Fig. 10).

Squid 2-9/10-00 #4 began its rapid increase in eye lens /)15N later than other squids, close . to 250mm ML, then showed an initial plateau followed by an increase and another

plateau (Fig. 7). Squid 2-11/12-01 #1 had the most deviant eye lens segment /)15N pattern (Fig. 11). This eye only showed a slight increase from 175mm ML to roughly 300mm

ML and then remained relatively constant (with slight variations) at low /)15N values to the final segment (Fig. 11). Although the initial value ("" 8%0 at roughly 150mm ML) seems high this is presumably due to the lack ofdata corresponding to smaller sizes.

Six squids each show a decrease of /)15N in their final (i.e.- outer) eye lens segment (Figs. 1, 2, 3, 7, 8, 9). The decrease is usually small ("" 1%0) except in two of the females (Figs. 1, 2) where the decrease to the outer segment was closer to 3%0. The larger decreases however, followed unusually high values. All of the decreases occurred while the squids were at or south of the subtropical front on their return migration. In three of these cases (Figs. 1,2,9) the eye lens segments get progressively lighter and have

endpoints close to the muscle and blood /) 15N values which may indicate that the eye, muscle and blood have fully accommodated to values in the new ecosystem. These three

squids (Figs. 1,2,9) all showed decreases in eye lens segment /)15N over the final 100 to

155 150mm ML and, assuming a 2mm growth per day, would indicate that these squids have been in the area for 50 to 75 days. Two squids did not decrease toward their final eye lens segments, although a plateau in the data had essentially been reached (Figs. 4, 6). These squid may be recent arrivals to the area where they were caught although only in one (Fig. 6) did the end points differ strongly from the muscle. The low and nearly flat curve of squid 2-11/12-01 #1 (Fig. I I) recalls a similar but very high eye-lens curve of S. oualaniensis (see below) and may indicate that, occasionally, some unaccounted for source ofbias interferes with accurate measurement of ,,15N values. The eleven curves show strong individual variation in the size and rate at which squid climb the trophic pyramid.

Ommastrephes bartramii (515N (Northern samples, sub-adults)

The final eye lens segments of all three northern squids have similar ,,15N values (Figs. 12-14) in agreement with the constancy ofmantle values seen in Chapter 2 for all squids taken at this locality. The variable immediate history (outer eye segment values) of these squid suggests that the constancy of the mantle ,,15N values was due to the locality (and its prey population), and that the squid were in transit through this area. Therefore, regardless of size, these squids incorporated the signal of the area they passed through which was fairly uniform over a wide range of sizes. The mantle values are high relative to final lens values and may indicate that lens values take longer to incorporate the signal.

Ommastrephes bartramii (515N (juveniles)

Three of the four juvenile O. bartramii eye lens segments did not show increases 15 in ,, N with mantle length, and all four had muscle ,,15N values that were greater than the outer eye-lens segment of each squid (Fig. 15-18). The lack of increase of ,, 15N is consistent with the initial plateaus seen in the initial portions of the eye lens curves of many of the adult O. bartramii. The higher ,,15N values of mantle muscle relative to outer eye-lens segments in these small squids is consistent with the possibility that the outer lens segments lag behind the mantle during periods of rapid change.

156 Sthenoteuthis oualaniensis 015N

S. oualaniensis does not undergo any known seasonal migration, and no geographical patterns in 015N were seen in the POM material during this study. The retrospective analysis of Ol5N in most S. oualaniensis sampled showed an exponential increase in eye-lens segment Ol5N with calculated ML (Figs. 20,21,23,24,26-32). Some individuals showed a steep and sudden increase in Ol5N with size (Figs. 20, 24, 26, 29). Other individuals exhibited a more gradual increase in Ol5N with size (Figs. 21, 30, 31, 32). The most gradual increase in eye lens segment Ol5N with size was shown by three caught in 2000 (Figs. 23, 27, 28). Three individuals showed marked deviations from this exponential increase in eye lens segment 015N with size, 2-14/15-99 #1,2-7/8-00 #1, and 2-12/13-00 #3 (Figs. 19,22,25). Squid 2-7/8-00 #1 (Fig. 22) shows on overall exponential increase, however there is a large peak in Ol5N for the eye lens at 166mm ML that is well above the general trend for this individual. Squid 2-14/15-99 #1 (Fig. 19) also showed an overall exponential increase, however the presence of an initial plateau, and a final plateau formed by the outer eye segment makes this retrospective eye lens curve resemble those obtained for many of the O. bartramii adults. Because squid 2-14/15-99 #1 (Fig. 19) was the largest S. oualaniensis caught (324mm ML) and overlaps the size range of many O. bartramii adults, somewhere above 300mm ML these squids may reach a size dependent Ol5 N plateau as seen in the O. bartramii Ol5N versus ML data, but not in the S. oualaniensis data. Another four S. oualaniensis (2-14/15-99 #3, 2-12113-00 #1,2­ 12/13-00 #4, and 2-13/14-00 #3) had patterns of eye lens segment Ol5N increase with calculated size that also showed sigmoidal attributes (Figs. 20, 23, 26, 28) but these individual curves still fit the exponential model better. A damped sigmoidal pattern could be typical for this species but it did not show up on the Ol5N versus ML data because the small sample size of squids> 300mm ML, and variability in the timing (size of ML) and extent of increase, obscured the pattern. The most bizarre pattern seen is from squid 2-12/13-00 #3 (Fig. 25) with all the retrospective values uniformly high. The muscle ol5N value was comparable to the eye and both were as high as any mantle or eye value obtained at any size. The blood value for squid 2-12/13-00 #3 also corresponds to values of muscle and blood found for other S. oualaniensis of this size. Although, 5 of the 9 squid (including paralarvae) from this same

157 station had unusually high values, no satisfactory explanation has been found for the trends observed in squid 2-12113-00 #3 (Fig. 25). It could be possible that this animal was under extreme nutrient stress throughout its lifespan, causing elevated 015N values due to catabolization of its own tissues, or that it had some biochemical deficiency that caused increased fractionation when forming its tissues.

Conclusions

The overall increase in 015N with size seen, both within a single individual and between individuals, for O. bartramii and S. oualaniensis is most likely controlled by ontogenetic trophic level increases. Although a geographical isotopic signal from a different ecosystem would have an effect on O. bartramii values, the increase in eye lens segment 015N with size for S. oualaniensis, a geographical non-migrator, emphasizes the importance of size alone. The hypothesis that increased 015N corresponds to increased size is also supported by the eye lens segment 015N comparisons between juvenile, northern sub-adult, and adult O. bartramii. The decrease of Ol5N in the outer-most eye lens segments of many O. bartramii may indicate a geographical Ol5N signal corresponding to the animals return to Hawaiian waters. Support for a geographical signal in O. bartramii is obtained from the largest S. oualaniensis captured. This large S. oualaniensis showed a plateau in the 015N eye lens at ca. 20/00 less than the median value for a similar sized O. bartramii. This difference may indicate a slight convergence of O. bartramii values towards those of S. oualaniensis as O. bartramii enters Hawaiian waters. Presumably the eye lens retrospective data mostly reflect a combination of an individual squids growth rate, feeding regime, and feeding behavior. Since trophic level and 015N values are, related to the size of the squid, then the individual growth rate of the squid could have an effect on the shape of the curve from the retrospective eye-lens data. Food quality (which will affect growth rate of the predator) can also have an effect on the eye-lens data. Fantle et al. (1996) found that blue crab, Callinectes sapidus, showed rapid growth with little 015N fractionation (0.1%0) when fed a high protein diet, and slow growth with substantial fractionation (up to 3.1%0) when fed low protein diets. Fantle et

158 al. (1996) postulated that the protein poor diets did not supply enough energy and the crabs catabolized their own tissues to meet metabolic requirements. Adams and Sterner (2000) also found an effect of dietary nitrogen content on 1)15N enrichment in daphnids. 15 They found that 1) N enrichment in Daphnia magna was linearly related to the C:N ratio of the food, and that enrichment ranged from 0 to 6%0. Finally, individual feeding specialization can also have a marked effect on 1)15N variation between individuals within a population, (Beaudoin et al., 1999). Age data, which can be determined by growth rings in the squid statoliths, could be beneficial in understanding retrospective 1)15N analyses. Anomalously old or young squids (very different times of hatching from the norm) would have different growth rates that could explain peculiar 1)15N retrospective signals. 15 There are several reasons for confidence that the major features of the <5 N data 15 provide useful and robust data: A general size relationship of increasing <5 N with increasing size was observed across several taxa. Consistent relationships between 1)15N 15 and ML were seen within both O. bartramii and S. oualaniensis. The <5 N data from the mantles and the eye-lens segments showed similar logistic patterns versus ML in 0. 1s bartramii. The <5 N data from the mantles, eye-lens segments, and blood showed similar 1s exponential patterns versus ML in S. oualaniensis. The <5 N data from the eye-lens 15 segments of individual adult O. bartramii fit strongly to logistic curves. The <5 N data from the eye-lens segments of individual adult and sub-adult S. oualaniensis fit strongly to exponential curves. With a few exceptions, the 1)15N patterns seem to reflect the feeding history of the squids. The exceptions, however, suggest that caution must be used in the interpretation of all of the data. The interactions between trophic level, feeding behavior, nutrient stress, tissue fractionation, and geographical variation on the stable isotopic signatures of an animal are still poorly understood. Separating and quantifying the effects of these factors requires controlled experimental manipulations. Such experiments should enable increased usage of stable isotopes for understanding a wide array of ecological questions.

159 References

Adams, T. S., and R. W. Sterner (2000). "The effect of dietary nitrogen content on trophic level 15N enrichment." Limnology and Oceanography 45(3): 601-607.

Altabet, M. A. (1989). "A time-series study ofthe vertical structure ofnitrogen and particulate dynamics in the Sargasso Sea." Limnology and Oceanography 34(7): 1181-1201.

Beaudoin, C. P., W. M. Tonn, E. E. Prepas, and L. I. Wassenaar (1999). "Individual specialization and trophic adaptability of northern pike (Esox lucius): an isotope and dietary analysis." Oecologia 120: 386-396.

Best, P. B., and D. M. Schell (1996). "Stable isotopes in southern right whale (Eubalaena australis) baleen as indicators of seasonal movements, feeding and growth." Marine Biology 124: 483-494.

DeNiro, M. J., and S. Epstein (1981). "Influence of diet on the distribution of nitrogen isotopes in animals." Geochmica et Cosmochimica Acta 45: 341-351.

Fantle, M. S., A. I. Dittel, S. M. Schwalm, C. E. Epifanio, and M. L. Fogel (1999). "A food web analysis of the juvenile blue crab, Callinectes sapidus, using stable isotopes in whole animals and individual amino acids." Oecologia 120: 416-426.

Gould, P., P. Ostrom, and W. Walker (1997). "Trophic relationships of albatrosses associated with squid and large-mesh drift-net fisheries in the North Pacific Ocean." Canadian Journal of Zoology, 75: 549-562.

Hesslein, R. H., K. A. Hallard, and P. Ramlal (1993). "Replacement of sulfur, carbon, and nitrogen in tissue of growing broad whitefish (Coregonus nasus) in response

160 to a change in diet traced by S34S, Sl3C, and S15N." Canadian Journal of Fisheries and Aquatic Sciences 50: 1993.

Hobson, K. A., and D. M. Schell (1998). "Stable carbon and nitrogen isotope patterns in baleen from eastern Arctic bowhead whales (Balaena mysticetus)." Canadian Journal ofFisheries and Aquatic Sciences 55: 2601-2607.

Hobson, K. A., R. T. Alisauskas, and R. G. Clark (1993). "Stable-nitrogen isotope enrichment in avian tissues due to fasting and nutritional stress: implications for isotopic analysis of diet." Condor 95: 388-394.

Kline, T. C. (1997). Confirming forage fish food web dependencies in Prince William Sound using natural stable isotope tracers. Forage Fishes in Marine Ecosystems, Alaska Sea Grant College Program.

Kline, T. C., W. J. Wilson, and J. J. Goering (1998). "Natural isotope indicators of fish migration at Prudhoe Bay, Alaska." Canadian Journal of Fisheries and Aquatic Sciences 55: 1494-1502.

Liu, K., and 1. R. Kaplan (1989). "The eastern tropic! Pacific as a source of 15N-enriched nitrate in seawater off southern California." Limnology and Oceanography 34(5): 820-830.

Mingawa, M. and E. Wada (1986). "Nitrogen isotope ratios ofred tide organisms in the East China Sea: A characterization of biological nitrogen fixation." Marine Chemistry 19(3): 245-259.

Miyake, Y. and E. Wada (1967). "The abundance ratio of 15N/14N in the marine environments." Recent Oceanographic Works from Japan 9: 32-53.

161 Mizutani, H., M. Fukuda, and Y. Kabaya (1992). "BC and 15N enrichment factors of feathers of 11 species of adult birds." Ecology 73(4): 1391-1395.

Mizutani, H., Y. Kabaya, and E. Wada (1991). "Nitrogen and carbon isotope compositions relate linearly in cormorant tissues and its diet." Isotopenpraxis 27: 166-168.

Shchetinnikov, A. S. (1992). "Feeding spectrum of squid Sthelloteuthis oualalliellsis (Oegopsida) in the Eastern Pacific." Journal of the Marine Biological Association ofthe United Kingdom 72: 849-860.

Takai, N., S. Onaka, Y. Ikeda, A. Yatsu, H. Kidokoro, and W. Sakamoto (2000). "Geographical variations in carbon and nitrogen stable isotope ratios in squid." Journal of the Marine Biological Association of the United Kingdom 80: 675-684.

Vander-Zanden, M. J., B. 1. Shuter, N. P. Lester, and J. B. Rasmussen (2000). "Within­ and among- population variation in the trophic position of a pelagic predator, lake trout (Salvelillus Ilamaycush)." Canadian Journal of Fisheries and Aquatic Sciences 57: 725-731.

Wada, E. and A. Hattori (1976). "Natural abundance of 15N in particulate organic matter in the North Pacific Ocean." Geochmica et Cosmochimica Acta 40: 249-251.

Wada, E. and A. Hattori (1991). Nitrogen in the sea: forms, abundances, and rate processes. Boca Raton: CRC Press.

