Comparative trophic ecology of yellowfin and bigeye tuna associated with natural and man-made aggregation sites in Hawaiian waters

Transcription

1 SCTB15 Working Paper YFT-6 Comparative trophic ecology of yellowfin and bigeye tuna associated with natural and man-made aggregation sites in Hawaiian waters R. Dean Grubbs 1, Kim Holland 1, and David Itano 2 1 Hawaii Institute of Marine Biology University of Hawaii 2 Pelagic Fisheries Research Program Joint Institute of Marine and Atmospheric Research University of Hawaii

2 INTRODUCTION Comparative trophic ecology of yellowfin and bigeye tuna associated with natural and man-made aggregation sites in Hawaiian waters. R. Dean Grubbs 1, Kim N. Holland 1, and David G. Itano 2 1 Hawaii Institute of Marine Biology University of Hawaii Coconut Island, Kaneohe, Hawaii University of Hawaii at Manoa, JIMAR Pelagic Fisheries Research Program Honolulu, Hawaii The formation of aggregations, groups of fishes attracted to a common resource (Freeman and Grossman 1992), is a common phenomenon among tropical tunas. These aggregations typically form in association with large physical entities that may be naturally occurring (e.g. cetacean pods, seamounts, floating debris) or man-made (e.g. weather buoys, drifting vessels, structures deployed specifically as Fish Aggregating Devices - FADs). Regardless of the associated structure, the underlying biological significance of these aggregation behaviors is poorly understood (Dagorn and Fréon 1999, Holland et al. 1999). These unknowns are compounded when one considers mixed-species aggregations. Theoretically, aggregating may be highly adaptive, imparting several advantages to group members such as decreasing the risk of predation, increasing foraging efficiency, and increasing reproductive success. This behavior has the maladaptive side effect, however, of increasing the susceptibility to exploitation by modern fishing gear due by locally increasing fish density. Indeed, the majority of tropical tunas captured by commercial and recreational fishers worldwide are associated. The construction of artificial fish aggregating devices (FADs) designed to increase catch rates of pelagic species began long ago and has become commonplace throughout the tropical Pacific (Brock 1985, Buckley et al.1989, Buckley and Miller 1994). The effects of the industrial-scale removal of aggregated tuna on tuna stocks and the ecosystem from which they are removed is largely unknown. Understanding the trophic dynamics of associated tuna is critical to properly evaluating these effects. In the spring of 21, we began an intensive investigation into the trophic ecology of yellowfin tuna (Thunnus albacares) and bigeye tuna (T. obesus) aggregations in Hawaiian waters. Hawai i offers an ideal setting to investigate the trophic ecology of tuna aggregations due to an abundance of aggregating sites, natural and artificial. and bigeye tuna occur in single-species and mixed-species aggregations around these structures. The primary aggregation sites of interest are four offshore NOAA weather buoys, the Cross Seamount (18 4' N, 158 1' W), and the multitude of nearshore buoys and FADs (Figure 1). To evaluate the potential trophic advantage of associating with these structures, we are investigating the trophic ecology of unassociated yellowfin and bigeye tuna as a control group. This is an ongoing research project. The results to date are presented here.

3 Honolulu, Hawaii, 2 th 25 th July 22 Figure 1: Sampling locations used to investigate the trophic ecology of associated bigeye and yellowfin tuna in Hawaiian waters. METHODS Samples are collected via port sampling and field sampling. Port samples are occasionally collected directly from returning commercial hand-line vessels, but most often port sampling is conducted with the cooperation of Honolulu seafood buyers. Samples are taken after fish are purchased at the United Fishing Agency auction. This allows the collection of samples from many different hand-line vessels fishing the various aggregation sites. Vessel captains are subsequently interviewed to obtain information concerning the location and date of capture as well as the bait and gear used. Port sampling also allows the collection of samples from unassociated fish caught by offshore longline vessels. Field sampling is conducted aboard the Fishing Vessel Double D, a commercial hand-line vessel. Samples are taken as soon as fish are brought aboard and immediately preserved in formalin to stop the digestion process. As with the port samples, the field samples allow the comparison of prey species taken by the two tuna species while associated with the Cross Seamount and the offshore NOAA weather buoys. In addition, the field samples may allow the determination and comparison of the timing of feeding in bigeye and yellowfin tuna. Additional information such as water depth and depth of capture is also available using field sampling. 2