162 List of Figures

FIGURE 1. /'; 15N VERSUS LENGTH OF EYE LENS SEGMENTS OF A FEMALE O. BARTRAMII. /'; 15N

OF MANTLE MUSCLE AND BLOOD PLOTTED AGAINST MANTLE LENGTH AT TIME OF

CAPTURE. MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OFEYE LENS RADWS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGTH NOTED IN BOTTOM RIGHT CORNER...... •.....•... 169

FIGURE 2. /';15N VERSUS LENGTH OF EYE LENS SEGMENTS OF A FEMALE O. BARTRAMII. /'; 15N

OF MANTLE MUSCLE AND BLOOD PLOTTED AGAINST MANTLE LENGTH AT TIME OF

CAPTURE. MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OF EYE LENS RADWS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGTH NOTED IN BOTTOM RIGHT CORNER 170

FIGURE 3. /'; 15N VERSUS LENGTH OF EYE LENS SEGMENTS OFA MALE O. BARTRAMl/. /'; 15N OF

MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE. MANTLE

LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE LENS

RADWS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER. 171

FIGURE 4. /';15N VERSUS LENGTH OF EYE LENS SEGMENTS OF A MALE O. BARTRAMII. /';15N OF

MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE. MANTLE

LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE LENS

RADllIS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER 172

FIGURE 5. /'; 15N VERSUS LENGTH OF EYE LENS SEGMENTS OF A MALE O. BARTRAMII. /'; 15N OF

MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE. MANTLE

LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE LENS

RADWS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER 173 15 FIGURE 6. /'; N VERSUS LENGTH OFEYE LENS SEGMENTS OF A MALE O. BARTRAMII. /'; 15N OF

MANTLE MUSCLE AND BLOOD PLOTTED AGAINST MANTLE LENGTH AT TIME OF

CAPTURE. MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

163 OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGTH NOTED IN BOTTOM RIGHT CORNER 174 15 FIGURE 7. /!, N VERSUS LENGTH OF EYE LENS SEGMENTS OF A FEMALE O. BARTRAMll. /!,15N

OF MANTLE MUSCLE AND BWOD PLOTTED AGAINST MANTLE LENGTH AT TIME OF

CAPTURE. MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGTH NOTED IN BOTTOM RIGHT CORNER '" ...... •...... 175 15 FIGURE 8. /!, N VERSUS LENGTH OF EYE LENS SEGMENTS OFA MALE O. BARTRAMlI. /!,15N OF

MANTLE MUSCLE AND BLOOD PLOTTED AGAINST MANTLE LENGTH AT TIME OF

CAPTURE. MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGTH NOTED IN BOTTOM RIGHT CORNER 176

FIGURE 9. /!, 15N VERSUS LENGlli OFEYE LENS SEGMENTS OF A MALE 0. BARTRAMll. /!,15N OF

MANTLE MUSCLE AND BLOOD PLOTTED AGAINST MANTLE LENGlli AT TIME OF

CAPTURE. MANTLE LENGlli FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGlli NOTED IN BOTTOM RIGHT CORNER 177

FIGURE 10. /!,15N VERSUS LENGlli OF EYE LENS SEGMENTS OF A FEMALE O. BARTRAMIl. /!,15 N

OF MANTLE MUSCLE AND BWOD PLOTTED AGAINST MANTLE LENGlli AT TIME OF

CAPTURE. MANTLE LENGlli FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGlli NOTED IN BOTTOM RIGHT CORNER 178

FIGURE 11. /!,15N VERSUS LENGlli OF EYE LENS SEGMENTS OF A FEMALE O. BARTRAMIl. /!,15 N

OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGlli AT TIME OF CAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGlli

NOTED IN BOTTOM RIGHT CORNER 179

FIGURE 12. /!,15 N VERSUS LENGTH OF EYE LENS SEGMENTS OF A MALE NORTHERN O.

BARTRAMll. /!,15N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGlli AT TIME OF

CAPTURE. MANTLE LENGlli FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

164 OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTI.E

LENGTH NOTED IN BOTTOM RIGHT CORNER '" 180

FIGURE 13. ",15N VERSUS LENGTH OF EYE LENS SEGMENTS OF A FEMALE NORTHERN O.

BARTRAMlI. ",15N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF

CAPTURE. MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGTH NOTED IN BOTTOM RIGHT CORNER 181

FIGURE 14. ",15N VERSUS LENGTH OF EYE LENS SEGMENTS OF A FEMALE NORTHERN O.

BARTRAMlI. '"15N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF

CAPTURE. MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGTH NOTED IN BOTTOM RIGHT CORNER...... ••....•.•...... •...... 182

FIGURE 15. '"15N VERSUS LENGTH OF EYE LENS SEGMENTS OF A JUVENILE O. BARTRAMlI.

'"15N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTI.E LENGTH

NOTED IN BOTTOM RIGHT CORNER 183

FIGURE 16. '" 15N VERSUS LENGTH OF EYE LENS SEGMENTS OF A JUVENILE O. BARTRAMlI.

'"15N OF MANTI.E MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OFEYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER 184

FIGURE 17. '" 15N VERSUS LENGTH OFEYE LENS SEGMENTS OF A JUVENILE O. BARTRAMlI.

'"15N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OFCAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OFEYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTI.E LENGTH

NOTED IN BOTTOM RIGHT CORNER 185

FIGURE 18. '" 15N VERSUS LENGTH OF EYE LENS SEGMENTS OF A JUVENILE O. BARTRAMlI.

'"15N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

165 LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER...... •...... 186 l5 FIGURE 19. d N VERSUS LENGTH OFEYE LENS SEGMENTS OF A FEMALE S. OUALANIENSIS.

d l5N OF MANTLE MUSCLE PWTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER 187

FIGURE 20. d l5N VERSUS LENGTH OF EYE LENS SEGMENTS OF A FEMALE S. OUALANIENSIS.

f!"15N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER 188

FIGURE 21. f!"15N VERSUS LENGTH OF EYE LENS SEGMENTS OF A FEMALE S. OUALANIENSIS.

d 15N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OFCAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER 189

FIGURE 22. d l5N VERSUS LENGTH OF EYE LENS SEGMENTS OF A FEMALE S. OUALANIENSIS.

d l5N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER 190

FIGURE 23. f!" l5N VERSUS LENGTH OF EYE LENS SEGMENTS OF A FEMALE S. OUALANIENSIS.

d 15N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER 191

FIGURE 24. d l5N VERSUS LENGTH OF EYE LENS SEGMENTS OF A FEMALE S. OUALANIENSIS.

d 15N OF MANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

166 LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANILE LENGlH

NOTED IN BOTTOM RIGHT CORNER...... •...... •...... •...... •...... •....•...... 192 15 FIGURE 25. 6. N VERSUS LENGlH OF EYE LENS AND GLADIUS SEGMENTS OF A FEMALE S. 15 OUALANlENSIS. 6. N OF MANTLE MUSCLE AND BLOOD PLOITED AGAINST MANTLE

LENGlH AT TIME OF CAPTURE. MANTLE LENGlH FOR EYE LENS SEGMENTS ESTIMATED

FROM REGRESSIONS OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID

NUMBER, AND MANTLE LENGlH NOTED INBOTTOM RIGHT CORNER 193 ls FIGURE 26. 6. N VERSUS LENGlH OF EYE LENS AND GLADIUS SEGMENTS OF A FEMALE S. 1S OUALANIENSIS. 6. N OF MANTLE MUSCLE PLOITED AGAINST MANTLE LENGlH AT TIME

OF CAPTURE. MANTLE LENGlH FOR EYE LENS SEGMENTS ESTIMATED FROM

REGRESSIONS OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER,

AND MANILE LENGlH NOTED IN BOTTOM RIGHT CORNER 194

FIGURE 27. 6. lSN VERSUS LENGlH OF EYE LENS SEGMENTS OF A MALE S. OUALANIENSIS. 15 6. N OF MANTLE MUSCLE PLbITED AGAINST MANTLE LENGlH AT TIME OFCAPTURE.

MANTLE LENGlH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGlH

NOTED IN BOTTOM RIGHT CORNER 195

FIGURE 28. II lSN VERSUS LENGlH OF EYE LENS SEGMENTS OF A FEMALE S. OUALANIENSIS.

A15N OFMANTLE MUSCLE AND BWOD PLOTTED AGAINST MANTLE LENGlH AT TIME OF

CAPTURE. MANTLE LENGlH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGlH NOTED IN BOTTOM RIGHT CORNER 196

FIGURE 29. A15N VERSUS LENGlH OF EYE LENS SEGMENTS OF A FEMALE S. OUALANIENSIS.

A15N OF MANTLE MUSCLE AND BLOOD PLOITED AGAINST MANTLE LENGlH AT TIME OF

CAPTURE. MANTLE LENGlHFOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGlH NOTED IN BOTTOM RIGHT CORNER 197

FIGURE 30. A15N VERSUS LENGlH OF EYE LENS SEGMENTS OF A FEMALE S. OUALANIENSIS. 15 A N OF MANTLE MUSCLE AND BLOOD PLOITED AGAINST MANTLE LENGlH AT TIME OF

CAPTURE. MANTLE LENGlH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

167 OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGTH NOTED IN BOTTOM RIGHT CORNER...... •...... •...... •...... 198 15 FiGURE 31. t1 N VERSUS LENGTH OF EYE LENS SEGMENTS OFA FEMALE S. OUALANIENSIS. . 15 t1 N OFMANTLE MUSCLE PLOTTED AGAINST MANTLE LENGTH AT TIME OF CAPTURE.

MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS OF EYE

LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE LENGTH

NOTED IN BOTTOM RIGHT CORNER 199 15 FiGURE 32. t1 N VERSUS LENGTH OFEYE LENS SEGMENTS OF A FEMALE S. OUALANIENSIS. 15 t1 N OF MANTLE MUSCLE AND BLOOD PLOTTED AGAINST MANTLE LENGTH AT TIME OF

CAPTURE. MANTLE LENGTH FOR EYE LENS SEGMENTS ESTIMATED FROM REGRESSIONS

OF EYE LENS RADIUS AND SQUID SIZE. DATE CAUGHT, SQUID NUMBER, AND MANTLE

LENGTH NOTED IN BOTTOM RIGHT CORNER 200

168 16 --.0-- Eye Mantle <>• Blood 14

12 <>•

Z 10 '"Go 8

6

4 2-14/15-99 #1 (570mm) 2 0 100 200 300 400 500 600 O. banramii mantle length (mm)

Figure 1. l)1SN versus length ofeye lens segments ofa female O. bartramii. l)15N of mantle muscle and blood plotted against mantle length at time of capture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

169 16 ---- Eye Mantle • Blood 14 <>

12 <>•

Z 10

~ '"<10 8

6

4 2-16/17-99 #5 (446) 2 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 2. olSN versus length of eye lens segments of a female O. banramii. o15N of mantle muscle and hlood plotted against mantle length at time of capture. Mantle length for eye lens segments estimated from regressions of eye lens radins and squid size. Date caught, squid nnmber, and mantle length noted in bottom right corner.

170 16 ----"- Eye • Mantle 14

12 •

Z 10 "'00 8

6

4

2-7/8-00 #1 (333mm) 2 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 3. a15N versus length of eye leus segments ofa male O. barlramii. a1'N ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions of eye leus radius and squid size. Date caught, squid number, and mantle leugth noted in bottom right corner.

171 16 -6- Eye • Mantle 14

12

Z 10 '"~oo

8

6

4 2-819-00 #1 (318mm) 2 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 4. liJ"N versus length ofeye lens segments ofa male O. barlramii. liJ5N ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estlmated from regressions of eye lens radius and squid size. Date caught, squid number, and mautle length noted in bottom right corner.

172 16 -- Eye • Mantle 14 •

12

Z 10 '"00 8

6

4 2-819-00 #3 (296mm) 2 50 100 150 200 250 300 350 O. bartramii mantle length (mm)

Figure 5.l)15N versus length of eye lens segments ofa male O. bartramii. l)15N of mantle muscle plotted against mantle length at time of capture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid uumber, and mantle length noted in bottom right corner.

173 16 ------Eye Mantle 0• Blood 14 • 0 12

Z 10 "'00 8

6

4 2·8/9·00 #4 (354mm) 2 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 6. ol'N versus length of eye lens segments of a male O. bartramii. o15Nofmantle muscle and blood plotted against mantle length at time of captnre. Mantle length for eye lens segments estimated from regressions of eye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

174 Figure 7. 81'N versus length ofeye lens segments of a female O. bartramii. 81'N of mantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius aud squid size. Date caught, squid number, and mantle length noted In bottom right corner.

175 16 --<>---- Eye Mantle 0• Blood 14

12

Z 10 '"<0 8

6

4 2-9/10-00 #5 (360mm) 2 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 8. 6lSN versus length of eye lens segments of a male O. bartramii. 615N of mantle muscle and blood plotted against mantle length at time of captnre. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

176 16 -- Eye Mantle <)• Blood 14

12

Z 10 "''"io 8

6

4 2-11/12-01 #2 (370mm) 2+----.,....------,-----,----,--'---,------/ o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 9. ol'N versus length ofeye lens segments ofa male O. bartramii. o15N ofmantle muscle and blood plotted against mantIe length at time of capture. MantIe length for eye lens segmeuts estimated from regressions of eye leus radius and squid size. Date caught, squid number, and mantIe length noted in bottom right corner.

177 18 --<0- Eye • 16 • Mantle

14

12

Z oJ) 10 -00

8

6

4 2-11/12-01 #5 (534mm) 2 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 10. 6lSN versus length ofeye lens segments of a female O. barlramii. 6lSN of mantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

178 16 -'>- Eye • Mantle <> Blood 14 •

12 <> Z 10 "'00 8

6

4 2-11/12-01 #1 (557mm)

2 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 11. lI15N versus leugth of eye leus segmeuts of a female O. bartramii. lI15N ofmantle muscle plotted against mantle leugth at time ofcapture. Mautle length for eye leus segmeuts estimated from regressions ofeye lens radius and sqnid size. Date caught, squid number, and mantle length noted in bottom right corner.

179 16 --->- Eye • Mantle 14 -

12

Z 10 • "'-00 8

6

4 Northern 0. bartramii #21 (220mm) 2+----r------,-----,---,----,------j o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 12. li1'N versus length of eye lens segments ofa male northern O. bartramii. li1'N of mantIe muscle plotted against mantle length at time of capture. MantIe length for eye lens segments estimated from regressions of eye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

180 16 ---<.- Eye • Mantle 14

12

Z 10 • '"-00 8

6

4 Northern O. bartramii #26 (201 mm)

. 2 , 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

15 15 Figure 13. (j N versus leugth ofeye lens segments of a female northern O. bartramii. (j N of mantIe muscle plotted against mantle length at time ofcapture. MantIe length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

181 16 --<-- Eye • Mantle 14

12 • .,Z 10 00 8

6

4 Northern 0. bartramii #28 (255mm) 2+----r-----,------,-----.----.------j o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 14. o15N versus leugth of eye lens segments of a female northern O. bartramii. o15N ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye leus segments estimated from regressious of eye lens radius and squid size. Date caught, squid number. and mantle length noted iu bottom right corner.