4 Honolulu, Hawaii, 2 th 25 th July 22 The species, fork length, and weight (if available) are recorded for each fish. Samples are collected from fish destined for retail markets; therefore, the collection method must cause minimal damage to the flesh of the fish. Stomach contents are collected by entering the peritoneal cavity through the opercular opening, severing the esophagus, and excising the entire stomach. Large stomachs are initially injected with 1% phosphate-buffered formalin. All stomachs are immersed in the formalin solution for at least 72 hours to ensure preservation. Samples are then stored in 5% ethyl alcohol until lab analysis. In the laboratory, stomachs are rinsed and all contents removed. All prey items are identified to the lowest possible taxon. The number, wet volume, and digestion state of each prey taxon is recorded. All data are entered into a parent database using Microsoft Access for subsequent analysis. Stomach fullness is estimated for each fish as the total volume of prey (ml) per kilogram of body weight. Fish body weight is estimated using length-weight relationships. PRELIMINARY RESULTS To date, 81 samples have been collected, 436 in port and 374 at sea (Table 1). Port samples were taken from fish landed by sixteen different commercial vessels. Sixty percent (N=483) of the tuna sampled were captured at the Cross Seamount and 32% (N=261) were captured near offshore weather buoys, all by the commercial handline fleet. The remaining eight percent were unassociated fish captured by the commercial longline fleet (N=56) or fish associated with nearshore FADs that were captured by our research group (N=1). Landings by the commercial handline fishery in Hawai i are dominated by bigeye tuna. This is reflected in our samples, as 71% are bigeye. Ninety-five percent of all samples have been analyzed in the laboratory. Table 1: Stomach sample summary Association Unassociated Cross Seamount Offshore Weather Buoys Nearshore FADs 1 TOTAL Total Analyzed

5 Honolulu, Hawaii, 2 th 25 th July 22 Stomach Fullness Stomach fullness was defined as the volume of prey in milliliters per kilogram of body weight. The significance of differences in stomach fullness within each tuna species was tested using and Analysis of Variance (ANOVA) with association as a single factor. For yellowfin tuna, stomach fullness was slightly higher for those associated with the Cross Seamount than either those associated with offshore weather buoys or those defined as unassociated (Figure 2). Statistically, the differences were insignificant however (Table 2: ANOVA; F=.45,? =.636). This suggests there is no trophic advantage to associating with the Cross Seamount or the offshore weather buoys for yellowfin tuna. For bigeye tuna, stomach fullness in those associated with the Cross Seamount was nearly four times that of unassociated bigeye tuna and forty-five times that of those associated with the weather buoys (Figure 2). These differences were highly significant (Table 3: ANOVA; F=4.27,? <.1) and suggest that, for bigeye tuna, there is a significant trophic advantage to associating with the Cross Seamount but no trophic advantage, and possibly a detriment, to associating with the offshore buoys. The significance of differences in stomach fullness between the two tuna species was tested for each association using a series of two-sample t-tests assuming unequal variances (Table 4). Stomach fullness for unassociated yellowfin and bigeye tuna were statistically similar (t=.39,? =.7). On the Cross Seamount, stomach fullness for bigeye tuna was nearly three times that of yellowfin (Figure 2), a difference that was highly significant (t=6.64,?<.1). For tuna associated with the offshore weather buoys, stomach fullness in yellowfin tuna was more than ten times higher than that of bigeye tuna (Figure 2). This difference was also highly significant (t=3.5,?=.31) Mean Stomach Fullness Seamount Offshore Buoy Unassociated Figure 2: Mean stomach fullness of yellowfin and bigeye tuna as a function of association. 4

6 Honolulu, Hawaii, 2 th 25 th July 22 Table 2: Summary of analysis of variance using stomach fullness of yellowfin tuna as the response variable and association (seamount, offshore buoy, or unassociated) as the single factor. Anova: Single Factor Source of Variation SS df MS F P-value Between Groups Within Groups Total Table 3: Summary of analysis of variance using stomach fullness of bigeye tuna as the response variable and association (seamount, offshore buoy, or unassociated) as the single factor. Anova: Single Factor Source of Variation SS df MS F P-value Between Groups E-17 Within Groups Total

7 Honolulu, Hawaii, 2 th 25 th July 22 Table 4: Results of two-sample t-tests comparing stomach fullness of bigeye and yellowfin tuna in different associations. t statistic? (sign.) N Mean Fullness N Mean Fullness µ 1 = µ 2 2-tailed Unassociated ml/kg ml/kg.39.7 Seamount ml/kg ml/kg 6.64** <.1 Weather Buoy ml/kg ml/kg 3.5**.31 The trophic advantage of associating may also be evaluated in terms of the proportion of stomachs containing prey. An empty stomach was defined as one containing less than.1 milliliters of prey per kilogram of body weight. Figure three shows the proportion of bigeye and yellowfin stomachs containing at least.1 ml/kg and 1. ml/kg of prey for each location. The results are similar to those using stomach fullness. For unassociated tuna, 75% of samples contained prey compared to only 56% of bigeye samples, however only 3% of yellowfin contained a minimum of 1. ml/kg compared to 41% for bigeye. On the Cross Seamount, more than 75% of both tuna species contained prey, but 57% of bigeye tuna contained at least 1. ml/kg compared to only 41% of yellowfin. For tuna associated with the weather buoys, 5% of yellowfin samples contained prey and 25% contained more than 1. ml/kg, however only 1% of bigeye contained prey and only 3% contained more than 1. ml/kg SF > 1. ml/kg SF >.1 ml/kg Seamount Offshore Buoy Unassociated Seamount Offshore Buoy Unassociated Figure 3: The proportion of bigeye and yellowfin tuna containing prey. (SF = stomach fullness) 6

8 Honolulu, Hawaii, 2 th 25 th July 22 Prey Diversity The taxonomic diversity of prey was compared for yellowfin and bigeye tuna associated with the Cross Seamount and the offshore NOAA weather buoys. Prey richness was estimated using the total number of prey families and diversity was estimated using Fisher s (Figure 4). The prey diversity and richness patterns mirrored that of stomach fullness. For both tuna species, prey diversity and richness were higher for those associated with the seamount than those associated with offshore buoys. Prey richness for seamount-associated tuna was much higher for bigeye (78 taxa) than for yellowfin (45 taxa), while on the weather buoys, prey richness was higher for yellowfin tuna (22 taxa) than for bigeye tuna (12 taxa). Diet diversity (Fisher s ) mirrored this pattern. This suggests that the forage base for tunas is much more diverse on the seamount than on the offshore buoys. It also suggests the two species either employ different feeding strategies or a portion of the forage base on the seamount is inaccessible to yellowfin tuna, while the opposite is true on the offshore buoys. Number of Families Diet Richness: Prey Families Seamount Weather Buoys 1 Diet Diversity: Fisher's Alpha Fisher's Alpha Seamount Weather Buoys Figure 4: Prey richness and prey diversity for yellowfin and bigeye tuna on the Cross Seamount and the offshore NOAA weather buoys. 7