182 16 ---<>- Eye • Mantle

14 -

12

Z 10 '"00 8 -

6

4 2-16/17-98 #25 (149mm) 2 0 100 200 300 400 500 600 o. battramii mantle length (mm)

Figure 15. l)I'N versus leugth ofeye lens segments of a juvenile O. barlramii. l)15N of mantle muscle plotted against mantle length at time of capture. Mantle length for eye lens segmeuts estimated from regressious ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

183 16 --...- Eye • Mantle 14 -

12

",2 10 <:0 8

6 •

4 2-16/17-98 #26 (151mm) 2 +-----,-----.,...------r-,------,,~---~----__1 o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 16. a15N versus length of eye lens segments of a jnvenile O. bartramii. al5N ofmantle muscle plotted against mantle length at time of capture. Mantle length for eye lens segments estimated from regressions of eye lens radius and sqnid size. Date caught, squid number, and mantle length noted in bottom right corner.

184 16 --- Eye • Mantle 14 -

12 -

Z 10 "'~oo 8 - •

6 -

4 2-11/12·99 #5 (161mm) 2 , , 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

15 Figure 17. o N versus leugth of eye lens segments of a juvenile O. bartramii. 015N of mantle muscle plotted against mantle length at time of capture. Mantle length for eye lens segments estimated from regressions of eye lens radius and squid size. Date caught, squid number, aud mantle length noted in bottom right corner.

185 16 --4--- Eye • Blood

14 -

12

Z 10 - '"00 8 • 6

4 2-11/12-99 #6 (150mm)

2 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 18. /;lSN versus length ofeye lens segments ofa juvenile O. bartramii. /;lSN ofmautle muscle plotted against mautle length at time of capture. Mantle length for eye lens segments estin)ated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

186 --'-- Eye 12 • Manlle • 10

z 8 "'00

6

4

2-14/15-99 #1 (324mm) 2-1----.---,----,-----.------.------.-----1 o 50 100 150 200 250 300 350 S. ouaJaniensis mantle length (mm)

Figure 19. /l15N versus length ofeye leus segments ofa female S. ouafnniensis. /l'5N ofmantIe muscle plotted against mantIe length at time of capture. MantIe leugth for eye leus segments estimated from regressions of eye leus radius and squid size. Date caught, squid number, and mantle length noted in hottom right corner.

187 --<>- Eye 12 • Mantle

10 •

z 8 '"70

6

4

2-14/15-99 #3 (200mm) 2+------,----,------,r---,------,----,----1 o 50 100 150 200 250 300 350 S. oualaniensis mantle length (mm)

Figure 20. Ii15N versus length of eye lens segmeuts of a female S. oualllnknsis. Ii15N of mantle muscle plotted against mantle length at time of capture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

188 12 -- Eye • Mantle • 10

z 8 "'/,0

6

4

2-14/15-99 #10 (223mm) 2+----,-----.-----,------,---,------r----j o 50 100 150 200 250 300 350 S. oualaniensis mantle length (mm)

Figure 21. &l'N versus length ofeye lens segments of a female S. oualaniensis. &15N of mantle muscle plotted against mantle length at time of capture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in hottom right corner.

189 12 -<>-- Eye • Mantle

10 • z 8 "'00

6

4

2-7/8-00 #1 (210mm) 2-+----,------,---,------,----,------,------1 o 50 100 150 200 250 300 350 S. oualaniensis mantle length (mm)

Figure 22. 015N versus length of eye leus segmeuts ofa female S. oualaniensis. 015N ofmautIe muscle plotted against mantIe leugth at lime ofcapture. MantIe length for eye leus segments estimated from regressions of eye leus radius and squid size. Date caught, sqnid number, and mantle length noted in bottom right corner.

190 12 ------Eye • Mantle

10

z 8 on • 00

6

4 -

2-12/13-00 #1 (209mm) 2+-----.-----,------,,----,------.-----,------1 o 50 100 150 200 250 300 350 S. oualaniensis mantle length (mm)

15 Figure 23. I) N versus length ofeye lens segments of a female S. ouakmiensis. 1)'5N of mantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

191 ---<-- Eye 12 • Mantle

10 • 8 onz -00

6

4

2-12/13-00 #2 (160mm) 2-+---~---~---~--~---,----~----l o 50 100 150 200 250 300 350 S. ouaJaniensis mantle length (mm)

Figure 24. lll'N versus length of eye lens segments of a female S. oualaniensis. lll'N of mantIe muscle plotted against mantIe length at time ofcapture. MantIe length for eye lens segments estimated from regressions ofeye lens radins and squid size. Date canght, squid nnmber, and mantIe length noted in bottom right corner.

192 12 -- Eye • Mantle <> Blood

10 - <>

8

6

4

2·12113·00 #3 (181mm) 2 +----....----,------,,----,--,---,--,----,--,----1 o 50 100 150 200 250 300 350 S. oualaniensis mantle length (mm)

Figure 25. 81'N versus length of eye lens and gladius segments of a female S. oualaniensis. 81SN of mantle muscle and blood plotted against mantle length at time of capture. Mantie length for eye lens segments estimated from regressions ofeye leus radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

193 12 -"'-- Eye • Mantle

10 •

z 8 Go'"

6

4

2-12/13-00 #4 (173mm) 2+-----,----,-----;,----,-----,----,----j o 50 100 150 200 250 300 350 S. oualaniensis mantle length (mm)

Figure 26. a1'N versus length of eye lens and gladius segments of a female S. oualaniensis. a1'N of mantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions of eye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

194 12 - --- Eye • Mantle

10

8 •

6

4

2-12/13-00 #6 (158mm) 2+------,---.,------;r-----,----,-----,-----j o 50 100 150 200 250 300 350 S. Dualaniensis mantle length (mm)

Figure 27. 1l'5N versus length of eye lens segments ofa male S.oualanknsis. 1115N ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

195 -<>- Eye 12 • Mantle <> Blood

10

z B • OJ <> 00

6

4

2-13/14-00 #3 (251mm) 2+------,---,----,-----,---,------,----1 o 50 100 150 200 250 300 350 s. ouaJaniensis mantle length (mm)

Figure 28. 615N versus length of eye lens segments ofa female S. oualaniensis. 61SN of mantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions of eye lens radins and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

196 -- Eye 12 • Mantle o Blood

10

8

6

4

2-12/13-01 #1 (225mm) 2-1---~---~---~--~---~--~-----1 o 50 100 150 200 250 300 350 s. oualaniensis mantle length (mm)

Figure 29. (I"N versus length ofeye lens segments ofa female S. oualaniensis. (l1SN of mantle muscle and blood plotted against mantle length at time of capture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

197 --<>- Eye 12 Mantle <>• Blood 10 • <>

Z 8 '"00

6

4

2-12/13-01 #11 (304mm) 2+------,----,----,r----r----.------,------j o 50 100 150 200 250 300 350 S. oualaniensis mantle length (mm)

Figure 30. li15N versus length ofeye lens segments of a female S. oualanumsis. li''N of mantIe mnscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in hottom right corner.

198 ---<>- Eye 12 • Mantle

10 •

Z 8 "'00

6

4

2-12/13-01 #12 (302mm) 2+-----,----...,....------.------,---,.-----.------1 o 50 100 150 200 250 300 350 S. ouaJaniensis mantle length (mm)

Figure 31. 3"N versus length of eye lens segments of a female S. oualanumsis. 315N of mantle muscle plotted against mantle length at time of capture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

199 -0- Eye 12 Mantle •<> Blood

10 <>

Z 8 "'00

6

4

2-12/13-01 #16 (308mm) 2-j-----,----r---r------r---,------,----j o 50 100 150 200 250 300 350 S. ouaJaniensis mantle length (mm)

Figure 32.IiISN versus length ofeye lens segments ofa female S. ouawa",a.i•. IiISN ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye leus segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length uoted in bottom right comer.

200 CHAPTER IV: ENERGETICS AND TROPHODyNAMICS 201

INTRODUCTION 201

MATERIALS AND METHODS ....•....•...... 203

RESULTS ...... •.. 209 Male S. oualaniensis 209 Female S. oualaniensis 209 Male O. bartramii ., 210 Female O. bartramii 211

DISCUSSION 212

CONCLUSIONS ...... •...... 213

REFERENCES 215

LIST OF FIGURES 221

Chapter IV: Energetics and Trophodynamics

Introduction

Ommastrephid squids play key roles in the open ocean ecosystem and are of commercial value as a food resource (Roper et al., 1984). Traditional finfish stocks have been steadily declining (Caddy and Rodhouse, 1998), while cephalopod landings have been increasing to a maximum of 3.3 mmt in 1997 (FAO, 2000). Researchers have speculated that in some areas depleted fish stocks are undergoing a widespread species replacement by squids (Pauly et al., 1998; Balguerias et al., 2000). Although squids are less efficient than fishes in some ways (e.g. locomotion is less efficient than the traditional undulatory motion of fishes) (O'Dor and Webber, 1986; O'Dor and Webber, 1991) squids can undergo rapid somatic growth (Forsythe and van Heukelem, 1987; O'Dor and Hoar, 2000). The ramifications on oceanic ecosystems of widespread replacement of fish by short-lived rapidly growing squids, with energetically costly locomotion, is unclear. One approach to understanding these effects is through the construction ofecosystem models based on energy flow. Unfortunately knowledge of many crucial factors necessary for models of oceanic squids, such as biomass, production, and life history parameters, are lacking (Piatkowski

201 et al., 2001). In additionl, because of the difficulty in capturing and rearing large ommastrephid squids, little data on energetics and growth of these organisms are available. Demont and O'Dor (1984) studied the effects of activity, temperature and mass on the metabolism of the ommastrephid Illex illecebrosus in the laboratory. Webber and O'Dor (1985) studied the metabolism of the same animal in a controlled recirculating water tunnel. Clarke et al. (1994) applied the consumption rates calculated for I. illecebrosus (DeMont and O'Dor, 1984) to a related ommastrephid Illex argentinus and constructed a preliminary energy budget based on growth and the biochemical composition of certain tissues. The trophic level concept is central to many ecosystem models ranging from the early days of ecosystem modeling (Lindeman, 1942) to more recent applications of global fisheries models (Pauly and Christensen; 1995). Some researchers argue against the trophic level concept (peters, 1980) in favor of models based on growth (Peters, 1983). Cousins' (1985), however, argues that a model ofmarine ecosystems based on a trophic continuum circumvents the failings ofdiscrete trophic levels and static food chains. Regardless of the nomenclature all ecosystem energy models rely on a trophic concept (e.g. the transfer of energy from its basal resource to some endpoint) and forms of species linkages (predation, parasitism, etc.). While ecological models use assigned trophic levels for their calculations (pauly and Christensen, 1995; MacFarlane et aI., 2002; Leonov and Kanatov, 2002) few, if any, seem to use isotopically-defined trophic levels. While many studies use stable isotopes to determine trophic levels and food web structure (Sholto-Douglas et aI., 1991; Gould et al., 1997; Kaehler et aI., 2000; and many others) few seem to model the concomitant energy fluxes between such linkages. This chapter takes an initial step in combining stable isotopic data with energy budgets to create simple energetics models for the ommastrephid squids Ommastrephes bartramii and Sthenoteuthis oualaniensis to aid in understanding the ecology of these squids.

202 Materials and Methods

The ecological impact of these squids was estimated using a combination of published energetics data, and growth models for other squids along with measured parameters for O. bartramii and S. oualaniensis obtained in this study. A general energy budget was constructed for both sexes of O. bartramii and S. oualaniensis.

Ie == Eg + Ex + G + R + SDA +Mb + Ma Ie == ingested energy Eg == egestion Ex == excretion G=growth R == energy into reproduction SDA == specific dynamic action Mb == basal metabolism Ma == active metabolism

Because the energy budget must be applied over the lifetime of the squid, a growth equation is needed. The growth of these squids was estimated using a growth equation derived for O. bartramii (Murata, 1990). The growth curve for O. bartramii was calculated by Murata (1990) from samples of catch data (150-470 mm ML). There are obviously problems with using an estimated growth rate, however the data of Murata seems as reliable as the other available data (Yatsu et al., 1997; Yatsu and Mori, 2000). This growth equation was also applied to S. oualaniensis in lieu of data for this species. The growth equation ofMurata (1990), ML, == 55.3I1I+e(1.084-o.OlSIt) (t == age in days), gives a logistic growth curve. Murata's equation was modified to yield a maximum squid mantle length equal to the maximum mantle lengths of males and females of O. bartramii and S. oualaniensis caught during this study. Using maximum mantle lengths to determine the growth curve more closely estimates the squids growth over their entire lifespan. The numerator was replaced by a constant (determined for each category) and another constant was subtracted from the entire equation to yield a mantle length of zero at time zero. These modifications resulted in an equation for each category of squid that started at zero and increased logistically to a maximum that corresponded to the maximum measured ML for each category:

203 ML, = ((all+e(l.084-0.015lt»)_b)

S. oualaniensis male: a= 28 b= 7 MLti..1 = 210 mm S. oualaniensis female: a = 44 b = 10 MLtinal =324 mm O. bartramii male: a = 53 b = 13 MLtinal = 394 mm O. bartramii female: a = 88 b = 22 MLtinal = 650 mm

Mantle lengths were calculated over 365 days and converted into grams of biomass (BM) using regressions of measured mantle length versus weight from cruises 5 1997 and 1999 for O. bartramii (BMo burtramii = 9xlO- *ML2,8178, ~ = 0.97) and 1999 only 2 for S. oualaniensis (BMs. ounIon!ensis = 5xlO-4*ML2.549, r = 0.81) (Figs. 1,2). Determining the total energy budget for these squids required four steps. Egestion (Eg) and excretion (Ex) were assumed to be small (Clarke et aI, 1994) and were ignored. This energy budget was then modified to take into account the energy lost at lower trophic levels. A total energy budget was calculated for each sex and species of squid.

Step 1: The energy required for somatic growth (G) and reproductive mass (R) was calculated in this step. Clarke et al. (1994) calculated the mean energy content of major tissues in both sexes (wet weight of testes, spermatophoric complex, ovary, nidamental gland, digestive gland, mantle, head, and viscera) of a confamilial species, Illex argentinus. The mean energy content of tissue for male (mean biomass 469 g) 1. argentinus (was 1.31 kcal/g and was 1.57 kcal/g for females (mean biomass 800 g). These values were detenuined for 1. argentinus near maturity but are applied here over the entire lifespan. Mean energy content was applied to the daily biomass increase for both O. bartramii and S. oualaniensis. Somatic growth and reproductive mass were assumed to account for the total daily biomass increase. Each daily biomass increase was then multiplied by the average energy content of male or female tissue.

G + R = (BMd+ 1 - B~) * Energy conenl(male or female)

Step 2:

204 Specific dynamic action (SDA) was calculated in this step. SDA was assumed to be 20% of energy incorporated into mass (O+R) based on previous findings for cephalopods and invertebrates (O'Dor and Wells, 1987; Parry, 1983).

SDA =(0 + R) * 0.20

Step 3a: The total metabolic requirements (Mtotal) for each squid were calculated at daily increments using the equation derived from Illex illecebrosus by Webber and O'Dor

(1985). Webber and O'Dor measured O2 consumption of squid of known size (200 to 550 g) in a swim channel with controlled flow speed and temperature (15 ± 0.20 C). They developed an equation of whole animal O2 consumption based on biomass (W) and speed (S): m102/hr = (1.58*Wo.725)*(1.016s). For this model swimming speed was approximated as 0.5 body lengths (ML) per second, which is an average speed calculated from three telemetered adult female O. bartramii (Nakamura, 1993). This assumes that the squids swim at a constant relative speed throughout their lives.