9 Honolulu, Hawaii, 2 th 25 th July 22 Feeding Strategy We used the method of Costello (199) to provide and initial evaluation of the feeding strategy used by yellowfin and bigeye tunas on the Cross Seamount and the offshore weather buoys. In this method, frequency of occurrence is plotted on x-axis versus percent abundance (numbers, volume, or weight) on the y-axis for each prey taxon. Feeding strategy can be inferred from the distribution of prey taxa and the relative importance of each taxon may be evaluated by its position on the plot. For example, for a predator that is highly specialized (Figure 5a), the frequency of occurrence and percent abundance would be high for a single prey type, therefore, that taxon would appear in the upper right hand corner of the graph. Rare prey would be represented in the lower left corner of the plot. Conversely, a predator that employs a generalized feeding strategy would have no dominant prey so the percent abundance would be low for all prey taxa while frequency of occurrence of prey taxa may be highly variable (Figure 5b). A 1 Specialized Feeding Strategy Rare Prey Dominant Prey B 1 Frequency of Occurence Generalized Feeding Strategy Wide Diet Breadth Frequency of Occurence Figure 5: Hypothetical Costello diagrams for (A) specialized and (B) generalized foraging strategies. 8

10 Honolulu, Hawaii, 2 th 25 th July 22 Costello diagrams are shown for yellowfin and bigeye tuna associated with the Cross Seamount in Figure 6. The y variable is percent abundance by numbers. These diagrams suggest that both species use remarkably similar, intermediate feeding strategies on the seamount. Note that the scale of the y-axis was truncated to 4 percent to improve visual clarity. For both species, no prey taxon constituted more than 4% of the total prey abundance, but each had one prey taxon constituting more than 3% and three taxa constituting between 1% and 2% of the total prey abundance. Of the total number of prey taxa (78 for bigeye and 45 for yellowfin) the four most abundant combined for more than 7% of the total prey for both tuna species. The most dominant prey for bigeye tuna were caridean shrimp of the family Oplophoridae, followed by peneoid shrimp of the family Sergestidae, myctophid fishes, cephalopods. The most dominant prey for yellowfin were sergestid shrimp followed by brachyuran crab megalopae, cephalopods, and oplophorid shrimp. Cross Seamount A Pandalidae Tuna Feeding Strategy Tuna Feeding Strategy Oplophoridae Myctophidae Bramidae Megalopae Pandalidae Howella Stomatopoda Peneidae Chiasmodontidae Stomatopoda Sergestidae Megalopae Cephalopoda Frequency of Occurence Myctophidae Oplophoridae Sergestidae Cephalopoda B Frequency of Occurence Figure 6: Costello diagrams for (a) bigeye tuna and (B) yellowfin tuna associated with the Cross Seamount. 9

11 Honolulu, Hawaii, 2 th 25 th July 22 Costello diagrams are shown for yellowfin and bigeye tuna associated with the offshore NOAA weather buoys in Figure 7. These also suggest that both species use similar feeding strategies, however they appear to be much more specialized than when associated with the seamount with each species specializing on a different taxon. Note that the scale of the y-axis was not truncated. More than 6% of prey items consumed by bigeye tuna were cephalopods and more than 8% of those consumed by yellowfin tuna were brachyuran crab megalopae. Offshore Weather Buoys A 1 Tuna Feeding Strategy Megalopae Cephalopoda B 1 Frequency of Occurence Tuna Feeding Strategy 8 Megalopae Chaetodontidae Thaliacea Cephalopoda Frequency of Occurence Figure 7: Costello diagrams for (a) bigeye tuna and (B) yellowfin tuna associated with the offshore NOAA weather buoys. 1

12 Honolulu, Hawaii, 2 th 25 th July 22 Diet Overlap The frequencies of occurrence of the most common prey taxa of tuna associated with the Cross Seamount are shown in Figure 8. The taxa included in these histograms are those that were present in at least five percent of the samples collected from the seamount for one or both species of tuna. Figure 8a includes only invertebrate taxa. With the exception of the cephalopods, all taxa on the left side of this chart are mesopelagic crustaceans. Most of these undergo diel vertical migrations, spending daylight hours below 3 meters (many below 5 meters) then moving shallower at night to feed. Without exception, the frequencies of occurrence of these taxa were higher in bigeye than yellowfin tuna stomachs. Those taxa on the right side of the chart include the urochordate salps (Thalliacea) that aggregated along the thermocline and various groups of epipelagic crustaceans that are rarely found deeper than the thermocline. Without exception, these taxa were more common prey of yellowfin tuna than bigeye tuna. The cephalopods occurred in nearly 1% of the stomach samples of both tuna species. These were lumped in the current analysis as more detailed taxonomic identification is not yet complete. To date, five families of cephalopods have been identified from yellowfin samples. All of these migrate vertically, spending the day deeper than 5 meters and migrating to surface waters at night. Thirteen cephalopod families have been identified in bigeye samples, including the five families preyed upon by yellowfin tuna. Of the eight families that were absent from yellowfin samples, four of these undergo vertical migrations and four are strictly mesopelagic. The same pattern was seen in the teleost prey. Thirteen mesopelagic families (Figure 8b) and ten epipelagic families (Figure 8c) were represented. The frequencies of occurrence of all mesopelagic families were higher for samples from bigeye tuna than yellowfin tuna. Seven of these families occurred in more than 2% of bigeye samples. The family Myctophidae (lanternfish) was most common, occurring in about 65% of all bigeye samples, followed by the families Percichtyidae (deep-sea basselets of the genus Howella) and Chiasmodontidae (swallowers) occurring in 47% and 42% of samples respectively. The myctophids were the only mesopelagic familiy that occurred in more than 2% of all yellowfin samples. The frequencies of occurrence of all epipelagic families were higher for yellowfin samples. The families Bramidae (pomfrets), Exocoetidae (flyingfish), and Acanthuridae (surgeonfish) occurred in more than 2% of all yellowfin samples. None of the epipelagic families occurred in more than 2% of bigeye samples. The data presented in Figure 8 suggests a significant degree of trophic separation between bigeye and yellowfin tuna on the Cross Seamount. They also suggest that both species forage on a wide range of prey taxa, with yellowfin tuna feeding more heavily on epipelagic taxa and bigeye tuna feeding more heavily on mesopelagic taxa. These data provide little information concerning the overall dietary importance of these taxa, however. These same taxa are represented in terms of percent abundance rather than frequency of occurrence in Figure 9. Percent abundance was defined as the proportion of the total number of prey represented by a given taxon. According to these data, very few of these taxa are numerically important. As indicated by the Costello diagrams, the shrimp family Sergestidae was the most important prey taxon of yellowfin, followed by crab megalopae, cephalopods, and the oplophorid shrimp. For bigeye tuna, the oplophorid shrimp were most important followed by sergestid shrimp and myctophid fishes (Figure 9a,b). The only numerically important mesopelagic fish family was the 11

13 Honolulu, Hawaii, 2 th 25 th July 22 Myctophidae, which represented 12.4% of bigeye prey and 6.7% of yellowfin prey. None of the epipelagic fish families were numerically important for either tuna species (Figure 9c). The numerical importance of the sergestid and oplophorid shrimp in both bigeye and yellowfin tuna samples suggests a higher level of trophic overlap between the species. Increasing the taxonomic and ontogenetic resolution of these taxa decreases this apparent overlap. Both families are mesopelagic; therefore they were expected to be more important in bigeye samples. There are two dominant species of oplophorids in offshore Hawaiian waters, Systelaspis debilis and Acanthephyra smithi. The distributions of both species are deeper than 5 meters during the day, however S. debilis migrates as shallow as 1 meters at night while A. smithi rarely occurs shallower than 2 meters. The abundance of S. debilis is reported to be much higher than that of A. smithi in Hawaiian waters. There are two dominant genera of sergestid shrimp in Hawaiian waters, Sergia and Sergestes. Adult Sergia are typically concentrated around 5 meters deep during the day and migrate to about 2 meters at night. The vertical distribution of adult Sergestes is slightly shallower than Sergia (4 meters in day, 15 meters at night). As juveniles, the vertical distributions of both Sergia and Sergestes are much shallower. During their nocturnal migrations, the juveniles of both species are found shallower than 1 meters and juvenile Sergia commonly migrate to the surface. Figure 1 shows the proportion each of these taxa contributes to the oplophorid/sergestid numerical total of each tuna species. For bigeye tuna, approximately 5% of the total was S. debilis, 25% were A. smithi, 15% were Sergia (adults and juveniles combined), and 1% were Sergestes (adults and juveniles combined). This probably mirrors the relative abundance of the taxa in mesopelagic waters. For yellowfin tuna, however, nearly 6% were the epipelagic juvenile stage of Sergia, and 2% were S. debilis, the shallowest and most common oplophorid species. 12