Mtotal = (1.58*Wo.725)*(1.016(o.47·ML»

Step 3b. Basal metabolism (Mb) was calculated by assuming S = 0 for each daily size.

Step 3c.

Active metabolism (Ma) was then calculated as Mtota1 minus Mb.

Step 3d.

205 Because Webber and O'Dor kept a constant temperature of roughly 15° C the Mb was modified to take into account the effect of temperature fluctuations encountered by O. bartramii and S. ouataniensis. Two temperature regimes were used in this model, tropical and temperate. The tropical regime consisted of a surface temperature of 25° C and 5° C at depth while the temperate regime consisted of surface temperature of 15° C and 5° C at depth. Both squid species were subjected to tropical surface temperatures until they reached 100mm ML at which point they began their diurnal vertical migrations and spent 12hrs at 25° C and 12hrs at 5° C. The start of vertical migration behavior was assigned to squids of 100mm ML based on the sizes ofjuvenile squids found in seabirds feeding near the surface (Ashmole and Ashmole, 1967; Seki and Hanison, 1989) which indicate that squids larger than 100mm ML are absent from surface waters. S. ouataniensis remained in the tropical regime throughout its life. O. bartramii moved out of the tropical regime and into the temperate regime by day 92, and then returned to the tropical regime by day 275, these time periods were chosen to roughly approximate the migratory pattern reported by Murata and Nakamura (1998) with the northerly migration occurring in May/June and the southerly migration occurring in NovemberlDecember. The temperature effect was calculated using the temperature modes and time periods described above coupled with an estimated QlO value of 2.6. This value was chosen as representative of that measured for a variety of animals (Schmidt-Nielsen, 1984; Hochachka and Somero, 2002), including cephalopods (Wells et at., 1983; Seibel et at., 1997), within their normal temperature range. Specifically, Seibel et at. (1997) found that a QlO value of 2.62 was applicable to some vertically migrating cephalopods. While some studies have found QlO values as high as 7, such values are generally considered an artifact of measurement outside the species normal range (see discussion in

Seibel et at., 1997). The temperature effect (T.) was then calculated as 2.6 times Mb for temperatures of 25° C and 0.38 for temperatures of 5° C.

Step 3e.

206 The whole animal oxygen consumption, modified for temperature, (Mtatal*) for each half day increment was obtained by adding in the daily active metabolism, and the modified basal metabolism.

Step 3f. The temperature modified whole animal metabolism calculated in step 6 was converted to an energy requirement in kcal/day by adding up the two 12 hI' periods in each day and multiplying by a standard conversion factor of 0.00463 kcal/mlOz (Elliot and Davidson, 1975). (MtataJ *)*(O.00463 kcal/mlOz)

Step 4. To obtain a daily energy requirement (kcal/day) based on somatic growth, reproductive growth, specific dynamic action, and metabolism (active and basal), the results of steps 1, 2, and 3 were combined. Daily energy requirement =step 1 + step 2 + step 3f

The next steps combine the isotopic data (see chapter 2) with the energy budget. The overall daily energy requirement calculated in step 8 was modified to take into account the energy needed at lower trophic levels to supply the squids energy requirement. This calculation is shown in the following three steps.

Step A.

The regressions of squid mantle length versus I)J5N determined from this study (See chapter 2) were used to assign a daily 1)1sN value to each squid.

AI. The 1)15N versus ML regression for S. oualaniensis was calculated as:

J5 ML I) Ns. aualan;.n,i, '" 4.2081 + (I.8276*I.0037 )

207 A2. The 815 N versus ML regression for O. bartramii was calculated using a Gompertz 4­ parameter sigmoidal curve:

-(ML-183.2625/38.0979) <15 . 6 -e uNo. bartramii = 6.4261 + .8482(e )

StepB.

Each daily Ol5N value calculated in step A was converted into a fractional trophic level (TL) for each different category of squid as follows: the Ol5N value of 2.8%0 was chosen as the baseline trophic level because of the values for POM found in chapter 2. A 815N value of 3.4%0 was used to define one trophic level step (see chapter 2). The trophic level was then defined as:

TL = ((815N- 2.8%013.4%0)+1)

Step C. Daily energy requirement (step 4) was modified by a trophic level factor representing the energy needed through lower trophic levels. The trophic level factor was calculated using the trophic level of the squid at that day (step B) and an ecological efficiency of 10% for each TL. The trophic level factor was calculated as an exponential of0.1 per trophic steps below that of the squid at each day (eg- TL of 4 equals 3 steps, or energy transfers).

Energy requirement(TL) = step4*(0.1(TL-I))

A daily consumption rate was also calculated for each squid. The daily consumption rate was calculated as a percent of body mass using the daily energy requirement (step 4) and an average energy yield of prey species per gram. An average energy yield (kcal/g) for seven species of myctophid prey of O. bartramii and S. oualaniensis (Bolinichthys longipes, Ceratoscopelus warmingii, Hygophum rheinhardti, Lampanyctus tenuifonnis, Lobianchia gemelleri, Myctophum selenoides, and

208 Taaningichthys minimus) was calculated as 1.0 kcallg from Childress (1990). This energy yield was then used to calculate the amount of average myctophid material in a percentage of squid predator body weight to meet an adult squids energy needs during each day of each model.

Results

Male S. oualaniensis

The modeled male S. oualaniensis grew to 210 mm ML and 0.42 kg in 365 days. Daily caloric energy requirement increased to day 95 when vertical migration began and there was a break in the graph showing energy requirement per day (Fig. 3). After day 95 the energy requirement values increased asymptotically to a maximum of 24 kcal/day (Fig. 3). The total lifetime energy requirement for a male S. oualaniensis was calculated at 5.7x103 kcal. Modifying the energy requirement for trophic level produced a similar, but accentuated, curve to the unmodified energy requirement curve (Fig. 4). The energy requirement modified for trophic level ofmale S. oualaniensis increased to day 97 where there was a break in the curve and then the values increased asymptotically to 906 kcalfday. The total metabolic impact of male S. oualaniensis converted for trophic level over 365 days was 1.7x105 kcal. The percent ofbody weight required to meet the energy demands of a male S. oualniensis started at 280% on day one and decreased exponentially to 5.6% on day 365 (Fig. 5).

Female S. oualaniensis

The modeled female S. oualaniensis grew to 324 mm ML and 1.3 kg in 365 days. Daily caloric energy requirement increased to day 66 when vertical migration began and there was a break in the graph showing the energy requirement per day (Fig. 6). After day 66 the energy requirement values increased asymptotically to a maximum of 56 kcal/day

209 (Fig. 6). The total lifetime energy requirement for a female S. oualaniensis was calculated at 1.4xlO4 kcal. Modifying the energy requirement for trophic level produced a similar, but accentuated, curve to the unmodified energy requirement curve (Fig. 7). The energy requirement modified for trophic level offemale S. oualaniensis increased to day 66 where there was a break in the curve and then the values increased asymptotically to 8.7x103 kcal/day. The total metabolic impact ofmale S. oualaniensis converted for trophic level over 365 days was 1.3xlO6 kcal. The percent of body weight required to meet the energy demands of a male S. oualniensis started at 398% on day one and decreased exponentially to 4.5% on day 365 (Fig. 8).

Male O. bartramii

The modeled male O. bartramii grew to 394 rom ML and 1.8 kg in 365 days. Daily caloric energy requirement increased to 10 kcal/day on day 55 and decreased to 7 kcal/day when vertical migration began (Fig. 9). After day 55 the energy requirement increased again to 22 kcal/dayon day 92 and decreased to 16 kcal/day when the northern migration crosses into temperate waters (Fig. 9). After day 92 the energy requirement increased to 44 kcal/day on day 275 and then increased to 74 kcal/day when the southern migration crosses into tropical waters. After day 275 daily energy requirement increased slightly to a maximum of 77 kcal/day. The total lifetime energy requirement for a male O. bartramii was calculated at IAxlO4 kcal. Modifying the energy requirement for trophic level produced a similar, but accentuated, curve to the unmodified energy requirement curve (Fig. lO). The TL modified energy-requirement of male O. bartramii increased to 112 kcal/day on day 55 and decreased to 84kcallday when vertical migration began (Fig. 10). After day 55 the TL modified energy requirement increased to 932 kcal/day on day 92 and decreased to 747 kcallday then the northern migration passed into temperate waters (Fig. 10). After day 92 the TL modified energy requirement increased to 5.5x104 kcal/dayon day 275 and increased to 8.5xl04 when the southern migration passed into tropical waters. After day

210 275 the TL modified energy-requirement increased slightly to 8.9x104 kcal/day. The total metabolic impact of male O. bartramii converted for trophic level oVer 365 days was 1.7x105 kcal. The percent of body weight required to meet the energy demands of a male O. bartramii started at 380% on day one and decreased exponentially to 4.1 % on day 365 (Fig. 11).

Female 0. bartramii

The modeled female O. bartramii grew to 650 mm ML and 7.6 kg in 365 days. Daily caloric energy requirement increased to 12 kcal/day on day 35 and decreased to 10 kcallday when vertical migration began (Fig. 12). After day 35 the energy requirement increased again to 79 kcal/day on day 92 and decreased to 63 kcal/day when the northern migration began (Fig. 12). After day 92 the energy requirement increased to 163 kcal/day on day 275 and then increased to 247 kcal/day when the southern migration began. After day 275 daily energy requirement increased slightly to a maximum of 250 kcaJ/day. The total lifetime energy requirement for a female O. bartramii was calculated at 5.0x104 kcal. Modifying the energy requirement for trophic level produced a similar, but accentuated, curve to the unmodified energy requirement curve (Fig. 13). The TL modified energy requirement of female O. bartramii increased to 135 kcal/dayon day 35 and decreased to III kcal/day when vertical migration began (Fig. 13). After day 35 the TL modified energy requirement increased to 6.9x104 kcal/day on day 92 and decreased to 5.6x104 kcal/day when the northern migration passed into temp waters (Fig. 13). After day 92 the TL modified energy requirement increased to 1.9xl05 kcal/day on day 275 and increased to 2.8x105 when the southern migration passed into tropical waters. After day 275 the TL modified energy requirement increased slightly to 2.9x105 kcal/day. The total metabolic impact of female O. bartramii converted for trophic level over 365 days was 5.7x107 kcal.

211 The percent of body weight required to meet the energy demands of a female O. bartramii started at 189% on day one and decreased exponentially to 3.3% on day 365 (Fig. 14).

Discussion

Based on the food requirement (in % body weight per day) the squids would need, the models overestimate their energy requirement for roughly the first 100 days for each squid. During this period the modeled consumption requirements are mostly likely greater than the physiological capabilities of the squids. However, after the initial period of about 100 days the consumption estimates drop off dramatically and stabilize at more reasonable values. Tung et al. (1973) and Tung (1976) classified stomachs of S. oualaniensis (100-270 mm ML) caught off of Taiwan as "medium" fullness with contents weighing::; 3% of body weight, and "full" stomachs with contents weighing> 3% of body weight. Therefore, according to these findings, one "full" stomach could come close to fulfilling the modeled energy requirements for S. oualaniensis males or females in the latter stages of the models. Shevtsov (1972), off the southeast of Japan, found 70% of O. bartramii (240·300 rom ML) stomachs with contents> 1% of body weight, but did not quantify the value further, and did not comment on stomach "fullness". According to Shevtsov's results most O. bartramii captured would have enough in their stomachs to satisfy roughly a third to a quarter oftheir energy requirements in the latter stages of the model. Given that feeding studies rely on baited capture methods (jigging etc.) and will most likely sample hungry squids, once would expect stomach content weights to be low. These results, therefore, are consistent with modeled results. The poor fit of the model to the early life history of these squids is not unexpected given the growth model used was derived from sub-adults and adults (Murata, 1990). The models indicate the effect that temperature could have on the energy demands of these squids. This effect is exemplified in a comparison between female S. oualaniensis and male O. bartramii. The size of the modeled male O. bartramii at 365 days was over 25% heavier than the modeled female S. oualaniensis at 365 days, yet their overall lifetime energy requirements were almost identical. This may be one contributing factor in the evolution of the migratory behavior of O. bartramii, and its larger body size,

212 in comparison to the smaller non-migratory S. oualaniensis. Although this model attempted to account for some aspects of behavior, such as diurnal vertical migration and geographical migration, beyond temperature it does not aecount for differential costs in vertical migration or horizontal migration versus nOimal swimming. These, and other more fine scale behaviors such as climb and glide swimming could affect energy demands. The models also indicate the effect that trophic levels could have on the amount of energy needed from the base ofthe foodweb. Comparing male O. bartramii and female S. oualaniensis, and taking into account the energy needed through lower trophic levels, the male O. bartramii required roughly 11 times the energy. In general, accounting for energy needed from lower trophic levels resulted in an almost 30 fold increase in energy required for male S. oualaniensis and a 100 fold increase in energy for female S. oualaniensis. Accounting for trophic-level needs resulted in an increase in roughly 1000 times over lifetime energy requirements for O. bartramii males and females. Without concrete values of population size, estimates of the ecological impact for these animals are difficult to make. Historically, the North Pacific fishery for O. bartramii was harvesting roughly 300,000 mtlyr (Murata and Nakamura, 1998). With a 1: 1 sex ratio of 650 mm ML females and 394 mm ML males, the overall energy required for this amount of O. bartramii (accounting for TL losses) in the North Pacific would be 1.3xlO15 kcal/yr. North Pacific production estimates range from 27 to 50xlO15 gC/yr (Behrenfeld and Falkowski, 1997) and assuming 10 kcal/gC (for a full explanation see Steele, 1975) the total energy in the North Pacific would be between 270 and 500xlO15 kcal/yr. Therefore using these energy models, O. bartramii could account for between 0.3 and 0.5% of the energy produced in the North Paefic. The population impact of these squids could actually be higher considering that many squids did not reach harvestable sizes due to natural mortality.

Conclusions

Combining measured metabolic functions and growth data, along with actual weight measurements into an energetics model provides a first estimate of the lifetime

213 energy requirements of these squids. The daily energy requirements of the latter stages of the models agree to an extent with published feeding values of these squids. Vertical migration, and in the case of O. bartramii horizontal migration, coincides with the time period when the squids energetic requirements begin to increase rapidly. The energy savings that O. bartramii obtains from crossing into colder northern waters is substantial and much of this energy is likely diverted into somatic and reproductive tissue growth. Given the lack ofevidence for direct competition between these similar squids in Hawaiian waters, this migratory behavior could be the result of the evolution of large body size and faster growth in O. bartramii to take advantage of a more energy rich but less energy demanding environment, and not predominantly as a means of avoiding competition with S. oualaniensis. Using the isotopicaly derived trophic level calculations within the framework of the energetics models gives an estimate of the total amount of energy these squids require from ecosystem. Given the debate on the ecological efficiencies within ecosystems, and the many assumptions used in the calculations, assessing the validity of the final energetic impact when modified for trophic level is difticult. Solid estimates of population sizes for these squids, along with fewer assumptions, would be required to make more accurate estimates of their ocean or worldwide impacts. The future of incorporating isotopically derived trophic information into ecological/energetic models is virtually unexplored. The benefits of a method to unequivocally assign trophic positions, that are not constrained by discrete levels, are enormous. Stable isotopes could provide modelers with solid indications of the amount of energy transfers between a certain species and its basal resource, which would be invaluable in accurate modeling of foodweb energetics. Validation of this techinique is necessary in a field setting however, and should be initially attempted in small, constrained ecosystems such as lakes or embayments. The ability to corroborate isotope data and actual measured energy transfers within a simple ecosystem is crucial to the widespread use of isotopically derived trophic information in energetic modeling.

214 References

Ashmole, N., and M. Ashmole (1967). "Comparative feeding ecology of sea birds of a tropical oceanic island." Bulletin of the Peabody Museum ofNatural History 24: 1-131.

Balguerias, E., M. E. Quientero, and C. L. Hernandez-Gonzalez (2000). "The origin of the Saharan Bank cephalopod fishery." ICES Journal of Marine Science 57: 15­ 23.

Behrenfeld, M. J., and P. G. Falkowski (1997). "Photosynthetic rates derived from satellite-based chlorophyll concentration." Limnology and Oceanography 21(1): 1-20.

Caddy, J. F., and P. G. Rodhouse (1998). "Cephalopod and groundfish landings: evidence for ecological change in global fisheries." Review of Fisheries Biology 8: 431­ 444.

Clarke, A., P. G. Rodhouse, and D. J. Gore (1994). "Biochemical composition in relation to the energetics of growth and sexual maturation in the ommastrephid squid Illex argentinus." Philosophical Transactions ofthe Royal Society of Landon B(344): 201-212.

Cousins, S. H. (1985). "The trophic continuum in marine ecosystems: structure and equations for a predictive model." Canadian Bulletin ofFisheries and Aquatic Sciences(213): 76-91.

Demont, M. E., and R. K. ODor (1984). "The effects of activity, temperature, and mass on the respiratory metabolism of the squid, Illex illecebrosus." Journal of the Marine Biological Association, U. K. 64: 535-543.

215 Elliot, 1. M., and W. Davidson (1975). "Energy equivalents of oxygen consumption in animal energetics." Oecologia 19: 195-201.

. FAO (2000). "Total production 1950-1998." FAO Yearbook of fisheries statistics "Catches and Landings".

Forsythe, J. W., and W. F. van Heukelem (1987). Growth. Cephalopod Life Cycles. P. R. Boyle. London, Academic Press. II: 135-156.

Gould, P., P. Ostrom, and W. Walker (1997). "Trophic relationships of albatrosses associated with squid and large-mesh drift-net fisheries in the North Pacific Ocean." Canadian Journal of Zoology 75:·549-562.

Hochachka, P. W., and G. N. Somero (2002). Biochemical adaptation: mechanism and process in physiological evolution. Oxford, New York, Oxford University Press.

Kaehler, S., E. A. Pakhomov, and C. D. McQuaid (2000). "Trophic structure of the marine food web at the Prince Edward Islands (Southern Ocean) determined by

ODC and 015N analysis." Marine Ecology Progress Series 208: 13-20.

Leonev, A. V., and G. A. Kanatov (2002). "Formalization of interactions between chemical and biological compartments in the mathematical model describing the transformation of nitrogen, phosphorous, silicon, and carbon compounds." Pacific International Council for the Exploration of the Seas (PICES) Report(20): 100-­ 107.

Lindeman, R. L. (1942). "The trophic-dynamic aspect of ecology." Ecology 67: 413-418.

McFarlane, G. A., A. S. Krovnin, B. A. Megrey, and M. J. Kishi (2002). "BASSIMODEL workshop to review ecosystem models for the subarctic pacific gyres." Pacific International Council for the Exploration of the Seas (PICES) Report(20): 1-5.

216 Murata, M. (1990). "Oceanic resources of squids." Marine Behavioral Physiology 18: 19­ 71.

Natsukari, Y. a. N. K. (1992). "Age and growth estimation of the european squid, Loligo vulgaris, based on statolith microstructure." Journal of the Marine Biological Association, U. K. 72: 271-280.

O'Oor, R. K., and D. M. Webber (1986). "The constraints on cephalopods: why squid aren't fish." Canadian Journal of Zoology 64: 1591-1605.

O'Oor, R. K., and D. M. Webber (1991). "Invertebrate athletes: Tradeoffs between transport effeciency and power density in cephalopod evolution." Journal of Experimental Biology 160: 93-112.

O'Oor, R. K., and J. A. Hoar (2000). "Does geometry limit squid growth?" ICES Journal of Marine Science 57: 8-14.

O'Oor, R. K., and M. J. Wells (1987). Energy and nutrient flow. Cephalopod Life Cycles. P. R. Boyle. London, Academic Press. II: 109-133.

Okutani, T., and 1. H. Tung (1978). "Reviews of biology of commercially important squids in Japanese and adjacent waters. 1. Symplectoteuthis oualaniensis (Lesson)." Veliger 21: 87-94,

Parry (1983). "The influence of cost of growth on ectothenn metabolism." Journal of Theoretical Biology 107: 453-477.

Pauly, D., and V. Christensen (1995). "Primary production required to sustain global fisheries." Nature 374: 255-257.

217 Pauly, D., V. Christensen, J. Dalsgaard, R. Froese, and F. Torres (1998). "Fishing down marine food webs." Science 279(860-863).

Peters, R. H. (1980). "Useful concepts for predictive ecology." Synthese 43: 257-269.

Peters, R. H. (1983). The ecological implications of body size. Cambridge, Cambridge University Press.

Piatkowski, u., G. J. Pierce, and M. Morais de Cunha (2001). "Impact of cephalopods in the food chain and their interaction with the environment and fisheries: an overview." Fisheries Research 52: 5-10.

Roper, C. F. E., M. J. Sweeney, and C. E. Nauen (1984). "Cephalopods of the world. An annotated and illustrated catalogue of species of interest to fisheries." FAO species catalogue 125(3): 1-277.

Schmidt-Nielsen, K. (1983). Animal Physiology: Adaptation and Environment. Cambridge, England, Cambridge University Press.

Seibel, B. A., E. V. Thuesen, J. J. Childress, and L. A. Gorodezky (1997). "Decline in pelagic cephalopod metabolism with habitat depth reflects differences in locomotory effeciency." Biological Bulletin 192: 262-278.

Seki, M. P., and C. S. Harrison (1989). "Feeding ecology of two subtropical seabird species at French Frigate Shoals, Hawaii." Bulletin ofMarine Science 45: 52-67.

Shevtsov, G. A. (1972). "Feeding habits of the squid Ommastrephes bartramii Leseur in the Kurile-Hokkaido region." Journal ofHydrobiologia 8(3): 77-80.

218 Sholto-Douglas, A. D., J. G. Field, A. G. James, and N. J. van der Merwe (1991). l2 "BC/ C and 15N/14N isotope ratios in the Southern Benguela ecosystem: indicators of food web relationships among different size-classes of plankton and pelagic fish; differences between fish muscle and bone collagen tissues." Marine Ecology Progress Series 78: 23-31.

Steele, J. H. (1975). The Structure of Marine Ecosystems. London, England, HarvardUniversity Press.

Tung, 1. H. (1976). "On the food habit ofcommon squid Symplectoteuthis oualaniensis (Lesson)." Report of the Institute ofFisheries Biology, Taipei 3(2): 49-66.

Tung, 1. R., L. Chi-Sheng, and H. Chan-Chin (1973). "The preliminary investigation for exploitation of common squid resources." Report of the Institute of Fisheries Biology for the Ministry of Economic Affairs and Nature, Taiwan University 3(1): 211-247.

Webber, D. M., and R. K. O'Dor (1985). "Respiration and swimming perfonnance of short-finned squid (Illex illecebrosus)." North Atlantic Fisheries Organization (NAFO) Scientific Council Study(9): 133-138.

Weihs (1973). "Mechanically effecient swimming techniques for fish with negative buoyancy." Journal ofMarine Research 31: 194-209.

Wells, M. J., R. K. O'Dor, K. Mangold, and J. Wells (1983). "Diurnal changes in activity and metabolic rate in Octopus vulgaris." Marine Behavioral Physiology 9: 275­ 287.

Yatsu, A, and J. Mori (2000). "Early growth of the autumn cohort of neon flying squid, Ommastrephes bartramii, in the North Pacific Ocean." Fisheries Research 45: 189-194.

219 Yatsu, A., S. Midorikawa, T. Shimada, and Y. Uozumi (1997). "Age and growth of the neon flying squid, Ommastrephes bartramii, in the North Pacific Ocean." Fisheries Research 29: 257-270.

220 List of Figures

FIGURE 1. 'THE RELATIONSHIP BETWEEN LENGTH AND WEIGHT FOR ALL O. BARTRAMII

CAUGHT DURING YEARS 1997 AND 1999 223

FIGURE 2. 'THE RELATIONSHIP BETWEEN LENGTH AND WEIGHT FOR ALL S. OUALANIENSIS

CAUGHT DURING 1999 224

FIGURE 3. 'THE TOTAL ENERGY REQUIREMENT (KCAL) FOR A MALE S. OUALANIENSIS (210

MM ML, 0.42 KG) OVER 365 DAYS 225

FIGURE 4. 'THE TOTAL DAILY ENERGY REQUIREMENT MODIFIED FOR ENERGY NEEDED

THROUGH LOWER TROPIDC LEVELS (KCAL) FOR A MALE S. OUALANIENSIS (21OMM ML,

0.42 KG) 226

FIGURE 5. 'THE AMOUNT OF FOOD NEEDED (% OF BODY WEIGHT) TO MEET THE MODELED

DAILY ENERGY REQUIREMENTS FOR A MALE S. OUALANIENSIS (210MM ML, 0.42 KG)...... 227

FIGURE 6. 'THE TOTAL ENERGY REQUIREMENT (KCAL) FOR A FEMALE S. OUALANIENSIS (324

MMML, 1.3 KG) OVER 365 DAYS 228

FIGURE 7. 'THE TOTAL DAILY ENERGY REQUIREMENT MODIFIED FOR ENERGY NEEDE

THROUGH LOWER TROPIDC LEVELS (KCAL) FOR A FEMALE S. OUALANIENSIS (324 MM

ML, 1.3 KG) 229

FIGURE 8. 'THE AMOUNT OF FOOD NEEDED (% OF BODY WEIGHT) TO MEET THE MODELED

DAILY ENERGY REQUIREMENTS FOR A FEMALE S. OUALANIENSIS (324 MM ML, 1.3 KG)...... 230

FIGURE 9. 'THE TOTAL ENERGY REQUIREMENT (KCAL) FOR A MALE O. BARTRAMII (394 MM

ML, 1.8 KG) OVER 365 DAyS 231

FIGURE 10. 'THE TOTAL DAILY ENERGY REQUIREMENT MODIFIED FOR ENERGY NEEDED

THROUGH TROPIDC LEVELS (KCAL) FOR A MALE O. BARTRAMII (394 MM ML, 1.8 KG)...... 232

FIGURE 11. 'THE AMOUNT OFFOOD NEEDED (% OF BODY WEIGHT) TO MEET THE MODELED

DAILY ENERGY REQUIREMENTS FOR A MALE O. BARTRAMII (394 MM ML, 1.8 KG)... 233

FIGURE 12. THE TOTAL ENERGY REQUIREMENT (KCAL) FOR A FEMALE O. BARTRAMII (650

MMML, 7.8 KG) OVER 365 DAYS 234

221 FIGURE 13. THE TOTAL DAILY ENERGY REQUIREMENT MODIFIED FOR ENERGY NEEDED

TIIROUGH LOWER TROPHIC LEVELS (KCAL) FOR A FEMALE O. BARTRAMII (650 MM ML,

7.8 KG) 235

FIGURE 14. THE AMOUNT OF FOOD NEEDED (% OF BODY WEIGHT) TO MEET THE MODELED

DAILY ENERGY REQUIREMENTS FOR A FEMALE O. BARTRAMII (650 MM ML, 7.8 KG) ...... , 236

222 7000

6000

5000

Oi 4000 <> .-:E y = 9E_OSx"B17B 0> .ij) R2 = 0.9718 3: 3000

2000

1000

0 0 100 200 300 400 500 600 700 o. bartramii mante length (mm)

Figure 1. The relationship between length and weight for all O. bartramii caught during years 1997 and 1999.

223 1000 900 800 0 700

~ 600 0 ~ Y=0.0005,('549 .<:. 2 -Ol 500 R =0.8141 "iii s: 400 300 200 100

0 0 50 100 150 200 250 300 S. oualaniensis mantle length (mm)

Figure 2. The relationship between length and weight for all S. ouakmiensis caught during 1999.

224 60 I • S. Qua/aniensis (male) I

~ 50 iii () ~ ~ ...c: CD 40 E CD .=:J CT 30 fI! >- Cl ~ CD c: 20 CD til -0 I- 10

O-f-~:""-_--r------,------,------J o 100 200 300 400 Days

Figure 3. The total euergy requirement (kcal) for a male S. oualaniensis (210 mm ML, 0.42 kg) over 365 days.

225 1000 i..:.: • S. oualaniensis (male) I ~ ....J I- ... 800 0 -"0 (])

~ 600 E t: -(]) E ~ 400 'S 0- ~ >- E> (]) 200 t: (]) iii {:.- 0 0 100 200 300 400 Days

Figure 4. The total daily energy requirement modified for energy needed through lower trophic levels (kcal) for a male S. ollhlaniensis (21Omm ML, 0.42 kg).

226 1000 • s. Dualaniensis (male) I ~ -0 ~ al Q.w -0-0 ale: ~

.r:~-iJ; .21 al ale: :=al >0- 35 10 .... .aE"8 o;!!.o _ "0 Ol ..Q

1 0 100 200 300 400 Days

Figure 5. The amount offood needed (% of body weigbt) to meet the modeled daily energy requirements for a male S. oualanien.i. (210mm ML, 0.42 kg).

227 60 • S. oualaniensis (female) I

~ 50 iii.., ~ ~ -c: 40 aI E aI ~ 'S 0- Q) 30 ~ ~ ~ Q) c: 20 aI iii -a I- 10

0-1o"""'------,------r------.....,------I o 100 200 300 400 Days Figure 6. The total energy requirement (kcal) for a female S. ouahmiensis (324 nun ML, 1.3 kg) over 365 days.