14 Honolulu, Hawaii, 2 th 25 th July 22 A 1 Major Taxa of Invertebrate Prey % of Samples B Cephalopoda 1 Sergestidae Oplophoridae Peneid Pandalidae Aegidae Thaliacea Euphausidacea Enoplemetophidae Major Families of Mesopelagic Prey Fish Phronimidae Stomatopoda Megalopae % of Samples C Myctophidae 1 Percichthyidae Chiasmodontidae Gonostomatidae Scopelarchidae Gempylidae Scombrolabracidae Paralapedidae Tetragonuridae Notosudidae Major Families of Epipelagic Prey Fish Melampheidae Bregmacerotidae Astronesthidae % of Samples Fistulariidae Blenniidae Ostraciidae Scombridae Monocanthidae Balistidae Chaetodontidae Acanthuridae Exocoetidae Bramidae Figure 8: Frequency of occurrence of prey taxa from tuna associate with the Cross Seamount. A) invertebrate prey taxa, B) mesopelagic teleosts C) epipelagic teleosts 13

15 Honolulu, Hawaii, 2 th 25 th July 22 A 4 Major Taxa of Invertebrate Prey B Cephalopoda 4 Sergestidae Oplophoridae Penaeidae Pandalidae Aegidae Thaliacea Euphausidacea Enoplemetophidae Major Families of Mesopelagic Prey Fish Phronimidae Stomatopoda Megalopae C Myctophidae 4 Percichthyidae Chiasmodontidae Gonostomatidae Scopelarchidae Gempylidae Scombrolabracidae Paralapedidae Tetragonuridae Notosudidae Major Families of Epipelagic Prey Fish Melampheidae Bregmacerotidae Astronesthidae Fistulariidae Blenniidae Ostraciidae Scombridae Monocanthidae Balistidae Chaetodontidae Acanthuridae Exocoetidae Bramidae Figure 9: Percent abundance of prey taxa from tuna associated with the Cross Seamount. A) invertebrate prey taxa, B) mesopelagic teleosts C) epipelagic teleosts 14

16 Honolulu, Hawaii, 2 th 25 th July Shrimp Prey Families Oplophoridae and Sergestdae Systelaspis debilis Oplophorus gracilirostris Oplophorus spinacauda Acanthephyra smithi Sergia young Sergia adult Sergestes young Sergestes adult Percent of Total Depth (meters) Night Day Figure 1: Relative percentage of oplophorid and sergestid shrimp in yellowfin and bigeye tuna associated with the cross seamount. Diel depth distributions for prey are also shown. The frequencies of occurrence of the most common prey taxa of tuna associated with the offshore weather buoys are shown in Figure 11. As with the Cross Seamount samples, the frequency of occurrence of cephalopods was very high for both tuna species. With the exception of cephalopods, these data indicate there was little trophic overlap between bigeye and yellowfin tuna. For yellowfin tuna, crab megalopae were present in 1% of the samples and thalliaceans (salps) were present in more than 3% of these samples. Sergestid shrimp, two families of amphipods, and three families of epipelagic fishes were present in more than 2% of yellowfin samples. For bigeye tuna, cephalopods (1%) and crab megalopae (33%) were the only taxa present in more than 2% of samples. In terms of percent abundance, crab megalopae were the dominant prey for yellowfin tuna and cephalopods were the dominant prey for bigeye tuna (Figure 12). 15