228 10x1OS m0 s. oualaniensis (female) .><: • I ~ I -I f- 3 a~ 8x10 -"0 .!!! "0 ""a 6x103 E c -Q) E 4x103 oS21 ~ >. ~ Q) 2x10' c: Q)

Figure 7. The total daily energy requirement modified for energy neede throngh lower trophic levels (kcal) for a female So oualaniensis (324 mm ML, 1.3 kg).

229 1000 • s. oualaniensis (female) I "0 e1 -3 0- e1CIJ >."0 . .cO) 0)'" --~~Q) >'05 "OQ) .8E 10 o~o_ -0 0) .3

1 0 100 200 300 400 Days

Figure 8. The amount offood needed (% ofbody weight) to meet the modeled daily energy requirements for a female S. ouafanknsis (324 mID ML, 1.3 kg).

230 100 • O. bartramii (male) I

~ III 80 0 .><: ~ • c -(J) E 60 ~ '5 a- ~ 40 S~ (J) c (J) III 0 f- 20

O__IIIE:------,------r------r------j o 100 200 300 400 Days

Figure 9. The total energy requirement (kcal) for a male O. barlramii (394 mm ML, 1.8 kg) over 365 days.

231 -; 100x10' ~ • O. bartramii (male) I ~ ~ -- ~ BOx10' ....0 "0 Ql :E "0 0 60x10' E C -Ql E Ql ~ 4Ox10' '5 cr l!! ;>, ~ Ql 20x10' c Ql

Figure 10. The total daily energy requirement modified for energy needed through trophic levels (kcal) for a male O. bartramii (394 mm ML, 1.8 kg).

232 1000 • O. barlramii (male) I "0Q) ~ 'S 0- Q)~"O'" >'c . Ol~ .- Q) Q) C == Q) >'a; "OQ) 10 .8Eo _ -;Ro -0 Ol 0 ....J -

1 0 100 200 300 400 Days

Figure 11. The amount offood needed (% ofhody weight) to meet the modeled daily energy requirements for a male O. bar/ramii (394 mm ML, 1.8 kg).

233 300 • O. barlramii (female) I

~ 250 ti'i .il ~ c: -Ql 200 E ...Ql '5 0- 150 ...Ql >, 2l (I) c 100 Ql ti'i -0 l- SO

0-f-'~------,------,------.------1 o 100 200 300 400 Days

Figure 12. The total energy requirement (kcal) for a female O. bartramii (650 mm ML, 7.8 kg) over 365 days.

234 -; 350xl ()3 J2 I • 0. bartramii (female) I -....I 300xl ()3 I- ..... 0 -"'C 250xl03 (Jl :E "'C 0 E 200x103 <: -(Jl E 150x103 .....(Jl ':; 0" l!? 1OOxl 03 >- .....OJ (Jl <: (Jl 50x1 ()3 (ij '0 I- 0 0 100 200 300 400 Days

Figure 13. The total daily energy reqnirement modified for energy needed through lower trophic levels (kcal) for a female O. bartramii (650 mm ML, 7.8 kg).

235 1000 Tr======:::::;------~ • O. bartramii (female) I

100

•

o 100 200 300 400 Days

Figure 14. The amount offood needed (% of body weight) to meet the modeled daily energy requirements for a female O. bartramii (650 mm ML, 7.8 kg).

236 CHAPTER V: CONCLUSIONS 237

SUMMARy 237

Chapter V: Conclusions

Summary

seA and SIA were used effectively to study the trophic ecology of 0. bartramii and S. oualaniensis, two ommastrephid squids, in Hawaiian waters. SeA and SIA can be used as complimentary techniques, each method alleviating certain weaknesses ofthe other. In this study each technique gave an incomplete depiction ofthe trophic status of these squids. Taken together these results set the framework for better understanding of each trophic niche these animals occupy in the pelagic ecosystem in Hawaiian waters. In addition the SIA was used retrospectively to infer historical trophic information. Using aspects from both the SCA and SIA portions ofthe study, a basic energy model was constructed for both sexes ofsquids. This energy model provided a reasonable first estimate ofthe energetic requirements ofthese squids and attempted to model the amount ofenergy needed from lower trophic levels to meet these demands. The SCA indicated a fundamental difference in the feeding habits ofthe O. bartramii and S. oualaniensis caught during this study. While both squids fed heavily on myctophid fishes and onychoteuthid squids there were substantial differences in feeding on prey even at the family level. SCA results indicated generalized feeding in 0. bartramii, and relatively more specialized feeding in S. oualaniensis. This feeding trend continued when the SCA results were analyzed at the species level for fish prey, and at the lowest possible taxa for cephalopod prey. Within the family Myctophidae, S. oualaniensis fed heavily on S. evermanni, C. warmingii, and H. proximumlrheinhardti, while 0. bartramii fed less frequently, and with less abundance, on a number ofmyctophid species. Although the results for squid prey were less pronounced than with fish prey, 0. bartramii still fed less frequently, and with less abundance, on a wider variety ofsquid prey than did S. oualaniensis, with the exception ofsimilar feeding on Onychoteuthis sp.

237 Using standard measures ofniche breadth (Levins' measure) and resource overlap (MacArthur and Levins) indicated the extent to which feeding on fish and squid prey differed between 0. bartramii and S. oualaniensis. There were slight indications that resource overlap ofS. oualaniensis on 0. bartramii increased with increasing size category ofS. oualaniensis. Within four categories ofshared myctophid prey, otoliths found in 0. bartramii were significantly larger than those found in S. oualaniensis. A similar situation was found between female and male S. oualaniensis, with larger otoliths recovered from female S. oualaniensis. Given their overlapping habitat, confamilial status, and size differences one might expect these squid species to show prey/predator interactions, or at least strong competitive interactions in areas ofsympatry. However, virtually no predation and little competition for food resources occurred between 0. bartramii and S. oualaniensis in this study. The proportions and frequencies ofitems in the diets of0. bartramii and S. oualaniensis, as well as the dietary measurements calculated from these values, indicate that, although S. oualaniensis feeds on a subset ofthe 0. bartramii diet, competition is low. Even considering the differences in feeding from the SeA results, assigning relative trophic positions to these squids within the ecosystem would still be highly subjective, since even the trophic status ofthe prey are poorly known. For this reason, SIA were performed in an attempt to determine quantitative trophic relationships for 0. bartramii and S. oualaniensis, both within the ecosystem and in relation to each other. In general the a15N data were, as expected, consistent with the broad categories of trophic levels generally assigned to various organisms. The POM was the lightest ofany category, while zooplankton were slightly higher. Average a15N values ofmyctophids and other mesopelagic fishes and squids were closer to average values ofS. oualaniensis than would be expected based on SeA. Ommastrephes bartramii had the heaviest a15N values ofany organisms sampled, and their values were roughly a trophic step (3.4%0) above those oftheir prey. Sampling methods ofsquid prey could have resulted in a 15 heavier 015N bias in the samples and may not accurately represent prey a N values for these squids. The a15N values indicated that, on average, 0. bartramii should be more

238 than I TL above S. oualaniensis. This is consistent with the different diets ofthese squid found in the SCA portion ofthis study. Body size (ML) was the greatest factor in detennining the a15N value for 0. bartramii and S. oualaniensis. The general trend of,, 15N values for squid muscle (mantle) showed different patterns for 0. bartramii and S. oualaniensis over the size ranges sampled. The ,,15N of0. bartramii muscle increased logistically with size and reached a plateau above 300 mm ML while the"15N ofS. oualaniensis muscle increased exponentially up to the largest size sampled. These different patterns are probably due to the 0. bartramii reaching a critical size where it's trophic level stabilizes while S. oualaniensis does not reach this threshold size. A possible geographical component to the different patterns cannot be ruled out. Data from individual retrospective eye lens tissue combined from all eyes sampled showed patterns of"15N increase against calculated ML that were similar to the general trends ofthe mantle data. The"15N values obtained from 0. bartramii eye tissue increased in a logistic pattern and those obtained from S. oualaniensis eye tissue increased exponentially, however the curve ofthe latter was below that described for mantle muscle. The reasons for the discrepancy between S. oualaniensis tissues are not clear, however the data were not corrected for the changes in refractive index that occur in the eye lens during growth. These changes may be responsible for some ofthe discrepancy observed. The results ofthe eye lens tissue ,,15N for S. oualaniensis allowed for categorization ofindividuals into "typical", "atypical", and "anomalous" groups. The

,,15N values ofthe typical and atypical eye lenses all increased with calculated ML, although the atypical group increased more rapidly. The atypical case was interpreted as feeding specialization that began at a small size and remained throughout the animals life until around the time ofcapture, unfortunately the SCA data could not confirm this. The anomalous group consisted ofthe eye lens from a single squid and the ,,15N were abnormally high and did not exhibit a discernible pattern with calculated ML. No explanation was found for this pattern. Blood samples taken from 0. bartramii and S. oualaniensis were also similar to the results obtained from muscle and eye lens tissue. Blood samples were only available for O. bartramii at the larger end ofthe size spectrum sampled, so total validation ofthe

239 blood /)15N pattern was not possible, however the blood samples ofthe larger individuals were in general agreement with the pattern seen in the muscle and eye tissue. Blood was sampled from a wider relative size range ofS. oualaniensis, and the /)lsN values increased exponentially with ML. The /)I~ results for the mantle muscle, eye lens tissue, and blood, showed large individual variation within the overall patterns established for each squid. Retrospective analyses ofeye-lens tissue for individual squid allowed some analysis ofthese causes of variation. In almost all squid sampled ofboth species, the mantle muscle was heavier than the blood, indicating that there may be an intrinsic fractionation between these two tissues. In general the mantle muscle was heavier than the outer eye segment also indicating that there is an intrinsic fractionation between these two tissues, although the exceptions were greater than in the comparison ofmuscle to blood. The individual analyses for the eye lenses ofeach squid revealed similar patterns of/)15N increase with calculated size seen when data from each species was combined and analyzed as a group, except that the timing and slope ofmajor changes showed great variation. The period ofmost rapid increase in /)15N for 0. bartramii eye lenses occurred between roughly 175 and 300 mm ML. Although most O. bartramii eye lenses sampled conformed to the logistic /)15N increase with size, a few exceptions did occur, and variability within the logistic curve was common. Several ofthe eye lens curves showed a decrease in /) lSN at the largest ML increment which may indicate a slight drop in TL as the squids move back into tropical waters during their southern migration. The eye lenses ofjuvenile O. bartramii less than 175 mm ML did not show an increase in /)lsN with calculated mantle length. This supported the mantle data, which suggested an unexpected TL plateau over this size range. Three sub-adult O. bartramii (225 to 255 mm ML) captured at 46° N explained a peculiar intermediate plateau in /)lsN muscle values versus ML for 30 squid taken at this latitude. The three sub-adult eyes sampled showed different histories that converged to the same final value indicating that the TL ofthese squid at this location was being dictated by local environmental factors, presnmably prey type and availability.

Except for the peculiar anomalous squid, the /)lsN values from individual eye lenses ofS. oualaniensis increased exponentially for almost all squids. Even so,

240 individual patterns showed high variability within the life ofone individual as we1l as between individuals, indicating that movement up the trophic pyramid is not a steady process. The pattern ofOl5N increase with ML from the eye tissue ofthe largest S. oualaniensis sampled (324 mm ML) had logistic qualities similar to the patterns seen in 0. bartramii eye lenses, however the pattern still fit we1l to an exponential curve. This pattern suggests that the largest S. oualaniensis may show a trophic plateau as was seen in 0. bartramii beginning at about the same size. Unfortunately insufficient numbers of large S. oualaniensis were caught to allow for confirmation ofthis trend. With a few exceptions, the 015N patterns seem to reflect the feeding history ofthe squids. There are several reasons for confidence that the major features ofthe 015N data provide useful and robust data: A general size relationship ofincreasing 015N with increasing size was observed across individual species ofseveral non-cephalopod taxa.

The 015N data from the mantles and the eye-lens segments both showed logistic patterns versus ML in 0. bartramii. The 01~ data from the mantles, eye-lens segments, and blood a1l showed exponential patterns versus ML in S. oualaniensis. Retrospective analysis ofindividual squids showed large variations in trophic position over fairly short time periods. Trophic position appears to be highly dynamic and is most likely a combination ofmany factors including, individual feeding behavior, prey availability, and fluctuations in food chain length. Retrospective analyses offer great potential for future studies in sorting out the causes ofindividual variation in TL.

Obtaining consistent 015N values, genera1lyreflective oftrophic level, for 0. bartramii and S. oualaniensis a1l0wed the SIA results to be incorporated into energetic models for these squids. Combining modifications ofexisting metabolic and growth equations for 0. bartramii into an energy budget produced daily energetic demands for 0. bartramii and S. oualaniensis that conform to their physiological capabilities and published feeding results in the later stages ofthe model. The model results indicated the probable impact ofmigration behavior (diurnal and geographical) on the energetic demands ofthese squids. The model also demonstrated the possibly large effects that increasing trophic position with age can have on the overa1l energy required from the ecosystem to sustain these squids. The model also indicated that the energetic costs ofa large geographical migration in 0. bartramii, relative to no such migration in S.

241 oualaniensis, is largely offset by the colder northern waters. In addition, richer feeding grounds in the north may make migration an energetically favorable option. This study indicates the importance ofusing complimentary techniques such as SCA and SIA when investigating difficult biological concepts such as trophic ecology. Using SIA in a complex trophic environment like the North Pacific Subtropical gyre remains challenging given the variables that can affect 515N, however the results ofthis study show that reproducible and ecologically relevant data can be obtained. The interactions between trophic level, feeding behavior, nutrient stress, tissue fractionation, and geographical variation on the stable isotopic signatures ofan animal are still poorly understood. Separating and quantifying the effects ofthese factors requires controlled experimental manipulations. Such experiments should enable increased usage ofstable isotopes for understanding a wide array ofecological questions. 15 The average 3.4%0 increase in 5 N per trophic level does not appear to apply to all organisms examined in this study. The S. oualaniensis sampled appear to be on average lighter than their stomach contents would suggest. This discrepancy can be only partially explained by sampling bias ofprey items. Other discrepancies arise when using the accepted increase, namely the high values obtained for paralarvae, some zooplankton, and for certain myctophids. Perhaps there is a discontinuity between smaller pelagic food webs and larger ones. This discontinuity, and the different energetic pathways that connect the two systems, may also lead to problems with using the accepted trophic level isotopic increase. Fractionation factors for lower trophic level organisms are almost certainly different from higher trophic level organisms; assimilation efficiencies certainly are. Also fractionation factors may change between different phyla ofhigher organisms, confounding the tidy picture ofa set numerical isotopic increase with trophic level. Clearly more work needs to be done before stable isotopic analyses reach their full potential as tools for elucidating ecosystem structure.

242 APPENDIX, AIle VALUES 243 OMMASTREPHES BARTRAMII IJ. "c (ADULTS) 243 OMMASTREPHES BARTRAMII IJ.13C (NORTHERN SAMPLES) 245 OMMASTREPHES BARTRAM II IJ.13C JlNENILES 246 OMMASTREPHES BARTRAM II IJ. l3C 247 STHENOTEUTHIS OUALANIENSISIJ. l3C (ADULTS) 247 STHENOTEUTHIS OUALANIENSIS Ol3C 249

Appendix: 6 13C Values

Ommastrephes bartramii 6 13C (adults)

2-14/15-99 #14 F (570): The Ii l3C ofthe tens segments oflhis squid ranged from-20.61 to-19.470/00; the mantle value was -21.00%0 and the blood was -20.47%0. The lens values increased initially from 159mm ML (-19.650/00) and then decreased to 239mm ML (20.47%0) where the values remained fairly constant to 314mm ML (­ 20.44%0) and then increased to 372mm ML (.19.50%0) followed by a rapid decrease to 417mm ML (. 20.260/00) and a more gradual decrease to 570mm ML (·20.50%0) (Fig. 36).

2-16/17-99 # 3 F (446): The Iil3C ofthe lens segments oflhis squid ranged from -20.55 to -19.270/00; the mantle value was -21.48%0 and blood was -20.890/00. The lens values generally decreased (with one peak at 152mm ML of­ 19.950/00) from 118mmML (-20.1%0) to 20lmmML (-20.37%0) then increased to 304mmML (-19.27%0) (with one peak at 245mmML, -19.55%0) and finally decreased to 446mmML (-20.55%o)(Fig. 37).

2-7/8-00 #1 M (333): The Ii l3C ofthe lens segments oflhis squid ranged from -20.95 to -19.440/00; the mantle value was -20.920/00. The lens vatues increased from the ftrst point at 125mm ML (-19.73%0) to 151mm ML (­ 19.44%0) and then decreased to the last point (with one peak at 224mm ML••19.76%0) at 333mm ML (­ 20.95%o)(Fig. 38).

2-8/9-00 #1 M (318): The Ii l3C ofthe tens segments ranged from -20.41 to -18.90%0; the mantle value was -20.460/00. The Ii l3C ofthe gladius segments ranged from -20.29 to -19.70%0. The lens values generally increased from 107mmML (-20.140/00) to 154mmML (-18.90) then decreased to 318mmML (-20.4I%o)(Fig. 39).

243 The gladius values increased to 40rnm ML (-19.95%.) and then remained relatively constant to 200rnm ML (-20.460/00) and finally increased to 318rnm ML (-19.70%0) (Fig. 39).

2-819-00 #3 M (296): The ,,"e ofthe lens segments ranged from -20.39 to -19.340/00; the mantle value was -20.63%0. The o"e ofthe gladius segments ranged from -20.54 to -19.620/00. The lens values showed an overall oscillatory decrease from 96rnm ML (-19.34%0) to 296rnm ML (-20.390/00) with two peaks around 150rnm ML (-19.37%0) and 221 (-19.53%0) (Fig. 40). The gladius values remained relatively constant to 70rnm ML (-19.97%0) theu decreased to 140rnm ML (-20.54%0) and finally increased to 296rnm ML (-19.620/00) (Fig. 40).

2-819-00 #4 M (354): The ,,"e ofthe lens segments rangedfrom-19.41 to -20.68%0; the mautle value was -20.45%0 and the blood value -20.29%•. The ,,"e ofthe gladius segments ranged from-20.34 to -19.320/00. The lens values iucreased slightly from 119 rnm ML (-20.430/00) to 209rnm ML (-20.25%.), and then increased up to 231rnm ML (-19.41) followed by a decrease to 354rnm ML (-20.680/00) (Fig. 41). The gladius values increased from 244 rnmML (-20.340/00) to a plateau between 329mmML (-19.75) and 339mmML (-19.71) followed by a fmal increase to 354rnm ML (-19.320/00) (Fig. 41).

2-819-00 #5 M (314): The "I'e ofthe gladius values ranged from-20.19 to -19.49%0; the mantle value was -20.460/00 and blood was 19.98%0. The gladius values initially increased to 40mm ML (-19.790/00) and remained relatively constant to 314rnm ML (-19.49%0) with one dip at 140mm ML (-19.69%0) (Fig. 42).

2-9110-00 #1 (322): The "I'e ofthe gladius values ranged from -20.39 to -18.890/00; the mantle value was -20.41 %0 and the blood was -20.21%•. The gladius values showed an overall oscillatory increase from 207mm ML (­ 20.26%0) to 322mmML (-18.91%0) (Fig. 43).

2-9110-00 #2 (321): The "Be ofthe gladius values ranged from -20.24 to -18.650/00; the mantle value was -20.54%0 and blood was -20.27%0. The gladius values showed an overall oscillatory increase from 206mm ML (-19.83%0) to 32lmm ML (-18.650/00) (Fig. 44).

2-9110-00 #4 (540):

244 The o13e ofthe lens values ranged from -20.97 to -19.28%0; the mantle value was -20.69%0 and the blnod was -20.43%0. The lens values showed an overall slight oscillating decrease from 140rnm ML (­ 19.43%0) to 407mm ML (-19.670/",,) followed by a rapid decrease to 540mm (-20.970/",,) (Fig. 45).

2-9/10-00 #5 (360): The o13e ofthe lens values ranged from -21.58 to .20.160/",,; the mantle value was -21.07%0 and blood was -20.850/"". The lens values showed an overall decrease (with substantial oscillations) to 270rnm ML (-20.530/",,) followed bya rapid decrease to 360mm ML (-21.580/",,) (Fig. 46).

2-11/12-01 #1 (557): The o"e ofthe lens values ranged from -21.79 to -20.030/",,; the mantle value was -21.68%0 and blood was -21.42%0. The lens values showed an initial increase from 150rnm ML (-21.010/",,) to 247mm ML (-20.030/",,) then decreased to 354mm ML (-21.520/",,) followed by a slight increase and plateau from 417 mmML (-21.28%0) to 503mmML (-21.33%0) with a final decrease to 557mm ML (-21.79%0) (Fig. 47).

2-11/12-01 #2 (370):

The 013e ofthe lens values ranged from -20.91 to -20.22%0; the mantle value was -20.930/00 and the blood was -20.57%0. The gladius values ranged from -21.50%0 to -18.69%0. The lens showed an initial plateau (with oscillations) from 97mm ML (-20.360/",,) to 217mm ML (-20.22%0) followed by a fmal decrease to 370mm ML (-20.91%0) (Fig. 48). The values showed an initial plateau from 255mmML (­ 20.36%0) to 295mm ML (-20.430/",,), a decrease to 310mmML (-21.50%0) and a final increase up to 370mm ML (-18.69%0) (Fig. 48).

2-11/12-01 #5 (534): The o"e ofthe lens values ranged from-21.27 to -19.910/",,; the mantle value was -20.60%0. The lens showed an overall decrease from 140rnmML (-20.140/00) to 340mmML (-21.27%0), with two peaks at 163mmML (-19.91%0) and 243mm ML (-20.020/",,), followed by a general increase to 534mmML (­ 20.420/00) (Fig. 49).

Ommastrephes bartramii 613C (northern samples)

#21 (220):

245 The aBC ofthe lens values ranged from -21.05 to -19.86%<>, and the mantle value was -20.08%•. The lens showed an overall increase from 130mm ML (-21.05%0) to 220mm ML (-19.860/00) with peaks at 153mm ML (-20.550/00) and I73mm ML (-20.380/00) (Fig. 50).

#25 (255): The aBC ofthe lens values ranged from -20.60 to -20.130/00, and the mautle value was -20.43%0. The lens showed an initial increase from 144mm ML (-20.360/00) followed by a decrease to a plateau between 168mmML (-20.58%0) to 224mm ML (-20.560/00) followed by a peak at 243mmML (-20.28%0) and a fmal decrease at 255mm ML (-20.430/00) (Fig. 51).

#26 (201): The aBC values ofthe lens ranged from -20.85 to -20.320/00, and the mantle value was -20.380/00. The lens showed a decrease from 117 mm ML (-20.450/00) to 153mm ML (-20.85%0), and then increased to 174mm ML (-20.320/00) followed by a plateau to 20lmm ML (-20.380/00) (Fig. 52).

Ommastrephes bartramii 613Cjuveniles

2-16/17-98 #25 (142): The aBC ofthe lens values ranged from -20.01 to -19.830/00, and the mautle value was -19.55%0. The lens showed a decrease from 94mmML (-19.83%0) to a plateau from 127mmML (-20.010/00) to 142mmML(-19.99) (Fig. 53).

2-16/17-98 #26 (151): The a"C ofthe lens values ranged from -20.23 to -19.670/00, and the mantle value was -20.03%0. The lens decreased from 96mm ML (-19.670/00) to I 26mm ML (-20.230/00) and then increased to 151mm ML (-20.03%0) (Fig. 54).

2-11112-99 #5 (161): The aBC ofthe lens values ranged from 118.88 to -18.69%0, and the mautle value was -18.72%0. The lens showed fairly constant values from 94mm ML (-18.88%0) to 161mm ML (-18.820/00) (Fig. 55).

2-11/12-99 #6 (ISO): The aBC ofthe lens values ranged from -19.25 to -18.780/00, and the mantle value was -19.82%0. The lens showed an initial increase from 82mm ML (-18.970/00) to 93mm ML (-18.78%0), a decrease to l11mm ML (-19.250/00) and an oscillatory increase to 150mm ML (-18.98%0) (Fig. 56).

246 Ommastrephes bartramii 6 13C

Preliminary analysis ofthe o"e values for the eye lens segments might indicate a geographical change in isotopic value ofan ecosystem. Although the factors governing the signals ofo"e aud ol5N in the envirorunent, and within an organism, are wholly separate the carbon signal could potentially indicate movement into a different ecosystem. The 0"e signals from the eye lenses aud the muscle ofthe four juvenile 0. bartramii caught in Hawaiian waters did not show auy distinctive patterns and exhibited small fluctuations overall (the largest chauge injuvenile eye lens segments was 0.560/00) (Figs. 53-56). The eye lens segments from the three northern 0. bartramii individuals also showed no distinctive general patterns with length, but did seem to start at lower 0"e values than the juveniles. The muscle values for the northern squids were close to the lower eud ofo"e values seen for the four juvenile 0. bartramii (z200/00) (Figs. 50-52). The 0"e signals ofthe eye lens segments from adult 0. bartramii also did not share a common general trend, aud most fluctuations ofthe 0"e were small, on the order of 1.50%0 (36-49). In general the o"e values ofthe muscle were lighter than the muscle for juveniles, while the eye lens aud gladius values fluctuated and overlapped the rauge for juveniles. The o"e values for the of2-9/10-00 #2 and 2-8/9-00 #4 both steadily increased over the size rauges sampled, and could indicate a steady movement into another isotopic regime (Figs. 44, 41). For 2-8/9-00 #1, #3, aud #5 showed very little change but all got heavier at the [mal segment (Figs. 39, 40, 42).

Sthenoteuthis oualaniensis 6 13C (adults)

2-14/15-99 #1 (324), The o"e values ofthe lens ranged from-19.31 to -18.83%0, and the mantle value was -19.700/00. The lens showed a decrease from 129mmML (-19.00%0) to 138mmML (-19.310/00), au increase to 196mm ML (-18.830/00) followed by a decrease to 324mm ML (-19.700/00) (Fig. 57).

2-14/15-99 #3 (200), The o"e values ofthe lens ranged from -19.32 to -18.83%0, and the mantle value was -19.19%0. The lens showed au increase from 71mm ML (-19.320/00) to 146mm ML (-18.83%0) aud then decreased to 200mm ML (-19.30%0) (Fig. 58).

2-14/15-99 #10 (223), The o"e values ofthe lens rauged from -20.30 to -18.81 %0, aud the mantle value was -19.170/00. The lens values increased from 77mmML (-19.25%0) to 140mm ML (-18.810/00), with one dip at 112mm ML (-19.27), followed by au decrease to 223mm ML (-20.300/00) (Fig. 59).

247 2-7/8-00 #1 (210): The 0Be values ofthe lens ranged from -19.00 to -18.070/00, and the mantle value was -18.66%0. The lens values showed an overall increase from 105nunML (-19.00%0) to 190nunML (-18.070/00) and then decreased to 210nunML (-18.38%0) (Fig. 60).

2-12/13-00 #1 (209): The /) Be values ofthe lens ranged from -18.87 to -18.25%0, and the mantle value was -18.65%0. The lens values remained fairly constant, with some variation, from 90nun ML (-18.75%0) to 209nun ML (­ 18.650/00) with a peak atl98nun ML (-18.250/00) (Fig. 61).

2-12/13-00 #2 (160): The oBe values ofthe lens ranged from -19.0910 -18.350/00, and the mantle value was -18.35%0. The lens values increased from 78nunML (-19.09%0) 10 84mm ML (-18.54%0), decreased to 94nun ML (­ 19.03%0), increased slightly to 130nun ML (-18.35%0), decreased again to 137nun ML (-18.60%0) and fmally increased to 160nun ML (-18.35%0) (Fig. 62).

2-12113-00 #3 (181): The oBe values ofthe lens ranged from -19.60 to -18.850/00, the mantle value was -18.610/00, and blood was -18.35%0. The lens values decreased overall from 94nun ML (-19.000/00) to 123nun ML (­ 19.600/00) then increased to 160rnm ML (-18.86%0) then remained constant to 181nun ML (-18.85%0) (Fig. 63). The values ofthe gladius showed an overall decrease from 106nun ML (-17.320/00) to 161nun ML (­ 18.230/00), then increased to a plateau between 166nun ML (-17.900/00) and 176mm ML (-17.84%0) and finally decreased to 181mmML (-18.20%0) (Fig. 63).

2-12/13-00 #4 (173): The oBe values ofthe lens ranged from -18.98 to -18.420/00, the mantle value was -18.28%0. The values ofthe gladius ranged from -17.66 to -16.93%0. The lens values remained relatively constant from 91mm ML (-18.85%0) to 122mm ML (-18.98%0), increased to 138nun ML (-18.420/00) then decreased overall to 173nun ML (-18.740/00) (Fig. 64). The values ofthe gladius remained relatively constant from 95mm ML (-17.50/00) to 145nun ML (-17.640/00), increased to 165nun ML (.16.93%0) then fmally decreased to (-17.27%0) (Fig. 64).

2-12/13-00 #6 (158): The OBe values ofthe lens ranged from -19.11 to -18.200/00, the mantle value was -18.55%0. The lens values increased from 85mm ML (-19.11%0) to 104nun ML (-18.20%0), decreased to 123nun ML (­ 19.05%0) and finally increased to 158mmML (-18.55%0) (Fig. 65).

248 2-13114-00 #3 (251): The o"e values ofthe lens ranged from -19.03 to -18.23%0, the mantle value was -18.34%0 and blood was -18.21%•. The lens values increased from 11 Imm ML (-19.030/00) to 200mmML (-18.38%0), decreased to 221mmML (-18.71%0), increased to 234mmML (-18.23%0) and finally decreased to 251mm ML (-18.60%0) (Fig. 66).

2-12113-01 #1 (225): The o"e values ofthe lens ranged from -19.81 to -18.26%0, the mantle value was -17.630/00 and blood was -17.63%0. The lens values remained relatively constant from 126mm ML (-18.48%0) to 210mm ML (-18.370/00) then decreased to 225mm ML (-19.810/00) (Fig. 67).

2-12/13-01 #11 (304): The 0"e values ofthe lens ranged from -19.68 to -17.90%0, the mantle value was -19.22%0 and blood was-18.7I%o. The lens values increased from 98mm ML (-18.74%0) to 11 Imm ML (-17.90%0), followed by an overall decrease to 304mm ML (-19.68%0) (Fig. 68).

2-12113-01 #12 (302), The o"e values ofthe lens ranged from-19.48 to -17.780/00, the mantle value was -18.58%•. The lens values increased slightly from 10Imm ML (-18.250/00) to 161mm ML (-17.990/00), followed by an overall decrease to 302mm ML (-19.48%0), with two peaks at 211mm ML (-17.78%0) and 243mmML (­ 17.95%0) (Fig. 69).

2-12113-01 #16 (308): The ol'e values ofthe lens ranged from -19.75 to -18.11%0, the mantle value was -18.88%0 and blood was -18.70%0. The lens values increased initially from 95mm ML (-18.840/00) to 112mm ML (­ 18.110/00) and remained relatively constant to 223mm ML (-18.240/00) followed by a decrease to 308mm ML (-19.750/00) (Fig. 70).

Sthenoteuthis oualaniensis st3c

The 0"e values for the eye lens segments ofS. oua/an/ensis stayed fairly constant at smaller sizes for many ofthe individuals, and decreased near the outer segments (Figs. 57-59, 61-64, 66-70). The overall variation in 0Be was low; all S. oualan/ens/s showed variations in the eye lens segments of< 20/00. Squid 2­ 12/13-00 #3, the squid with the high ol'N values for muscle and eye lens segments, showed a differing OBe

249 pattern for eye lenses and for gladius values (Fig. 63). The gladius O"C values showed an overall decrease while the eye lenses showed an initial decrease followed by an increase to a plateau. The gladius and eye lens segments for sqnid 2-12/13-00 #4 both showed similar O"C patterns (Fig. 64). Ths difference between squids 2·12113-00 #3 and #4 could indicate that there were physiological differences in #3 that may have caused the high o15N values (Figs. 63, 64).

250 -18 -r;======;------,

-A- Eye • Mantle o Blood

-19

-21 •

2-14/15-99 #1 (570mm) -22 +-----,------.------r------,,-----,------j a 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 36. Ii"e versus length ofeye lens segments ofa female 0. bartramii. Ii I'e ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

251 -18 Tr====::::;------~ -- Eye • Mantle <> Blood

-19

-21 <> • 2-16/17-99 #5 (446) -22 +----,----,----,----,------,------j o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 37. o"e versus length ofeye lens segments ofa female 0. bartramii. o"e ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

252 -18 F====,------~ ~Eye • Mantle

-19

u ~ -20 00

-21

2-7/8-00 #1 (333mm) -22 -l------,------.------r------,----,------j o 100 200 300 400 500 600 o. bartramii mantle length (mm)

Figure 38. a13e versus length ofeye lens segments ofa male 0. bartramii. ol3e ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

253 ·18 T;:::=====;------~ ---'- Eye ~ Gladius • Mantle

-19

-20

-21

2-8/9-00 #1 (318mm) -22 -f----r------r----,------,------,------j o 100 200 300 400 500 600 O. barlramii mantle length (mm)

Figure 39.Il 13C versus length ofeye lens segments ofa male 0. bartramii.Il 13C ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

254 -19 ~------~

-20

U '"~eo •

-21

-.>- Eye ~ Gladius Mantle • 2-8/9-00 #3 (296mm) -22 0 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 40. o"e versus length ofeye lens and gladius segments ofa male 0. bartramii. one ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

255 -18 T;:::===:::::::;------, ~Eye ~ Gladius • Mantle <> Blood -19

() ~ -20 GO

-21

2-8/9-00 #4 (354mm) -22 +------,-----r------r------,----,------i o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 41. 1\ l3e versus length ofeye lens and gladius segments ofa male 0. bartramii. 1\ l3e ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

256 -18 ,-;::::===::::::;------~

-<>- Gladiu5 • Mantle ¢ Blood

-19

¢ •

-21

2-8/9-00 #5 (314mm) -22 -I-----.-----.-----.------.------.------i o 100 200 300 400 500 600 O. bartramii mantle length (mm)

l3 Figure 42, 0l3e versus length ofgladius segments ofa mate 0. bartramii. o e ofmantle muscle and blood plotted against mantle length at time ofcapture, Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size, Date caught, squid number, and mantle length noted in bottom right comer.

257 -18 ---<>- Gladius Mantle <>• Blood

-19

f () -20 N '"~oo <> J •

-21

2-9/10-00 #1 (322mm) -22 -+----,----,----,----,----,------1 o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 43. IiBe versus length ofgladius segments ofa male 0. bartramii. Ii Be ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, sqnid number, and mantle length noted in bottom right corner.

258 -18 T;:::======;------~ ~ Gladius • Mantle <) Blood

-19

",.u -20 ~oo •

-21

2-9/10-00 #2 (321mm) -22 -j----,------,------,----,----,------j o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 44. o"e versus length ofgladius segments ofa male a. bartramii. one ofmantle muscle and blood plotted against mantle length a( time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

259 -18

-J>- Eye Manlle <>• Blood -19

;? -20 ~

-21

2-9/10-00 #4 (540mm) -22 +----,.------,------.---,----,...------1 o 100 200 300 400 500 600 o bartramii mantle length (mm)

Figure 45. oBe versus leugth ofeye lens segments ofa female 0. bartramii. OBe ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

260 -18 Tj====::::;------~

---'>- Eye • Mantle <) Blood

-19

"'() -20 ~GO

<) -21 •

2-9/10-00 #5 (360mm) -22 +----r------,----,------,-----,------1 o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 46. o"e versus length ofeye lens segments of a male 0. bar/ramii. o"e ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

261 -18 T.====::::::;------~ 2-11/12-01 #1 (557mm) -'>- Eye • Mantle <> Blood

-19

-21

<>

-22 +----,-----,-----,-----,-----,------j o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 47. OBC versus length ofeye lens segments ofa female 0. bartramii. OBC ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom rigbt comer.

262 -18 T;:::=====;------, ~Eye ~ Gladius • Mantle <> Blood -19

-20

-21

2-11/12-01 #2 (370mm) -22 +----,------,------,------.-----,------1 o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 48. OBe versus length ofeye lens and gladius segments ofa male 0. bartram;;. oBe ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

263 -18 T.====:::::::;------~

~Eye • Mantle

-19

-20 9~

2-11/12-01 #5 (534mm)

-22 +-----,------,------,------,----,------1 o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 49. o"c versus length ofeye lens segments ofa female 0. bartramii. OBC ofmantle mnscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

264 -18 -rr=====;------, -<>- Eye • Mantle

-19

-20

-21

Northern O. barlramii #21 (220mm) -22 +----,------,--~--,___---___"r----__,,----___1 o 100 200 300 400 500 600 O. barlramii mantle length (mm)

Figure 50, ol3e versus length ofeye lens segments ofa male northern 0. bartramii. ol3e ofmantle muscle plotted against mantle length at time ofcapture, Mantle length for eye lens segments estimated from regressions ofeye lens radius aud squid size. Date caught, squid number, and mantle length noted in bottom right comer.

265 -18 T;====::;------i

--<>- Eye • Mantle

-19

() '" -20 -00 ~

-21

Northern O. bartramii #22 (255mm) -22 +------r----,r-----,------r----,------1 o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 51. o"C versus length ofeye lens segments ofa female northem 0. bartramii. O"C ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted io bottom right comer.

266 -18 Tr====:::;------i ~ Eye • Mantle

-19

(.) '"~ -20 00 • V -21

Northern O. bartramii #26 (201 mm) -22 +----,------.------.------,----,------1 o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 52. ~ l3e versus length ofeye lens segments ofa female northern 0. bartramii. ~l3e ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

267 -18

---A- Eye • Mantle

-19 • () -20 ~ '"~oo

-21

2-16/17-98 #25 (149mm) -22 +----,-----,-----,-----,-----,------1 o 100 200 300 400 500 600 O. bartramii mantle length (mm)

Figure 53. /j Be versus length ofeye lens segments ofa juvenile 0. bartramii. /j13e ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught. squid number, and mantle length noted in bottom right comer.

268 -18 -r,======;------, ~ Eye • Mantle

-19

() M -20 ~c<>

-21

2-16/17-98 #26 (151mm) -22 -t----.,------,------,----,------,------J o 100 200 300 400 500 600 o. bartramii mantle length (mm)

Figure 54. {j"e versus length of eye lens segments ofa juvenile 0. bartramii. {jl3e ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

269 -18 ,------,

-19

u .., -20 ~oo

-21

--<-- Eye • Mantle 2-11/12-99 #5 (161mm) -22 +'=====r==!.------.-----,------,-----,------i o 100 200 300 400 500 600 O. barlramii mantle length (mm)

Figure 55. i)"e versus length ofeye lens segments ofa juvenile 0. bartramii. i)"e ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid numher, and mantle length noted in bottom right comer.

270 -18 -,------,

-19

()

'"~

-21

...... >-- Eye • Mantle 2-11/12-99 #6 (150mm) -22 , , 0 100 200 300 400 500 600 o. bartramii mantle length (mm)

Figure 56. o13e versus length ofeye lens segments ofa juvenile 0. bartramii. ol3e of mantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

271 ·16 Tr====:::::;------~

--<>- Eye • Mantle

·17

·18 <..l '"~<.O ·19 •

-20

2-14/15-99 #1 (324mm) ·21 +---~---__,__--___,---_._---._--____r--_____1 o 50 100 150 200 250 300 350 S. oualaniensis mantle length (mm)

Figure 57. Ii 13C versus length ofeye lens segments ofa female S. oua/aniensis. Ii 13C ofmantle omscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

272 -16 -r;=====,------, ~ Eye • Mantle

-17

-18

-19

-20

2-14/15-99 #3 (200mm) -21 o 100 200 300 s. Dua/aniensis mantle length (mm)

Figure 58. ol3e versus length ofeye leus segments ofa female S. oua/an/ensis. o13e ofmantle muscle plotted against mantle length at lime ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer

273 -16 T;:::===::::;------~

-<.- Eye • Mantle

-17

-20

2-14/15-99 #10 (223mm) -21 -l------.------~------___,_----l o 100 200 300 S. oualaniensis mantle length (mm)

Figure 59. ,,"e versus length ofeye lens segments ofa female S. oualaniensis. "I'e ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer

274 -16 Tr=====J------~ ~ Eye • Mantle

-17

-18 • -19

-20 -

2-7/8-00 #1 (210mm) -21 +------.,------r------,-----' o 100 200 300 S. Dua/aniensis mantle length (mm)

Figure 60. Ii"e versus length ofeye lens segments ofa female S. oualaniensis. Ii 13C ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

275 -16 Tr======:;------j --- Eye • Mantle -17

-18

() • '"~oO

-19

-20

2-12/13-00 #1 (209mm)

-21 o 100 200 300 s. oualaniensis mantle length (mm)

Figure 61. OBe versus length ofeye lens segments ofa female S. aua/aniensis. OBe ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

276 -16 Tc====::::;------~

-.0.- Eye • Mantle

-17

-18 •

-19 ~

-20

2-12/13-00 #2 (160mm) -21 o 100 200 300 s. oualaniensis mantle length (mm)

Figure 62. 613C versus length ofeye lens segments ofa female S. oualaniensis. 613C ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

277 -16 T.=====;------~ ~ Eye - Gladius • Mantle -17 <> Blood

-18 <>• -19

-20

2-12/13-00 #3 (181mm) -21 o 100 200 300 S. Dualaniensis mantle length (mm)

Figure 63. Ii Be versus length ofeye lens and gladius segments ofa female S. oualaniensis. IiBe ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

278 -16

-0- Eye ~ Gladius Mantle -17 • ~

-18 - () • '"~oo ~ -19

-20

2-12/13-00 #4 (173mm) -21 +------,------,------r------' o 100 200 300 S. Dua/an/ensis mantle length (mm)

Figure 64. ol3e versus length ofeye lens and gladius segments of a female S. oua/aniensis. ol3e ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

279 -16 Tr====::::::;------~

-<>- Eye • Mantle

-17 -

-18 • () '"~oo

-19

-20

2-12/13-00#6 (158mm) -21 o 100 200 300 S. ouaJaniensis mantle length (mm)

Figure 65. ol3c versus length ofeye lens segments ofa male S. oua/aniensls. OBC ofmantle muscle plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radins and squid size. Date caught, squid number, and mantle length noted in bottom right corner.

280 -16 -r;::===::::;------'------~

--<>-- Eye • Mantle <> Blood -17

-18

-19

-20

2-13/14-00 #3 (251mm) -21 o 100 200 300 S. oualaniensis mantle length (mm)

Figure 66./i13C versus length ofeye lens segments ofa female S. oualaniensis. /i 13C ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

281 -16

-<>- Eye Mantle • Blood -17 <> • -18 () "'-- '"~"" -19

-20

2-12/13-01 #1 (225mm) -21 o 100 200 300 S. oualaniensis mantle length (mm)

Figure 67.li13C versus length ofeye lens segments ofa female S. oualaniensis.li 13C ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

282 -16 Tr====:::;------~

-->- Eye • Mantle <> Blood -17

-18

() '" ~"" <> -19

-20

2-12/13-01 #11 (304mm) -21 +------___.------..,---~--___.----' o 100 200 300 S. oualaniensis mantle length (mm)

Figure 68. o13e versus length ofeye lens segments ofa female S. oualaniensis. o13e ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom rigbt corner.

283 -16 T;:=====;------, ---<>- Eye • Mantle

-17

-18 • -19

-20

2-12/13-01 #12 (302mm) -21 +------,------,------,------' o 100 200 300 S. oualaniensis mantle length (mm)

Figure 69. O"C versus length ofeye lens segments ofa female S. oualaniensis. O"C ofmantle nrusc1e plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid number, and mantle length noted in bottom right comer.

284 -16 T;:::===:::::::;------~ ~ Eye • Mantle <> Blood -17

-18

() '"~ "" <> -19 •

-20

2-12/13-01 #16 (308mm) -21 o 100 200 300 S. oua/an/ensis mantle length (mm)

Figure 70. o"e versus length ofeye lens segments ofa female S. oua/aniensis. one ofmantle muscle and blood plotted against mantle length at time ofcapture. Mantle length for eye lens segments estimated from regressions ofeye lens radius and squid size. Date caught, squid nurtiber, and mantle length noted in bottom right comer.

285