17 Honolulu, Hawaii, 2 th 25 th July 22 1 Major Taxa of Invertebrate Prey % of samples Cephalopoda Gymnosomata Euphausidacea Callianidae Flabellifera Enoplemetophidae Phronimidae Hyperiidea Sergestidae Thaliacea Megalopae 1 Major Families of Prey Fishes % of samples Bramidae Lophotidae Trachipteridae Chiasmodontidae Paralapedidae Ostraciidae Myctophidae Echeneidae Ogcocephalidae Alepisauridae Molidae Acanthuridae Gempylidae Exocoetidae Tetraondontidae Chaetodontidae Figure 11: Frequency of occurrence of prey taxa from tuna associate with the offshore NOAA weather buoys. A) invertebrate prey taxa, B) mesopelagic teleosts C) epipelagic teleosts 16

18 Honolulu, Hawaii, 2 th 25 th July 22 1 Major Taxa of Invertebrate Prey Cephalopoda Gymnosomata Euphausidacea Callianidae Flabellifera Enoplemetophidae Phronimidae Hyperiidea Sergestidae Thaliacea Megalopae 1 Major Families of Prey Fishes Bramidae Lophotidae Trachipteridae Chiasmodontidae Paralapedidae Ostraciidae Myctophidae Echeneidae Ogcocephalidae Alepisauridae Molidae Acanthuridae Gempylidae Exocoetidae Tetraondontidae Chaetodontidae Figure 12: Percent abundance of prey taxa from tuna associated with the offshore NOAA weather buoys. A) invertebrate prey taxa, B) mesopelagic teleosts C) epipelagic teleosts 17

19 Honolulu, Hawaii, 2 th 25 th July 22 Overlap Indices Three overlap indices were used to evaluate the degree of trophic overlap between associations and between tuna species (Table 5). Percent abundances of prey families (or nearest possible taxon) were used in the analyses. For each of these indices, a value of.6 (or 6% overlap) is considered significant. No significant overlap was found between bigeye tuna associated with the Cross Seamount and those associated with the offshore buoys. Similarly, no significant overlap was found between yellowfin samples as a function of association. In other words, the prey taxa of seamount-associated tunas are different from the prey taxa of tunas associated with offshore buoys. On the Cross Seamount, the trophic overlap between bigeye and yellowfin tuna was significant for one of three indices. This is due to the overlap in the families Oplophoridae and Sergestidae. As discussed above and shown in Figure 1, more detailed resolution of these taxa eliminates much of this overlap. No significant overlap was found between bigeye and yellowfin tuna associated with the offshore buoys. Table 5: Indices of trophic overlap for bigeye and yellowfin tuna associated with the Cross Seamount and the offshore NOAA weather buoys. Comparison of Interest Jaccard Sorensen s Abundance Moriseta /Horn : Seamount vs. Weather Buoys : Seamount vs. Weather Buoys Seamount: vs * Weather Buoys: vs

20 Honolulu, Hawaii, 2 th 25 th July 22 CONCLUSIONS AND FUTURE DIRECTIONS The vertical distributions of yellowfin and bigeye tuna are known to differ (Holland et al 199). tuna generally select deeper, colder waters (15 to 25 meters) while yellowfin are known to be concentrated near the thermocline and the mixed surface layer, or less than 75 meters deep (Brill et al. 1999, Holland et al. 199). Telemetry data suggests that this pattern breaks down near FADs and the two species overlap considerably (Holland et al. 199), though residence times differ (Holland et al. 1999). In fact, both species are primarily caught in nearsurface aggregations at the offshore weather buoys and the Cross Seamount. The preliminary analysis of these data suggests that the separation in vertical distribution is maintained during feeding. tuna fed primarily on mixed-layer prey while bigeye tuna fed on deep scattering-layer prey. These data also suggest that associating with the Cross Seamount may impart a significant trophic advantage to bigeye tuna but may offer little or no advantage to yellowfin tuna. This result may explain why Holland et al. (1999) found that the residence time of bigeye tuna on the Cross Seamount was much longer than that of yellowfin. Preliminary analyses (not presented here) suggest, however, that this trophic advantage may be highly dynamic and seasonal and connected primarily with the life cycles of oplophorid and sergestid shrimp. In addition, these results indicate that association with the offshore weather buoys offers no trophic advantage to either tuna species and may actually be detrimental to bigeye tuna. More samples are being collected to thoroughly evaluate this possibility. If true, this would suggest that detailed evaluations of the ecological impact of man-made structures in tropical waters might be warranted. The forage base of both species is extremely diverse. In this study, 34 families of fishes and at least 14 invertebrate taxa were identified as prey of yellowfin tuna while 53 families of fishes and at least 25 invertebrate taxa were identified as prey of bigeye tuna. This is comparable to other published studies. Despite this diversity, the percent abundance data indicate that relatively few taxa are numerically important in the diet. Percent abundance data emphasizes the trophic importance small, abundant prey, such as crab megalopae in the current study. We also plan to analyze the prey diversity data in terms of percent volume. This method emphasizes the importance of large, less abundant prey such as the teleosts that may be numerically unimportant but are very important volumetrically. For instance, large epipelagic fishes such as Ranzania laevis and Alepisaurus ferox are major contributors to the mean stomach fullness of yellowfin tuna associated with the offshore weather buoys. These taxa were not recognized in the current analysis, however, because they are numerically less important than small invertebrate taxa. Including these analyses will provide better insight into the trophic ecology of these tuna. 19

21 Honolulu, Hawaii, 2 th 25 th July 22 AKNOWLEDGMENTS The Pelagic Fisheries Research Program (PFRP) provided funding for this project. This project would not be possible without the cooperation of the captains of the Hawaiian commercial handline fleet. We especially thank Captain Joe Dettling of the F/V Double D for allowing us to sample aboard his vessel. We also thank the United Fishing Agency for cooperation and the use of their facilities and the Honolulu seafood buyers, especially Jed Hirota of Seafood Hawaii, for allowing us to collect samples from purchased fish. LITERATURE CITED Brill, R. W., B. A. Block, C. H. Boggs, K. A. Bigelow, E. V. Freund, and D. J. Marcinek Horizontal movements and depth distribution of large adult yellowfin tuna (Thunnus albacares) near the Hawaiian Islands, recorded using ultrasonic telemetry: implications for the physiological ecology of pelagic fishes. Marine Biology 133: Brock, R. E Preliminary study of the feeding habits of pelagic fish around Hawaiian fish aggregating devices or can fish aggregating devices enhance local fisheries production? Bull. Mar. Sci. 37(1): Buckley, R. M., D. G. Itano, and T. W. Buckley Fish aggregating device (FAD) enhancement of offshore fisheries in American Samoa. Bull. Mar. Sci. 44(2): Buckley, T. W. and B. S. Miller Feeding habits of yellowfin tuna associated with fish aggregating devices in American Samoa. Bull. Mar. Sci. 55(2-3): Costello, M. J Predator feeding strategy and prey importance: A new graphical analysis. Journal of Fish Biology 36(2): Dagorn, L. and P. Fréon Tropical tuna associated with floating objects: a simulation study of the meeting point hypothesis. Can. J. Fish. Aquat. Sci. 56: Freeman, M.C. and G. D. Grossman Group foraging by a stream minnow: Shoals or aggregations? Anim. Behav. 44: Holland, K. N., R. W. Brill, and R. K. C. Chang Horizontal and vertical movements of yellowfin and bigeye tuna associated with fish aggregating devices. Fish. Bull. 88: Holland, K. N., P. Kleiber, and S. M. Kajiura Different residence times of yellowfin tuna, Thunnus albacares, and bigeye tuna, T. obesus, found in mixed aggregations over a seamount. Fish. Bull. 97: