W&M ScholarWorks
Dissertations, Theses, and Masters Projects Theses, Dissertations, & Master Projects
1980
Aspects of vertical distribution and ecology of the dominant meso- and bathypelagic fishes from the Norfolk Canyon region
John V. Gartner College of William and Mary - Virginia Institute of Marine Science
Follow this and additional works at: https://scholarworks.wm.edu/etd
Part of the Fresh Water Studies Commons, Marine Biology Commons, and the Oceanography Commons
Recommended Citation Gartner, John V., "Aspects of vertical distribution and ecology of the dominant meso- and bathypelagic fishes from the Norfolk Canyon region" (1980). Dissertations, Theses, and Masters Projects. Paper 1539617501. https://dx.doi.org/doi:10.25773/v5-hh66-jz44
This Thesis is brought to you for free and open access by the Theses, Dissertations, & Master Projects at W&M ScholarWorks. It has been accepted for inclusion in Dissertations, Theses, and Masters Projects by an authorized administrator of W&M ScholarWorks. For more information, please contact [email protected]. ASPECTS OF VERTICAL DISTRIBUTION AND ECOLOGY OF THE
DOMINANT MESO- AND BATHYPELAGIC FISHES FROM THE
NORFOLK CANYON REGION
A Thesis
Presented to
The Faculty of the Department of Ichthyology
The Virginia Institute of Marine Science and
School of Marine Science of the
College of William and Mary
In Partial Fulfillment
Of the Requirements for the Degree of
Master of Arts
by
John Vo Gartner, Jr.
1980 APPROVAL SHEET
This thesis is submitted in partial fulfillment
the requirements for the degree of
Master of Arts
/7 /
/ /
^ 'Autrror ^
Approved, December 1980
[usick, ComniiFtee Chairman
Evon P. Ruzecki
r i David E. Zwerner
Robert H. Gibbs, JjfZ Curator of Fishe; United States Museum of Natural History Washington, D.C. TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS ...... iv
LIST OF T A B L E S ...... ix
LIST OF FIGURES ...... xiv
ABSTRACT ...... xvi
GENERAL INTRODUCTION ...... 2
GENERAL METHODS AND MATERIALS ...... 3
PART I. VERTICAL DISTRIBUTION OF THE MIDWATER FISHES IN NORFOLK CANYON ...... 13
PART II. FEEDING HABITS AND TROPHIC ECOLOGY OF SELECTED DOMINANT MESO- AND BATHYPELAGIC FISHES FROM THE NORFOLK CANYON REGION ...... 98
PART III. THE PARASITE FAUNAS OF SELECTED DOMINANT MESO- AND BATHYPELAGIC FISHES OF NORFOLK CANYON, WITH NOTES ON THE ECOLOGICAL IMPLICATION OF THEIR PRESENCE .... 199
CONCLUSIONS ...... 272
APPENDIX A ...... 278
APPENDIX B ...... 282
APPENDIX C ...... 285
LITERATURE CITED ...... 287
VITA 300 ACKNOWLEDGEMENTS
Completion of this study would not have been possible without the help of many people who gave freely of their time, advice, and expertise throughout the course of my work.
I first wish to thank my committee members, Dr. John A. Musick,
Dr. Robert J. Diaz, Dr. Evon P. Ruzecki, and Mr. David E. Zwerner of
VIMS, and Dr. Robert H. Gibbs, Jr. of the United States National
Museum of Natural History, Washington, D. C., for their critical review of the manuscript. I am grateful to all of them for answering many barrages of questions, as well as for their patient sifting through the many pages presented in early drafts.
These and other people also provided help in specific sections of the thesis:
Part I, Vertical Distribution- Drs. Jack Musick and Bob Gibbs provided much valuable information on midwater fish distributions, while Dr. Bob Diaz answered many questions regarding statistical analyses, and helped me set up the NPAR1WAY program. Dr. John Ruzecki supplied extensive information concerning the hydrographic regime around Norfolk Canyon, as well temperature and salinity data for the three R/V Gilliss cruises dealt with in this study.
I also wish to thank Mr. Ken Tighe of Ecological Analysts, Inc.,
Forked River, N.J., and Mr. Dave Bengston of the University of Rhode
Island for the use of distributional data from their theses on
Serrivomer species and Nemichthys scolopaceus, respectively. Dr. George Grant and John Olney provided VIMS Planktology data on shallow water near-shore captures of Benthosema glaciale and
Ceratoscopelus maderensis in the Norfolk Canyon area.
Part II, Feeding Habits- A great many people aided me in this portion of the study through the generous loan of their unpublished data, and helped in making or confirming identifications, as well as answering many questions and providing keys and/or reference specimens for comparison with prey items.
Special thanks are due to Dr. George R. Sedberry, now at the
South Carolina Marine Resources Center, Charleston, S.C., for supplying extensive information regarding types of food habits analyses, and for allowing me to use unpublished stomach content data for six specimens of Nessorhamphus ingolfianus and 17 specimens of
Derichthys serpentinus from cruise CI-73-1Q.
I also thank Dr. Jack Musick and Mr. Ken Sulak for submersible observations related to feeding, and Mr. Roy Crabtree for unpublished macrourid stomach data.
Drs. Richard H. Backus and James E. Craddock of the Woods Hole
Oceanographic Institution kindly loaned me 18 specimens of pelagically captured Scopelogadus beanii for use in comparative feeding studies, while Dr. William L. Fink and Mr. Karsten Hartel of the Museum of
Comparative Zoology, Harvard University allowed me to dissect an additional 16 specimens of S. beanii. A number of people at VIMS helped me with identification of various prey items. Ms. Marcia Bowen (Ostracods), Mr. Gary Gaston
(Polychaetes), Dr. George Grant (Gelantinous Zooplankton and
Copepods), Ms. Cathy Meyer (Copepods), Mr. Karl Nilsen (Bivalve
Mollusc), Mr. Peter Smythe (Euphausiids), Mr. Jacques van Montfrans
(Penaeid and Caridean shrimps), Dr. Michael Vecchione (Cephalopods), and Ms. Cathy Womack (Hyperiid Amphipods) identified, provided literature for the identification of, or confirmed identifications of the mentioned prey groups, and I thank them for their time and interest.
Dr. Thomas E. Bowman of the Department of Invertebrate Zoology,
United States National Museum of Natural History, Washington, D.C., supplied information regarding the vertical distribution of
Parathemisto gaudichaudii near Norfolk Canyon, for which I am grateful.
Part III, Parasitology- I wish to thank the entire Parasitology
Department at VIMS for their tremendous helpfulness and assistance throughout this portion of my study.
Dr. Frank 0. Perkins graciously allowed the use of materials and technician time for preparation of histological sections. In addition, he provided information regarding fungal stains and morphology. Ms. Patrice Mason ably performed numerous histological preparations for me. Dr. Gene Burreson, Mr. Dave Zwerner, Mr. Dave Benner, and Mr. Joe
Sypek all pitched in with great enthusiasm and contributed a lot of time to teach me parasitological techniques, provide references and available literature, and aid in identification of various parasite groups. I am especially indebted to Dave Zwerner for the many kindnesses and continual support he showed me throughout the course of this study.
Mr. Tom Munroe also provided data and information on deep-sea parasites, and staining techniques that were helpful for working on trematodes.
Dr. Robin M. Overstreet, Gulf Coast Research Lab, Ocean Springs,
Miss., kindly took the time to confirm identifications of the digeneti trematodes. Dr. Ronald A. Campbell of Southeastern Massachusetts
University has also been most helpful, confirming the adult cestode as a new genus, examining the trypanorhynch larvae, and also commenting on the manuscript.
I also want to thank my fellow graduate students at VIMS for their part in making my stay at VIMS enjoyable both during and after school. I especially want to thank the Ichthyology crew; Bill Raschi,
Ken Sulak, George Sedberry, Roy Crabtree, Tom Munroe, Jacques Carter, and Eric Anderson, for helping to maintain a balanced atmosphere of academia and insanity.
A special thanks is extended to the long-suffering typists and word processing people who managed to decipher my writing and get everything in print; Barbara Taylor and Nancy Peters in Ichthyology,
and Annette Stubbs, Cheryl Ripley, and Ruth Edwards of Word
Processing.
My parents, John and Mary, and the rest of my family have always
been a source of love and support for me in this and all my endeavors,
and for this I am especially grateful.
Finally, I would like to thank Joby Hauer for her patience and
understanding throughout the long haul. LIST OF TABLES
Part I. Vertical Distribution
Table Page
1 List of Meso- and Bathypelagic Fishes Captured around Norfolk Canyon with Numbers of Specimens and Stations . . . 22
2 CPUE (No. Specimens/0.5 hr.) by Depth Strata for each Season and Overall (Specimens from Water Hauls Excluded) . 25
3 Vertical Distribution of Nemichthys scolopaceus based on CPUE (Spec. /0.5 hr.) ...... 34
4 Day-Night CPUE (Spec./0.5 hr.) for Nemichthys scolopaceus . 37
5a Vertical Distribution of Serrivomer beanii based on CPUE (Spec./0.5 h r . ) ...... 38
5b Vertical Distribution of Serrivomer brevidentatus based on CPUE (Spec./0.5 hr.) ...... 41
6 Mean Haul Depth of Trawls Capturing Serrivomer vs. S_. brevidentatus ...... 42
7a Vertical Distribution of Nessorhamphus ingolfianus based on CPUE (Spec./0.5 hr.) ...... 44
7b Vertical Distribution of Derichthys serpentinus based on CPUE (Spec./0.5 hr.) ...... 46
8 Vertical Distribution of Eurypharynx pelecanoides based on CPUE (Spec./0.5 hr.) ...... 48
9a Vertical Distribution of Gonostoma elongatum based on CPUE (Spec./0.5 h r . ) ...... 51
9b Vertical Distribution of Gonostoma bathyphilum based on CPUE (Spec./0.5 hr.) ...... 55
10a Vertical Distribution of Argyropelecus aculeatus based on CPUE (Spec./Q.5 hr.) ...... 59
10b Vertical Distribution of Stenoptyx diaphana based on CPUE (Spec./0.5 hr.) ...... 61
11a Vertical Distribution of Chauliodus sloani based on CPUE (Spec./0.5 hr.) ...... 63
ix Table Page
lib Vertical Distribution of Stomias boa ferox based on CPUE (Spec. /0.5 hr.) .... ” I ^ 7 ” T T ^ T ...... 66
12 Day-night CPUE (Spec./0.5 hr.) for Stomias boa ferox .... 68
13a Vertical Distribution of Benthosema glaciale based on CPUE (Spec./0.5 hr.) ...... 72
13b Vertical Distribution of Ceratoscopelus maderensis based on CPUE (Spec./0.5 h r . ) ...... 79
14 Day-Night CPUE (Spec./0.5 hr.) forCeratoscopelus maderensis 83
15 Vertical Distribution of Lampanyctus macdonaldi based on CPUE (Spec./0.5 hr.) ...... 85
16 Vertical Distribution of Scopelogadus beanii based on CPUE (Spec. /0.5 h r . ) ...... 87
17 Distributional Groups of Dominant Midwater Fishes in the Norfolk Canyon Region ...... 91
Part II. Feeding Habits
la List of Species Examined with Percentage Empty Stomachs and Stomach Content Information ...... 102
lb Ranges and Mean Lengths of Midwater Fishes Examined for Stomach Contents ...... 102a
2 Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Nemichthys scolopaceus ...... 109
3 Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Serrivomer beanii ...... 114
4a Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Derichthys serpentinus ...... 120
x Table Page
4b Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Nessorhamphus ingolfianus ...... 125
5 Size and Depth of Capture of Eurypharynx pelecanoides in relation to Prey Type I n g e s t e d ...... 128
6a Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Gonostoma elongatum ...... 130
6b Change in Diet Composition of Euphausiids and Hyperiid Amphipods (Major Prey Items) with Season for Gonostoma elongatum ...... 135
7 Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Argyropelecus aculeatus ...... 138
8a Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Sternoptyx diaphana ...... 143
8b Change in Diet Composition of Euphausiids and Hyperiid Amphipods (Major Prey Items) with Season in Sternoptyx diaphana ...... 144
9 Depth of Capture vs. Percent Empty Stomachs Compared between Chauliodus sloani and Stomias boa ferox ...... 151
10 Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Ceratoscopelus maderensis ...... 156
11 Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Lampanyctus macdonaldi ...... 158
12a Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Scopelogadus beanii Captured by 45' Otter Trawl ...... 161
xi Table Page
12b Percent Frequency Occurrence (F), Percent Numerical Abundance (N), Percent Volumetric Displacement (V), and Index of Relative Importance (IRI) of Food Items in the Stomachs of Scopelogadus beanii from W.H.O.I. Collections (Captured by Midwater T r a w l ) ...... 163
13a Change in Hyperiid Amphipod Prey Composition with Season (Identifiable Species Only) in Scopelogadus beanii .... 165
13b Change in Diet Composition of Hyperiid Amphipods and Gelatinous Zooplankton (Major Prey Items) with Season in Scopelogadus beanii ...... 165
14 Percent Frequency Occurrence (F), Percent Numerical Abundance (N), and Percent Volumetric Displacement (V) of Identifiable Prey Taxa Recovered from the Stomachs of All Predators ...... 176
15 List of Predators with Respect to Feeding Strategy and Major Prey Type ...... 181
16 Depth of Capture vs. Percent Empty Stomachs for Some Dominant Fish Species (Peak Feeding Range in Parentheses) . 191
17 Proposed Feeding Strategies of Dominant Midwater Fishes around Norfolk Canyon ...... 196
Part III. Parasitology
1 Dominant Host Species and Numbers Examined for Parasite Analysis ...... 207
2 Major Parasite Groups and Species Recovered. Percent Infection Data by Taxonomic Grouping of Hosts ...... 208
3 Percentage of Infection in Host Species by Depth Zones . . 253
4 Parasite Group Infection Levels (No. of Infected Specimens) in Meso- and Bathypelagic Host Species (Transitional Species Excluded) ...... 255
5 Liver Comparisons Between Normal and Infected Individuals of Scopelogadus beanii by Size Group ...... 259
xii Table Page
6 Comparison of Weights of Scopelogadus beanii Specimens. Normal vs. Infected (Ichthyophonus type fungus) Individuals ...... 260
7 Distribution of Fungus for Size Range of Scopelogadus beanii Examined ...... 262
xiii LIST OF FIGURES
Part I. Vertical Distribution
Figure Page
1 Trawl Stations, Cruise CI-73-10 (Spring) ...... 5
2 Trawl Stations, Cruise GI-75-08 (Summer) ...... 7
3 Trawl Stations, Cruise GI-74-04 (Fall) ...... g
4 Trawl Stations, Cruise GI-76-Q1 (Winter) ...... -q
5 Catch-Per-Unit-Effort (CPUE) in Specimens/0.5 hr. by Season within each Depth Stratum (Bottom Depths) ...... 28
6 Frequency Curves and Mean Sizes of Benthosema glaciale Captured in Trawls from Shallower than 1000 m Bottom Depth (Solid Line) and Deeper than 1000 m Bottom Depth (Dashed Line) ...... 74
7 Frequency Curves and Mean Sizes of Ceratoscopelus maderensis Captured in Trawls from Shallower than 1000 m Bottom Depth (Solid Line) and Deeper than 1000 m Bottom Depth (Dashed
Line) • • ...... • . • o . o . o ......
Part II. Feeding Habits
1 Feeding Intensities in Nemichthys scolopaceus over a 24 hr. Period ...... Ill
2 Feeding Intensities in Serrivomer beanii over a 24 hr. Period ...... 116
3 Feeding Intensities in (a) Derichthys serpentinus and (b) Nessorhamphus ingolfianus over a 24 hr. Period . . . . . 121
4a Gonostoma elongatum. Percent occurrence by high taxonomic groups by Number, Volume, and Frequency of Occurrence . . . ^ 2
4b Feeding Intensities in Gonostoma elongatum over a 24 hr. P e r i o d »««.««.eoeo.oo»o..o«e.oo 132
5 Feeding Intensities in (a) Argyropelecus aculeatus and (b) Sternoptyx diaphana over a 24 hr. Period ..... 140
6 Feeding Intensities in (a) Chauliodus sloani and (b) Stomias boa ferox over a 24 hr. Period ...... 146
xiv Figure Page
7 Feeding Intensities in Scopelogadus beanii over a 24 hr# P e r i o d ...... - ...... 169
8 Relative Percentages of Frequency of Occurrence, Numerical Abundance, and Volumetric Displacement of Major Prey Taxa . 177
Part III# Parasitology
1 Ichthyophonus type fungal spores from (a) Gonostoma bathyphilum, (b) Scopelogadus beanii, (c) Derichthys serpentinus, and (d) Derichthys serpentinus ...... 232
2 Ichthyophonus type Fungal Hyphae from Scopelogadus beanii, Fig. 2 a - 2 c . 234
2 Ichthyophonus type Fungal Hyphae from Scopelogadus beanii, Fig. 2d-2f ...... 236
3 Ichthyophonus type Fungal Infections from Scopelogadus beanii, Fig. 3a-3c ...... 238
3 Ichthyophonus type Fungal Infections from Scopelogadus beanii, Fig.3d - 3 f ...... 244
4 Ichthyophonus type Fungal Infections from Scopelogadus beanii, Fig. 4a-4b ...... 245
4 Ichthyophonus type Fungal Infections from Scopelogadus beanii, Fig.4c - 4 e ...... 247
xv ABSTRACT
Seasonal otter trawling in the region of Norfolk Submarine Canyon over the continental slope and rise from 1973 through 1976 yielded nearly 1800 specimens of meso- and bathypelagic fishes. Approximately 1600 specimens belonging to 23 species were considered dominant, i.e., captured during all four seasons.
This study was divided into three primarily descriptive sections; the bathymetric distributions of the 23 dominant species, and feeding habits and internal parasite composition of 18 selected dominant species.
Vertical distributions were analyzed from catch-per-unit-effort data and treated statistically by Kruskal-Wallis analysis for significant changes in species abundances with depth or season. Five main bathymetric patterns of distribution were delineated; Upper Mesopelagic (100-750 m, 2 species), Mesopelagic (201-1000 m, 5 species), Lower Mesopelagic (501-1000 m, 3 species), Lower Mesopelagic/Upper Bathypelagic Transitional (751-1500 m, 2 species), and Bathypelagic (>1000 m, 8 species). Three species, Benthosema glaciale, Ceratoscopelus maderensis, and Argyropelecus aculeatus were found to exhibit a bimodal distribution pattern, with most specimens taken in nets trawled at bottom depths of 750 m or less, but with a significant number of captures in nets which fished bottom at greater than 1000 m, indicating a deeper dwelling or further offshore component of these species. Overall catch statistics for all species captured revealed peak catch rates during the spring with a smaller peak in fall reminiscent of spring-fall plankton blooms.
Feeding habits were described, using the index of relative importance to indicate major prey taxa. Changes in food habits with depth, season, sizes of both predators and prey, and diel period were shown to occur to varying degrees among the species examined. Selectivity vs. generality of diet was also examined. Feeding strategies were proposed based on theoretical activity levels and apparent degree of food selectivity calculated.
Identifications of the internal parasite faunas associated with the digestive system indicated that adult parasites were more common than had previously been found in midwater fish hosts. In addition, evidence was presented on the presence of a pathogenic fungus in the liver of a bathypelagic host, Scopelogadus beanii. Ecological implications of the harboring of other parasitic forms by midwater fishes were also discussed.
Data presented in each of these sections suggested the possibility of increased near-bottom presence of the dominant fish species rendering them more susceptible to capture by otter trawl. Effects of this phenomenon on vertical distribution, feeding, and internal parasite faunas were discussed.
xv i ASPECTS OF VERTICAL DISTRIBUTION AND
ECOLOGY OF THE DOMINANT MESO- AND BATHYPELAGIC
FISHES FROM THE NORFOLK CANYON REGION GENERAL INTRODUCTION
The Norfolk Submarine Canyon is located at the edge of the continental shelf approximately 120km east of Norfolk, Virginia,
U.S.A. An intensive study of demersal fish assemblages was conducted in and around this Canyon on the continental slope and rise during four seasonal cruises from 1973 to 1976. Even though the principal collecting gear used during this study was an otter trawl (a bottom sampling trawl), numerous meso- and bathypelagic (midwater) fishes were captured in addition to demersal fishes. The purpose of the present study is to describe the seasonal bathymetric distribution patterns of 23 meso- and bathypelagic species that are dominant in the trawl collections, and to provide information on the food habits and parasitology of 18 of these dominant species. In addition, the present study examines the hypothesis that midwater fishes regularly occur near the sea floor on the continental slope and rise where the oceanic meso- and bathypelagic zones impinge upon the bottom.
2 GENERAL METHODS AND MATERIALS
Data Collection
Most of the specimens for this study were captured during four cruises made from 1973-1976 in the region of the Norfolk Submarine
Canyon (approx. 36°58’ - 37°10* N, 73°0Q’-74°45TW) and an adjacent
"open" continental slope area to the south (approx. 36°34I-36°46,N,
73°00!-74°45'W). These four cruises, CI-73-10 (1-20 June 1973),
GI-74-04 (11-25 Nov. 1974), GI-75-08 (5-20 Sept. 1975), and GI-76-01
(22-31 Jan. 1976), roughly comprised the seasons spring, fall, summer, and winter respectively. On all cruises, a 13.7 m (45 ft.) semi-balloon otter trawl with 4.45 cm (2 in.) stretch mesh in the wings and body, and 1.27 cm (1/2 in.) stretch mesh liner in the codend was used. For a detailed description of gear, see Musick, 1979.
On all cruises a stratified random sampling design was incorporated such that the sampling areas were divided into strata, whose boundaries were the 75 m, 150 m, 400 m, 1000 m, 2000 m and 3000 m isobaths. The 2000-3000 m depths were not sampled during CI-73-10.
The < 75 m stations were excluded from analysis in this study, as no midwater fish species occurred at such shallow depths. CI-73-10 included 39 successful trawl stations between 87 and 1720 m. Midwater fishes were captured at 30 of these stations, of which 4 were "water hauls," stations at which the net did not strike bottom but captured pelagic species. During CI-73-10 this occurred at deeper stations
(1190 to 1876 m bottom depths, Fig. 1). GI-74-04 included 34
3 4 successful trawls between 78 and 2642 m, of which 31 hauls contained midwater species. One "water haul" occurred at 1935 m bottom depth
(Fig. 2). Thirty-nine successful stations were completed during
GI-75-08 (85 to 3083 m) , with pelagic fishes captured at 26 of them.
In addition, two "water hauls" were made at 1842 and 1943 m bottom depths (Fig. 3). During GI-76-01, 41 stations were successfully occupied from 83 to 2920 m. Twenty-eight hauls captured midwater fishes. Two additional stations at 1910 and 2242 m bottom depths were water hauls (Fig. 4).
Depths of stations were averages compiled from precision echo-sounding recorder (PESR) traces. Depths were entered into the log taken every three minutes from set to haul during 0.5 hr. tows
(stations less than 2000 m depth), and every six minutes during 1 hr. tows (deeper than 2000 m). All pelagic fishes taken were identified, counted, weighed (to nearest 0.1 g ) , and measured (mm standard or total length). Specimens were fixed in a 10% seawater formalin solution at sea, then washed out in water and transferred to either
70% ethanol or 40% isopropyl alcohol. 5
Figure 1. Trawl Stations, Cruise CI-73-10(Spring) 6
in cn
LU “ ujQ- “
10.0 I Hauls At Which Only Pelagic Species Were Captured O-1 - Water iS
u» c
Q. GO
I ro Is-i o h-O too 7
Figure 2. Trawl Stations, Cruise GX-75-08(Summer) Go
U~i CO J J Water Water Hauls At Which Only Pelagic Species Were Captured CO CO ; §2
CD E E 3 CO 00 o I m r- i o <5 o - o - f--0 (DO to o ro«* 9
Figure 3. Trawl Stations., Cruise GX~74-04(Fall) 1C
O o
CO CO J J z iu uijc I o0.0 —^
CO CO "Water Hauls" At Which Only Pelagic Species Were Captured
o Li_
o I sj- r- I o o - o r>- o fO CM 11
Figure 4. Trawl Stations, Cruise GI-76-01(Winter) 12
CO CO i J Water Water Hauls" At Which Only Pelagic Species Were Captured CO CO
o I CD h- i
CD o - o o VERTICAL DISTRIBUTION PATTERNS OF THE MIDWATER FISHES IN NORFOLK CANYON INTRODUCTION Distributional patterns of deep-sea pelagic fishes have proven to be both complex and variable. In addition, depths below 1000 m have been inadequately sampled due to financial and technical restraints. Also, biases due to sampling gear and design make much of the existing information difficult to interpret. The mesopelagic zone, defined as including depths between approximately 200 and 1000 m in oceanic waters (sensu Marshall, 1971) has been the primary object of study of deep-sea fish distributions. Generally, a picture has emerged that shows this zone as having well stratified fish populations, whose species compositions and distribution varies with changes in hydrography, season and time of day (Badcock and Merrett, 1977; Hopkins and Baird, 1977). Physical and biological parameters both play crucial roles in determining spatial distributions. Many species of midwater fishes have geographic (i.e. horizontal) ranges which are delimited by various hydrographic features (Ebeling, 1962; Ebeling and Weed, 1963; Backus et al., 1965; 1969, 1970, 1977; Robertson et al., 1978). Vertically, however, there is little evidence to suggest spatial conformity to hydrographic parameters. Badcock and Merrett (1977) review this subject in detail, therefore, it will not be dwelt upon here. Light intensity at depth, and its’ variability with season, time of day, and sea surface conditions appears to be the main factor 14 15 determining at least the upper depth limits of vertical distribution (Marshall, 1971). Biological factors such as size of fish, maturity, intra- and interspecific competition may also determine distribution patterns (both horizontally and vertically), and may be more important than physical constraints (Badcock and Merrett, 1977). It is obvious that there are many factors that must be considered when examining vertical distributions of midwater fishes. Because of a lack of data for many of the variables (e.g. D.O., light penetration), this section of the present study presents a general description of the bathymetric distributions of dominant meso- and bathypelagic fish species collected by an otter trawl over four seasons in a small locale, the Norfolk Submarine Canyon and adjacent continental slope area of the Western North Atlantic Ocean. These collections were all made within the confines of the southern extreme of the "Slope Water" water mass (Ruzecki, pers. comm.). The species examined comprised the dominant portion of the midwater fish community, if community was defined as "a group of populations co-occurring in time and space" (McNaughton and Wolf, 1979). Catch-per-unit-effort (CPUE) for each species within defined depth strata (see Methods and Materials) was used to determine overall depth of distribution and peak capture ranges for 23 dominant species of fish. Apparent differences in CPUE with depth were analyzed for statistical significance. Differences between day and night CPUE's 16 were examined for the more abundant species. Physical parameters and their possible influences on fish distributions were examined, although no attempt was made to define regulatory mechanisms. Less abundant species were briefly discussed. METHODS AND MATERIALS Data Analyses For this study, the depth strata were redefined to fit the conceptual habitat or zonation definitions that have been created for pelagic fishes. Three main strata were recognized; the epipelagic (75-200 m) , mesopelagic (201-1000 m), and bathypelagic (below 1000 m). These depths correspond to those presented by Marshall, 1971. The 75 m isobath was selected as the shallowest because it coincided with a stratum boundary in the otter trawl sampling design (Musick, 1979), and because it delimited the shallowest stratum in which midwater fishes were found. The meso- and bathypelagic zones were each subdivided into three sections in order to more clearly delineate any trends in fish distribution. The divisions were; upper mesopelagic (201-500 m), lower mesopelagic (501-750 m, 751-1000 m), upper bathypelagic (1001-1500 m) , and lower bathypelagic (1501-2000 m, > 2000 m ) . The lower mesopelagic was divided into two subzones because several bathypelagic species frequently occurred in the 751-1000 m range, and the two subzones of the lower bathypelagic were created because no stations were located > 2000 m during CI-73-10. Otter trawl captures of pelagic fishes present at best a semi-quantitative data base. In addition, midwater fishes exhibit a non-normal distribution pattern (Ebeling et al., 1970). For these reasons, and because many species were captured in low numbers, 17 18 non-parametric statistical analyses were used to determine vertical distributions of species. All raw catch data were converted to catch-per-unit-effort (CPUE) exressed as specimens captured per 0.5 hr tow. Comparisons of data sets among cruises (seasons) were able to be made directly because all four cruises utilized nearly identical sampling apparatus, and each followed the same sampling design. The mean CPUE's of all species captured were pooled and tested for statistically significant changes with depth strata within each season. Then, each depth stratum was tested over the four seasons for significant differences in CPUE. A second series of analyses examined essentially the same things as above, but with respect to individual species. Only those species characterized as "dominant” were selected for this analysis. For the purposes of this study, dominance was defined as the occurrence of a given species during all four seasons of the year. This definition of dominance allowed for the inclusion of rare year-round species, as well as those species whose frequency of occurrence varied widely among seasons. Both series of analyses were conducted using the Kruskal-Wallis non-parametric test. This test ranks observations from lowest to highest after pooling all groups and considering them a single sample. 19 The statistic H (Sokal and Rohlf, 1969) was then calculated by the expression: n . a 1 2 a wn = ------12 E (E R) i - 3 (En. ! + 1) a a n . (En.)(En. + 1 ) 1 11 7 a where £ n^ is the sum of the sample sizes of the entire analysis (n^), i 12 and 3 are constants, (e R)j_ is the sum of the ranks for the ith group, and a represents the number of groups. The null hypotheses tested in this study were that depth does not affect CPUE within a cruise, and that CPUE within a given depth range does not change with season. H is approximately distributed as chi-square (x^) with a-1 degrees of freedom. In data where ties exist between observations, H must be divided by the correction factor D, expressed as: m ET. 3 D = 1 ______a a a (En. - 1)E n . (En. + 1) ™m 1 1 i where ETj is the sum of tabled values T given for the number of variates tied in the jth group of ties. For the first series of tests (mean CPUE's of pooled totals all species captured) the Kruskal-Wallis tests were calculated by hand. Thereafter a Statistical Analysis System (SAS) program using an NPAR1WAY procedure, Wilcoxon option, was used. For multiple samples, the Wilcoxon option performs the Kruskal-Wallis test. The probability 20 level was set at p=0 .1 (two-tail test) in order to discern any general trends in CPUE change with depth or season. Highly significant statistical differences were considered to have occurred at p<0.05 level for a two tail test. Although CPUE’s were computed for all 23 dominant species, only 14 were subjected to computer analysis. The remaining 9 species either had very small sample sizes, or were restricted in their distribution to a single depth stratum. Differences in abundance of individual species were tested for each main depth stratum (7 5-200 m, 201-500 m, 501-1000 m, 1001-2000, > 2000 m). If significant differences were found, then 501-1000 m and 1 0 0 1 - 2 0 0 0 m were subdivided into four substrata for further analysis. RESULTS AND DISCUSSION All Species A total of 1,795 specimens, comprising 7 orders, 38 families, 57 genera, and 76 species of meso- and bathypelagic fishes was captured during the four cruises. Table 1 presents a list of species, with numbers of specimens captured and stations per cruise. Examination of CPUE for each depth stratum (Table 2) revealed a catch maximum in the lower mesopelagic (501-1000 m). Within this depth range, CPUE was highest between 501-750 m in the spring, but shifted to the 751-1000 m strata the other three seasons. If CPUEfs were pooled for all four cruises, the overall catch maximum remained between 501-750 m, although it did not differ significantly from the CPUE for 751-1000 m. Kruskal-Wallis analysis of each season revealed that CPUE’s did differ significantly with depth strata. Spring, summer and winter values were highly significant at p=0.05 (two-tail test), while fall was marginally significant at .01 (Table 2). This marginal significance was attributed to the reduced numbers of samples at the 75-200 m and 751-1000 m strata. When the 501-750 and 751-1000 m, and 1001-1500 and 1501-2000 m strata were joined to provide comparable numbers of samples among strata, all four seasons showed highly significant differences among CPUE's with depth strata. 21 22 TABLE 1 LIST OF MESO-AND BATHYPELAGIC FISHES CAPTURED AROUND NORFOLK CANYON WITH NUMBERS OF SPECIMENS AND STATIONS 73-10 75-08 74-04 76-01 TOTAL SPRING SUMMER FALL WINTER 116 ORDER-FAMILY SPECIES 28Sta. 28Sta. 3lSta. 3lSta. Sta. Angui Hi formes Nemi. chthyidae Nemichthys scolopaceus** 241(25) 28(11) 74(18) 27(12) 370(66) Serrivomeridae Serrivomer beanii,f/' 12 ( 7) 5( 5) 15 ( 8) 6{ 6) 38 (26) S. brevidentatus-* 2( 2) 2( 2) 4( 3) l( 1) 9( 8) Derichthyidae Derichthys serpentinus' 21^ 12) 3< 3) 3( 3) 4( 4) 31( 22) Nessorhamphus ingolfianus'/OC 10( 7) 2< 2) 4( 4) 14( 8) 30 f21) Saccopharyngidae Saccopharynx flagellum l( 1) K 1) Eurypharyngi dae Eurypharynx pelecanoides* 4( 4) 5( 4) 2( 2) K 1) 12(11) Salmoni formes Bathylagidae Bathylagus euryops* 1( 1) 1( 1) 3 f 3) 3( 2) 8( 7) B. compsusf?) 2( 1) 3( 3) 1( 1) 6( 5) B. longirostris K 1) K 1) 1^ 1) 3( 3) Alepocephalidae Xenodermichthys copei K 1) 3 ( 3) 2( 2) 6( 6) Searsiidae Mentodus rostratus 1( 1) 1( 1) Photi chthyidae Polymetme corythaeola* I7r 3) 2( 1) K 1) 12( 2) 32f 7) Gonostomati dae Gono s toma elonga tum'fW 13< 9) 10( 7) 3or 15) 22r 9) 75 (40) G . b a thyph i lum5' 5 ( 3) 3( 3) 3< 3) isr 8) 26 (17) Cyclothone microdon'' 123( 8) 3( 2) 7< 5) l( 1) 134^ 16) C. braueri V 1) 1( 1) G. pallida 2( 1) 2( 1) Vinciguerria sp. 1( 1) K 1) Sternoptychidae Argyropelecus aculeatus5* 9( 8) 5 ( 5) 5( 5) If 1) 20(17) A. gigas K 1) 2< 2) 1( 1) 4( 4) Sternoptyx diaphanax,v 12( 6) 7 ( 6) 8( 6) 3( 3) 30(21) Polyipnus asteroides 7 ( 3) 1( 1) 5( 4) 13( 8) Maurolicus muelleri' 266( 3) 4( 2) 2( 2) 21 ( 4) 293^ 11) Chauliodontidae Chauliodus sloani** 21( 13) 3( 3) 17(11) 15(11) 56( 38) C. danae K 1) K 1) Stomiati dae Stomias boa ferox'°v 24( 12) 8( 7) 44(16) 8( 6) 84(41) S. affinis If 1) 1( 1) Astronesthidae Borostomias antarcticus 1< 1) 1( 1) Rhadinesthes decimus K 1) I' 1) Melanostomiatidae Melanostomias spilorhynchus 2 ( 2) 2( 2) 4( 4) Echiostoma barbatum K 1) 1( 1) Eustomias bigelowi 1( 1) K 1) Trigonolampa miriceps 1( 1) 1( 1) Grammatostomias flagellibarba V 1) 1( 1) 23 TABLE l(cont'd) ORDER-FAMILY SPECIES SPRING SUMMER FALL WINTER TOTAL Malacosteidae Photostomias guernei* 1( 1) 1( 1) 2( 2) K i) 5( 5) Malacosteus nigerVv 5( 4) 3( 2) 3( 3) 2 ( 2) 13(11) Myctophiformes Myctophidae Benthosema glaciale'f'f 87(18) 19( 9) 28(11) 5( 3) 139(41) Lobianchia gemellari 1( 1) 2( 2) K 1) 4( 4) L. dofleini 5( 1) 5( 3) 1( 1) 11( 5) Diaphus dumerilii H( 3) 1( 1) 12 ( 4) Lampadena speculigera 1( 1) 1( 1) 2( 2) Ceratoscopelus maderensis”* 42(12) 6( 6) 29(15) 7( 4) 84(37) Lampanyctus macdonaldi* 1( 1) 10( 5) 6( 4) 2( 1) 19(11) L. crocodilus 1( 1) 2( 1) 1( 1) 4( 3) L. ater 1( 1) 1( 1) 2( 2) L. cuprarius 1( 1) K 1) Lepidophanes guentheri 1( 1) i( 1) Gonichthys coccoi 1( 1) l( 1) Myctophum affine 2( 2) 2( 2) M. punctatum 1( 1) 1( 1) Hygophum hygomii 14( 4) 14( 4) Notoscopelus resplendens 1( 1) 1( 1) 2( 2) Symbolophorus veranyi 1( 1) K l) Neoscopelidae Neoscopelus macrolepidotus 1( 1) 4( 1) 5( 2) Evermanne11i dae Evermanne1la balbo 1( 1) 1( 1) Scopelarchidae Scopelarchus analis K 1) 1( 1) Gadiformes Zoarcidae Melanostigma a11anticum** 14 ( 5) 30( 8) 7( 5) 2( 2) 53(20) Lophiiformes Melanocetidae Melanocetus johnsonii K 1) 1( 1) Ceratiidae Ceratias holbolli l( 1) 1(1) Cryptopsaras couesi 2( 2) 2( 2) Beryciformes Me lampha idae Scopelogadus beanii,f,f 10( 5) 18 ( 9) 27(10) 13( 7) 68(31) Poromitra capito 1( 1) 1( 1) P. crassiceps 1( 1) 1( 1) 1( 1) 3( 3) Scopeloberyx opisthopterus 10( 4) 2( 1) 10( 2) 22( 7) S . robustus 1( 1) 2( 2) 1( 1) 4( 4) Melamphaes suborbitalis 1( 1) 1( 1) Anoplogasteridae Anoplogaster cornuta 1( 1) 1( 1) Rondeletiidae Rondeletia loricata K 1) 1( 1) 2( 2) Cetomimidae Cetomimus gilli K 1) K 1) 24 TABLE L(cont1d) ORDER-FAMILY SPECIES SPRING SUMMER FALL WINTER TOTAL Perciformes Chiasmodontidae Chiasmodon sp.A 1( 1) 1( 1) 2( 2) Chiasmodon niger 3( 2) 3( 2) Gempylidae Promethichthys prometheus 2( 2) 2( 2) Nesiarchus nasutus 1(1) 1(1) Trichiuridae Benthodesmus simoyi 2( 2) 2( 2) B. atlanticus 1( 1) 1( 1) -Stations in parentheses; no. of specimens and stations include water hauls. *-Year-round resident (dominant) species. Numerically abundant dominant species. TABLE 2. CPUE (No. Specimens/0.5 hr.) BY DEPTH STRATA FOR EACH SEASON AND OVERALL (SPECIMENS FROM WATER HAULS EXCLUDED) - 3 r-l •H Q) D CDCD M C 0 0 o r*- O C O C O M C M C CO —CMm M C 1— 1— —1 1— CM i— 1— 1 T— M C O v M C O C 1 t— M C —! i— M C M C ^ r © v M C M C 5 0 o < o u M C O o O ! CO > X O m o t o O C o —1 r— i— H \ o - t-*- —i i— —1in i 1 r— p i i— O C H r m M C O liO O H ) 1 X O O C C". M C O C o n o r-- CM o 5 0 O C O C H r M C H r H r H r O C O C CM O C M C M C H i M C O C cr» H r H r O r^, LT) O 1 o \ r < n i O C © 5 < O C * O O U O U O M C l 1 N O O U O C o u m M C O o O H i o u O C iH o O o o O U o u o H r H r r*^ 1 H r O C —1 1— H r o N O o O o o p O U rH —11— H 1 0 \ CM 0 0 O rH uo p 25 zi cd H ’TO r© H i O O O O o •H T3 H i O O rH in o iH uo rH CM O O u O f^*- r*. rH o o ^s /^ H i o u o o u H i H i H r H r x: H r •H •H 4-> 4-1 4-> ■5 CD H o B d Comparison of each depth stratum over the four seasons (Table 2) revealed that significant differences in CPUE occurred only in the lower mesopelagic (501-1000 m ) , p=.05 for two-tail test. Analysis of the individual 501-750 m and 751-1000 m subdivisions showed that only the former differed significantly with the season, p=0 . 1 (two-tail test). An item worthy of note with respect to seasonality was the fluctuations of catch over the seasons. The peak CPUE at most depths occurred in the spring. The CPUE then dropped to quite low levels during the summer, rose again in fall, and dropped in winter. This spring-fall increase in fish abundance parallels the phytoplankton-zooplankton blooms, of which the larger is during the spring (Raymont, 1963). The relationship between these patterns may be direct, indirect, or simply fortuitous. The high CPUE in the uppermost (75-200 m) depth range during spring was due to a catch of 249 individuals of Maurolicus muelleri. The CPUE in this stratum normally was quite low, because the 200 m contour approximates the shelf break, which has been long been recognized as an area of sharp faunal change (Raymont, 1963 for plankton; see Musick, 1976 and citations therein for benthos). Although several species of midwater fishes were found to associate with the shelf break, no species was regularly found shelfwards of the 1 0 0 m isobath. 27 The CPUE values for spring and summer, after accounting for the high 75-200 m value, both showed a rapid rise to peak catch rates (501-750 m in spring, 751-1000 m in summer) followed by a steady decline, i.e., unimodal catch peaks. Fall and winter values however, showed bimodal peaks (Fig. 5). There was a principal peak in the 751-1000 m stratum and a lesser one between 1501-2000 m. This was attributable to the increased presence of bathypelagic species during the fall and winter seasons in this stratum. Ten species of bathypelagic fishes were captured each season during fall and winter, as compared to 4 and 6 species in the spring and summer respectively. The change in patterns of abundance in the 501-1000 m stratum as indicated by the significant difference in CPUE within and among seasons was difficult to correlate with any existing physical data for the Norfolk Canyon area. Temperature, salinity and dissolved oxygen data taken in Norfolk Canyon area (Ruzecki, pers. comm.) and also examined from Deepwater Dumpsite 106, a region over the slope to the north of Norfolk Canyon (Warsh, 1975) revealed that the density gradient formed by the permanent thermocline, and the oxygen minimum layer remained around 300 m throughout the year. Jahn (1976) also presented data supportive of this. Light may partly account for this peak concentration of fishes. Many midwater species are known to maintain certain depth levels in response to ambient light levels, and adjust their depth with 28 Figure 5 Catch-per-unit-effort (CPUE) in specimens/0.5 hr. by season within each depth stratum (bottom depths) O O O O O • • 73-10 (SPRING) X X 75- 08 (SUMMER) A 74-04 (FALL) O CJ — CVJ rO ( JU s*o/-oads ) 3fld0 cx x " o " m o OJ in DEPTH STRATA (m) 30 fluctuations in illumination (Kampa, 1970). Penetration of light in the deep-sea at levels detectable by midwater fishes can occur to depths greater than 1000 m in the clearest waters (Clarke, 1970). Attenuation of light can readily occur in water with high particulate loads of both a biological (animate and inanimate) and non-biological nature, causing large fluctuations in depth of penetration. Thus it may be that in spring, when the major phyto- and zooplankton blooms occur, much of the downwelling light is blocked, so that fishes move to the shallower strata to maintain themselves at some particular isolume. In summer, fall and winter, when light penetration is greater (due to increased sunlight in summer, and decline in plankton for all three seasons), species descend. This is most pronounced in the winter, when water is clearest due to low particulate (i.e. plankton) loads (Raymont, 1963). A biological factor which may play a role in the peak CPUE between 500-1000 m is the vertical distribution of zooplankton. According to Raymont (op. cit.) and Vinogradov (1970), in the northwest Atlantic Slope Waters (ca. 39°N, 65-73°W), a plankton biomass maximum exists, independent of seasonal fluctuation, at about 600 m (800 meters in the Gulf Stream) due to the transport of northern plankters at depth in the intermediate waters. Fish species may congregate in the depths around this potentially rich prey source. Of the fish groups represented in this study, several were important, being numerically dominant at various times of the year. 31 Salmoniform and myctophiform fishes contributed heavily to the catches. Salmoniform fishes comprised 46.0% of the total catch for all seasons (Table 1). The myctophiforms comprised 17.1% of the total catch. Anguilliform fishes also were well represented, amounting to 27.3% of the total, and including all known midwater eel families except the Cyemidae and the Monognathidae. It is probable that this large contribution was due to the greater catch efficiency of the large otter trawl when compared to smaller midwater trawls, but may also in some measure be the result of increased availability of eels in the slope area, or possible increased frequency of occurrence of eels near-bottom. The beryciform, gadiform, lophiiform and perciform contributions were low, 5.7%, 3.0%, 0.2%, and 0.6% respectively. These are typical percentage returns and are due to two factors: first the number of pelagic species in the gadiform and perciform families is low (only 1 species, Melanostigma atlanticum represents pelagic gadiforms); second the beryciform, lophiiform and perciform representatives are largely deeper dwelling or, in the case of the latter family, very rapid swimmers. Thus, incomplete sampling of habitat or escape may have caused low catches. Dominant Species According to the definition presented in the Methods and Materials section, a total of 1,632 specimens from 5 orders, 21 genera 32 and 23 species, were classified as dominant members of the pelagic fauna (Table 1). The 23 dominant species represented 90.9% of the total catch for the four seasons. The 12 most frequently captured (captured at a minimum of two stations per cruise, Table 1) species comprised 65.0% of the dominants (59.1% of the total catch). The results are presented by species according to the taxonomic scheme of Nelson (1976). Order - Anguilliformes Suborder - Anguilloidei Family - Nemichthyidae Nemichthys scolopaceus Nemichthys was the most abundant species captured during all seasons. Although 266 Maurolicus muelleri were taken during the spring as opposed to 241 _N. scolopaceus, Nemichthys was taken at 25 stations, as opposed to 3 stations for Maurolicus. A total of 370 specimens was captured, of which 241 were taken in the spring, 7 4 in fall, 27 and 28 in winter and summer respectively. Of these totals, 16 specimens came from water hauls (14 spring, 1 summer, 1 fall). Nemichthys appeared to conform to the spring-fall abundance pattern noted in general for all species. Beebe and Crane (1944) noted that although they sampled April through September off Bermuda, half their specimens were captured in September. However, Bengston (1973; pers. comm.), found no appreciable changes in 33 catch-per-unit-effort with season in the Bermuda Ocean Acre study area. Examination of catch per unit effort data (Table 3) showed that N. scolopaceus was ubiquitous throughout the depth strata, although over 90% of the specimens occurred where the mesopelagic zone (201-1000 m) intercepts the bottom. The highest CPUE occurred within the 501-750 m spring through fall but was found in the lower mesopelagic (751-1000 m) in winter. Also, Nemichthys appeared to avoid the upper water layers (75-500 m) in summer, when plankton levels were low, and illumination levels were probably high. Differences among CPUE's per depth stratum were highly significant (x2 = 22.2, p<.G005; x2 = 13.2, p<.0395; x2 = 12.2, p<.0569, df = 5) during the spring, fall, and winter seasons respectively, but only marginally so in summer (x2 = 10.4, p < .1071, df = 5). This lower significance in summer was attributed to the relatively even CPUE in the two lower mesopelagic strata. There was also a statistically significant difference among seasons in the 501-750 range caused by the sharp decline in winter catches (x2 = 8.3, p<.0397, df = 3). In the lower depth stratum (>1501 m) there was a slight increase in CPUE at the 1501-2000 m strata spring through fall. This was more heavily reinforced if water haul samples were included. Leavitt (1935, 1938), conducted discrete depth sampling for 300 zooplankton to great depths (3000 m and deeper) in Slope Waters and consistently Table 3. Vertical distributions of Nemichthys scolopaceus based on CPUE (spec./0.5 hr •iH i—4 i—l & [H CO CO •H 4J c w • • m o o CM n* t—, O CM i— 1 l m CM in o Oc n c0 CO i—i o o m MCM CM CM CM o M< CM G> Numbers in parentheses denote calculations including water haul samples. 34 * * - CPUE converted to specimens/0.5 h r s . ++ - 3 seasons only. 35 found two plankton abundance maxima which were independent of sample bias: a primary maximum at ca. 800 m, and a second smaller maximum between 1400-1600 m. Between the depths of 1200 and 1400 m a plankton minimum was noted. It may be that the distribution of Nemichthys and those of other species to be discussed correspond vertically to these plankton maxima. The depth information presented here is in agreement with those few works that have been done on Nemichthys. Roule and Bertin (1929) concluded that "Nemichthys occupies a considerable vertical expanse of marine waters but the majority are found above 1000 meters." These captures were primarily in the tropical region of the North Atlantic. Beebe and Crane (1944) found that _N. scolopaceus occurred between 1000 and 2000 m off Bermuda, but noted small specimens much shallower in the water column (bathysphere observations). Bengston's (1973) discrete-depth data indicated the bathymetric range of juveniles (the largest size class caught) was from 50 m to below 1200 m in the Bermuda area. Nielsen and Smith (1978) while not providing any data, stated that whereas snipe eels have been taken to below 2000 m, "It would probably be more correct to call them mesopelagic than bathypelagic." The specimens captured ranged from juvenile through adult stages, based on Bengston's (1973) classification. Adult males were not captured at any time. Recent captures of ripe male Nemichthys over the Blake Plateau area (Musick, unpub. data) may point towards a more southerly spawning ground than Norfolk Canyon. Bengston (pers. comm.) 36 noted that the Ocean Acre specimens he examined were all larval (postlarvae) through juvenile stages. Metamorphic specimens and juveniles tended to occur deeper (700-900 m) while coefficients of dispersion showed postlarvae to be highly clumped in the epipelagic. The CPUE was also examined by season for day-night differences (Table 4). Nemichthys showed an increased number of specimens per 0.5 hr. at night in shallow waters in spring and fall, which were the two strongest data sets. From this it appeared that limited diel up-slope migrations may have occurred. Family - Serrivomeridae Serrivomer beanii The second most abundant eel in the Norfolk Canyon area was . beanii, with a total of 38 specimens captured. This species apparently was bathypelagic in the study area because CPUE analysis (Table 5a) showed it to occur primarily in strata below 1 0 0 0 m, although one specimen was taken in the fall where the bottom depth was approximately 700 m. However, some of these data were from water hauls occurring in deep waters. Because of the low numbers of specimens, Kruskal-Wallis analyses showed no significant difference in CPUE within or among seasons, although CPUE in depth stratum 5 (1501-2000 m) approached statistical significance among seasons. DAY-NIGHT CPUE(spec./0.5hr.) FOR Nemichthys scolopaceus o o o o o o o o o o o o o 1 1— o o o o o o m o CM o o o I''. 1 9 • • • • 9 • • • • • 9 —l I— I— 1 —t i— O C o o m o O CM CM o o CM m CM o CO i • • • • • o c r-H vO —1 T— r—i O r^. o o CM o vO CO rH m r'-. m O CO CO 1 1 1 1 t l 1 • • • • • • • o o i i— i MCM CM CM CM r-H CM m —l l— —1 1— o m o o o o m o o O o o i— — 1 1 1 1 8 • • • o • • • i CM o o c CM rH O o —i i— o o o o o o o 1^. o m iH o 1 1 1 i • 9 • • • • • • • 00 o —1 I— o O rH CM o o o CM CO o o —1 1— CM o m o o O CM m CM 1 • • • 9 9 9 CM O v o o o CM o o /V O CM o m I 1 I 1 1 1 I 1 • • • • 37 Table 5a. Vertical distribution of Serrlyomer beanii based gn. CPUE (spec,/0,5 hp.). Numbers in parentheses denote calculations including water haul samples. * - CPUE converted 'to specimens/0.5 h r s . ++ - 3 seasons only. 39 As with Nemichthys, mostly juvenile through adult specimens, according to Tighe's (pers. comm.) classification were captured. Twenty-four specimens examined for food habits (next section) ranged from 249-724 mm T.L. Specimen captures occurred most often in spring and fall, although there was no marked difference between catch rates for these seasons. The shallowest captures (ca. 700-950 m) were of specimens of smaller size. Analysis of mean specimen size with depth strata revealed a larger-deeper trend for S. beanii (501-750 m, x = 400.0 mm T.L.; 751-1000 m, x = 429.5 mm T.L.; 1001-1500 m, 5 = 437.6 mm T.L.; 1501-2000 m, x = 488.5 mm T.L.; >2000 m, x = 515.8 mm T.L. Distribution of smaller, younger specimens is known to be shallower (Tighe, pers. comm.). The apparent distribution pattern found for IS. beanii in the present study is well supported by other studies to date. Roule and Bertin (1929) noted that 60% of the Serrivomer species caught (both beanii and brevidentatus) were taken with 1000-3000 meters of wire out (equivalent to approximately 500-1500 meters depth fished), while another 32% were captured with more than 3000 mwo. Beebe and Crane (1936) determined that the mean depth of capture ranged from 550 m for adolescents to 800 m for adults. Tighe (1975), using discrete depth IKMT data from Bermuda Ocean Acre noted that "juveniles were concentrated at depths of 550-1000 m with the greatest concentration... at 601-650 m during the day and 951-1000 m at night... Comparisons of cumulative-size frequency curves indicate 40 that larger S_. beani (sic) occur deeper in the water column both day and night." The center of distribution for Serrivomer beanii appears to be southwards of Norfolk Canyon, as both Beebe and Crane (1936) and Tighe (pers. comm.) found it to be quite abundant in the Bermuda area. Jahn (1976) noted that S ° beanii occurred in both Slope Water and Gulf Stream areas, but appeared to be slightly more abundant in Slope Water. Analysis of CPUE for day-night differences showed no apparent trends. Although Badcock (1970) presumed Serrivomer to be a strong migrator, Tighe (1975) indicated that there is no evidence for this based on his Ocean Acre samples, and suggested Badcock*s interpretation may be due to contamination of shallow hauls from earlier, deeper tows. Serrivomer brevidentatus Serrivomer brevidentatus was a rare year-round species. Only 9 juvenile to adult (247-635 mm T.L.) specimens were taken throughout the sampling period. The distribution of this species was similar to that of S^. beanii in that it appeared to be bathypelagic (Table 5b) . Comparison of mean depths of capture for both species however, showed that S. brevidentatus was found deeper than S. beanii in the strata where they co-occurred except for the winter, when only one specimen of S. brevidentatus was taken (Table 6 ). Table 5b. Vertical distribution of Serrivomer brevidentatus based on CPUE (spec./0.5 hr.) Numbers in parentheses denote calculations including water haul samples. * * - CPUE converted to specimens/0.5 hrs. ++ - 3 seasons only. MEAN HAUL DEPTH OF TRAWLS CAPTURING Serrivomer beanii VS. Serr. brevidentatus E rs M w hJ CO ffl P4 <5 p S3 CO Q H W ; £ CO g Pd M o P3 h H ,p X rO co| •H •H rP ,P ,.Q CO CO CO CO co| •H •H •H h e •H •H CO CO I d U o CN r-~ in o o 1 CN t-i o o m 1 r— 1 o O CTv o vO O r-~-m as in 1 « . i—i oo 00 00 r4 av CO o o o CN T— O rx. in 00 O l . . I CN "vf 00 Os o o O o 00 CN o 'd- CO co o o i—( i i— o 1 i—l i m o 1—1 o 1—1 T—1 I 1—1 LO — —f — I • • o • . 1 1 iH I o CN VO i—i O r—1 oo o o 1— 1 o VO CO «— 1 r-x m 1—1 Ov 1 CO rx i m r-x vO CN OV m i m r^. —i — —i — I ■ . • . • . 1 1 CN CN CN o CN o CN Beebe and Crane (1936) and Tighe (1975) both noted the apparent deeper distribution pattern for this species. In addition, both commented on its rarity. This fish may be nowhere common in the North Atlantic, or it may have a much deeper center of distribution than is commonly sampled. Marshall (1954) interpreted the difference in abundances between these two species of Serrivomer to be the result of competition. Family - Derichthyidae Nessorhamphus ingolfianus This eel also appeared to be common in the Norfolk Canyon area because 30 specimens were captured over the four seasons. Although capture data was scanty, CPUE indicated this species to be a lower mesopelagic inhabitant, found primarily between 751-1000 m most of the time (Table 7a). The availability of Nessorhamphus was highest in the spring and winter seasons. For these periods Kruskal-Wallis analyses showed that CPUE within seasons approached but never reached statistically significant levels. No significant differences were noted among seasons. Sedberry (1975) hypothesized that N. ingolfianus may occur regularly near the bottom in the slope area because it is often taken in otter trawls. This species is a rare constituent of the catch of midwater trawls (pers. observation) and Beebe (1935a) noted that Nessorhamphus "is rare among the deep-sea fish captures off Bermuda, Table 7a. Vertical distribution of Nessorhamphus ingolfianus based on CPUE Cspec.^0,5 hp.) Numbers in parentheses denote calculations including water haul samples. * * - CPUE converted 'to specimens/0.5 h r s . -H- - 3 seasons only. 45 only once occurring every 52.5 nets drawn between 400 and 1000 fathoms..." Beebe1s findings indicate that Nessorhamphus spawns in spring and early summer, in the Sargasso Sea area, with metamorphosis of young occurring at depth from August to October. This may account for the virtual absence of N_. ingolf ianus in the study area during the summer (September cruise). The small amount of data available on day-night movements indicated no migratory activity. Derichthys serpentinus Thirty-one specimens of this species, all of which appeared to be maturing to mature adults (T.L. range 170-270 mm, n = 18), were taken during the sampling periods. A highly significant difference in CPUE with depth was noted in the spring, the season of peak availability (Table 7b) (x2 = 17.63, p<.0034, df = 5). The vertical range of I), serpentinus strongly overlapped that of Nessorhamphus. The two frequently were found in the same trawl hauls. Derichthys however, showed a significantly different CPUE in the 501-750 m strata with season (x2 = 7.5, p<.0567, df = 3). As with N_. ingolfianus, this species appeared to be an inhabitant of the lower mesopelagic zone. Presumably, as its morphology and frequent appearance in otter trawls indicate, it may occur near the slope face on a regular basis as Sedberry (1975) suggested for Nessorhamphus. 47 Beebe (1935b) found Derichthys occurred at similar depth ranges, and with nearly equal frequency as Nessorhamphus off Bermuda. Castle (1970) reported that Derichthys is very widespread geographically, but probably is nowhere common. His findings also indicated the probable spawning locale as somewhat south of Bermuda. Suborder - Saccopharyngoidei Family - Eurypharyngidae Eurypharynx pelecanoides Eurypharynx appeared to be a common constituent of the deep bathypelagic zone around Norfolk Canyon. It occurred regularly in the nets throughout the study period, although total catch figures were quite low. Some references to this species (Marshall, 1954, Idyll, 1964) suggested that it is an inhabitant of the deep bathypelagial, but presented little data. Grey (1956), stated that the center of distribution in the western Atlantic lies between 1400-2800 m. Jahn (1976) noted that Eurypharynx was much more abundant in Slope Water than Northern Sargasso Sea water in samples to 1000 m depth. The distribution of specimens taken in Norfolk Canyon showed a strong similarity to Grey's (1956) depth analysis (Table 8 ). Examination of CPUE shows that the majority of specimens occurred at 1501 m and deeper. However, two of the larger specimens (ca. 550 mm T.L.) were captured at much shallower stations (742 m in spring, 906 m 49 in summer). This is of interest as Bohlke (1964) indicated that gulper eels appear to be poor swimmers. It has been suggested that Eurypharynx may occasionally occur near bottom (Marshall, 1954) based on stomach content evidence. This idea is supported by the results of the food habits section of this study (see next section). Order - Salmoniformes Suborder - Argentinoidei Family - Bathylagidae Bathylagus euryops Little can be said for this species other than it was a rare member of the community in the study area (a total of 8 specimens captured), occupying the mid- to lower bathypelagic zone (general depth of capture below 1400 m). One 108 mm S.L. specimen was taken at approximately 900 m. This species was tentatively identified as B. euryops following Cohen (1964). The bathylagids however, remain taxonomically confused because of their fragility and lack of specimens (Cohen, pers. comm.), therefore the identification of the specimens is tentative. Cohen (1964) noted that B. euryops (identified as B. benedicti) occurred from 700-1800 m off Bermuda, but that northern Bathylagus occurred between 500-1500 m. This coincided in part with the findings of this study. Deeper extension of the range however, could have been due to deeper sampling. 50 Suborder - Stomiatoidei Family - Photichthyidae Polymetme corythaeola ]?. corythaeola was represented by 32 specimens. It appeared to be a species which occurred in low to moderate abundance, as the depth stratum it inhabited (201-500 m) was heavily sampled. Polymetme was an upper mesopelagic fish in the study area. All but one capture came from the 201-500 m stratum, at an average depth of about 300 m. The single capture outside this stratum came from 600 m. The largest numbers of specimens, 17 and 12, occurred in spring and winter respectively; perhaps this species responded to cooler water temperatures. Grey (1964) reported this genus as benthic, because no records of midwater captures exist. It is difficult to accept this, in view of the morphology of the body and jaws, which strongly resemble Gonostoma elongatum and other Gonostoma species and Gonostomatid genera. Indeed, this species was placed in the Gonostomatidae for many years until Weitzman (1974) considered them sufficiently aberrant to be separated along with several other genera in a new family, the Photichthyidae. It is more probable that this is a "pseudooceanic" (sensu Nafpaktitis, et al., 1977) or "pseudopelagic" species (sensu Baird, 1971), occurring in close proximity to the bottom, near the shelf break area. Depth of distribution, agrees well with Grey (1964) who stated "They inhabit shallow depths, usually about 300-500 m." Table 9a. Vertical distribution of Gonostoma elongatum . ased on CPUS (spec./0,5 hr. Numbers in parentheses denote calculations including water haul samples. * * - CPUE converted to specimens/0.5 hrs. ++ - 3 seasons only. 52 Day-night capture information revealed that the specimens were taken only at night. The reason for this cannot be determined until more work has been done on this species. Family - Gonostomatidae Gonostoma elongatum The fifth most abundant species of midwater fish captured around Norfolk Canyon was G^. elongatum with a total of 75 specimens ranging from 80-280 mm S.L. Analysis of CPUE data indicated that this species had a broad depth range, inhabiting all but the shallowest stratum (75-200 m) throughout the year. Peak catches were generally in the 501-750 m range, spring and summer, and shifted to 751-1000 m in fall and winter (Table 9a). The CPUE differences among depth strata within a season was significant at p = 0.1 (two-tail test, x2 = 9.9, p<.0783, df = 5) for the spring cruise, while significant differences among seasonal CPUE were found in depth stratum 4 (751-1000 m), also at p = 0.1 (x2 = 6.4, p<.0938, df = 3). Gonostoma elongatum is geographically a widely distributed species, but is perhaps most common in the western Atlantic (Grey, 1964). Jahn (1976) noted that it appeared equally abundant between Slope and Sargasso Sea water masses. Krueger et al. (1975) reported it to be the "most abundant mesopelagic predator" in their study area, Deepwater Dumpsite 106, over the continental slope to the north of Norfolk Canyon (ca. 38-39.5°N, 71-73.5°W). They also stated "Both 53 larvae and adults are present, and this species undoubtedly spawns in the Deepwater Dump area." Grey (1964) indicated that principal captures occurred below 500 m, and theorized two areas of concentration, between 300-1100 m and 1400-1700 m. She listed two discrete captures, 1200 m off Bermuda, 850-950 m off S. Africa. Krueger and Bond (1972) examined Bermuda Ocean Acre material and showed evidence for extensive vertical migrations, with daytime captures between 600 and 1 0 0 0 m, and night captures between 50-350 m, with larger individuals deeper. Clarke (1974) found a similar pattern off Hawaii, a 560-725 m concentration by day, rising to 60-265 m at night. Krueger and Bond (1972) concluded that a portion of the population does not undergo migration, whereas Clarke (1974) suggested non-migratory habits for adults. The findings of the present study agree with these earlier results. Both the shallow and deep concentrations noted by Grey (1964) were indicated by capture data of G. elongatum in spring and summer. Krueger and Bond (1972) suggested a late spring (June) and summer spawning period for G^. elongatum off Bermuda. This may account for shallower occurrence of the species during these months in the present study area. Day-night capture data showed little evidence of migratory activity. This may be due to the capture of mostly adult specimens, which probably do not undertake migration (Clarke, 1974). 54 Gonostoma bathyphilum This species was a common year-round member of the bathypelagic fauna, and was represented by 26 specimens taken at 17 stations. Except for the capture of 2 specimens from about 700 m in winter (both taken at night) , (}. bathyphilum was captured in the 1501 m and deeper strata (Table 9b). Although no significant differences in CPUE were found within or between seasons. The number of specimens captured in the winter, increased three-fold over the next highest capture season (spring). It may be that this increased availability resulted from an upslope migration of G^. bathyphilum in winter. Depth of capture information agrees well with published accounts. Grey (1956, 1964) noted that the species is more common below 2000 m than above, and further stated, "The reduction in number and the rudimentary nature of the photophores, as well as the absence of a swim bladder, are correlated with depth of habitat." Krueger and Bond (1972) captured 9 specimens at depths between 1200-1900 m off Bermuda. Although the shallowest specimen captures were at night, not enough data were available to delineate day-night trends. Diel migration would not be expected for a bathypelagic species (Marshall, 1954, 1971). An increasing size with depth trend was noted for this species. The mean size for the two specimens taken at 700 m was 111.5 mm S.L. However, from 1501-2000 m, x = 138.7, 2001-2500 m; x = 148.5, and 2501-3100 m, x = 153.5. 56 Gyclothone microdon This species also occurred primarily in and below the 1501-2000 m depth stratum, although several specimens were taken in shoaler waters (ca. 501-750 m) in the spring. Generally, abundances were low except in the spring. Of 134 total specimens, 123 were taken during spring sampling. The "rarity" at other times of the year may have been due to overlooking their presence in the net. Cyclothone frequently "hangs up” in great numbers in the wings, and may easily be overlooked (or ignored) if the meshes are not carefully and thoroughly examined. Cyclothone are quite abundant throughout the world oceans, as noted by many accounts (e.g. Grey, 1964; Mukhacheva, 1964; Badcock, 1970; Krueger and Bond, 1972; Kobayashi, 1973; Krueger et al . , 1975, 1977; Jahn, 1976; Badcock and Merrett, 1977). G. microdon appears to be the most abundant species of the genus in slope waters (Jahn, 1976, Krueger et al. 1977). This abundance may however be biased because this species is one of the largest of the genus attaining a size of >60 mm S.L. Other abundant, but smaller species may have escaped through the net meshes. Kobayashi (1973) listed a maximum abundance of (3. microdon taken by Dana expeditions as lying between 1500 - 1599 m with most of the population at 800 m and deeper. He also noted a lack of vertical migration for this species. The depth information in published accounts agrees well with the present findings, however, data was too scanty to speculate on vertical migration. 57 Family - Sternoptychidae Maurolicus muelleri Maurolicus was represented by a total of 293 specimens. Of these, 266 were taken during the spring of which 249 came from a single station at 123 m. This species exhibited a very narrow depth range, mostly between 100-300 m. One specimen was taken during the fall at a 700 m station and may represent a contaminant from an earlier station. This species has been reported to be an inhabitant of near-shelf waters (Okiyama, cited by Krueger et al., 1977) and appears to have a "pseudoceanic” distribution. Grey (1964) reported specimens from deep trawls, but stated that "The species probably does not normally descend far below 300-400 m." She also noted that it often occurs in tremendous schools near the surface. Jahn (1976) found Maurolicus to have a strong preference for Slope Water. Musick (1973) included this species as a permanent year-round resident in the Gulf of Maine. M. muelleri was the shallowest regularly caught midwater species in this study, and was assigned to the upper mesopelagic group. All captures of Maurolicus occurred at night, as with Polymetme corythaeola, a species of apparently similar habits. In addition, the most frequent captures were spring and winter. 58 Argyropelecus aculeatus A. aculeatus was one of the less common dominant species. Only 20 specimens ranging from 33-83 mm S.L. were captured. Krueger et al. (1977) captured 27 specimens 11-56 mm S.L. around Deepwater Dumpsite 106, and included this species as a "spawner" in that area. Its rarity around Norfolk Canyon may point to a center of distribution and spawning area outside the study area. The capture data for this species among bathymetric strata presented an anomalous pattern that was difficult to interpret. Table 1 shows that this species was most abundant in spring and steadily declined in numbers summer through winter. Table 10a indicates that the distribution appears to be split into a shallow (201-750 m) and deep (>1000 m) group. There was no significant difference in CPUE with depth or season. Published accounts (Baird, 1971; Badcock and Merrett, 1977) showed A. aculeatus to be an upper mesopelagic form whose center of distribution was around 200-550 m by day with upward migrations to 50-200 m at night. It has been captured by dip net at the surface at night (pers. observation). The deepest record for the species is ca. 1000 m. It is difficult to interpret the capture of 14 out of 20 specimens in the deeper stations. A rarely sampled part of the population may occur at depths > 1 0 0 0 m or the species may avoid the shallower strata where their usual bathymetric habitat impinges on the 60 slope. A trend for decreased mean size with depth was noted, although sample sizes were too small to draw any conclusions regarding this point. Sternoptyx diaphana Sternoptyx was represented by 30 specimens taken mostly in the lower mesopelagic in the study area. The species abundance remained relatively constant over the seasons except during the winter when it declined sharply. CPUE data (Table 10b) showed fluctuating availability with depth which was not significant. No significant seasonal trends were observed. Baird (1971) gave a depth range for this species of 400-1200 m with peak concentrations between 700-900 m, and no evidence for vertical migration. This is supported by the results of this work. Again, however, there appeared to be a deeper dwelling or more oceanic contingent for whose presence there is no explanation. The narrow range of sizes (26-34 mm S.L. except for one specimen 44 mm S.L.) suggests a single year class was present in the canyon area at all times. Perhaps only juveniles or sub—adults inhabit the canyon region. Krueger et al. (1975) present evidence for Sternoptyx diaphana spawning populations to the north. Jahn (1976) noted a slight increase in abundance this species in the Northern Sargasso Sea waters, as compared to Slope Waters. 62 Family - Chauliodontidae Chauliodus sloani This is the seventh most abundant dominant species of the study area, with 56 specimens captured. The fact that no specimens were less than approximately 100 mm S.L. suggests that this species is recruited to the study area although it may spawn near Norfolk Canyon. Kruskal-Wallis analysis of CPUE revealed no significant differences within or among seasons. However, certain trends were apparent (Table 11a). Specimen capture was relatively even except for the summer when a large drop in numbers was noted (Table 1). The only report of spawning for this species (Clarke, 1974; off Hawaii) indicates a spring-summer activity. Perhaps C. sloani migrated out of the study area for spawning purposes in the summer. Specimen captures were generally from deeper waters. Only one specimen was collected above the 751-1000 m depth stratum. The majority of specimens occurred in the lower mesopelagic and upper bathypelagic (1001-1500 m) strata. This deep displacement of the population may be to avoid competition with Stomias boa ferox another common predator whose depth range overlaps C^. sloani (see Feeding Habits section for full discussion). Morrow (1964) listed the vertical range of this species in the Western North Atlantic as 1000-1800 m during the day, with extensive vertical migration into the upper 800 m at night. Badcock (1970) also noted this migration pattern but set the upper day depth at 500 m. 64 Clarke (1974) located it between 450-825 m by day, and 45-225 m at night off Hawaii, but noted that larger specimens (>120 mm) stayed below 600 m by day and 175 m at night. In addition, he collected very few large fish (>185 mm) whereas nearly 60% of the specimens taken in this study were >_185 mm S.L. This larger-deeper phenomenon was earlier reported by Ege (1948) in his monograph on Chauliodus. He noted "...we notice at once, that the larger specimens from a length of 1 0 0 mm, were not taken with less wire out than 300 m, but chiefly with 1000 metres wire." For the DANA expeditions, the ratio of meters of wire out to fishing depth was 2 : 1 (Gibbs, 1969; Sulak, pers. comm.). Ege (1948) concluded that 2000 meters wire out (ca. 1 0 0 0 m) produced the largest catches of C^. sloani. These figures substantiate the present findings. In addition, there was some evidence for vertical migration from the Norfolk Canyon data, but more information is needed before making any definite statements. Family - Stomiatidae Stomias boa ferox Stomias was the third most abundant species overall. Eighty-four specimens were captured (size range 90-378 mm S.L.). This species showed a vertical distribution pattern similar to, but broader than that of Chauliodus sloani. Most of the specimens again were found in the mesopelagic, but shallower than Chauliodus (Table lib). CPUE analyses showed significant (x^ = 11.2, p<.0469; = 13.8, p<.0321, df = 5) differences among depth strata during the spring and fall, which were the periods of highest abundance. Fall, seemed to be the season in which S^. J). ferox was most available for capture, because nearly twice as many specimens were caught at that time as were taken in the spring (Table 1). No significant difference in CPUE between seasons was noted however. Based on CPUE, the 501-1000 m depth stratum was the peak depth of capture. Catch rates were equivalent between 501-750 m and 751-1000 m in the spring, and nearly so in fall and winter although peaks were in 501-750 m stratum during these seasons. During the summer, however, peak CPUE was in the 751-1000 m stratum. Chauliodus sloani, which showed considerable distributional overlap with S. b. ferox was most abundantly captured between 751-1000 m bottom depths. There exists little published information regarding the depth ranges occupied by this species, and that which exists is confusing. Ege (1934), indicated that S. b. ferox is largely restricted to shallow depths, theorizing that deep catches of small specimens arose from contamination. Examination of his data reveals that specimens >_200 mm S.L. formed 0.4% of the total catch, whereas in the present study specimens >200 mm size comprised 70.2% of the total. Morrow (1964) stated that S. b. ferox appears to be most abundant in the 67 upper water layers, 300 m depth. His data were apparently based largely on Ege's work. Beebe (1937) listed a total catch of 96 "Stomias ferox" in meter-net hauls from depths of 300-1,000 fathoms off Bermuda. Gibbs (1969) in his monograph on Stomias reanalyzed the Dana material of Ege (1934) and postulated a daytime occurrence of the species below 500 m, with a night time concentration in the upper 200 m. Krueger et al. (1975) occupied ten stations over Deepwater Dumpsite 106 to the north of Norfolk Canyon, using a 10 ft. Isaacs-Kidd midwater trawl (IKMT). All sampling was done at night. Stomias boa ferox were taken at eight stations with maximum depths between 615-790 m. Two other stations, both extending to 550 m contained no Stomias. The present data agrees with the findings of the works by Beebe (1937) and Gibbs (1969) although none of these studies adequately sampled deeper waters. A day-night difference in depth of capture was noted in fall (the strongest data set), and to a lesser degree in spring (Table 12). Too few specimens were collected in summer and winter for any clear trend to be discerned. Family - Malacosteidae Malacosteus niger This species was rare in the canyon area with 13 specimens captured at 11 stations. There is little that can be reported regarding this species. So few specimens were captured that CPUE variations presented little information. Only three specimens, 68 o o CM o o o o o o ffj o o m o o o o o o o o o co CO oo CO * o o o o a o o O o o o i «••• • • i o o <—1 CM CM rH i o o o O OO o o o m O o O o CM m r"~ O m O o 1- 1 r—1 CM CM ffi H ■1 in irH i—i 1-1 r—1 T""f £ j3 r'. o o in o o A CM m r^» o in § CO I—1 i—i 69 however, were captured outside the 1001-2000 m depth strata. One of these was very shallow, at 376 m, and probably represented contamination from an earlier tow. Clarke (1974) reported that M. niger off Hawaii occupied a range from 520-900 m, with most taken from 600-700 m. The size range of most of his material overlapped that of Norfolk Canyon specimens. Beebe (1937) listed 26 M. niger captured between 500-1000 fathoms off Bermuda. Badcock (1970) reported but a single specimen, captured off Fuerteventura between 875-700 m in a closing IKMT. This species, which has a broad distribution worldwide may be found in the lower mesopelagic in some areas and the bathypelagic in others. It appeared to be a solitary form well segregated from other individuals of the same species. Photostomias guernei Photostomias guernei was the rarest midwater species to be considered a dominant member of the Norfolk Canyon midwater community. A total of 5 specimens was captured. Records (Beebe, 1937; Badcock, 1970; Clarke, 1974) indicate that Photostomias seems to be more common than Malacosteus. Beebe (1937) listed a total of 99 specimens (compared with 26 M. niger) from 500-1,000 fathoms depth. Badcock (1970) captured 16 specimens off Fuerteventura, 13 of which occurred between 350-950 mm by day and night. One specimen was taken near the surface (<70 m) at night, 70 while the remaining 2 individuals came from day tows shallower than 200 m, and could have been contaminants. Clarke (1974) presented data on 159 specimens from Hawaii, and showed a mesopelagic (350-800 m) distribution by day, and migration into the epipelagic (15-300 m) at night. Two specimens each of F_. guernei occurred in 751-1000 m (818 and 828 m), and 1001-1500 m (1408 and 1488 m) strata hauls. Such close agreement (the two shallow captures were in summer and winter, the deeper in fall and spring respectively) suggest that the species lives deeper in fall and spring, than summer and winter. The fifth specimen was taken at 2624 m in fall. This species also is considered to have a bathymetrically transitional pattern until more information can be gathered. Order - Myctophiformes Family - Myctophidae Benthosema glaciale This species was the second most abundant species captured in the Norfolk Canyon area, with a total of 139 specimens taken from 41 stations. JB. glaciale is considered the most abundant myctophid species in the North Atlantic, dominating subpolar-temperate myctophid catches (Nafpaktitis et al., 1977). Its geographic range extends down to about 40°N on both sides of the Atlantic (somewhat further south on the western side), and into the Mediterranean Sea, with an apparently 71 isolated population in a small area in the upwelling region off North Africa (Badcock and Merrett, 1977). Norfolk Canyon CPUE data showed a broad vertical range with two abundance peaks (Table 13a). No significant difference in catch rates from different depth strata was observed, but CPUE among seasons was significantly different (x^ = 8.2, p<.0822, df = 3) for the 751-1000 m depth range. Abundances of Benthosema appeared to be divided into two distinct groups, one centered in strata < 1 0 0 0 m, the other, below (Table 13a). Peak catches for each group were in the strata from 751-1000 m (spring-fall), and 501-750 m (summer-winter) for the upper group, and 1501-2000 m (all seasons) for the lower. Except for two specimens captured by water haul, no Benthosema were taken in the stratum below 2 0 0 0 m. Computation of mean sizes vs. depth strata indicated a trend of decreasing mean fish size with increasing depth, a trend which would probably not occur if contamination was the cause of the apparent deep captures. Mean fish length in each depth stratum, calculated using data pooled from all seasons (N=129) was; 201-500 m (n = 37) x = 50.5 mm, 501-750 m (n = 16) x = 53.2, 751-1000 m (n - 15) x = 47.3, 1001-1500 m (n = 17) x = 46.4, and 1501-2000 m (n = 44) x = 43.0. The data were divided into two groups, the shallow (<1000 m) and the deep ( > 1 0 0 0 m) catches (sample sizes of 6 8 and 61 respectively). Figure 6 presents frequency curves and mean size for each group. A 73 Wilcoxon two sample test, a non-parametric test which ranks the data from smallest to largest was run according to the methods of Sokal and Rohlf (1969). Because the n ’s were large, an expression which approximates normality was used, and a t-value generated. The null hypothesis was that the mean size of fish captured above and below 1000 m was not significantly different. This is what would be expected if a) a single population of individuals was being sampled, and b) contamination was the reason for many of the apparently "deeper" captures. The null hypothesis was found to be false, at a high level of significance (t = 3.63, p<.001). This pattern may be the result of several factors. Smaller fishes may occur deeper than larger ones. This situation is not uncommon in midwater fishes (e.g. ceratoid larvae and young, Bertelsen, 1951), but it does not seem to be true for Benthosema. Halliday (1970) found that in slope waters off Nova Scotia, "0- group fish ... were taken in both the shallowest and deepest day and night tows." He also found that by day, mean size showed an increase at the 201-250 f depths vs. shallower hauls, and then decreased again in the underlying 250-300 f stratum, a situation seen in Norfolk Canyon. At night, mean length increased with depth, the opposite to what was observed from the canyon data. Another argument against the smaller-deeper theory is that VIMS planktology data from opening/closing bongo trawls at Station L 6 , a station located at the northwestern end of Norfolk Canyon showed a wide range of B. glaciale 74 Figure 6 . Frequency curves and mean sizes of Benthosema glaciale captured in trawls from shallower than 1 0 0 0 m bottom depth (solid line) and deeper than 1 0 0 0 m bottom depth (dashed line) »—- <— » E •E E E in GO o ro in IK IK ' ------ o □ UJ UJ cr IK Z> Z> ^ t— t- — in Q. 10 a . 10 h- < •' < " O C o c - .. CO o !39) z E Z E n- UJ O UJ O ID 2 o 2 O ID o 2 5 2 m >-1 UJ ^ UJ ^ up L u ll Q. CL id * CO (0 o 0 10 1 1 10 T1 in 1 o in in in O in i 10 in 5" o ID fO in O —---- ro O r O lD CM m CM SIZE RANGE ( mm SL) Benthosema glaciqle (n = o = (n mm glaciqle SL) Benthosema RANGE ( SIZE CVII iS in " T ~ "T" ~ r ~ O lO O m m r o CVJ OJ HD1VD 1V101jo % 76 sizes (9.2 - 36.7 mm) in <200 m of water, although specimens were not abundant. A second possibility is that the captures of small 15. glaciale further offshore and/or deeper represented a deep group of expatriated individuals, carried southwards by the submergence of boreal surface waters. A number of factors lend credence to this hypothesis. Several accounts have noted the presence of deep, apparently non-migratory groups of 15. glaciale at the southern ends of the species* range. Beebe (1939) reported two small (26-18 mm) specimens of "Myctophum glaciale" off Bermuda from 700-900 fms depth. Gibbs et al. (1970) found a number of J5. glaciale during Bermuda Ocean Acre studies with maximum abundance calculated to be 1000 m. They estimated most of the captured specimens to be in the year class one size range, and suggested that this species spawned north of the Gulf Stream system and did not occur in the Ocean Acre area as adults. They also theorized that Benthosema made its way to Bermuda at depth via the North Atlantic Central waters. Badcock and Merrett (1977) noted that at the southern end of its range in the Eastern North Atlantic (ca. 40°N), B. glaciale exhibits a split vertical distribution with a large, deep (750-1000 m) non-migratory fraction of individuals. Additional evidence suggestive of southerly transport of Benthosema was collected during cruises 49 (July-August 1978) and 55 (February-March 1979) of the R/V Oceanus from Woods Hole. Deep discrete samples (800-1000 m) made using a 10 m M0CNESS sampler (Wiebe 77 et al., 1976) in Western Boundary Undercurrent (WBUC) off Cape Romaine, S.C. captured large numbers of Benthosema glaciale and Sergestes arcticus (during Oceanus 49 the catch bucket was full of nothing but these species; pers. observation). The WBUC may transport these boreal species southwards, and could be one of the mechanisms by which Benthosema reached Bermuda (Backus, pers. comm.). Another argument for the expatriation theory is examination of the species sizes in the Norfolk Canyon deep waters with respect to Hallidays' (1970) analysis of age and growth. He found that 15. glaciale spawned off Nova Scotia in early spring, with post larvae 10-25 mm being found in July. The only captures of specimens within this size range near Norfolk Canyon were in spring (June) and summer (September) when a 21 and 18 mm specimen were captured in nets fished between 1501-2000 m. Halliday (1970) also found Benthosema distributed over a wide temperature range from 0°-18°C, with peak abundances between 4-16°C. The 4° isotherm in Norfolk Canyon occurs around 1400 m (Wenner, 1978). Leavitt (1938) noted a plankton maximum between 1400-1600 m in Slope Waters, 1.5 to 2 times greater than that of immediately shallower or deeper layers. Thus biological and physical features appear to be potentially favorable for a deep dwelling group of Benthosema near Norfolk Canyon. The mean sizes of the shallow and deep groups (Figure 6 ) indicated about a one year difference in age, according to Hallidayfs 78 figures. The shallow group appeared to be composed mostly of older individuals about 3.5 years old (range about 1.5-4), while the deep group averaged about one year younger (range from young of year to 4). The vertical distribution pattern for B^. glaciale in the depth strata above 1000 m agrees well with existing data (Halliday, 1970; Badcock and Merrett, 1977; Nafpaktitis et a l . , 1977). Benthosema was virtually absent from winter samples. This did not appear to be a measure of sampling effort (or lack of). It could be related to spawning which occurs in winter and early spring farther north (Halliday, 1970). This phenomenon was noted for the other common myctophid species Ceratoscopelus maderensis, and the hatchet fishes Argyropelecus aculeatus and Sternoptyx diaphana, and is at present inexplicable. Both Ceratoscopelus and Argyropelecus exhibited the smaller-deeper trend shown by Benthosema. Ceratoscopelus maderensis A total of 84 specimens of this species was captured, making it the second most abundant myctophid, and fourth most abundant dominant species. CPUE analyses showed no significant differences with depth or season. This species occurred throughout the mesopelagic (201-1000 m) zone, and primarily occurred between 201-750 m (Table 13b). A number of similarities between this species and Benthosema were noted. First, the phenomenon of decreasing mean fish size with 80 distance offshore and/or increasing depth was noted. Calculations of mean size with depth strata showed a generally decreasing trend. Wilcoxon two sample test of < 1 0 0 0 and > 1 0 0 0 m depth strata groups (n = 61 and n = 11 respectively) again showed a highly significant difference (t = 2.43, p<.02) in mean size of individuals. Figure 7 is a graph of size frequency for each group. All specimens were from larger size ranges. Krueger et al. (1975) noted this pattern at Deepwater Dumpsite 106 and stated, "We suspect that this species does not reproduce in the study area, and that the present specimens are expatriates from colder waters." Due to the small sample size of the deep group, little more will be said about it. Mesopelagic captures of maderensis in the < 1 0 0 0 m depth strata agreed well with published information. Backus et al. (1968), Musick (1973), Jahn (1976), Krueger et al. (1977) and Nafpaktitis et al. (1977) all note that C. maderensis is abundant in the cold slope water, with vertical ranges of 225-1000 m with maxium abundances between 300-600 (Backus et al., 1968; Nafpaktitis et al., 1977). Day-night differences in depth stratum (shallower at night) are clearly seen in Table 14. Backus et al. (1968) reported a mean size of 62.4 mm for 774 specimens captured from enormous schools of Ceratoscopelus observed by D.S.R.V. Alvin in slope waters off southern new England. This figure agrees well with the 64.2 mm mean size found for the 61 mesopelagic jC. maderensis in the present study. 81 Figure 7. Frequency curves and mean sizes of Ceratoscopelus maderensis captured in trawls from shallower than 1 0 0 0 m bottom depth (solid line) and deeper than 1 0 0 0 m bottom depth (dashed line) 82 o UJ CO lf> If) 10 (0 If) If) If) If) _ (0 fO maderensis Ceratoscopelus RANGE SL) (mm SIZE o in O m o m m <3- ro rO C\J CM H01V0 1V101 P o/o DAY_NIGHT CPUE(spec./0.5 hr.) FOR Ceratoscopelus maderensis w i P a: o o o o o o c> o o o o o o o m CM 1 1 * . ♦ • • « • • • • • « • • • • • • • o o o CM CM CO o l—l r-H m o o m o ON CM o 1 » o o n* i— i i o o o CO CM o r—1 m co 1 1 1 1 • r—1 m o o o o o CO m o o p'* I—I o o o o o l—l in i i * « • CO o o o 1—1 t— i m l—l o o o o CO m o 1 1 1 o 1 « e • * • o o o in o i— i CO i— l o o CM m o o o o o « o • • 84 Lampanyctus macdonaldi Only 19 specimens of this large myctophid were taken (74-154 mm S.L.) making it one of the rarer dominant species in the study area. It was probably bathypelagic in the canyon area (Table 15), with only one specimen captured shallower (809 m) than 1000 m. This species was most common in summer (10 specimens captured). L. macdonaldi has been noted by submersible observers within 2 m of the bottom (Craddock, pers. comm.). Norfolk Canyon otter trawl captures may be of individuals from near bottom. There are few published records of the distribution of this species. Jahn (1976) listed it as having strong Slope Water preferences. Nafpaktitis et al. (1977) set distributional limits between 550 to greater than 1000 m by day, and about 850 m at night for adults, although juveniles occurred much shallower. Order - Gadiformes Suborder - Zoarcoidei Family - Zoarcidae Melanostigma atlanticum This species was the eighth most abundant. A total of 53 specimens was captured. Aspects of the distribution and ecology of this species in Norfolk Canyon have been studied and discussed by Wenner (1978) and Markle and Wenner (1979). M. atlanticum was a mesopelagic inhabitant, and was most abundant in the summer. Peak 86 CPUE was between 501-1000 m spring and summer, and 201-750 m fall and winter. Order - Beryciformes Suborder - Berycoidei Family - Melamyphaidae Scopelogadus beanii A total of 6 8 specimens of S_^ beanii was obtained from 31 stations. This species was the sixth most abundant dominant species found in the study area. Catch-per-unit-effort data (Table 16) shows that this species was probably bathypelagic around Norfolk Canyon. Peak abundances occurred in the 1501-2000 m or deeper strata. Other bathypelagic species, e.g. Gonostoma bathyphilum, Eurypharynx pelecanoides, and Lampanyctus macdonaldi were represented by far fewer specimens. Only 10 of the specimens were not adults (_<8 Q mm S.L.). These 10 were all half grown according to Ebeling and Weeds classification (1963). The other 58 specimens ranged in size from 80-118 mm S.L. Ebeling and Weed (1963) listed the upper limit of vertical distribution for this species as 800-1000 m for adults, 500-600 m for half grown. They also noted that vertical migrations probably do not occur. Except for one adult specimen taken at ca. 650 m depth, no captures occurred shallower than 800 m for either adults or half grown. Half grown specimens were not taken at stations deeper than 1850 m. Table 16. Vertical distribution of Scopelogadus beanii based on CPUE (spec./0.5 hr Numbers in parentheses denote calculations including water haul samples. * * - CPUE converted to specimens/0.5 hrs. ++ ~ 3 seasons only. 88 Jahn (1976) found S_, beanii to be far more abundant in Slope than Sargasso Sea waters. The large number of specimens taken during the present study may have been the result of the presence of this species near the bottom. Stomach content and parasitological information (see other sections of this study) indicated a possible benthic interaction. High CPUE (up to 3.0 spec/0.5 hr.) showed greater availability to otter trawl capture than any other bathypelagic species. The largest numbers of specimens were captured in the summer and fall, with as many as 7 specimens taken in a single haul. CPUE, however, was not significantly different with depth or season. Non-Dominant Species A total of 53 species of meso- and bathypelagic fishes were represented by captures from three or fewer cruises. A number of the species were represented by a single specimen. From published accounts, some species, which were probably under-represented in the catches, may be year round members of the Norfolk Canyon midwater fish community. Appendix B provides a taxonomic listing of these species, the cruises during which they were taken, and the depth that the net fished. 89 General Discussion and Conclusions Bias of Data The major problem in analyzing collections of pelagic fishes from non-closing nets is that no statement can be made as to where the fishes were taken in the water column. Open net catches may be contaminated, especially by shallow dwelling species, during net descent and/or ascent (Pearcy 1964; Harrison, 1967; Ebeling et a l . , 1970; Halliday, 1970; Clarke, 1973, 1974, Krefft, 1970). This problem may be minimized by fishing the net at a given depth for a period of time far exceeding that of descent and ascent (Pearcy, 1964; Halliday, 1970), or by assessing transition of species distribution by upper depth of capture limits (Clarke, 1973, 1974). Comprehensive studies of distribution are often confounded by such factors as the variability of depth distribution with geographic locale (e.g. the vertical distribution of Benthosema glaciale in the N.W. Atlantic is 275-850 m by day, 0-225 at night; in the Mediterranean Sea, 375-800 m day, 12-200 m at night, but off Bermuda _B. glaciale is found between 750-1200 m by day, (Nafpaktitis et al., 1977), and the apparent worldwide distribution of a number of deeper-dwelling species (e.g. Chauliodu sloani, Morrow, 1964; Eurypharynx pelecanoides, Bohlke, 1966). The diel vertical migratory habits exhibited by many fish species (Pearcy and Laurs, 1966; Gibbs, 1969; Gibbs et al., 1970; Clarke, 1973; Clarke and Wagner 1976; Krefft, 1976, and many others) also make interpretations of data 90 difficult especially when attempting to delineate patterns for a number of species. Recent evidence (Gibbs et a l . , 1970; Pearcy et al., 1979) suggests that not all members of a population migrate all the time. The otter trawl used in this study was a bottom sampling device. Pelagic fishes may have been captured during descent or ascent of the net or they may have been captured near the bottom. Contamination of the net during descent or ascent was probably not significant for dominant, abundant species. Upper limits of capture of species generally agreed with published accounts. The number of individuals of a given species captured over a particular depth range was consistent among seasons in many cases. Statistically significant differences in catch rates were found for a given species among different depth strata. Examination of species data on the upper depth of capture, CPUE for depth strata, and broad depth strata as indicators of distributional transitions between species groups allowed for the definition of species groups by faunal zone. Table 17 lists the dominant species and the depth zone (e.g. mesopelagic, bathypelagic, etc.) which they inhabit. Comparison of depth of capture data vs. published accounts for many species showed excellent agreement in all but three species, Benthosema glaciale, Ceratoscopelus maderensis, and Argyropelecus aculeatus which apparently occurred much deeper than most published accounts have noted. These occurrences could have been due to net contamination upon descent or ascent of the net. However, a DISTRIBUTIONAL GROUPS OF DOMINANT MIDWATER FISHES IN THE NORFOLK CANYON REGION ”>>-C —i —i i— i i— r—I i—i •H A £ h PQ h Q P •rH i— l r—H i—I •H •r ctitb H • S I S S fc1" t 4-1 . P P b bO H * i—l o H M C o r—1 i—H o o *—I o n i _ a E e B O P 0 1 i n } ■> i— 1 —l r— P s 1 3 i H * w|c^)|cn|cj| •H *H P i — — 4-t •H o P P p P P p aj •• • t> 1 p iP P MH 1— 1 H • W H • MH •H i— l S3 ■H s 4-1 a & Cti P QJ bC P p o p p P X p o P P bC p p p o p U o • S | o | o P P *rH ,p TJ tH ’H c/U •H 4-1 CO p Cti p QJ CO QJ > P p P P O p p p- o o i H • 4-4 — P P bO P p O O P u O £ E P . 1 t i— l rP ♦H H • 0 4-1 — 4-> 4-1 P 6 P B P P P a P P • . . i |cq|i-l|cn| p p o o p P CO i—l P p P P o p B . H • Q U H • P P P . CD QJ a o cu 0 CO o a) bC 1—1 P U p P 1 p g P p p Cti P QJ CO P P p CO co p p CO O P QJ Cti CO o CO p CO 1 1 I • • #s i P3 r—1 p p p P p o o P co P £ PU • 91 92 statistically significant difference in mean size between individuals captured in the shallower and deeper depth strata (smaller fishes found deeper) for the first two species, and a similar trend for the third species was found, indicating the possible presence of separate deeper or further offshore dwelling groups of individuals of these species. In addition, very few published midwater studies have conducted extensive sampling below 1 0 0 0 m, so there is little information about deep distribution patterns. Net efficiency in capturing midwater fishes was assessed qualitatively. Comparisons of "water hauls," tows which were made in midwater rather than on bottom due to an insufficient scope of cable, showed much higher capture rates of both species and specimens in comparison to successful benthic tows, suggesting poor efficiency of capture in midwater during descent or ascent of net that successfully sampled the bottom. Capture ratios are approximately 3:1 water haul, benthic haul respectively, (12.2 species, 36.2 specimens per water haul, 4.8 species, 10.9 specimens per benthic haul excluding one catch of 257 specimens in a shallow haul). There is increasing evidence in the literature for the near-bottom presence of midwater fish species. Many of the dominant species found in this study have been implicated in this near-bottom phenomenon. Marshall (1954) reported a sea urchin in the stomach of Eurypharynx pelecanoides and concluded that it was scavenged off the bottom. Additional evidence of Eurypharynx ingesting benthic prey was indicated in this study by the presence of a holothurian in the 93 stomach of one of the specimens examined for food habits (see next section). Grey (1964) in her species account of the family Gonostomatidae indicated that Maurolicus muelleri, Gonostoma elongatum, and CJ. bathyphilum spend time quite close to the bottom. In the same account, she considered Polymetme corythaeola to be a benthic species. Submersible observations (Musick, pers. comm.) and presence in the stomach of benthic piscivores (in Bathysaurus ferox; pers. observation) confirm the presence of G. bathyphilum near bottom. Submersible observers have also noted the regular presence of Lampanyctus maconaldi and other myctophids (Craddock, pers. comm.) within 2 m of bottom. Sedberry (1975) noted the consistent capture of Nessorhamphus ingolfianus in the otter trawl samples from Norfolk Canyon and concluded that it may spend time on or near bottom. In addition, Sedberry and Musick (1978) recorded specimens of Ceratoscopelus maderensis, Nemichthys scolopaceus, Sternoptyx diaphana and Cyclothone sp. from the stomachs of Synaphobranchus kaupi (first two species) and Phycis chesteri (latter two species), both inhabitants of the demersal zone of Norfolk Canyon. These prey items occurred in specimens captured during the day, and appeared to have been consumed while alive. Sedberry and Musick (1978) concluded that the lower end of the vertical range would bring these pelagic species close to the bottom (especially during the day for vertical migrators) where predation by 94 benthic or demersal species then occurred. This conclusion is reinforced by the findings of Marshall and Merrett (1977; cf. 494). Markle and Wenner (1979) found that the near-bottom abundance of Melanostigma atlanticum and Xenodermichthys copei varied seasonally. Based on CPUE, food studies, and internal parasite faunas they postulated that these species approach bottom at certain times of the year to utilize the two-dimensional slopewater interface for reproductive aggregation. Other species of fishes have been suggested to have near bottom affinities. Devany (1969) defined the species Neoscopelus macrolepidotus as benthic or engybenthic. Clarke (1973, pers. comm.) noted the consistent captures of larger numbers of five species of myctophids and two species of stomiatoids in bottom trawls when compared with midwater hauls off Hawaii. The larger-deeper phenomenon has been noted by most workers dealing with vertical distribution of midwater fishes (most citations within this section of study mention this phenomenon). Clarke (1973 pers. comm.) noted that the bottom dwelling myctophids and stomiatoids were larger in size and heavier bodied than those individuals of the same or related species taken in midwater hauls. Grey (1964) stated with respect to Gonostoma bathyphilum captures, "...many of the larger specimens have been caught in bottom trawls." The Lampanyctus macdonaldi (also Lampadena sp.) observed by Craddock were all 95 approximately 6 in. in length (pers. comm.), i.e. the adult size range for these species. Harrisson (1967) correlated the presence of large specimens of myctophids with areas overlain by high standing crops of zooplankton, "...along the edges of continental shelves and mid-oceanic ridges or near island groups," which is precisely where pseudooceanic species are found. The presence of only larger specimens of the dominant species from Norfolk Canyon in the otter trawl is probably due to their presence near-bottom and resultant availability of these species to the net. Sampling bias through loss of small specimens does not appear to have been a major factor because small Cyclothone and myctophids down to 18 mm S.L. were consistently present in catches. A total of 23 dominant species were captured from the Norfolk Submarine Canyon and adjacent "open" continental slope to the south. Analysis of CPUE delineated five main distributional groups: the upper mesopelagic (201-750 m) characterized by Maurolicus muelleri and Polymetme corythaeola, the mesopelagic group (201-1000 m) consisting of Nemichthys scolopaceus, Stamias boa ferox, Chauliodus sloani, Gonostoma elongaturn, and Melanostigma atlanticum; the lower mesopelagic (501-1000 m) group of Derichthys serpentinus, Nessorhamphus ingolfianus, and Sternoptyx diaphana; a lower mesopelagic/upper bathypelagic transitional group consisting of the malacosteids Photostomias guernei and Malacosteus niger, and the 96 bathypelagic species Eurypharynx pelecanoides, Serrivomer beanii, S. brevidentatus, Cyclothone microdon, Gonostoma bathyphilum, Bathylagus euryops, Lampanyctus macdonaldi, and Scopelagadus beanii. The remaining three species, Benthosema glaciale, Ceratoscopelus maderensis and Argyropelocus aculeatus, had bimodal vertical distributions, of which the shallow group was primarily found in the upper mesopelagic, whereas the deeper individuals occurred in the bathypelagic zone. It was found that the two groups were significantly different in size (and probably age) for Benthosema and Ceratoscopelus. This trend was noted for A. aculeatus, but due to small sample size, was not statistically significant. Thus, a total of 10 mesopelagic, 2 transitional, 8 bathypelagic and 3 bimodal species distribution, were found to comprise the dominant species assemblages of the Norfolk Canyon area. Many workers have noted that hydrographic properties of water masses play an important role in determining geographic range of midwater fishes (Ebeling, 1962; Ebeling and Weed, 1963; Backus et al. 1965, 1969, 1970, 1977; Haedrich, 1972; Robertson et al., 1978). None of the dominant species found in the present study are ecologically limited to the Slope Water, although several may be most abundant in this water mass. Most of the species are best characterized as subarctic through subtropical in distribution, although a few such as Eurypharynx pelecanoides, Gonostoma bathyphilum, and Chauliodus sloani are cosmopolitan. 97 Many midwater species may be little constrained vertically by temperature or salinity changes, especially those with diel migratory habits. Light, however, appears to play an important role in determining the upper limit of depth distribution (see Marshall, 1971, cf. 37, and Nafpaktitis et al., 1977, cf. 18 for discussion and references). While light may limit the upper end of vertical distribution, it need not affect the lower limit, as noted for plankton by Leavitt (1938). The abundance of a number of species might be linked to the high productivity of slope waters as opposed to more central ocean waters (Be et al., 1971). Upwelling from deep waters may occur over the continental slope. (Ruzecki, 1979). Such increased productivity undoubtedly supports larger numbers and biomass of midwater fishes than those in less productive areas. More work (preferably with bottom sampling closing gear) needs to be done at all depths to determine the extent of midwater fish interactions with the near-bottom waters over continental slope and other submerged topographic features (e.g. seamounts). PART II. FEEDING HABITS AND TROPHIC ECOLOGY OF SELECTED DOMINANT MESO- AND BATHYPELAGIC FISHES FROM THE NORFOLK CANYON REGION INTRODUCTION Because of the numerous logistical and biological problems inherent in making in situ observations of feeding habits and laboratory maintenance of specimens for trophic studies, little information exists on the trophic ecology of living midwater fishes (Robison, 1973; Childress and Meek, 1973; McCosker and Anderson, 1976; Belman and Anderson, 1979). However, the food habits of various species of midwater fishes, particularly those that migrate vertically to feed in shallow waters at night (Marshall, 1960), have been studied e.g. myctophids and gonostomatids (Paxton, 1967, Holton, 1969; Collard, 1970; Gjosaeter, 1973; Baird et al. 1975; Kinzer, 1977; Clarke, 1978; Pearcy et. al . , 1979; Worner, 1979 and others). As a result, much important preliminary trophic information has been obtained for these upper mesopelagic species. Hopkins and Baird (1973) described the diet of Sternoptyx diaphana, a non-migratory, lower mesopelagic hatchet fish. Apart from this work, little quantitative information exists on feeding in lower meso- and bathypelagic fishes (sensu Marshall, 1971) other than material of a general nature (Beebe, 1935a, 1935b; Beebe and Crane 1936, 1947; Bertelsen, 1951, Marshall, 1954, Bohlke, 1964; Grey, 1964, and some others). Knowledge of feeding strategies and trophic structure of and among midwater fishes is sparse. Merrett and Roe (1974) analyzed food selectivity in addition to describing feeding patterns of several 100 species of mesopelagic fishes. DeWitt and Cailliet (1972) compared feeding habits and strategies for two species of the genus Cyclothone. Cailliet (1974) examined feeding strategies of Leuroglossus stilbius (Bathylagidae) and Stenobrachius leucopsarus (Myctophidae) in relation to plankton ecology. Legand and Rivaton (1969) examined the food habits of 11 species of midwater fishes (8 myctophid, 1 bregmacerotid, 1 gonostomatid, and 1 chauliodontid), and erected a two-level trophic structure which they related to active or passive feeding strategies, based on prey-to-predator body weight ratios, and percentage of empty stomachs. The purpose of this section of the study is to present information on the feeding ecology of 18 dominant midwater fish species taken from a small geographic locale, i.e. , the region of Norfolk Submarine Canyon in the Western North Atlantic, and an adjacent continental slope area to the south. Prey items were identified and relative importance of the various prey to diet were analyzed. Apparent dietary shifts, based on seasonal, diet, depth, or size related changes were examined. Possible feeding strategies of the various fish species, based on gross examination of body and jaw morphology, prey types, and previous accounts were suggested, as well as possible prey selectivity by predator species. The relationships of prey items to prey abundances were examined. Hypothetical trophic patterns among the dominant species were constructed. METHODS AND MATERIALS Fish species and numbers of specimens examined for stomach content analysis are listed in Table la. The mean sizes and size ranges for the species examined are listed in Table lb. Some of the material utilized came from the Norfolk Canyon study area, but from cruises other than CI-73-10, GI-74-04, GI-75-08, or GI-76-01. These ancillary specimens are also listed in Table 1, with station capture and cruise information presented in Appendix C. The stomach contents of specimens of Nessorhamphus ingolfianus and Derichthys serpentinus from cruise CI-73-10 were previously examined and reported by George Sedberry (1975). This data was incorporated into the information accumulated on these species during the present study. Additional specimens of Nemichthys scolopaceus and Scopelogadus beanii captured from regions outside of the canyon area, by methods and apparatus other than those used in the Norfolk Canyon study, were also examined. Two male specimens of N. scolopaceus, captured over the Blake Plateau region of the Western Atlantic during cruise 213-3 of the FFS Anton Dohrn were examined for stomach contents. As males are extremely rare, it was thought that these might provide additional information. Eighteen specimens of j>. beanii were obtained from Woods Hole Oceanographic Institution (WHOI) collections and sixteen from the Harvard University Museum of Comparative Zoology (MCZ) and the stomach 101 LIST OF SPECIES EXAMINED WITH PERCENTAGE EMPTY STOMACHS AND STOMACH CONTENT INFORMATION 8^ < •sJO O O EH X O HO EH <5 W EH pH EH O X O H > p § r~T~| . —l I— CO CO o CO P-l Q E w EH c/3 S O O Q M EH w s EH o M Ed CO o CO o w n Pu w e h v©- QGCON^HOOinONONtyvNOOOvl-invOvO oo X vo - cooo -c o-d mvooo m m m — o n t H r ^ r j s c i - r ^ o o o o f N i n i i—i o v f v X M3 ^htoNnOLvc MHooj H nM LnvocM nvO i^chvtoorN •f“H CO -M 4-1 CO cd DCD CD u u d 5 CD > CD m • X csj m x: *rl co co ^m mI" i—i i—i <—i <(■ <— a 4-1 4-J CO a ✓— P CO V-O co CO I—H VO cr» m o in CM CM X T— ’*o CM CO CM •rH M3 •rH CO - cd d CD cd bC CO CD O d cd O O P • • • • I s \ ;«n r-H CO m vO o vO co 00 CM m CT\ oo co CM CM r—I <± r-H V oo 1 E 4-1 _ — CO O 1 1 1 1 h / 1 X M3 X M3 M3 M3 X f i— 3 M •rH 3d M3 •rH •H - 4-1 4-1 i - CD 4-1 4-1 4-1 — CO cd CO CO u d o B S3 o CO HC - CD 4-1 4-1 CD CD JH CO o cd CD o d CD B 5 CD CO 0 CD o CD cd CD > CD CD o £ cd d O P d B CD U •H i m i X X M3 jd M3 •rH PM 4-1 4-1 4-1 - CD 4-1 O CD 4-1 4-1 cd CO CO cd £ CD CO J CD o CD p CD CD }H CO o cd CO CD X p d o O CD CD CD CD d B P a d o 0 o o B d > o d 0 h • - X CM X i •H •rH TJ T— •rH r Ln •rH -Jc -Jc 4-1 — - 4-1 4-1 CO dd cd JM CO JH 5 d P CD d 5-i cd 0 ID d o CO co CO o CO CD U cd bO CO cd H » B a> CD B CD CD CD CD O d P P £ i m v% i 1 t—l M3 l T— r—l *r4 s r Ed 4-1 CO a CO o O O a) o a O • X! -9 MM M3 X! •rH MM •rH X ' 4-1 4-1 4-1 - O 4-1 4-1 4-1 dcd cd cd CD 5 DCDO CD CO CD CD d CO a 5 d d O TABLE lb RANGES AND MEAN LENGTHS OF MIDWATER FISHES EXAMINED FOR STOMACH CONTENTS SPECIES SIZE RANGE MEAN LENGTH N. scolopaceus 335~925mm T.L. 590mm D. serpentinus 170-290mm T.L. 2 2 1 mm N. ingolfianus 174-353mm T.L. 243mm S. beanii 249-724mm T.L. 478mm S. brevidentatus 207-635mm T.L. 370mm E. pelecanoides 187-550mm T.L. 440mm G. elongatum 105-227imn S.L. 158mm G. bathyphilum 115-162mm S.L. 139mm A. aculeatus 32- 67mm S.L. 47mm S . diaphana 25- 41mm S.L. 30mm S. b. ferox 70-315mm S.L. 2 2 1 mm C. sloani 108-264mm S.L. 186mm L. macdonaldi 72-142mm S.L. 1 0 1 mm C. maderensis 50- 72mm S.L. 61mm B. glaciale 37- 65mm S.L. 54mm M. niger 66-167mm S.L. 1 1 1 mm P. guernei 93-114mm S.L. 103mm S. beanii (O.T.) 42-112mm S.L. 97mm S. beanii (M.W.T.) 37-112mm S.L. 1 0 0 mm 103 contents were analyzed. These specimens, all captured by pelagic trawls, provided data comparison with the otter trawl material. Capture information for these specimens is also contained in Appendix C. All specimens were identified, measured (mm standard or total length) and weighed (nearest 0.1 g). Then an incision was made from the pharyngeal region to the anus, and the abdominal cavity opened. Prior to the removal of the viscera, the abdominal cavity and internal organs were examined under a dissecting stereo-microscope for parasitic organisms (See Part III, parasitology for detailed methods). After this, a cut was made through the anterior-most portion of the esophagus, with another circumscribing the anus, and the viscera were removed. The organs were placed in a petri dish containing 40% isopropyl alcohol (the preservative in which the specimens were kept). The stomach and intestine were then separated by cutting through the intestine at the pyloric junction, and removed to separate dishes. The stomach was opened and the contents washed into the dish. The material was gently sorted through while being examined under the microscope for internal parasites. Upon removal of parasites, the food material was placed in vials of 40% isopropyl, labelled and stored. The same procedure was followed with the intestine. Although food and fecal contents were not actually identified or analyzed, 104 often recognizable parts (e.g. amphipod exoskeletons) were intact enough to aid in prey identification. Although many studies estimate degree of stomach fullness (Paxton, 1967 and following citations, see page 99), this is a rather subjective process, and was not done in this study, other than to note extremely full (distended) or completely empty stomachs. Information on degree of digestion was recorded using three qualitative categories: freshly ingested to slightly digested, where the prey material was virtually intact; partly to half digested, where obvious digestion had occurred but parts of the prey were still recognizable; and well digested, where usually nothing identifiable remained, although some recognizable hard parts of fishes, cephalopods, or crustacea may have been present. All prey items were identified to the lowest possible taxon and counted. Frequently, difficulties in counting resulted from digestion of the organisms. Because of this, well digested remains usually were considered a single organism. There were some exceptions. Fish abundances were determined from hard parts (otoliths, skulls), cephalopod counts were made from the number of beaks present, numbers of decapods and euphausiids were evaluated by counting eyes and dividing by two, and amphipod number came from counting heads that were present. Several analyses were carried out to determine how various prey items contributed to the diet. In order to avoid bias that might 105 arise because of large volume displacement of one or a few organisms, e.g. fish or large crustacea, or high numbers of prey, e.g. euphausiids, (Pinkas et al., 1971; Windell and Bowen, 1978), an overall index of relative importance developed by Pinkas et al. (1971) and utilized in other feeding studies (McEachran et al., 1976; Sedberry and Musick, 1978) was employed. The index of relative importance (IRI) incorporates several factors in its calculation: percent frequency of occurrence, the number of stomachs in which a specific prey item was found divided by all non-empty stomachs for a given fish species; percent numerical abundance, the total number of the specific prey item divided by the total of all prey items found in a given predator (also known as the percentage composition by number, Windell and Bowen, 1978); and percent volume displacement, the volume of the specific prey item divided by volume of all prey items found for the given fish species. Volume displacement was calculated by the methods of McEachran et al. (1976). Food items were blotted to remove excess moisture, then placed in a vial calibrated to a certain volume (5, 10 or 20 ml). A self levelling burette was used to titrate liquid into the vial up to the calibration mark. The amount of liquid left in the burette was subtracted from the volume of the vial used. After calculation of these factors, the IRI was computed as follows: IRI = F(N + V) 106 where F is % frequency of occurrence, N is % numerical abundance, and V is % volume displacement. Shifts in diet and feeding periodicity were analyzed for those species whose sample sizes were deemed sufficient for analysis. Feeding was examined in various size groups by percent empty stomachs and prey type. Size divisions were arbitrary but tended to segregate animals according to developmental stages. Large eels, such as Nemichthys or Serrivomer were divided in 100 mm size increments, as their size ranges generally spanned >400 to 500 mm. Intermediate size species (e.g. Stomias, Derichthys, Nessorhamphus) were divided into 50 mm classes, while smaller fishes (e.g. Scopelogadus, Argyropelecus) whose overall size ranges did not exceed 50 mm were grouped in 10 mm increments. Depth of capture, feeding intensity and prey type were also examined. Depth strata were divided into upper (< 200 m, 201-500 m) and lower (501-750, 751-1000 m) mesopelagic, and upper and lower bathypelagic (1001-1500 m, 1501-2000 m, > 2000 m) zones. The division of the lower mesopelagic into two sections was necessary because several bathypelagic species were found in the 751-1000 m depth range. Seasonal changes in feeding habits were determined by comparing the number of stomachs containing a certain prey group with the season of sampling. The additional specimens from cruises listed in Appendix C were placed in the season that the time of sampling occurred. 107 Diel feeding was analyzed. The 24 hr. diel period was divided into four 6 hr. sections: 0001-0600, 0601-1200, 1201-1800, 1801-0000. The times 0600 and 1800 were approximate averages for sunrise and sunset over the four seasons. Since no samples in any season fell within that seasons time of sunrise or sunset, the data for all cruises was pooled together to determine diel trends. RESULTS AND DISCUSSION The stomach analysis of each dominant midwater fish species is presented in this section, along with a discussion of feeding habits and possible feeding strategies. A general discussion examining trends in feeding (e.g. seasonal, diel), prey abundance and selectivity, and trophic structure follows. Order - Anguilliformes Suborder - Anguilloidei Family - Nemichthyidae Nemichthys scolopaceus Stomachs from 2 54 specimens were examined. Only 114 (44.9%) contained food. Nemichthys appeared to be stenophagic in the Norfolk Canyon area. Only two crustacean groups, sergestid shrimps and euphausiids were found, of which the former predominated (IRI of 16981 and 121 respectively). The recognizable sergestid remains were of Sergestes arcticus. Identifiable euphausiids were Nematoscelis sp. Table 2 lists the prey items and their relative importance to the d i e t . Selectivity in Nemichthys has been attributed to its specialized jaw structure. Mead and Earle (1970), from observations made from D.S.R.V. Alvin, noted that Nemichthys hangs vertically (head upward) in the water column. From these observations, and examination of stomach contents, they speculated that Nemichthys captures its prey, sergestid shrimps, by entangling the long downward-trailing antennae 108 PERCENT FREQUENCY OCCURRENCE (F), PERCENT NUMERICAL ABUNDANCE (N), PERCENT VOLUMETRIC DISPLACEMENT (V), AND INDEX OF RELATIVE IMPORTANCE (IRI) OF FOOD ITEMS IN THE STOMACHS OF -3 r3 —I r— •H 4-1 cd o co CO 3 CO 3 G a o 3 o o P- M C to i P M O 25 E h O 4-1 3 3 3 Mi Q a co CM 0} O O a>33 3 2 —1 i— i CO rH CM CM n- —i T— CO CM m MvO CM O C — H • 4-1 4-1 3tH3 W3 CO CU co CO cd Mo CM st vO vO vO CM CM O r X ' H r H • O C 4-1 - 3 4-J —s 3 3 u 3 CO 3 O 3 3 3 CO to U O 3 ! M C cu h • • • » • H r H r 1— 3 3 rH 1 co •H co co LO CO — 4-1 to CO 3 3 U 3 3 3 • • « • 1 I 3 r ON vO st OvO vO CO o CO ON 3 3 3 3 ON vO m 3 3 CM H • H • •rH 3 3 3 H • O U u 3 3 3 4J 4 M C 3 G 3 G 3 CM CO — • • » • ■ * • / O r 3 3 X V—r H r 3 3 3 - o u 3 0 b 3 Mi 4-1 M C s t VO VO CM oo 00 rH rH ON i— 1 o LO CM ON H — 3 O • 1 X H r l rH o •rH I''- H r CM •H NON ON 25 W 4-1 —1 3 3 3 O 3 3 3 3 cd CO cd cd 3 o 3 3 3 g CM CM • • • co CO r^- M 3 , r-'. 3 3 •rH •rH W 3 3 3 3 M C 3 ♦ • St St St 1—1 »M 00 oo ON 1 - EH 4-1 3 o • • 1 CM O 33 n- 33 •rH •iH !=> 4-1 3 3 3 3 I i l l I 1 l I i • St St ON 33 P 4-1 u a) u 3 CJ 3 m • 109 110 of the shrimps (which also hang vertically head upward in the water) in its jaws. The present study supports this view. Many of the slightly digested shrimps whose heads were still relatively intact had only a small portion of the antennae remaining. The major part of the antennae appeared to have been snapped off, which would occur as the eel swallowed if the antennae were wrapped around the jaws. Most of the shrimp had been ingested tail-first; the only free end able to be swallowed if the head was tied up around the fish's jaws. The only euphausiid identifiable was Nematoscelis sp. which possesses a long trailing pair of second thoracic legs (Mauchline and Fisher, 1969) which also could become entangled in Nemichthys "beaks” . The high number of empty stomachs might be explained by this passive feeding strategy. The probability that a prey item would contact and be entangled by Nemichthys in the three-dimensional midwaters might be lower than that of contact with an actively pursuing predator or an "ambush strategist," a predator which waits for prey to come within a certain range, and then lunges (a type of behavior probably precluded in Nemichthys due to its extremely elongate caudal filament which would be a hindrance in developing thrust). Figure 1 shows feeding intensities over a 24 hour period in the form of percent stomachs with food vs. time. There is little variation in the percentages. This suggests a fairly constant frequency of occurrence and availability of prey. Ill Figure 1. Feeding intensities in Nemichthys scolopaceus over a 24 hr. period % STOMACHS WITH FOOD IOO-. 40- 70- 90- 80- 50- 60- 30- 1801 =5 0 4 N= 63 N = 44 N = 60 N=85 0000 n = 37 - 1 0 0 0 n =27 N. scolopoceus 0600 PERIOD ( hrs.) PERIOD - 0601 n =24 1200 - 1201 n = 29 1800 - 112 113 Sergestes articus is a diel migrator (Omori, 1974). The findings of this study (see Part 1) indicated that Nemichthys was also capable of diel migration, thus enabling it to remain with its food source throughout the entire diel cycle. Depth and size of fish showed no significant changes in the percent of Nemichthys with food, or prey type. Two of the 254 specimens examined were males taken over the Blake Plateau (2 9°11*N, 77°07’W) during cruise 213-3 of the FFS Anton Dohrn. These were examined to determine whether the regression of the jaws in adult males (Nielsen and Smith, 1978) caused cessation of feeding. The stomachs were empty, and seemed to be atrophying, because they were very small and difficult to find. The entire abdominal area was filled with ripe testes, as noted by Mead and Rubinoff (1966). Apparently maturation processes for male Nemichthys result in the termination of feeding activities. Family - Serrivomeridae Serrivomer beanii Food items were found in the stomachs of 14 (58.3%) of 24 specimens examined. Table 3 shows the prey types and their importance in the diet. Serrivomer beanii apparently feeds on fishes, euphausiids and cephalopods. Fishes were the most important (IRI=2288) by virtue of volume displacement. However, euphausiids (IRI=!885) were more important numerically and occurred in more stomachs. Data from Beebe and Crane (1936) showed that from 23 114 M o n 0 0 0 0 m r"- I*—1 0 0 vO 0 0 P£j COVO o o i n o o r— 1 CMCM o o M CM vO ON co o o CM CO r--- CM i—i CM O U XJ M r*N Pei 3 CO T“l 0 0 oo vO i—4 CM vO CM 0 0 ••• ••• • ♦ ••• H co CM vO O o co CMCO <*• i-H m cu T“"H 3 CO vO vO T“—1 OO CM O i—i vO vO r^- CM 00 •••• 0 • • •••• 3 m in !—l r*- CM o **-4 m m vO CM •r4 T— 1 CM CM m i—l CM co CO 3 CU 6 r— 1 1—4 CO vO CO co CO r-4 ,— i vO CO o • • • •• o • •a • •• o r^. < ± OO < ± CO r'- r-4 OO OO 3 l— 1 CM i— I co flH CM CM m 3 3 4-J CO a 3 •rH > cd •r4 •rl B CO JO ^*1 cu 3 t o 3 u i— i cd 3 w u a 3 3 3 C u cd •r4 > CO 3 ft, a 3 CO CO CU 3 X) cu CO 3 •rH Cd *rl Cd 4-J 0) CO o 3 T O ,— 1 T J CO *p4 3 CO u o CU • H 3 H-4 cd •i-4 3 a . o • H l-l "r"l p 3 CU O CO CO CJ 4-J O 4-> 3 -u i—1 o 3 3 4-J O o cd '— i 4-J cd <— i • 3 CO • 1“—I 4-J • u 3 3 cd 3 cd x ) XI O XI 3 -M CO •H a . - u e C U 4-J * H •H 3 •r4 4-J 3 3 3 cu o cu 3 0 3 3 O 3 O 3 3 O O H 5 5 W H 3 3 O 3 H i— i O c u 3 ■vC o CU 3 o o O CM 3 -U O CO XI 3 3 CU 3 X 3 o •r4 f"“*l O 3 CO CU O •rH o <44 4-J X) cd o 3 3 cd a> CM 3 3 XI o 4-J 3 o cd CO •rl O <\ C u cd » H 3 £ 3 cd o a> co O I 3 a r— 4 o 3 3 4 J CO CO ctf cd cd •H CO 3 3 3 4-j ,3 4-J 3 3 X) r— 1 P m CO a , 3 O o 3 T“H a> 5 3 i— l CO U 6 o o M W 3 *r4 3 X O e> cu CU 115 stomachs found to contain food, 19 had euphausiids or shrimp-like crustacea, and 3 had fish remains. Few of their specimens were over 300 mm T.L. while all but one specimen examined in this study were >350 mm T.L.. Apparently, although euphausiids are still important in the diet, fishes become more significant dietary components in larger S . beanii. Cephalopods were rare (2 stomachs), but were intermediate in volume displacement. This was due to the presence of a fairly intact specimen of a transitional larval-subadult of Chiroteuthis veranyi (?) (identified by M. Vecchione). Although nemichthyid-like in general appearance, Serrivomer has a much more muscular body and an abbreviated caudal filament. This body configuration, coupled with prey type information, suggests that it is a fast moving predator. Beebe and Crane (1936) confirmed this from bathysphere observations. As with Nemichthys, there was no apparent depth, size, or time relation to prey type or feeding intensity. Lack of feeding periodicity, or depth related change in feeding may be associated with the restricted captures of Serrivomer beanii in the lower mesopelagic, and primarily in the bathypelagic zones. Tighe (1975) has indicated that contrary to Badcock's (1970) analysis, Serr. beanii migrates little or not at all. Thus, being confined to this energy-poor zone, predatory activities would probably not be periodic in nature, but rather unceasing, as shown in Fig. 2. 116 Figure 2. Feeding intensities in Serrivomer beanii over a 24 hr. period 1801 - 0001- 0601 - 1201 - 0000 0600 1200 1800 PERIOD (hrs.) Serr. beanii 118 The size of the gape in Serrivomer beanii is large enough to swallow a wide variety of prey. Thus, there could conceivably be much overlap in prey types among Serrivomer of various sizes. Serrivomer brevidentatus Although only 8 specimens of this species were collected, with 5 having empty stomachs (37.5% with food), the stomach contents of the 3 with food were compared with the findings for Serrivomer beanii. Only euphausiids were found in the stomachs of S^. brevidentatus. The only identifiable material was identified as Stylocheiron sp., the vertical distribution of which generally is deeper than that of the Nematoscelis sp. found in S^. beanii (Mauchline and Fisher, 1969). Most of the S. brevidentatus were captured from deeper stations than beanii. Beebe and Crane (1936), in their review of the family Serrivomeridae, noted that they are all fast swimmers with similar food habits. The stomachs of _S ° brevidentatus that they examined contained crustacean material. Competition for food and space may be lowered by modal differences in the depth ranges occupied by the two species. Family Derichthyidae Derichthys serpentinus Seventeen specimens were examined for this study. In addition, the data from 17 specimens examined from cruise CI-73-10 and reported 119 by Sedberry (1975) were added to the present findings, and the combined data were analyzed. Of the 34 specimens, 23 (67.6%) had food items present. The diet consisted largely of Sergestes arcticus (38.4% frequency, 71.6% volume, Table 4a) as Sedberry (1975) noted previously. Beebe (1935a) also reported the presence of Sergestes and shrimp remains in five Derichthys he examined. The only other prey found were euphausiids (Meganyctiphanes norvegica and Stylocheiron sp.) which came from the stomachs of Derichthys <200 mm in length. Depth of capture did not appear to influence prey type, but time of day seemed to have an effect. Sedberry (1975) noted indications of possible daytime feeding in Derichthys. He found freshly ingested shrimps in the stomachs of day-captured specimens, but only well digested material in the night captured animals. Figure 3a shows that higher percentages of stomachs with food were found at night and early morning (0 0 0 1 - 1 2 0 0 ), with the lowest values occurring between 1801-0000. Despite this, only partly to well digested prey were found at night. Freshly ingested shrimp were only found in specimens taken between 1100-1900 hrs. Based on the digestive states of prey found, feeding probably occurred during the day, with slow digestion taking place during the night. The percentage of empty stomachs during the day may be correlated with increased escape by prey due to higher illumination levels. The lowest percentage of stomachs with food (1801-0000 hrs.) w i— I vO CO O rH pcj CM CM CTv m rH LO CO M vO CM VO oo co m CM vO oo <± vQ LO rH in > • • • • ® • • O i—i CM rH CM <± CO T-100 rH CM CM CO CO LO CO 00 Figure 3 Feeding intensities in (a) Derichthys serpentinus, and (b) Nessorhamphus ingolfianus over a 24 hr. period N = I0 N = 6 N = 11 (a) D. serpentinus 100 90- o 80- O O 70- n= 7 60- £ CO X 50- o < n = 4 2 40 O f— CO 30- 20 - 10- 0 1801- 0001- 0601- 1201 - 0000 0600 1200 1800 (b) N. ingolfianus 100- N = 14 o O O u. X £ in x o < 2 o 4 0 - f— C/5 1801 - 0001- 0601- 1201- 0000 0600 1200 1800 PERIOD (hrs.) 123 was probably due to upward migration of Sergestes out of the vertical range of Derichthys. Derichthys appears to be a rapidly swimming predator which visually cues on its prey. The heavy muscular body and large eyes are evidences for this type of feeding pattern. In addition, they appear to be rarely taken by conventional midwater nets (Beebe 1935a, Castle, 1970, personal observation) indicating either an ability to escape capture, or a more frequent near-bottom presence, away from the nets' fishing range. Intact specimens of Sergestes were all found to have been swallowed tail-first, which may indicate that Derichthys minimizes prey loss by cutting off the escape route for the prey, as shrimps normally dart backwards in response to disturbance (Barnes, 1974). Because Sergestes is known to hang vertically in the water column in a head-upward position (Mead and Earle, 1970), Derichthys most probably attacks from below. Submersible observations (Musick, pers. comm.) support this theory. During a dive in D.S.R.V. Alvin in Norfolk Canyon (37°03.1'N 74°41.4'W, 656 m bottom depth, 1524 hrs), Musick observed "a Derichthys serpentinus ca. 12" long in midwater at depth of 585 m...feeding, snapping head back with violent motion every couple of minutes.. .eel was more or less vertical in water column." 124 Nessorhamphus ingolfianus Euphausiids and sergestids were the only identifiable food items, found in 23 (71.9%) of 32 stomachs examined (Table 4b). Euphausiids, however, were the principal food, comprising the bulk of the percentage in all categories (frequency 87.0%, numerical abundance, 86.3%, and volume, 69.2%, Table 4b). Four species of euphausiids were identifiable. These were, in decreasing order of importance, Nematoscelis sp. (IRI=1413), Euphausia krohnii (IRI=820), Stylocheiron sp. (IRI=47), and Meganyctiphanes norvegica (IRI=19). Those that were relatively intact all measured 20 mm in length, indicating that N. ingolfianus may be size selective towards prey. Nessorhamphus appears quite similar in overall body structure to Derichthys, and both probably exhibit an active foraging strategy. Also, both have well developed eyes and appear to be only rarely captured by midwater trawls (Beebe, 1935b, personal observation). Jaw structure appears quite different in the two eels, however. Nessorhamphus has a much smaller gape and throat than Derichthys and would conceivably have a difficult time swallowing the larger sergestids. The single Sergestes arcticus found was from a specimen 270 mm long, which is towards the upper end of the size range for Nessorhamphus. Beebe (1935b) found euphausiids (Thysanopoda sp.) and a small Sergestes in 6 stomachs he examined. Differences in feeding habits may allow these two species to broadly overlap in vertical and horizontal distribution, with minimnal competition for prey. PERCENT FREQUENCY OCCURRENCE (F), PERCENT NUMERICAL ABUNDANCE (N), PERCENT VOLUMETRIC DISPLACEMENT (V), AND INDEX OF RELATIVE IMPORTANCE (IRI) OF FOOD ITEMS IN THE STOMACHS OF X X i—I <4-4 •H ■rl CO co 1 CO M cU CO (D o u P P 0 § Pd P O S3 H m o •U CU *H u M C CO cU cU O CO CU P P w P T“M o c O C O C O C o c o o M C o —1 T— o o X -id X H • X •i— 1 •i-4 W Mi cU CU CO cU P o CU cU P P C O * l t M P . N O i-4 o r-4 o c 4 r— N O m •rH •rH 4-1 a cU ) W cU O Mi P p a ucu cu p ± • —1 i— Sedberry (1975) from whom some of the Nessorhamphus data (6 specimens) was received, noted that N. ingolfianus may come quite close to the bottom because it often appears in otter trawl catches. This is also true of Derichthys, and both frequently occurred in the same hauls together. Their bodies appear heavier than those of typical midwater eels (e.g. Serrivomer). Their proximity to the bottom may also be related to the high percentage of specimens taken with food during the day (Fig. 3b). Observers in submersibles have noted that large numbers of sergestid- and euphausiid- like shrimps, as well as other crustaceans may be concentrated near the bottom along the continental slope (Musick, pers. comm.). This concentration may be especially pronounced during the day, when the zooplankton are at the deepest point of their vertical migration (Marshall and Merrett, 1977). This abundant prey probably ascends at night out of the depth ranges of the largely non-migratory Derichthys and Nessorhamphus. Suborder Saccopharyngoidei Family Eurypharyngidae Eurypharynx pelecanoides As would be expected with a true bathypelagic predator, the prey items ingested were of a wide variety, and the percentage of empty stomachs was high, with only 6 (37.5%) of 16 stomachs containing any food. Prey items included unidentified copepods, an unidentifiable ostracod, teleost remains, a caridean shrimp (Acanthephyra sp.), and a 127 holothurian (Family Molpadidae?). No prey preference could be determined. Ebeling and Cailliet (1974) noted that in a bathypelagic, food-scarce environment, the feeding strategy should be such that any item encountered is eaten. This appears to be so with Eurypharynx. Examination of size and depth of capture and prey composition shows that both may influence the diet of Eurypharynx (Table 5). The smaller, more shallowly captured Eurypharynx ingested the smaller food items (copepod, ostracod), while a wider variety of material occurred in stomachs of larger fish captured deeper. The data, however, were very scanty. Diel feeding analysis revealed empty stomachs between midnight and noon. As the overall sample size was small, this may reflect sampling error. The presence of the molpadid (?) holothurian in one stomach suggests that Eurypharynx may feed on the bottom. Marshall (1954) noted the presence of a sea urchin in the stomach of a Eurypharynx stating that it "... must have been picked off of the sea floor." The mechanism by which Eurypharynx could pick animals off the bottom is a mystery, as the jaws are enormous and extremely fragile (net feeding is precluded for this latter reason because all but one specimen came to the surface with totally mangled jaws). Creation of suction by opening the mouth is a possibility, although Robins (pers. comm.) feels that the jaw musculature is far too weak to create a suction. The presence of an Acanthephyra sp. in a stomach may also point to near bottom feeding by Eurypharynx, as Omori (1974) showed that 128 TABLE 5 SIZE AND DEPTH OF CAPTURE OF Eurypharynx pelecanoides IN RELATION TO PREY TYPE INGESTED. SIZE (mm) NO. EXAMINED NO. EMPTY______PREY TYPES < 2 0 0 2 1 Copepod, Ostracod 201-300 1 1 301-400 2 1 Crustacean fragments 401-500 10 7 Acanthephyra sp., fish remains, Holothurian (Family Caudinidae?), unidentified material DEPTH (m) NO. EXAMINED NO. EMPTY PREY TYPES < 1500 1 1 1501-2000 9 7 Crustacean fragments, copepod, ostracod 2001-3000 5 2 Acanthephyra sp., fish remains, Holothurian (Family Caudinidae?), unidentified material 129 Acanthephyra approach the bottom as they grow older, and the genus is abundant in otter trawl hauls in the present study area (Wenner, 1979). Order - Salmoniformes Suborder - Stomiatoidei Family - Gonostomatidae Gonostoma elongatum Of all fishes examined (?. elongatum had the second highest percentage of stomachs with food. Forty-eight specimens were examined of which 42 (8 7.5%) contained food. Although Gonostoma fed on a wide variety of prey items (Table 6 a), only the euphausiids (76.2% frequency of occurrence) and hyperiid amphipods (38.1% frequency of occurrence) were significant components of the diet (Fig. 4a). The euphausiids were the most important in all categories (7 6.2% frequency, 80.3% numerical abundance, and 63.2% volume), with Nematoscelis sp. (one of which was identifiable as N_« megalops) , Meganyctiphanes norvegica, Euphausia krohnii, and Thysanopoda sp. being found in decreasing order of abundance and importance. Although numerical abundance of Nematoscelis sp. was far greater than Meganyctiphanes (47.1% to 8.5%) M. norvegica displaced more than half of the volume of Nematoscelis (14.4% vs. 26.7%) as the specimens were much larger than Nematoscelis (23-34 mm vs. ^ 20 mm). All of the amphipods found were members of the family Hyperiidae. The species included Parathemisto gaudichaudii, Vibilia armata, and PERCENT FREQUENCY OCCURRENCE (F), PERCENT NUMERICAL ABUNDANCE (N), PERCENT VOLUMETRIC DISPLACEMENT (V), AND INDEX OF RELATIVE IMPORTANCE (IRI) OF FOOD ITEMS IN THE STOMACHS OF T~1 o 4-> 4-> cd O O o CO cd H O o B cu O P P B Pi 04 n F P F E w s 04 M h h CJ 4-J cd cd cu ?H CO CJ p 0oo a o n m j o o | u~> o oo d * o l r—1 H • H • H • PM d (U d u d (U . a n o'Jin tn X 4-J d O cu d o t o , d •H H r d J T H • 4-J 4-J CO CO O d cu d CU • H r CM EH 4-J O 00 TU T3 •iH •rH •iH d> J 4 CU d d d> xl iw •d oo m r- •rH PM ■u u J-to cu d cu o o 131 132 Figure 4. (a) Gonostoma elongatum. Percent occurrence by high taxonomic groups by number, volume, and frequency of occurrence. (b) Feeding intensities in Gonostoma elongatum over a 24 hr. period 133 A - PISCES (a) B - EUPHAUSIACEA 100 C - AMPHIPODA ( Hyperildae) 80- D- DECAPODA E- COPEPODA 60 - m 2 3 40 1.9 20H 1.0 0 . 7 I i 0 A a . C D 0 . 9 20 - E 11.9 38.1 7.1 4.8 2 40- 3 _l O 6 0 - > B 80- 76.2 100 % FREQUENCY (F) G. elongatum N = 14 N= 13 N = 6 N = 15 (b) n= 13 a o n = 10 O co X o 1801- 0001- 0601- 1201 - 0 0 0 0 0 6 0 0 1200 1800 PERIOD (hrs.) G. elongatum 134 one very large (35 mm) specimen of Phronima sedentaria. There was no significant difference in the importance of the species consumed (Table 6 a). Although amphipods frequently were found in stomachs (38.1% overall), numerical abundances (0.3-2.7%), volumes (0.5-3.4%) and frequencies of occurrence of individual species (2.4-7.1%) were quite low. Table 6 b shows the presence of prey items in stomachs during various seasons. Euphausiids dominate in the spring and winter, amphipods in summer, and both occur in the fall. This may indicate a dietary shift in Gonostoma elongatum based on changes in prey abundances with season. Several Gonostoma captured in the spring had obviously distended stomachs which contained from 28-41 Nematoscelis sp . Decapods (Sergestes sp.) and fishes were occasional components of the diet. The fishes included a Cyclothone sp., and unidentifiable myctophid, and a paralepidid. The presence of the paralepidid, as well as the amphipods (most of which were in the larger size ranges for their respective species), both noted for their ability to swim rapidly (Rofen, 1966; Bowman and Gruner, 1973), indicates G. elongatum is a fast moving predator. Size of (S. elongatum and depth of capture showed no relation to prey type, but there appeared to be a diel change in feeding. Figure 135 TABLE 6b CHANGE IN DIET COMPOSITION OF EUPHAUSIIDS AND HYPERIID AMPHIPODS (MAJOR PREY ITEMS) WITH SEASON FOR Gonostoma elongatum SEASON______NO. EXAMINED NO. EMPTY % AMPHIPODS % EUPHAUSIIDS Spring 11 0 18.2 90.9 Summer 6 0 83.3 33.3 Fall 22 5 47.1 82.4 Wittteir 8 1 28.6 85.7 136 4b shows that Gonostoma feeds most heavily from midnight through noon, with a decline through the afternoon and evening period. The period 1801-0000 comprises time in which vertically migrating animals are ascending towards the surface. G^. elongatum is known to be a constituent of this migratory activity (Grey, 1964; Krueger and Bond, 1972; Clarke, 1974) and individuals with empty stomachs may represent new arrivals to the surface waters, animals still migrating upwards, or those remaining at depth and not feeding. This last possibility could be the explanation, because the specimens with empty stomachs during this period all came from trawls fished around 2 0 0 0 m depth. Gonostoma bathyphilum Gonostoma bathyphilum is the only species of all those examined which apparently regularly regurgitated its food upon capture. Eighteen specimens were examined of which 6 (33.3%) still had food in their stomachs. Of the remaining 12, approximately 7 showed signs of forcible evacuation of stomach contents, i.e., food remains in mouth, and inward contraction of stomach. Of the intact stomachs, 2 had fish remains (one of which was Ceratoscopelus maderensis), 3 had unidentifiable material, and one had echinoderm fragments. The presence of the echinoderm material is interesting because G. bathyphilum is known to occur quite close to the bottom. Observers in submersibles have noted it within several meters of the bottom (Musick, Sulak, pers. comm.), and a freshly 137 ingested specimen was found in the stomach of Bathysaurus ferox (personal observation), a usually sedentary benthic piscivore, which occasionally darts off the bottom a short way ( ^ 1 m) to capture prey (Musick, pers. comm.). Because (2. bathyphilum possesses the typical weakly ossified, poorly muscled, watery body found in bathypelagic fishes (Marshall, 1971) it probably is a lethargic predator, feeding only when prey come within range. It may augment this sparse diet by feeding on animals near or on the sea floor. Family - Sternoptychidae Argyropelecus aculeatus Fifteen specimens were examined, of which slightly more than half (8 specimens, 53.3%) had food in their stomachs. Three specimens however, had their stomachs everted through their mouths due to expansion of the swim bladder, and were excluded from further analysis. Thus, 66.7% of Argyropelecus with intact stomachs had food. Those specimens preyed on a limited range of prey items; euphausiids, gelatinous zooplankton, and hyperiid amphipods were the most important. One unidentifiable cephalopod was also found. The euphausiids were the most important component in the diet, but the jelly plankton (primarily salps in this case but the term also includes siphonophores and hydromedusae) were nearly as important on a volumetric basis as the euphausiids (36.8% vs. 38.7%, respectively) and occurred in a number of stomachs (37.5%, Table 7). PERCENT FREQUENCY OCCURRENCE (F), PERCENT NUMERICAL ABUNDANCE (N), PERCENT VOLUMETRIC DISPLACEMENT (V), AND INDEX OF RELATIVE IMPORTANCE (IRI) OF FOOD ITEMS IN THE STOMACHS OF —I r— < 4-1 CO cd Ci bC Cl cd QJ N > o cu d) o CO O d d CL d p E o P S M P >-i E w h h 4 m H t T“—f S cd co o O d r-H CO M C CO CM rd vo 00 m in in X H X rd •rH rH X o o cd o cd o Cu Cu CD PJ cu cd d Cu O o cd CM ♦ • • • • • • • • • • • CO O C vo rH O — ii— i—i i i— i M00 CM XI ^ r r-- in oo m H i MH X •rH •rH •rH H • IS! 4-1 4-1 a — 4J CO dX cd cd d cu d o d d o O PL, d d d (U CU o n • • • i o 4-> d X cd Cd Cd u CL, CO cu o CL, O d d j < vO CM m o o o CM d r Si m *rH •rH m p 4-1 4-J cd id dcd cd u N CU M a cd 0 ) H r 1— 1 rH 00 —i i— rH co vo o CM 00 o r-^ 4-1 dcd cd O • • rd i vO I''. vO i i s H CM o m CO •rH •rH •rH m — S i — w 4-1 PC O co cd cu cd cu cd d co >i cu u CU a cd (U C d d > d d CL, o l 1 * * - Gelatinous Zooplankton is an inclusive term encompassing salps, hydro 138 medusae, and siphonophores. 139 Although limited, an examination of seasonal diet data indicated that hyperiid amphipods were most important in the summer. Summer was also the only time during which jelly plankton were consumed. Hyperiid amphipods and salps are known to exhibit a commensal to parasitic relationship (Harbison et al. 1977; Madin and Harbison, 1977). Data for diel feeding were also weak, but it appeared that peak feeding occurred at night (1801-0000 hrs., Fig. 5a) which is understandable, as A. aculeatus is a common component of the surface waters at night (Schultz, 1964, Baird, 1971, personal observation). The findings of this study differ from those of Merrett and Roe (1974) who examined 42 A_. aculeatus from the Eastern North Atlantic, and found them to be selective for conchoecid ostracods. Regional variation in plankton abundances is the most probable reason for the discrepancy, as copepods and ostracods were common in Merrett and Roe's study area, but do not appear to be so around Norfolk Canyon. A. aculeatus may be capable of rapid pursuit of prey. This is evidenced by its ability to migrate over quite large vertical distances (300 m or more, Baird, 1971), and by the presence of a cephalopod in one stomach. Sternoptyx diaphana Stomachs of 19 specimens were examined, of which 1 was found to be everted as a result of swimbladder expansion. Sixteen of the 18 specimens with intact stomachs (88.9%) had food in them. Thus, S. 140 Figure 5. Feeding intensities in (a) Argyropelecus aculeatus and (b) Sternoptyx diaphana over a 24 hr. period 141 N = 5 N = 3, (a) A. aculeatus □ o o co x o < 2 O ►— CO 1801- 0001- 0601- 1201- 0000 0600 1200 1800 N = 6 N = I N = 5 N = 6 (b) S. diaphana 100 9 0 - Q o 8 0 - o 7 0 - 6 0 - CO X o 50- < 2 o 40 - 30- 20 - 10- 0- 1801- 0001- 0601- 1201- 0000 0600 1200 1800 PERIOD (hrs.) 142 diaphana had the lowest percentage of empty stomachs of all species examined. The euphausiids (Nematoscelis megalops and N_. sp.) were the major constituents of the diet of £L diaphana, comprising 71.9% of the volume, and 58.2% of the number of prey items (Table 8 a). The hyperiid amphipod, Parathemisto gaudichaudii, was the second most important prey item. Seasonally, it equalled the euphausiids in importance in the summer (Table 8 b). Ostracods (Family Halocypridae, Subfamily Conchoecinae) and fishes were minor constituents of the d iet. Sternoptyx appeared to feed most heavily between noon and early morning (1200-0600 hrs. Figure 5b). The reason for the higher percentage of empty stomachs between 0601-1200 hrs. may be explained by small sample size. Hopkins and Baird (1973) indicated little feeding periodicity in Sternoptyx and it is most probable that this is the case in Norfolk Canyon. Hopkins and Baird (1973) examined 13 stomachs of j^. diaphana from regions they labelled "NW Atlantic Pocket", which includes the present study area, and 21 specimens from the "WN Atlantic Central", roughly the Northern Sargasso Sea. Dietary analyses showed an extremely wide range of prey. In the NW Atlantic Pocket, fish, hyperiid amphipods (of a number of genera other than Parathemisto), copepods, and ostracods formed the major prey groups, with euphausiids and other prey groups in relatively low abundance. The WN Atlantic Central 143 CM o CM oo oo oo CM P i ON ON ON r-'- i-l CM m oo in m CM i—i CM ON P i ON v o o ON ON CM • •• •••• i—1 OO 00 o u o T-1 CM I— 1 CM I''. CO ON •••*•••CO C M v o C M C O v O C M v O o 1— 1 00 1—1 COCMCO in m O CM CM oo o CM ON • ••• « •• • CM o v O i— 1 CO in NO OO i— 1 m CO 00 •r-l •i—! n o Cl cd CO CO cd x CU Ci oo o o •iH •r4 r—1 cd H n o cd B X d W 3 CU PQ cd cu CU Pi < M 6 co H 4 J cu O CO co CU CO cd 4-1 *p^ *rl cd O P-l n o CO r H t~—1 n o CU •H•H cu cu •H I— 1 Pi B o o °iH cu CU cu CO COCO 4-1 >-> X o o d o 4-1 4-1 4-1 cd r—t • o cd cd crt X cd no r—1 Pi B g cu 4-1 •iH cd cd <0 no O O 1+-I cd cu X o 4-1 cd cd cd •H Cd no nO •r4 £ EH a> o O CO o o CU d 4J cd cd •r^ cd CO Ci 4J X a) cu co CU Cu o O d 3 CO Pi Pi o w •iH CU o P-l Pu 144 TABLE 8 b CHANGE IN DIET COMPOSITION OF EUPHAUSIIDS AND HYPERIID AMPHIPODS (MAJOR PREY ITEMS) WITH SEASON IN Sternoptyx diaphana SEASON______NO. EXAMINED NO. EMPTY % AMPHIPODS % EUPHAUSIIDS Spring 15 0 20.0 100.0 Summer 8 2 66.7 66.7 Fall 3 0 33.3 100.0 Winter 2 0 0.0 50.0 145 showed an even wider variety of prey (expected in the less productive, more diverse, Sargasso Sea area), with diet preference shifting to amphipods, chaetognaths, and ostracods. The authors thus concluded that Sternoptyx is a random feeder of "limited pursuit capability" which appears to take fewer, but larger food items, with the overall diet being lower in diversity in regions overlain by cooler water masses, because diversity of prey is lower there. While Sternoptyx may have a short-range pursuit strategy since it is a largely non-migratory, lower mesopelagic inhabitant, it appeared to have an even more restricted diet (a result of selectivity?) in the Norfolk Canyon area than was found by Hopkins and Baird. Family - Chauliodontidae Chauliodus sloani Forty-two specimens, ranging in size from 60-242 mm were examined, of which 22 (52.4%) contained food. The only prey found in C^. sloani were myctophid fishes. The only identifiable remains were all Ceratoscopelus maderensis. There was no change of diet noted over the size range of the fish, depth of capture, or season. Over a diel cycle, however, there seemed to be two peak periods, 0601-1200 and 1801-0000, during which the fish fed (Fig. 6 a). This may be linked to the upward and downward migrations of myctophids which may be exposed in greatest numbers to Chauliodus at these times. Figure 6 . Feeding intensities in (a) Chauliodus sloani and Stomias boa ferox over a 24 hr. period 147 N = 8 N = 14 N = 9 N = 11 (a) C. sloani I00-, 90 - o O 8 0 - o u. X 70- n = 6 £ 60- n= 8 to X o 50 < 2 o 40 ►— CO 30- 20 1801 - 0 0 0 1 - 0601- 1201 - 0000 0600 1200 1800 N = 10 N = 15 N = 23 (b) S. b. ferox 1 0 0 - o O 8 0 - O co x o < o f- CO 1801 - 0 0 0 1 - 0601- 1201- 0000 06 00 1200 1800 PERIOD (hrs.) 148 Borodulina (1972), reported on feeding of stomiatoid fishes (chauliodontids, stomiatoids, idiacanthids) and noted that they exhibited some diet specialization being primarily ichthyophages. In the present study no more than the remains of a single fish was found in any Chauliodus stomach. Those remains in which head and tail could be distinguished were taken headfirst. Tchernavin (1953) theorized that a light organ at the tip of the elongate first dorsal ray, and additional light organs in and around the mouth served to attract prey, which would then be taken head first. Legand and Rivaton (1969) examined C. sloani and 10 other species of midwater fishes from the Indo-Pacific, and on the basis of feeding habits, stomach contents, and quantitative prey-to-predator body weights erected a two level trophic structure; level A included the 10 other species, characterized by low percentage of empty stomachs and small prey items as indicators of high trophic activity, and level B, occupied by C^. sloani, which took fewer, larger prey less often, exhibiting in essence a "snake-like" feeding strategy. In the present study, the percentage of empty stomachs was much lower than that found by Legand and Rivaton (1969) (47.6% vs. 69.5%). In addition, extreme selectivity of prey (apparently one species of myctophid, Ceratoscopelus maderensis) seems to occur in Chauliodus around Norfolk Canyon, as it is known to take a wider variety of prey in other areas (Morrow, 1964a; Legand and Rivaton, 1969, Borodulina, 1972). Presumably both of these factors are related to prey 149 abundance, because C^. maderensis may be very abundant in Slope Water (Backus et al., 1968, Krueger et al., 1977). With respect to prey abundance and availability, Ceratoscopelus maderensis occurs in huge schools and large numbers in Slope Water. Backus et al. (1968) reported immense schools (10-15 individuals/m^ of water) from submersible observations made in slope waters south of New England (39°N 70°W). Krueger et al. (1977) estimated a concentration of roughly 3.5 million individuals per square kilometer around 39°N, 72-73°W. The findings of this study (see Part I) show _C. maderensis to be the second most abundant myctophid and third mostabundant midwater fish species overall around Norfolk Canyon. Family - Stomiatidae Stomias boa ferox Of the 59 specimens of _b. ferox examined, 36 (61.1%) had prey in their stomachs. The stomach contents of Stomias were similar to those of Chauliodus in that the only items found were myctophid fishes. A number of these were identifiable as Ceratoscopelus maderensis which ranged in size from 5 5-70 mm standard length. Two myctophids taken from stomachs of Stomias captured in the fall were found to be Diaphus dumerilii (50 and 55 mm). As with Chauliodus, all prey were taken head first, and in all but two specimens only a single prey item was found per stomach. In the other two, vertebral remains were found in the gut along with freshly ingested prey. 150 Borodulina (1972) examined 57 specimens of S. b. ferox and found only three to contain any food, all of which were myctophids. He theorized that this low percentage was due to regurgitation of stomach contents in the net, a situation which rarely occurred in the Norfolk Canyon collections. Borodulina also suggested that Stomias, Chauliodus and other "ichthyophages" migrated with the organisms of the deep scattering layer (primarily myctophids) and then lay in wait until prey came near enough to capture. This feeding strategy is similar to the "snake-like" activity characterized by Legand and Rivaton (1969) for their "trophic level B" predator, Chauliodus sloani. Stomias appeared to feed in the same manner. Because Stomias and Chauliodus had broadly overlapping vertical ranges, and fed on nearly identical prey, the data were analyzed to see if there was any mechanism which might reduce direct competition for food between the species. An examination of depth of capture and number of empty stomachs (Table 9) reveals that most of the Chauliodus occurred in deeper water than Stomias. Chauliodus peak abundance was between 751-1000 m while most Stomias were found from 501-750 m. Comparison of percent empty stomachs for the two species, indicated that Stomias had a higher percentage of empty stomachs (50.0%) in the 751-1000 m range than in its peak abundance range (501-750 m, 23.3% empty). Chauliodus showed the exact opposite trend, 77.7% empty between 501-750 m, and only 38.5% empty from 751-1000 m. In addition, comparisons of empty stomachs at other depth levels showed that Stomias had a higher incidence of empty stomachs in deeper water, 151 TABLE 9 DEPTH OF CAPTURE VS. PERCENT EMPTY STOMACHS COMPARED BETWEEN Chauliodus sloani AND Stomias boa ferox ______DEPTH OF CAPTURE (m)______SPECIES______<200 201-500 501-750 751-1000 100tti-1500 1501-2000 >2000 C. sloani 50.0* 77.7 38.5 50.0 40.0 25.0 S. b, ferox 100.0* 33.3 23.3 50.0 75.0 71.4 0.0* Percentages based on two or fewer specimens 152 whereas Chauliodus empty stomachs are more often found in shoaler waters. In addition to this apparent spatial partitioning of feeding, there also seems to be a temporal separation between the two species. Figure 6 a shows the diel feeding pattern of C_. sloani with peaks at 0601-1200 and 1801-0000 hours. Stomias (Figure 6 b) primarily fed between 1201-1800 and 0001-0600 hrs. This temporal difference in feeding periods between species suggests that Chauliodus feeds as the prey move up and down, while Stomias feeds after the prey have established themselves at their day or night depth levels. Clarke (1974) noted that no individuals of Chauliodus under 120 mm occurred within the upper 175 mm of water off Hawaii. Perhaps Chauliodus undertakes a more limited vertical migration and so may commence feeding while Stomias, which enters shallow water layers at night (Gibbs, 1969), is still migrating. Family - Malacosteidae Both malacosteid species examined in this study, Malacosteus niger and Photostomias guernei are represented by very few specimens. Because very little is known about the food habits of either, they were included in the stomach analyses. Malacosteus niger Thirteen specimens of Malacosteus were captured, of which 10 were available for examination. Only 4 (40.0%) of these contained food. 153 The only identifiable food items were copepods, Paraeuchaeta norvegica, found in 3 stomachs, with a euphausiid, Nematoscelis megalops in the fourth. P^. norvegica is a relatively large (ca. 7 mm) copepod which is commonly found in the slope waters around Norfolk Canyon (Grant, pers. comm.). Little can be said of the feeding habits of M. niger due to small sample size, and a lack of previous accounts. Although Borodulina (1972) indicated that because of the expansible jaw apparatus, Malacosteus should be able to swallow large prey, he examined no specimens, and did not further elaborate on his statement. From the jaw and body morphology, Malacosteus probably belongs to the stalking predator or ambush group with Stomias and Chauliodus. Its success at attracting prey may be lessened by lack of photophores or luminescent barbels found in other stomiatoids (Morrow, 1964b), and by its non-migratory habits (Clarke, 1974). Photostomias guernei Even fewer specimens of this species (5) were captured in comparison to Malacosteus. However, 4 of the 5 (80.0%) contained food. In all stomachs, only the remains of sergestid shrimps were found. In the stomach of a 93 mm specimen, a 60 mm Sergestes arcticus was found, demonstrating the ability of this species to engulf very large prey in relation to its body size. In another stomach, a shrimp identifiable as Sergestes sp. was also found. 154 Photostomias may be more successful at capturing prey than Malacosteus, because Photostomias undertakes vertical migrations from daytime depths in the lower mesopelagic to quite near the surface at night (Badcock, 1970; Clarke, 1974). This would allow it to utilize a far richer prey source than Malacosteus, which does not migrate (Clarke, 1974). Like Malacosteus, P. guernei possesses median fins set well back on the body with large ventrals located on the posterior half of the body (Morrow, 1964b). This general body morphology suggests that Photostomias like Malacosteus utilizes a wait and strike feeding strategy. Order - Myctophiformes Family - Myctophidae Benthosema glaciale Although Benthosema is the most common myctophid found in slope waters (Nafpaktitis et al., 1977), and was represented by 139 specimens captured during the four cruises, only 1 2 were available for stomach analysis (a large number of specimens apparently were lost). Less than half (5 specimens, 41.7%) contained food. Only 2 contained identifiable prey, an ostracod of the suborder Halocypriformes, and a hyperiid amphipod, Parathemisto gaudichaudii. A number of published accounts on feeding in Benthosema exist. These indicate that calanoid copepods and ostracods are the principal prey items, but that euphausiids appear more frequently in larger fish (> 30 mm S.L.), and that B. glaciale feeds most intensively at night 155 (Gjosaeter, 1973; Kinzer, 1977; Worner, 1979). Of these studies, only Kinzer noted the occasional presence of amphipods even though Worner1s study material came from the same geographic location, indicating that predation on amphipods is probably related to temporal abundances. Ceratoscopelus maderensis Ceratoscopelus was the second most abundant myctophid in the Norfolk Canyon area, with a total of 84 specimens taken during the four cruises. A large percentage of these apparently were lost or discarded, as only 6 specimens were available for stomach analysis. However, 12 specimens from the Norfolk Canyon area captured during earlier cruises were also examined and included for analysis. Thirteen of 17 (7 6.5%) examined had stomachs containing food. The 18th stomach was everted due to swimbladder expansion. Euphausiids were the major dietary component (IRI=1430) as seen in Table 10. Fish and copepod remains were also found. Worner (1979) determined that the main food items of C. maderensis in the coastal upwelling zone of NW Africa were calanoid copepods, and that feeding was most intensive at night. Differences in prey types found by Worner and the present study may be attributed to the size of the fish examined; less than 25% of the Ceratoscopelus examined by Worner (1979) were > 50 mm in length while all but one specimen in this study were > 50 mm. Thus diet change with increasing size, as occurs in Benthosema may be a factor, PERCENT FREQUENCY OCCURRENCE (F), PERCENT NUMERICAL ABUNDANCE (N), PERCENT VOLUMETRIC DISPLACEMENT (V), AND INDEX OF RELATIVE IMPORTANCE (IRI) T—1 as o o as x) CM x> •H u O T o *3 3 CL) 3 o o D-i 3 ex, cd o CD (X o a, • • • • ■ « • • o o •i ,3 X •rl °rl T“H •H vO !— T •H 00 o 00 Om LO r-- w W - CO cd cd cd CO cd cd a 3 O 3-i 3 o a) 3 3 3 (X 3 • • • • • 1 i VO i OV CM H r o o o r-^ r-. r^. •P-l :z; •M -1 CO ca ca a O CD CD g c X X) o cr. H r H • m CTV CO CT\ ov CM -r4 r-~- M cd a3 ca cd although geographic differences in plankton abundances may also be a factor. Too few stomachs were analyzed to determine seasonal or diel trends. However, more shallowly captured specimens had a lower percentage of empty stomachs than did those taken deeper. Lampanyctus macdonaldi A total of 19 specimens of this species were captured, 14 of which were examined for stomach contents. Eight specimens (57.1%) had food present in the stomachs. The variety of prey items ingested was low (Table 11). The major prey taxon was the hyperiid amphipod Parathemisto gaudichaudii, which was found in 75.0% of the stomachs, and comprised 62.5% of the total number and 82.4% of the total volume of prey items found. An unidentified copepod and an ostracod were also found in separate stomachs. Bowman (1960) noted that the genus Parathemisto is often the overwhelmingly dominant component of the cool-water amphipod fauna. Grant (1979) found P. gaudichaudii to be the numerically dominant amphipod species in the surface waters at a station along the northern edge of Norfolk Canyon. At Deepwater Dumpsite 106 (approx. 38°40'-39°00'N, 72o00,-72°30’W), Bowman (pers. comm.) has found that although most of P^. gaudichaudii are found between 7 5-150 m, individuals may reach depths of 2000 m. Also specimens from the surface in boreal areas may be carried to these depths by downwelling, southward flowing currents. PERCENT FREQUENCY OCCURRENCE (F), PERCENT NUMERICAL ABUNDANCE (N), PERCENT VOLUMETRIC DISPLACEMENT (V), AND INDEX OF RELATIVE IMPORTANCE (IRI) OF FOOD ITEMS IN THE STOMACHS OF -U cd cd CO S O B - P 3 PM s >-• 1-1 E-t w S CO JJ dcd cd *-1 cd a> CO o 3 i I—I O'! OvO vO MCM CM o o O CM m m O T O T — TO H•H ■H OO CO u cd pi o a) o 3 a) o m c o CM CM PM • • • • • • • • • • • 1 i— I i—1 O CM MCM CM CM O T TO ■U o u CO JM cd cd cd CO cd 3 dcd cd o o 3 o u o o • • VO CM oO r—1 O vO X m O in 00 00 O T X X O T TO •iH ♦rl H • H • •H <3 P u 4-» cd cd CO tc cd u M Nafpaktitis et al. (1977) indicate that L^. macdonaldi occurs in the lower mesopelagic into the bathypelagic as adults. Data show a possible limited vertical migration into the lower mesopelagic. Thus, Lampanyctus probably feeds at depth, because only one specimen (with an empty stomach) was taken above 1 0 0 0 m. Lampanyctus macdonaldi has been seen quite close to, and even in contact with the bottom (Craddock, pers. comm.). L. macdonaldi unlike other typical bathypelagic species (e.g. Gonostoma bathyphilum) is a relatively large muscular fish and apparently is capable of more extensive swimming activity. However, because of the reduced pectoral fin structure, it may utilize a wait-and-lunge mode of feeding, as the pectorals are important for braking and turning (Marshall, 1975). Possession of small pectoral fins may render effective pursuit difficult, especially with fast moving prey such as Parathemisto (Bowman and Gruner, 1973). Order - Beryciformes Family - Melamphaidae Scopelogadus beanii Sixty-nine stomachs from beanii were examined, of which 52 (75.4%) were found to contain food. The main prey items of Scopelogadus were hyperiid amphipods and gelatinous zooplankton. Other prey included a bivalve mollusc, polychaete annelids, copepods, halocyprid ostracods and fish remains (Table 12 a). Three genera of amphipods were identified, Phronima, Vibilia, and Parathemisto, of 160 which the first two were of approximately equal importance and about twice as important as Parathemisto. The gelatinous zooplankton were largely unidentifiable due to their fragile nature, however, Grant (pers. comm.) determined that the majority appeared to be hydromedusae, with some salps and one possible siphonophore. Seasonal changes in diet composition were noted, both among amphipod species (Table 13a) and overall prey type (Table 13b). Vibilia propinqua, the most common of the two Vibilia species found, occurred spring through fall, but most often in the summer. Vibilia armata was found only in the spring. Parathemisto gaudichaudii occurred in the winter, spring, and summer and was most common in the latter season. Phronima atlantica also occurred spring through fall, but dominated the fall stomach contents. Overall prey seasonality (Table 13b) shows that amphipods were common spring through fall, with a sizeable decrease during the winter period, while "jelly" plankters were abundant only in summer and fall samples. Although approximately the same number of specimens were examined in spring and winter (12 vs. 13 respectively), very few amphipods or gelatinous plankton were found in stomachs during the winter, indicating a possible paucity of these prey during that time. The species of amphipods taken are known to form parasitic (Phoronima and Vibilia sp.) or commensal relationships (Parathemisto) with various gelatinous zooplankton (Madin and Harbison, 1977; Harbison et al., 1977). In fact, several Phronima atlantica and 161 i—t co CMO CO co CO CM COCM <1- ON CO rH n> T—1 o 1—1 NO rH oo rH in pj ON m CM m co NO CO NO j> CM CO <± ON H± rH o <± CO CO <1- •«•••• • • • ••• ♦ • o i—1 ON o o CM o m rH rH m o I—1 CM i—H CM rH NO ON ON OON n. t—1 CO ON ON r^. CM ON ON a <± O rH o r—( COCMCM o CO rH oo rH o cn i—t i—i I—1 m o n oo o CO ONCM co ON CO CM CM ON oo •••••••••••••• o i—1 co o CO rH ON co 1—1 CO ON ON 1—1 m m T—1 rH rH m d •1“1 EH •rH •H •rH d pc< d cd W d cd EH cd cd X EH rd CM O o CO •H cd PM Q — d w cd d o W d m L-J d d • •H ♦ id cd H cd cr cd PM P M d cd d cd M CO d CM O >■< Kj ✓—s o d •HP cd CO <—I PQ Mlr* c^. d CJ o a d o o rH CD '•—^ PM • w2 a) cd o e P p P d & Q p 1 • CD •r4 Mi CD Mi Ml PM co CO cd CD o O pj PM cd MH P PM a cd CO •H •H cd Ml O s CO d •|“l CO O 6 S cd d *H CO p •rH P O O cd cd cd CD CD g •H p H cd o d ♦H •H *rl XI rd •H •H o H d CD •• T—1 i—1 i—l p P d Mi d rH <3 d d d d •rl *r4 cd cd o CD d cd o o tr—1 •r-4 •rH •H M3 X Me Mi Me PM CD p d o d d d •H •H •rl cd cd rd CO o a PM d d d > > > PM PMPM td PM EH d o P 'd cd rH PH O o CM cd cd d p co cd cd o CD d cd d cd d cd PM cd cd o CD o d O o rO d r d d o CJ o pm CO O •H o •rl cd cd PM •r-t d CD rH tn P p Mi CD Xi rH T— 1 CD rH cd CO p CM PM l“ l CD d o rH d co O o CM d PM <± m c o 1 Pi r—1 oo 1 r- 1 r^. 1 o 1 5m i~l CM 00 CM 1 • 9 • • 1 rH m rM O 1 v O #» CO £X 1 r —1 O'* < t a -* av 1 cd • • • • 1 CO rH m o o 1 VO •rHS’ CO CO oo v O a v a v m cd Pm • • • • • • cx CO O'! T— 1 1- 1 CO m B i n i-1 n - o o cud a 5m CU c q 4J d •rH CU co > s d •rH a) •rl CO u d 5SJ r-l pa d cu CJ H cd 5m d cd cu •rl CM c j 4-> >* cd CO d •u o cd pa cn CU »-a Pm d I—I co CO PQ 5m CU 00 •rH < O 4-> 00 H cu od TO TO .d 4J • •H •rH CO CO d d •rl d CU d d cd u i-H o cx r d oP oh CM od CO r d d P h t o o •rl o d CO cd o •rH cu m 4J TO CJ cd d cd TO rd cd ss 4-» CU ■U (U o CO •rH •rH e> rv d UM £ CU o OO OO I''- X 12 12 30 PS 34 00 52 m CM 912 546 r-~ 229 rH 9237 1159 OfJ m NO 00 Mt 00 NO 00 ON o ON CO i—i •«••• •••• • • •• • • a w t-H rH NO o r^. o o rH n CO rH m r-'. C". o X rH X X CO w o XI ON co CO ON r-H CO rH rH NO CO rH co Q •• » ••• • • •• • • •• 'd" r"» td- rH CM CM CM o rH CM Ht NO CM a X CM CO X CM 00 1Eh g TABLE 13a CHANGE IN HYPERIID AMPHIPOD PREY COMPOSITION WITH SEASON (IDENTIFIABLE SPECIES ONLY) IN Scopelogadus beanii SEASON NO.EXAMINED NO.EMPTY %V.propinqua 70V.artnata %Pa.gaudichaudii 70Ph. atlantica Spring 12 5 28.6 28.6 14.3 28.6 Summer 17 1 31.2 0 . 0 18.8 5.9 Fall 27 8 15.8 0 . 0 0 . 0 36.8 Winter 13 3 0.0 0.0 10.0 0.0 TABLE 13b CHANGE IN DIET COMPOSITION OF HYPERIID AMPHIPODS AND GELATINOUS ZOOPLANKTON (MAJOR PREY ITEMS) WITH SEASON IN Scopelogadus beanii SEASON______NO. EXAMINED NO. EMPTY 7> AMPHIPODS % "JELLY*1 PLANKTON Spring 12 5 100.0 14.3 Summer 17 1 6 8 . 8 6 8 . 8 Fall 27 8 47.4 47.4 Winter 13 3 20.0 30.0 166 Vibilia propinqua were found within the remains of gelatinous plankton. In contrast to Phronima and Vibilia, Parathemisto was ingested in relatively low numbers. Madin and Harbison (1977) noted that P^ gaudichaudii rides on the outside of salps, and often swims away when disturbed. Although amphipods and "jelly" plankters were consumed in roughly equal proportion summer through winter (Table 13b), during spring amphipods were found far more often than gelatinous animals. Several benthic organisms were found in Scopelogadus stomachs including one specimen of a nut clam, Nucula sp., (identified by K. Nilsen) and two polychaete worms, of either the family Polynoidae or Aphroditidae (Gaston, pers. comm.). Thus, S^. beanii must occur quite near to or have contact with the bottom at times. Scopelogadus beanii has a narrow jaw with extremely small teeth. The gape, however, is quite large. Of all the members of the genus, this species seems to be the most completely adapted for bathypelagic living (Ebeling and Weed, 1963). It possesses typically reduced ossification and musculature, and probably exhibits a lower metabolism and lower activity levels. Judging from jaw morphology and suggested activity levels, j^. beanii probably uses suction to engulf prey, either by lying in wait for prey or slowly maneuvering to within a proper striking distance. This second strategy is more probable, as "jelly" plankton are probably very limited in their ability to escape from predators. 167 Suction feeding is hypothesized by Bertelsen (1951) for ceratioid angler fishes, and must be more energy efficient in comparison to an actively pursuing or lunge type of feeding strategy. Suction feeding would allow Scopelogadus to be successful in feeding on the gelatinous plankton, and could also enable S_. beanii to ingest epibenthic organisms (such as the polychaetes). Thirty-four stomachs from specimens of Scopelogadus beanii captured by midwater trawls away from the bottom were examined. Twenty-two (64.7%) contained food. Although dietary components were similar, the relative importances changed (Table 12b). Copepods were the most important (IRI=1159), and comprised the most numerous prey (31.9%). Fishes (of the genus Cyclothone), although still a minor numerical constituent in the diet made up most of the volume (42.5%). Euphausiids were also found. Hyperiid amphipods of the same genera as found in the Norfolk Canyon material and gelatinous zooplankton still were important components in the diet. Benthic prey items were completely absent from the stomachs. Perhaps at depths > 1000 m heaviest concentrations of gelatinous plankton and amphipods might occur near the bottom rather than in the water column. Thus, near-bottom Scopelogadus which feed heavily on these prey items may be able to exploit these prey more fully while Scop. beanii away from bottom influence must augment their forage with other common pelagic prey (i.e. copepods, euphausiids). Musick (pers. comm.) observed that gelatinous zooplankton (including medusae and salps) were much more abundant within 30 m of the bottom than in the 168 water column during a dive on the D.S.R.V. Alvin adjacent to the Norfolk Canyon area. These observations were made during a transect made by the submersible from 1900 to 1600 m depth. As would be expected from a passive bathypelagic feeder, diel feeding in Scopelogadus from near Norfolk Canyon, occurs at approximately equal intensity throughout the 24 hr. period, with percentages of stomachs with food being slightly higher during the day (Figure 7). Diel feeding data for the pelagically captured Scop, beanii is available for only 18 specimens. Of 10 specimens captured in the daytime, 6 (60.0%) had food in the stomachs, while at night, 6 of 8 specimens (75.0%) contained prey. While these data were insufficient to evaluate quantitatively it is possible that prey (euphausiids, copepods) migrating upwards at night become more available for capture. 169 Figure 7. Feeding intensities in Scopelogadus beanii over a 24 hr. period >/„ STOMACHS WITH FOOD IOO- i N = 11 1801 0000 - - 4 N = 22 N = 14 0001 n= 10 EID (hrs.) PERIOD 0600 Scop, beanii - - 0601 n = 18 1200 - N =22 n = 17 1201 1800 - - 170 GENERAL DISCUSSION Bias of Data Two different types of bias must be taken into account when examining food habits; feeding by captured fishes while in the net, and regurgitation of stomach contents. Although some workers have found net feeding to occur in high percentages in some midwater fishes (Lancraft and Robison, 1980; Anderson, pers. comm.), net feeding was highly improbable for all specimens examined in this study. This judgement is based on two factors. First, capture by otter trawl generally resulted in severe damage to the delicate specimens of midwater fishes. Although mud was often found in the mouths and gills of many specimens examined, it was virtually absent in stomachs, testimony for swift onset of death. Holton (1969) cited the similar rapid death of Lampanyctus (= Triphoturus) mexicanus as evidence for lack of bias from net feeding. Second, larger midwater species with pronounced dentition (e.g. Chauliodus, Nemichthys) frequently were found with their teeth tangled in the wings of the net (preventing feeding). Third, net mesh size would allow most items of prey-size to pass through (except for piscine prey items). Regurgitation tends to bias feeding intensity and periodicity data. Although it is a problem, it is often easier to quantify due to presence of regurgitated material and everted stomachs in mouths. In this study, other than stomach eversion caused by swimbladder expansion in several hatchet fish and myctophids, obvious signs of 171 172 regurgitation were found only in Gonostoma bathyphilum. Gonostoma bathyphilum was noted frequently with the stomach contracted, and food remains in the mouth consequently, therefore, only statements on prey type were made for this species and those specimens of other species with everted stomachs were excluded from analyses. Prey Abundances and Selectivity of Feeding Although plankton were collected in the Norfolk Canyon area concurrently with the otter trawl collections, the plankton from these collections has not been analyzed as of this writing. Comparing published accounts of zooplankton abundance and distribution in Slope Water with the stomach content data gathered from this study may provide a general picture of plankton abundance and availability near Norfolk Canyon. In addition, synthesis of these various data may allow inferences to be made regarding possible selective feeding in some of the fish species examined. Studies by a number of workers (Bigelow and Sears, 1939; Clarke, 1940; St. John, 1958; Grice and Hart, 1962; Be et al. 1971) in the upper waters (< 275 m) of the Western North Atlantic have shown that as one samples from the shelf through the slope to central ocean (Sargasso Sea) waters, zooplankton diversity increases while numerical abundance and biomass decrease. Zooplankton community structure in Slope Water varies with season, but generally shows intermediate diversity values with abundance and biomass decreasing as depth increases. 173 Below 275 m, Vinogradov (1970) noted that in the Western North Atlantic slope waters, despite the general decrease in abundance with water column depth, a distributional and "enrichment" (= biomass) maximum occurs around 600 m water depth with a distinct minimum between 1200-1400 m, below which biomass once again generally increases. Hopkins and Baird (1973) indicated that patterns of diversity of the plankton in upper water layers in the North Atlantic Ocean were similar to diversity patterns exhibited by the potential prey of Sternoptyx diaphana, a non-migrating, lower mesopelagic inhabitant. The results of these studies suggests that a moderate diversity and abundance of prey may be available to midwater predators in the slope area. The planktonic taxa found in collections made in Slope Water were analyzed by Grice and Hart (1962) and Be et al. (1971). Their findings agreed closely. Copepods and chaetognaths formed major constituents of the plankton both numerically and volumetrically. Ostracods appeared to be variable, being second in importance in the latter study, but considered unimportant in the collections of the former. This may be due to geographic or temporal variability. Euphausiids, amphipods, and decapods were less abundant, possibly reflecting gear selectivity. Salps and other gelatinous plankters also were variable in that they frequently formed a major percentage of sample volume, but occurred sporadically. In addition, difficulties in analyzing numerical and volumetric parameters usually resulted in gelatinous plankton being excluded from analysis. 174 The stomach contents of planktivorous fishes, especially generalized or random feeders, may reflect the relative abundances of various prey taxa, i.e., they may feed on the organisms most easily obtainable. This has been documented by a number of workers examining dietary components of midwater fishes (Cailliet, 1972; Gjosaeter, 1973; Hopkins and Baird, 1973; Baird et al . , 1975; Kinzer, 1977; Gorelova, 1978; Pearcy et al., 1979). The faunal composition of the food of fishes examined in this study were very different from published accounts of the composition of zooplankton from Slope Water (Grice and Hart, 1962; Be et al., 1971). The three major planktonic groups found in the Slope Water, the copepods, ostracods, and chaetognaths were unimportant or non-existent in the diets of the dominant midwater fish species examined. This is somewhat puzzling if it is assumed that the majority of these species conform to the "catch whatever you can" type of feeding strategy. The complete absence of chaetognaths was especially peculiar since many accounts (Hopkins and Baird, 1973; Merrett and Roe, 1974; Baird et al., 1975; Kinzer, 1977; Gorelova, 1978; Pearcy et al., 1979; Worner, 1979) list chaetognaths as midwater fish prey, ranging from unimportant to relatively common components of the diet, depending on fish species and geographic locations. Vinogradov (1970) and Grant (pers. comm.) describe chaetognaths as an important element of the slope plankton, with vertical distribution of species well overlapping those of the midwater fishes examined for this study. 175 Mobility and size of chaetognaths were probably not factors in their absence, because relatively small, slow-moving predators (such as Sternoptyx diaphana, have been found to have them in their stomachs (Hopkins and Baird, 1973). Thus, the complete absence of chaetognaths as dietary components of midwater fishes in Norfolk Canyon is inexplicable at present. Table 14 presents a listing of identifiable prey taxa recovered from all fish stomachs examined in this study with percentages of frequency of occurrence, numerical abundance and volumetric displacement for each group. Crustacean items predominate both in frequency and numerical abundance (47.1 and 6 7.8%), whereas fish remains constitute the bulk of the volume (61.2%). If the two totally piscivorous species, Chauliodus sloani and Stomias boa ferox, have their stomach contents removed from consideration, then the crustaceans overwhelmingly dominate in all categories. Figure 8 illustrates the percentages of the frequency of occurrence, numerical abundance and volumetric displacement of the major prey taxa relative to one another. The euphausiids make up the bulk of the crustacean prey in all categories but volume displacement. Amphipods (Hyperiidae) were second in frequency of occurrence, while decapods (Sergestidae) were first in volume but third in the other two categories. Copepods and ostracods formed very small percentages of the diet and were never important in any single species, save for Malacosteus niger and pelagically captured specimens of Scopelogadus beanii, in which copepods (Paraeuchaeta norvegica) dominated. 176 TABLE 14 PERCENT FREQUENCY OCCURRENCE (F), PERCENT NUMERICAL ABUNDANCE (N), AND PERCENT VOLUMETRIC DISPLACEMENT (V) OF IDENTIFIABLE PREY TAXA RECOVERED FROM THE STOMACHS OF ALL PREDATORS TAXON F N V Mollusca Pelecypoda 0 . 2 0 . 1 < 0 . 1 Cephalopoda 0 . 8 0.3 1.3 Total 1 . 0 0 . 6 1.3 Annelida Polychaeta 0.5 0 . 2 < 0 . 1 Echinodermata 0 . 2 0 . 1 < 0 . 1 Crustacea Ostracoda 1 . 8 1.3 < 0 . 1 Copepoda 2.3 3.2 0 . 2 Amphipoda (Hyperiidae) 16.6 14.6 3.4 " (Gammaridea) 0 . 2 0 . 2 < 0 . 1 Euphaus iacea 2334 43.4 8 . 1 Decapoda (Penaeidea) 1 0 . 8 5.2 13.8 " (Caridea) 0 . 2 0 . 1 1 . 2 Total 47.1 67.8 27.3 Gelatinous Zooplankton 7.8 3.6 1 . 1 Pisces Teleostei 19.1 9.2 61.2 (4.3)** (2 .1 )** (3.6) * - Gelatinous Zooplankton is an inclusive term encompassing salps, hydromedusae, and siphonophores. ** - Values after totally piscivorous species (C.sloani and S . b.ferox) are removed. 177 Figure 8 . Relative Percentages of Frequency of Occurrence. Numerical Abundance, and Volumetric Displacement of Major Prey Taxa. 178 FREQUENCY OF OCCURENCE > CD c z > o □ : r 3 -o < CO > o m c I- rn o (Caridea) ( Hyperiidae) ( COPEPODA AMPHI PODA PODA AMPHI % after removing EUPHAUSIACEA TOTAL GELATINOUS DECAPODA OSTRACODA ANNELIDA ZOOPLANKTON DECAPODA bars indicate CRUSTACEA ECH IN ODER MATA ODER IN ECH PISCES (Interna I AMPHI PODA PODA AMPHI (Gammaridea) (Penaeidea) piscivores) MOLLUSCA PREY TAXA 179 Gammaridean amphipods and caridean decapods were insignificant dietary components• Selectivity may account for the discrepancies between prey type and apparent prey abundances. Feeding selectivity may occur with respect to size of prey ingested, whereas taxonomic selectivity is indicated when specific prey taxa are eaten in quantities disproportionate to their abundances (Hopkins and Baird, 1977). Morphological specializations of predator species, especially with respect to their jaw structures, may result in a narrow range of prey types being eaten (e.g. Mead and Earle, 1970, on Nemichthys). Such selective strategies in midwater fishes have been demonstrated. Several studies (Paxton, 1967; Samyshev and Schetinkin, 1971; Gjosaeter, 1973; Hopkins and Baird, 1973; Kinzer, 1977; Gorelova, 1978; Pearcy et al., 1979; Worner, 1979) have found a positive correlation between prey size and fish size. Merrett and Roe (1974) examined plankton abundances, diversity, vertical distribution and feeding habits of several species of mesopelagic fishes in the Eastern North Atlantic. Based on relative plankton abundances from plankton trawls and percentages of the different plankton groups in fish stomachs, they concluded that three species, Valenciennellus tripunctulatus, Argyropelecus aculeatus, and Lampanyctus cuprarius, were selective for copepods, ostracods, and amphipods respectively, while a fourth species, Notolychnus valdiviae actively selected against ostracods. Two other species, A. hemigymnus and Lobianchia dofleini, appeared to feed randomly. Samyshev and Schetinkin (1971) 180 found that Maurolicus muelleri, Diaphus dumerilii, Lepidophanes guentheri and D. taaningi showed prey size and taxonomic selectivity off northwest Africa. Size selectivity, i.e., larger individuals of a predator feeding on larger prey items, appeared to occur in most species examined in this study. Relatively few small fishes of any species were captured by the otter trawl, because they either escaped through the meshes of the trawl, or were absent (see Table lb for ranges and mean sizes of fishes examined). Thus, euphausiids, amphipods and sergestids may have been important food items in the study area despite their lower relative abundances in Slope Water because the midwater fishes captured were mostly larger individuals. The identifiable copepod species, Paraeuchaeta norvegica, usually was represented by specimens 8 - 1 0 mm in length which seems to support the idea that larger zooplankton may have been selectively preyed upon. Table 15 lists the fish species examined with apparent feeding pattern and major prey types• Size selection of some prey species also appeared to occur. Most of the intact euphausiids (Nematoscelis megalops, Euphausia krohnii, and Meganyctiphanes norvegica) appeared to be in the uppermost size ranges for their respective species (20-35 mm). Amphipods (Parathemisto gaudichaudii, Vibilia propinqua, V^. armata, and Phronima atlantica) found in the stomachs of Scopelogadus beanii were usually small in size (< 10 mm). This was the size range reported by Madin and Harbison (1977) and Harbison et al. (1977) for these species, when 181 TABLE 15 LIST OF PREDATORS WITH RESPECT TO FEEDING STRATEGY AND MAJOR PREY TYPE SPECIES______STRATEGY______MAJOR PREY TYPE N. scolopaceus Selective Sergestes arcticus Serr. beanii , Random Euphausiids, fishes, cephalopods Serr. brevidentatus' ------Euphausiids N. ingolfianus Selective Euphausiids D. serpentinus Selective Sergestes arcticus E. pelecanoides Random Fishes,invertebrates G. elongatum f Selective Euphausiids G . bathyphilum“ ------Fishes, invertebrates A. aculeatus Random Euphausiids, ampphipods, gel.plankton S. diaphana Random Euphausiids, amphipods C. sloani Selective Ceratoscopelus maderensis S.b.ferox Selective Caratoscopelus maderensis M. niger' , ------Paraeuchaeta norvegica P . guernei'------Sergestes arcticus C. maderensis' Euphausiids B. glaciale ------Mixed L. macdonaldi Selective Parathemisto gaudichaudii Scop. beanii Selective Amphipods, gelatinous zooplankton - Denotes small sample size from which no strategy can be determined. 182 associated with gelatinous zooplankton. In Lampanyctus macdonaldi, which in the present study fed solely on P. gaudichaudii, and Gonostoma elongatum, stomachs of which contained P. gaudichaudii, Vibilia sp., and Phronima sedentaria, the mean size of the amphipods was approximately twice as large as that of specimens ingested by Scopelogadus. Size selection of prey also apparently occurred in the piscivores Stomias boa ferox and Chauliodus sloani. The two prey species recovered from Stomias, Ceratoscopelus maderensis and Diaphus dumerilii, ranged from 55-70 mm and 50-55 mm respectively, with a mean length of 61.9 mm for C^. maderensis, while the only relatively intact prey item in Chauliodus, a C^. maderensis, was approximately 50 mm. This size selection may be a function of the numerical abundance of prey of this size. Backus et al. (1968) commenting on vast schools of C. maderensis in Slope Waters off New England, noted "There was very little variation in the size of the fish in the schools...had a mean length of 62.4 mm...range of 52-74 mm." These figures closely agree with the range and mean length of C^. maderensis in Stomias stomachs. Morphological modifications of the jaw structure in several of the fish species examined was probably a contributing factor in the prey type ingested. The specialized feeding of Nemichthys has already been discussed elsewhere in this study and will not be repeated here. Other species exhibiting such jaw and feeding specialization were Chauliodus sloani, Stomias boa ferox, and possibly Nessorhamphus ingolfianus. 183 Chauliodus and Stomias both have obviously exaggerated dentition, used for piercing and holding large prey items. In addition, both have a specialized skeleton which allows the head and jaw to be pivoted in such a way that large prey can be ingested (Tchernavin, 1951; Morrow, 1964). Such adaptations certainly play roles in both species' primarily piscivorous dietary habits. There is apparently no published evidence that indicates specialized feeding mechanisms or habits for Nessorhamphus, however, as mentioned in the feeding account for this species in the present study, the gape is not large, and the throat is quite narrow when compared with other midwater eels (pers. obs.). It may be that this allows feeding primarily on smaller prey items such as the euphausiids. The habits of the plankton themselves may also allow fishes to more easily prey upon them; selection for a prey source requiring a minimal energy expenditure for capture. For example, the gelatinous zooplankters are not very rapid swimmers, which may enable slowly moving predators such as Scopelogadus beanii to easily approach and ingest them. The head upward-antennae downward orientation of sergestids enhances contact and subsequent entanglement in the jaws of Nemichthys (Mead and Earle, 1970). Vertical movements of various organisms may allow for increased predation as they ascend and/or descend through the habitats of various predators (Merrett and Roe, 1974; Marshall and Merrett, 1977). 184 A second explanation which may account for the difference between apparent plankton abundances and stomach contents is that of the near-bottom interaction of predators feeding on richer sources of pelagic food occurring close to the bottom. Sedberry (1975) stated that "Turbidity currents may indirectly play an important role in the nutrition of fishes which feed on pelagic prey by resuspending detritus particles and increasing the food supply for planktonic animals near the bottom..." Although this statement was made with respect to increased availability of pelagic prey for benthic predators, pelagic predators living near the Continental Slope would also have access to an enriched prey source. Omori (1974) noted that shrimps occur where food concentrations are greatest. This near-bottom particulate layer could serve as a richer food source than would the depauperate water column. Submersible passengers (Sulak, pers. comm.) have observed concentrations of sergestid-like shrimps and unidentified euphausiids within a short distance of the bottom. A number of studies have shown that euphausiids may congregate near bottom, especially during the day. Mauchline and Fisher (1969) provided evidence for the common presence of Meganyctiphanes norvegica near the bottom. McEachran et al. (1976) noted M. norvegica as the only euphausiid species found in the stomachs of four species of benthic skates (Raja spp.). Marshall and Merrett (1977) found that the two most common species of euphausiids of NW Africa, Nematoscelis megalops and Euphausia krohnii, 185 approached close to the bottom. These three euphausiid species were the most common prey species found in the present study. Pelagic shrimps are known to undergo ontogenetic descent, with adults generally occurring deeper than larvae or juveniles (Omori, 1974). Vertical migration of adult Sergestes arcticus may allow them to closely approach the bottom over the Continental Slope. Crabtree (pers. comm.) found that stomachs of the demersal macrourid Nezumia bairdii, which were captured from a wide geographic range of stations from New England to Cape Romaine, S.C., and including the Norfolk Canyon region, were filled with S^. arcticus remains. Thus, taxonomic selection of certain prey species (e.g. Sergestes, Nematoscelis) could have been due to the more frequent presence of these species near the bottom. Near-bottom concentrations of plankton may allow fishes that do not migrate to congregate along this layer to utilize enriched food resources. During the day, the downward vertical migration of fish species may also bring them near the bottom in slope areas (Marshall and Merrett, 1977). Various species of midwater fishes have been noted near the bottom by observers in submersibles. This is especially true of bathypelagic, non-migratory species such as Lampanyctus macdonaldi (Craddock, pers. comm.) and Gonostoma bathyphilum (Musick, pers. comm.). Presence of midwater fishes near bottom had been corroborated by their presence in benthic or benthopelagic fish stomachs. An intact specimen of G. bathyphilum was recovered from the benthic piscivore Bathysaurus ferox (pers. obs.), 186 and Sedberry and Musick (1978) noted the presence of freshly ingested myctophids and Nemichthys scolopaceus in the stomachs of the demersal eel Synaphobranchus kaupi. The presence of benthic prey items in the stomachs of several midwater fishes examined in this study is clear evidence for the near-bottom interaction of individuals of pelagic fish species on at least an occasional basis. Although not found in many species or individuals (3 species, 6 specimens), benthic animal remains in midwater fish stomachs establish the link to the utilization of food from near the bottom. Interestingly, the three species of fishes found with benthic prey in the stomachs were all bathypelagic non-migrators (Scopelogadus beanii, Eurypharynx pelecanoides, and Gonostoma bathyphilum). It is unlikely that the food was picked off the bottom because of the jaw structure of these species. As Sedberry (1975) suggested, perhaps strong bottom currents, suspend material (detritus plus occasionally some benthic inhabitants) and near-bottom midwater fishes are able to ingest the suspended benthic animals. Prey Abundances and Dietary Trends Dietary patterns were examined with respect to season of the year, fish size, depth, and diel periodicity. a) Season Seasonal variation in prey composition was noted in several species of midwater fishes examined; Gonostoma elongatum, Sternoptyx diaphana, and Scopelogadus beanii. It is possible that other 187 euryphagic species also changed their diet with season, but in many cases, not enough specimens were examined to clearly delineate trends. G_. elongatum and S^. diaphana are euryphagic primarily feeding on euphausiids, but utilizing hyperiid amphipods as secondary components in the diet. In the summer period, however, amphipods are the more abundant prey item (Gonostoma, Table 6 b) or are consumed equally as often as euphausiids (Sternoptyx, Table 8 b). Presumably, this shift in abundance is due to increased presence of hyperiids in the study area during the summer survey (cruise GI-75-08). Bowman (1960) noted that Parathemisto often completely dominates the cold-water amphipod faunas. Grant (1979) found that Parathemisto gaudichaudii was a dominant member of the neuston in fall, winter and spring samples taken from 1976-1977 at station L 6 , located at the northwestern edge of Norfolk Canyon (approx. 37°05*N, 74°35fW). During the summer, however, it was apparently only occasionally taken. Increased presence of J?. gaudichaudii in Gonostoma and Sternoptyx during the summer may be due to a downward migration of the amphipods in response to increased surface temperatures. Parathemisto only occurred in G. elongatum taken deeper than 700 m. Gosner (1971) listed Vibilia spp. and Phronima spp. as primarily Gulf Stream forms. Transport of these amphipods to the Norfolk Canyon area may occur through the formation of anticyclonic eddies or warm core rings, created by the pinching off of portions of Gulf Stream meanders (Richardson, 1976). These eddies, which generally move southward along the continental slope, commonly transport warm water 188 organisms with them (Haedrich, 1972). Ruzecki (1979) noted that the hydrography of Norfolk Canyon during cruise GI-75-08 was marked by the presence of Gulf Stream Water transported to the canyon area by an anticyclonic eddy. Thus, increased predation on amphipods may reflect additional presence of transported material. The diet of Scopelogadus beanii was composed of hyperiid amphipods and gelatinous zooplankton. Presence of these items may also have reflected seasonally varying abundance of the prey types, as amphipods dominated in the spring, and were found in an equal number of stomachs as the gelatinous zooplankton during the summer and fall. Both groups were poorly represented in winter. The "jelly" plankton were found in greatest abundance in stomachs in the summer (Table 13b). This is similarly reflected in the diet of Argyropelecus aculeatus which fed on euphausiids, hyperiid amphipods, and gelatinous zooplankton. The gelatinous plankton were only found in stomachs of fish captured in the summer. Several workers have documented alterations in prey composition with season in midwater fish species. Gjosaeter (1973), examining feeding in Benthosema glaciale, found that copepods dominated the diet in summer, while euphausiids were an important component in winter. He found that this directly correlated with abundances of these prey groups, thus prey availability paralleled prey utilization. Cailliet (1972) found that the diet of Leuroglossus stilbius in the Santa Barbara and Santa Cruz basins off California related to 189 availability of salps and larvaceans, the preferred prey. He noted that in the Santa Barbara basin, where salps and larvaceans were abundant year round, the diet remained stable. In the Santa Cruz basin, the abundances of these prey declined during the warming summer period, and smaller copepods became the dominant dietary constituents. Collard (1970) and Cailliet (op. cit.) examined seasonal variation in diet for Stenobrachus leucopsarus and found that selection of prey type varied with the abundance and diversity of prey groups during different times of the year. b) Size Although difference in sizes of fishes has been found by many workers to influence the type of prey selected (Paxton, 1967; Collard, 1970; Gjosaeter, 1973; Merrett and Roe, 1977; Kinzer, 1977; Clarke, 1978; Pearcy et al., 1979; Worner, 1979), such trends were not found in this study, because of the limited size range of fishes sampled, or small sample sizes of some species. c) Depth Depth also appeared to have little effect on prey type. This may be due to extended vertical ranges of plankton over the slope due to variable currents and nutrient levels or may reflect predator activities. Depth appeared to play a role in the feeding activities of the predator species. When depth of capture was compared to percent 190 empty stomachs, many species appeared to have low percentages of empty stomachs in certain depth strata. Table 16 lists a number of species and their apparent preferred feeding strata. One species, Gonostoma elongatum shows a broad range of feeding depths. The most likely explanation is that it is the only migratory species listed in the table and is able to feed throughout its migration ranges. The sudden marked increase in empty stomachs below 2 0 0 0 m suggests that specimens of Gonostoma taken at greater than 2000 m are outside of their optimal depth range. Other studies of fishes that migrate upward at night (Gjosaeter, 1973; Merrett and Roe, 1974; Baird et al., 1975; Kinzer, 1977; Ozawa, 1977; Worner, 1979) have found that feeding in these species occurs at night. DeWitt and Cailliet (1972) found that Cyclothone signata, optimally fed most often within restricted depth layers in its’ habitat. Pearcy et al. (1979) found that in Stenobrachius leucopsarus, portions of the population did not ascend into surface layers to feed, but rather, remained at depth and fed there. Although many prey taxa were similar, the rank of various prey between shallow and deep-dwelling specimens differed. In addition, deeper living prey were also found in stomachs of deeper-dwelling Stenobrachius. This suggests that optimal feeding depths for midwater fishes may not be unimodal. This would apply to Gonostoma elongatum apparently. DEPTH OF CAPTURE VS. PERCENT EMPTY STOMACHS FOR SOME DOMINANT FISH SPECIES* ( ( PEAK FEEDING RANGE IN PARENTHESES ) —i i— t —1 T— t i—i 1-1 i— M i o O M C o o o o o m in o !"■» o o o o o -d" 0 0 o M C o n c o n c m M H g »o n» i g 1 i vo O v co V-/ <1*o M C 1 1— o o o O C m m r-* M C n i m o 3 O C I3| . . a • • • « * • * • • 9 • • • ° 9 o o O C o ' M C o —1 1— o O O _ U U g m O v —N/— N /— O V —1 1— o O O o c o M C O C O C m o o o o o «IS| g g i i i i i i . . . ^ ^ M i g O v o o o m o M C m r'. m • 9 s -c - <1- —1 r— M C o 0 0 o \ /— r". M | 0 | M| C O i g i g g g a • . • . ■ t H r—1 o —g i— o o ± < O C s /— o o — . o • o 1 2* . -]C O r— 1 '^S m o c o o o o o - n m g O v O g v g g . - f * —1 t— o o o —•,/— O v o o O v O v —1 1— r^. o c o O ' — CM • a •3'— CM M-t X) M| •H CO| Ol Q 4J +J ■M d CO CO co O u a) > o o u o d d o a. d § cu o d d u u d cu Cu 0 O X o d ) M cr\ h e CM •H n J 4 d h CO CO h 191 192 d) Dlel Feeding Periodicity Hopkins and Baird (1977) have reviewed various aspects of feeding ecology in midwater fishes. Many of the vertical migrators show distinct diel feeding periods, with intense feeding at night, and little or no feeding during the day. Many of the species analyzed for feeding periodicity in the present study also show cyclical feeding patterns. The vertical migrators Gonostoma elongatum (Figure 4b), Argyropelecus aculeatus (Figure 5a), Chauliodus sloani (Figure 6 a ) , Stomias boa ferox (Figure 6 b) and Ceratoscopelus maderensis all showed unimodal (G^. elongatum, A. aculeatus, C_. maderensis) or bimodal (C. sloani, S_» _b. ferox) feeding periodicity. These peaks generally occurred at night (1801-0000 and 0001-0600 hrs.). Non-migratory species such as Derichthys serpentinus (Figure 3a), Nessorhamphus ingolfianus (Figure 3b), and Lampanyctus macdonaldi exhibited periodicity as well. Peak predatory activities in these species probably corresponded to movements of migratory prey through the predatorsr habitats. Other species such as Nemichthys scolopaceus (Figure 10 and Scopelogadus beanii (Figure 7) showed little or no diel periodicity. Both of these species appeared to be selective and it may be that the prey they utilized occurred at a relatively constant level of abundance. Trophic Structure and Feeding Strategies Legand and Rivaton (1969) constructed a two level trophic classification for 11 species of mesopelagic fishes. Trophic level A 193 was composed of 1 0 species characterized generally by small body size, low percentage of empty stomachs (which was presumed to indicate high trophic activity) and a low prey-to-predator body weight ratio. Trophic level B, comprising the single species Chauliodus sloani displayed opposite characteristics. One difference between their work and that of the present study is that jC. sloani specimens examined by Legand and Rivaton (1969) were all 1 0 0 mm in length or smaller with roughly half of the specimens only around 20 mm long. The small sizes of these specimens could have accounted for a high percentage of empty stomachs due to lack of food of a proper size. Chauliodus sloani from the present study had a much lower percentage of empty stomachs (47.6% compared to Legand and Rivaton*s 69.5%). Only 2 out of 42 specimens examined were 100 mm or smaller in length. A general trend was noted of decreasing percentage of empty stomachs with increasing size. An additional difference is that the species examined by Legand and Rivaton (1969) were trawled from 210 m or less. Analysis of results obtained from the present study suggest optimal depth levels for feeding activities. Large C^. sloani were found to have fed more intensively deeper in the water column. This may be true for smaller specimens also and could explain the high percentage of empty stomachs that was found by Legand and Rivaton. Thus, conclusions concerning trophic position or feeding strategy based on percentages of empty stomachs may be fallacious. Inferences 194 made regarding feeding based on prey type and prey abundance, and functional morphology of the predator species probably better reflect actual trophic activities. The validity of morphological assumptions seems to be supported by the few studies that have been done on midwater fishes maintained in aquaria. Childress and Meek (1973) studied the feeding activities of Anoplogaster cornuta, the "Fangtooth". This is a heavy bodied non-migratory beryciform found in the lower mesopelagic zone. Examining the gross morphology of the species, one notes the presence of small eyes, a small caudal peduncle and tail in relation to body size, and an open lateral line system. This would seem to indicate a fish which detects prey (or predators) through tactile stimuli rather than by visual cues. The most outstanding features of Anoplogaster are large bony head and big mouth full of long fang-like teeth, and its large pectoral fins. The inferences drawn from these characters denote a short range predator that may ingest large prey, such as fishes or crustaceans. Childress and Meek (op. cit.) suggested that Anoplogaster apparently utilizes chemotactile stimuli to initiate feeding. The stimuli eliciting a positive feeding response were those received in the head and mouth region (when stimulated, a short lunge was the response). Negative stimuli, causing movement away from the source, were those received in the caudal or lateral line regions, indicating possible response to predator activities. 195 Feeding habits of a second species of midwater fish maintained in aquaria, the mesopelagic zoarcid Melanostigma pammelas were studied by Belman and Anderson (1979). M. pammelas is a small eel-like fish with very large eyes, a small mouth, and reduced pectoral fins. In addition, the body is enveloped within a gelatinous "envelope" of undifferentiated squamons epithelial cells (Anderson, pers. comm.). From these features, M. pammelas probably feeds on small plankters that it locates visually. The presence of the envelope and small pectorals indicate low levels of swimming activity (cf. Stomias, Chauliodus). The results of Belman and Anderson (1979) indicated that Melanostigma used both a sight-lunge (visually cueing on prey and striking at it from a distance of 1 body length or less), and contact-suck (prey comes in contact with head region, and is engulfed by suction) method of feeding, of which the former was more successful than the latter. The general swimming activity was either motionless in the water (or on the bottom of the aquarium), or by a sinusoidal undulation of the body. Based on prey types, anatomical and morphological appearances of the predators, and previous published accounts, the fish species examined for the present study showed two main feeding patterns active foraging or passive (either drifting or wait-and-lunge). Within both the active and passive feeding patterns are species which seem to selectively or randomly utilize prey. This may be function of predator morphology, prey availability and other factors. Table 17 shows tentative feeding strategies for the species examined in this THEORETICAL STRUCTURE OF FEEDING PATTERNS AND STRATEGIES OF DOMINANT MIDWATER FISHES AROUND NORFOLK CANYON #> O C s q < 3 CO w >* M W 3 O O C W rG W O M > EH EH Ej M O >H « EH w 2 1 {— a E-J H w a S Q W Xl Q r •iH •H X) Q C H • CO •rH I M o i a i Q u*H ■u 4-1 -u cd CO G CO (U u u u dcd cd CO P 0) G uCO cu G 3 cu p cu CU > p cu G a) a 3 • • • H r H M r-H •iH 1 < •rH x» l O +J CO w d3 cd cd G G o G 3 CO cd cu P G CQ CO o cu 3 & 3 • • • • • rH 30 4-1 M cd O e J X —1 p— > x •rH Crane, 1936 for Serrivomer). The proposed pattern indicates feeding strategies and activities; it does not attempt to delineate trophic levels (i.e. energy levels). In addition, the placement of species in certain categories is not fixed. Activity levels and feeding strategies may change with concomitant changes in predator size and prey availability. In summary it appeared that size selection (larger fish feeding on larger prey) was the primary factor determining food habits of most of the midwater species examined. In addition, there appeared to be selection for larger individuals of the various prey species by a number of the midwater predators. Morphological specialization of the mouth and jaw structure evidently controls prey utilization in some species. Taxonomic selection also appeared to occur in several fish species, but may have been a function of prey abundance. The small range of prey species eaten may be due to increased presence of these species near the bottom in the Norfolk canyon area. This concentration of plankton may represent an enhanced prey source for midwater fishes which may then aggregate near bottom to utilize this prey source. Benthic prey items, found in several stomachs, suggested at least the occasional near-bottom presence of midwater fishes. 198 Depth was found to influence feeding in that many fishes appeared to have optimal foraging depths. Seasonality played a role in several species feeding habits, causing them to shift their diet in apparent response to fluctuations in prey abundance. Diel feeding periodicity was demonstrated in many of the species, with either a unimodal or bimodal peak of feeding intensity noted. Several species however, exhibited uniform diel feeding intensity. PART III THE PARASITE FAUNAS OF THE DOMINANT MESO- AND BATHYPELAGIC FISH SPECIES OF NORFOLK CANYON, WITH NOTES ON THE ECOLOGICAL IMPLICATION OF THEIR PRESENCE. INTRODUCTION Noble (1973) stated that "...parasites are a part of the environment of the fish, and the host is the immediate environment of its internal parasites." Thus, a host and its "parasite-mix" (sensu Noble, 1960) form an interacting unit. Little work has been done to examine these possible interactions in deep-sea pelagic fishes. The few studies detailing host-parasite relationships (Collard, 1968, 1970; Noble, 1973) focused primarily on how midwater fishes affect their parasites via growth changes, feeding habit changes related to growth, different sexes and maturity of the host, etc.. Collard (1970) made passing reference to pathology and disease, but no study to date has made a detailed examination of effects of the parasite-host interactions in midwater fish hosts. Although the major emphasis of these works was on the mesopelagic fish faunas, all provide a general account of parasitism in fishes of the deep-sea pelagial. Much of the other literature available on the parasites of deep-sea fishes is limited to taxonomic descriptions of new species and their hosts(e.g. Noble and Collard, 1970; Noble and Orias, 1970, 1975; Orias, Noble, and Alderson, 1978). Earlier records are recounted by Collard (1968). Typically, midwater fish species exhibit a lower incidence of infection, both in numbers of fishes parasitized and types of parasites found, than do epipelagic, deep-sea benthic, or inshore species. In addition, adult helminth parasites are extremely rare (Noble and Collard, 1970). Later studies indicated that meso- and 200 201 bathypelagic fishes each have a different parasitological makeup, mesopelagic species being more heavily infected with larval nematodes, while bathypelagic hosts show a higher incidence of digenetic trematodes and cestodes (Noble, 1973). In this section of the study are identified the parasite faunas primarily located in the digestive systems and associated organs of 18 dominant midwater fish species ( 1 2 mesopelagic, 6 bathypelagic) collected around Norfolk Canyon. The parasites recovered were listed for each host species. These accounts were compared to any past studies in which the host species (or species closely related to the hosts) were examined. The possible effects of various parasites on their host(s) was explored and possible modes of infection via feeding habits or other activities were discussed. Habitats of the host(s) (meso- vs. bathypelagic) were examined with respect to the parasites found and compared for similarities, differences, or both. Finally, the trends noted by Collard (1968) in his study on Pacific midwater fish parasites were compared to those noted in the Atlantic hosts of the present study. METHODS AND MATERIALS Most of the fish specimens utilized for the feeding habits portion of this study were also subjected to parasitological examination. Fish species and numbers examined are presented in Table 1 . Primary necropsy of the specimens involved only those organs associated with the gastro-intestinal system, i.e. liver, esophagus, stomach, intestine, and their attached mesenteries and surrounding coelomic cavity. Data presented by Collard (1968) indicate these are the primary sites for helminthological infection. In rare cases, the heart and gall bladder were examined. Gill and mouth regions frequently sustained heavy damage and were subject to much flushing action, therefore, they were ignored. The external surface of the body also received much damage and usually was not examined, however, one specimen each of Photostomias guernei and Benthosema glaciale had parasitic copepods attached externally, and these were removed and included in the analysis. After opening the abdomen, a brief gross examination of the viscera and abdominal cavity was made using a dissecting stereo- microscope. The viscera were then removed by cutting at the pharynx and around the rectum. Following this, a more thorough examination of the coelomic cavity and peritoneum and outside of viscera was made, and any obvious or suspected parasites were noted as to number and location and removed. The stomach and intestine were then dissected 202 203 and washed out with 70% ethanol (for a complete discussion of this procedure, refer to the food habits section) into a Petri dish. Suspected parasites were removed and labeled as to location and number. All parasite material was stored in separate vials containing 70% ethanol and 5% glycerin. Preparation of the parasite material was as follows: -Nematodes. Temporary mounts of larval forms were made in either lactic acid or lactophenol. Permanent mounts of larvae were made by dehydrating to 1 0 0 % ethanol, clearing in methyl salicylate, and mounting in Permount. Adult nematode mounts were made in essentially the same manner, except that the worms were stained with Semichon?s acetocarmine during the dehydration series. -Digenetic Trematodes. Several different techniques and stains were used. This was necessary as the relatively long period of time (up to 6 years) that the worms spent in preservative lessened the ability of the internal structures to take up stain. There were two basic techniques; first, the worms were hydrated to water and stained with either Reynold's double stain, Van Cleave's hematoxylin, or alum cochineal, dehydrated to 1 0 0 % ethanol, cleared in methyl salicylate or creosote, and mounted in Permount. In the second method, alcohol stains (70% ethanol), either Semichon's acetocarmine or Mayer's paracarmine were used, followed by dehydration to 1 0 0 % ethanol, then cleared and mounted as above. In both methods, the trematodes were 204 initially hydrated to water and placed in 5% sodium bicarbonate to adjust internal pH. -Cestodes. All larval cestodes were treated as above, with the primary stains being either alum cochineal or Semichon’s acetocarmine. The adult cestodes had proglottids removed at various stages of development, with the different sections being stained by either Reynold's double stain, Van Cleave’s hematoxylin, Harris' hematoxylin (aqueous stains), or Semichon’s acetocarmine and Mayer’s paracarmine (alcohol stains). Clearing and mounting as for trematodes. Adult cestode proglottids were serially sectioned for taxonomic purposes. The sections were stained with Harris' hematoxylin and eosin (HHE). -Copepods. Copepods were temporarily mounted whole in lactic acid for study. When detailed examination was required, copepods were stained in picric acid-fuchsin, dissected in lactic acid, and the parts were permanently mounted in Hoyer's mounting medium. -Acanthocephala. Bullock’s (1963) technique using borax carmine stain was used. -Fungi. Whole mounts of fungi were prepared using Semichon’s acetocarmine during dehydration as previously outlined. Cotton blue in lactophenol and lactic acid were used to make temporary mounts. Serial sectioning of fungal cysts was carried out using McManus' PAS 205 technique outlined in Luna (1970), and also using a standard Harris1 hematoxylin and eosin stain. Microphotographs of the serial sections were made using a Leitz Ortholux I compound microscope equipped with phase contrast and a Leica 35 mm camera. A variety of light and phase settings, and lens objectives were used to obtain maximum clarity for the micro photographs . All parasites were identified to the lowest possible taxonomic level. In almost all cases, as the bulk of the material was larval, identification was restricted to generic or higher levels. RESULTS Six hundred sixty-eight specimens comprising eleven families, sixteen genera, and eighteen species were examined for parasites (Table 1). Of these, two hundred sixty-one specimens (39.1%) were found to have some parasite or cyst. All of the families and genera (100.0%) and all but one species (93.3%), the bathypelagic eel Serrivomer brevidentatus, were parasitized. Table 2 lists the major parasite groups and species within each group that were found in the hosts examined for this study, with percent infection data also given. The results are presented in two parts. Section A gives the results in the taxonomic order of the hosts, listing the following: host data [no. hosts examined/no. hosts parasitized/percent of hosts parasitized, no. of parasite species recovered], and parasite data [infection incidence and percentage, location(s) of parasite(s) in host, and host reaction, if any was observed]. In addition, any previous studies involving the host or species related to the host are briefly mentioned. 206 207 TABLE 1 LIST OF DOMINANT HOSTS CAPTURED. NUMBERS EXAMINED FOR PARASITE ANALYSIS. HOST SPECIES NO. CAPTURED NO.EXAMINED ADDITIONAL Nemichthys scolopaceus 380 252 Serrivomer beanii 38 24 S. brevidentatus 9 .8 Nessorhamphus ingolfianus 38 20 6 Derichthys serpentinus 32 11 7 Eurypharynx pelecanoides 12 11 5 Gonostoma elongatum 75 48 G. bathyphilum 26 18 Argyropelecus aculeatus 20 9 6 Sternoptyx diaphana 31 19 Chauliodus sloani 56 36 Stomias boa ferox 105 60 Malacosteus niger 13 10 Photostomias guernei 5 5 Benthosema glaciale 139 12° Ceratoscopelus maderensis 84 6° 12 Lampanyctus macdonaldi 19 14 Scopelogadus beanii 70 69 1152 632 36 - Additional specimens are from Norfolk Canyon region, but taken on cruises other than CI-73-10, GI-74-04, GI-75-08, GI-76-01. For listings, see catch data in Appendix. ° - Due to a lack of storage space during cruises, much of this material was discarded. MAJOR PARASITE GROUPS AND SPECIES RECOVERED. PERCENT INFECTION DATA BY TAXONOMIC GROUPING OF HOSTS M C/2 O E C M w CO PM CO W g w g w C CO W w M H CO g W P M _ _ /2 /2 m m m « ' S'S S'S S'S B^ o g O o O P3 23 CJ > W Pi o pa • . • oo vo vO CM vO v o 0" o o CO CM o o CO 0 0 CM v O " 6-2 . . • . • •.. . . •. o o CM n c m o O o o o O o CO CM »i—i S M 1—1 <3- r-H o o a v CM r-H i n CM i n i— ! CJ O < n CO i— 1 W S3 i H , Ph CO /•"N vo CM O CM VD v O v O v O r ^ - VO r—H OO B-2 «. • • «• . «. ». « r—1 m vo CM O c m n u o n m v O i n r H v 1—1 L O CM i—H i H CO w M • i—! CO CTv < r i—i i— i i—i i—i CO i— i CM CJ O 1 3 < 1 EH W S3 C Ph Q CO T3 E-* r'-o. CM oo CM CM O CM CM CM CM o o CM n on vO 6-2 •. . * . . . • ». . « O i—1 vO oo •H v o m v O v O v O VO 0 0 VO CM 33 I—l o o n c m 1 1 !—1 S ■. I—l co o -d " r-H i— 1 i— 1 I—1 CO r H CM w o 1 3 S3 S3 w w CJ (J pq ___ S 1—1 1—1 r-H co o r^> co I—1 r—1 i— l 1—1 CO r - l CM w •• • • • • •• • • •• o> I— o c o co c o a v I"- COOO CO CM o 7 2 2 7 CM i— 1 H 1—1 I—i )-3 • I—l OO f —I OO CO i—i i— i i— i i— I CO T—1 Osl I—1 O pH CO Q W W e- i pe CM CO in M W •jc cx cx CO > cu cd ••• • • CO CO H •H t o C O 0 0 o o 00 CO CM CM CM o f • • » ••• O 1— I o CO o OO O vO C M C M v O CM CO o o CO -cj- H 53 PM CO vO r^» vO VO VO VO CO S'S LO vO u o O uo UO U 0 r-. uO CM CO w I—l CO U0 o o CO H CM CM CM UO CM CM CM CO oo VO &-S ••• o VO oo VO v O CM v O V O v O UO T d pel UO i— 1 CM CO CM i—1 CO i—i OO CM i—1 i—1 &-e ••••••• CTv r^- CTV i—1 00 OV O'. vO cH 00 1—1 c o i—1 CO i— 1 o> CM rH 1—1 Em 3 CO Q cu H W 00 CM rH CM EH CM H Pi CO CO M W cu cd C O > 3 m i CU CO 6 CO cd co PM HI fM Ml 3 CO oo VO wVO 6>S M • CO CO U O • ✓-V CO oo 5-s • r—1 r-^ CO CM CO CO 1—1w uo o o i—1 <1 w a H Pm <3 Q CO H /—N o o CO vO B>S • • h3 O i—l UO o PC 'w' CM o I—1 2 . '■d- vO w o I—1 525 S5 W CM O W pq i—l < 1 i—l ■ Q) <1) CX cd tO 43 4-1 cx to CO CO 43 k P 3 <3 C 03 H O d O 43 cd H cx o Cfl to cd 43 }M 4-1 o 43 cx U cn M 212 Section A. Host - Parasite List Order - Anguilliformes Suborder - Anguilloidei Family - Nemichthyidae HOST 1 Nemichthys scolopaceus [252/59/23.4%] No. parasite species recovered = 7 Parasites Pseudopoecelus sp. adult (Trematoda, Digenea, Opocoelidae) Incidence - 24/252, 9.5% Location(s) - Free inside lumen of intestine. Host Reaction - none apparent Echinorhynchus sp. adult (Acanthocephala, Echinorhynchidea, Echinorhynchidae) Incidence - 1/252, 0.4% Location(s) - Penetrated intestine from inside. Host Reaction - Worm penetrated through intestinal wall from inside. Large hole in wall. Nybelinia sp. 1 postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 3/252, 1.2% Location(s) - outside stomach wall, inside stomach wall Host Reaction — tissue deformation Phormobothrium sp. (?) plerocercoid (Cestoda, Tetraphyllidea, Phyllobothriidae) Incidence - 6/252, 2.4% Location(s) - Free inside lumen of upper intestine. Host Reaction - Obstruction of intestine by body of worm. Johnstonmawsonia sp. adult (Nematoda, Spirurida, Rhabdochonidae) Incidence - 6/252, 2.4% Location(s) - Free inside lumen of intestine. Host Reaction - none apparent Contracaecum type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 2/252, 0.8% Location(s) - Free in coelom, free inside intestine. Host Reaction - none apparent Tetraphyllidean plerocercoid (Cestoda, Tetraphyllidea) Incidence - 22/252, 8.7% Location(s) - Free inside intestine. Host Reaction - none apparent Previous Studies Collard (1968) examined 1 specimen of Nemichthys from the Pacific Ocean and reported no parasites. 213 Munroe (pers. comm.) also examined 1 specimen captured in Hudson Canyon in the Western North Atlantic and found 1 unidentified nematode. Machida (1975) described a new genus and species of rhabdochonid nematode, Johnstonmawsonoides nemichthyos, from specimens of _N. scolopaceus taken in Suruga Bay near Japan. Family - Serrivomeridae HOST 2 Serrivomer beanii - [24/4/16.7%] No. parasite species recovered = 4 Parasites Anisakis type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 1/24, 4.2% Location(s) - Free in coelom. Host Reaction - none apparent Johnstonmawsonia sp. adult (Nematoda, Spirurida, Rhabdochonidae) Incidence - 1/24, 4.2% Location(s) - Free inside intestine. Host Reaction - none apparent Pseudophyllidean plerocercoid (Cestoda, Pseudophyllidea) Incidence - 1/24, 4.2% Location(s) - Free inside intestine. Host Reaction - none apparent Tetraphyllidean plerocercoid (Cestoda, Tetraphyllidea) Incidence - 1/24, 4.2% Location(s) - Free inside intestine. Host Reaction - none apparent Previous Studies Collard (1968) examined 7 specimens of the related species S^. sector and found 5 to be infected with 3 genera of nematodes. HOST 3 S. brevidentatus [8/0/0.0%] Family - Derichthyidae HOST 4 Nessorhamphus ingolfianus [26/19/73.1%] No. parasite species recovered = 4 214 Parasites Pseudopoecelus sp. adult (Trematoda, Digenea, Opoecoelidae) Incidence - 4/26, 15.4% Location(s) - Free inside colem of intestine. Host Reaction - none apparent Nybelinia sp. 1 postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 13/26, 50.0% Location(s) - Attached to mesentery outside intestine, free inside intestine, encysted on mesentery outside stomach, free in coelom. Host Reaction - tissue deformation Thynnascaris type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 6/26, 23.1% Location(s) - Free in coelom (near rectum). Host Reaction - none apparent Tetraphyllidean plerocercoid (Cestoda, Tetraphllidea) Incidence - 1/26, 3.8% Location(s) - Inside intestine, encysted on mesentery outside intestine. Host Reaction - none apparent Fibrous tumor (cause and type unknown) Incidence - 1/26, 3.8% Location(s) - Encapsulated in wall of coelom. Host Reaction - Tissue deformation, (fibrous encapsulation). Previous Studies None. HOST 5 Derichthys serpentinus [18/16/88.9%] No. parasite species recovered = 3 Parasites Nybelinia sp. 1 postlarva (Cestoda, Trypanorhycha, Tentaculariidae) Incidence - 1/18, 5.6% Location(s) - Encysted on mesentery outside intestine. Host Reaction - tissue deformation Thynnascaris type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 1/18, 5.6% Location(s) - Free in coelom. Host Reaction - none apparent Anisakid nematode larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 1/18, 5.6% Location(s) - Free in coelom. Host Reaction - none apparent 215 Ichthyophonus type fungal spores and hyphae (Fungi, Zygomycetes, Entomophthorales) Incidence - 17/18, 94.4% Location(s) - Encysted in mesentery outside stomach, encysted in mesentery outside intestine, encysted on wall of coelom, in mesentery, encysted in liver. Host Reaction - Although more material is needed to demonstrate pathogenicity, it appears the fungus either directly attacks pancreatic tissue or elicits a host reaction that ultimately destroys the tissue. Previous Studies Beebe (1935a) examined 18 specimens of Derichthys captured near Bermuda and reported a single specimen with "...a number of parasitic worms embedded in the stomach wall." Suborder - Saccopharyngoidei Family - Eurypharyngidae HOST 6 Eurypharynx pelecanoides [16/2/12.5%] No. of parasite species recovered = 1 Parasites Pseudophyllidian adult gen. et sp. n. (Cestoda, Pseudophyllidea, Triaenophoridae) Incidence - 2/16, 12.5% Location(s) - Inside upper intestine attached to intestinal wall. Host Reaction - Blockage of intestine by the coiling of the worm proglottids. Tissue deformation of intestinal wall due to gripping action of bothria. Distension of intestine to several times normal diameter. Previous Studies None. Order - Salmoniformes Suborder - Stomiatoidei Family - Gonostomatidae HOST 7 Gonostoma bathyphilum [18/3/16.7%] No of parasite species recovered = 2 Parasites Lecithophyllum sp.? adult (Trematoda, Digenea, Hemiuridae) Incidence - 1/18, 3.6% Location(s) - Attached outside stomach wall. Host Reaction - none apparent Ichthyophonus type fungal spores (Fungi, Zygomycetes, Entomophthorales) Incidence - 1/18, 5.6% Location(s) - Free inside stomach. Host Reaction - Zoospores encapsulated in fibrous tissue (?)• Cyst (cause and type unknown) Incidence - 1/18, 5.6% Location(s) - Attached to wall of pyloric caeca. Host Reaction - none apparent Previous Studies Orias, Noble, and Alderson (1978) examined 6 specimens of _G. bathyphilum from the Irish Sea and found no parasites. HOST 8 G_. elongatum [48/21/43.8%] No of parasite species recovered Parasites Nybelinia sp. 1 postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 2/48, 4.2% Location(s) - Encysted in mesentery outside stomach. Host Reaction - none apparent Nybelinia sp. (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 3/48, 6.2% Location(s) - Encysted in mesentery outside stomach, encysted in mesentery. Host Reaction - All three specimens were in varying stages of destruction from infiltration of cellular material into the cysts. Anisakis type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 5/48, 10.4% 217 Location(s) - Encysted in mesentery, encysted outside stomach, free in coelom. Host Reaction - none apparent Anisakid larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 1/48, 2.1% Location(s) - Encysted in mesentery. Host Reaction - Fibrous tissue around nematode, infiltration of cellular material into cyst. Ichthyophonus type fungal spores and hyphae (Fungi, Zygomycetes, Entomophthorales) Incidence - 2/48, 4.2% Location(s) - Encysted in wall of coelom, encysted in mesentery outside stomach, encysted in mesentery. Host Reaction - Fibrous tissue around spores and hyphae. Degenerate cell material inside(?). Fibrous tumor (cause and type unknown) Incidence - 1/48, 2.1% Location(s) - Encapsulated in wall of coelom. Host Reaction - Tissue deformation, infil tration of tumor by fibrous material. Cysts [cause(s) and type(s) unknown] Incidence - 10/48, 20.8% Location(s) — Encysted in mesentery, encysted in liver, encysted in mesentery outside stomach Host Reaction - Encapsulation in fibrous tissue, infiltration of dense cellular material. Last stages of host destruction of fungi similar to I), serpentinus? Previous Studies None. Family - Sternoptychidae HOST 9 Argyropelecus aculeatus [15/2/13.3%] No. of parasite species recovered = 1 Parasites Nybelinia sp. 3 postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 1/15, 6.7% Location(s) - Free in coelom. Host Reaction - tissue deformation Cyst (cause and type unknown) Incidence - 1/15, 6.7% Location(s) - Encysted in mesentery. Host Reaction - Encapsulation in fibrous tissue, infiltration of cellular material. 218 Previous Studies Collard (1968) examined 5 specimens of A. lychnus, 12 specimens of A. pacificus and 1 specimen of A. olfersi from the Pacific. He found only 2 ,A. pacif icus to be infected, one with a trematode (Monilocaecum sp.) and the other with a larval acanthocephalan. HOST 10 Sternoptyx diaphana [19/10/52.6%] No. of parasite species recovered = 5 Parasites Nybelinia sp. 2 postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 2/19, 10.5% Location(s) - Free in coelom, encysted outside stomach. Host Reaction - none apparent Tentacularia sp. 1 postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 1/19, 5.3% Location(s) - Free in coelom. Host Reaction - none apparent Tetraphyllidean plerocercoid (Cestoda, Tetraphyllidea) Incidence - 3/19, 15.8% Location(s) — Free and encysted inside intestine, encysted in wall of pyloric caeca. Host Reaction - None apparent in those hosts which only harbored one or two worms. One host had over a hundred encysted larvae in the pyloric caeca and throughout the upper intestine which undoubtedly caused impediment of material passing through the intestine. Terranova type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 1/19, 5.3% Location(s) - Free in coelom. Host Reaction - none apparent Anisakis type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 1/19, 5.3% Location(s) - Free inside stomach. Host Reaction - none apparent Digene (Trematoda, Digenea) Incidence - 1/19, 5.3% Location(s) - Encapsulated in gonads. Host Reaction - Encapsulation by fibrous tissue, in process of being destroyed by infiltration of cellular material. Cyst [cause(s) and type(s) unknown] Incidence - 3/19, 15.8% Location(s) - Encysted outside stomach, encysted in mesentery, encysted inside intestine. Host Reaction - All were encapsulated in fibrous material. Cyst outside stomach partially melanized. Tumor (cause and type unknown) 219 Incidence - 1/19, 5.3% Location(s) - Encapsulated in wall of coelom. Host Reaction - Tissue deformation, infiltration of tumor by fibrous material, fibrous encapsulation. Previous Studies Collard (1968) examined 17 diaphana and reported 2 with parasites. One fish possessed 5 phyllobothriid cestode larvae, while the other possessed hundreds of what Collard believed were proteocephalan larvae embedded in the intestine. Family - Chauliodontidae HOST 11 Chauliodus sloani [36/22/61.1%] No. of parasite species recovered - 7 Parasites Hirudinellid digene (possibly Botulus microporus) immature (Trematoda, Digenea, Hirudinellidae) Incidence - 1/36, 2.8% Location(s) - Free inside upper intestine. Host Reaction - Digene was very large, possibly causing intestinal blockage. Tentacularia sp. 2 postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 3/36, 8.3% Location(s) - Imbedded under external skin, free in coelom. Host Reaction - tissue deformation Nybelinia sp. 4 postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 5/36, 13.9% Location(s) - Free in coelom, encysted in mesentery outside stomach, free inside stomach, encysted in mesentery outside intestine, imbedded under external skin. Host Reaction - tissue deformation Tetraphyllidean plerocercoid (Cestoda, Tetraphyllidea) Incidence - 9/36, 25.0% Location(s) - Encysted in liver, free and encysted inside intestine, encysted outside stomach, free in hepatic duct, encysted in walls of pyloric caeca. Host Reaction - Blockage of hepatic duct by larvae. Pseudophyllidean plerocercoid (Cestoda, Pseudophyllidea) Incidence - 7/36, 19.4% Location(s) - Encysted outside stomach, free inside lumen of intestine. Host Reaction - none apparent 220 Anisakis type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 3/36, 8.3% Location(s) - Free in coelom, free outside stomach. Host Reaction - none apparent Contracaecum type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 1/36, 2.8% Location(s) - Free in coelom. Host Reaction - none apparent Cysts [cause(s) and type(s) unknown] Incidence - 3/36, 8.3% Location(s) - Encysted in liver, encysted outside stomach. Host Reaction - displacement (replacement?) of liver Previous Studies None. Family - Stomiatidae HOST 12 Stomias boa ferox [60/50/83.3%] No. of parasite species recovered = 4 Parasites Nybelinia sp. 5 postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 2/60, 3.3% Location(s) - Encysted within stomach wall, free inside Stomach. Host Reaction - tissue deformation Tentacularia sp. 2 postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 1/60, 1.7% Location(s) - Free in coelom. Host Reaction - tissue deformation Anisakis type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 10/60, 16.7% Location(s) — Free in coelom, encysted in mesentery, encysted in mesentery outside stomach. Host Reaction - none apparent Tetraphyllidean plerocercoid (Cestoda, Tetraphyllidea) Incidence - 45/60, 75.0% Location(s) - Free inside stomach, encysted in stomach wall, free and encysted inside intestine, encysted in hepatic duct, encysted in walls of gall bladder. Host Reaction - Possible blockage of hepatic duct. 221 Previous Studies Collard (1968) examined 15 specimens of the related species S^. atriventer from the Pacific, and reportd a 60.0% infection. From the 9 parasitized specimens he listed the following; 6 specimens with phyllobothriid plerocercoids, 1 with an unidentified cestode larva, and 4 with unknown gill cysts. Family - Malacosteidae HOST 13 Malacosteus niger [10/1/10.0%] No. of parasite species recovered = 0 Parasites Cyst (cause and type unknown) Incidence - 1/10, 10.0% Location(s) - Encysted outside stomach. Host Reaction - Encapsulation in fibrous material, infiltration with dense material. Some melanization. Previous Studies None. HOST 14 Photostomias guernei [5/2/ not applicable due to small sample size] No. of parasite species recovered = 3 Parasites Sarcotretes sp. adult (Copepoda, Lernaeidae, Lernaeenicinae) Incidence - 1/5, not applicable because of small sample size Location(s) - Embedded in left side of jaw. Host Reaction - Entrance wound in jaw. Displace ment of tissue internally. Deep penetrating lesion, probable loss of blood. Rhabdochonid nematode adult (Nematoda, Spirurida, Rhabdochonidae) Incidence - 1/5, not applicable Location(s) - Free inside intestine. Host Reaction - none apparent Tetraphyllidean plerocercoid (Cestoda, Tetraphyllidea) Incidence - 1/5, not applicable Location(s) - Free inside stomach. Host Reacton - None Apparent 222 Previous Studies N o n e . Order - Myctophiformes Family - Myctophidae HOST 15 Benthosema glaciale [12/4/33.3%] No. of parasite species recovered = 3 Parasites Sarcotretes sp. adult (Copepoda, Lernaeidae, Lernaeenicinae) Incidence - 1/12, 8.3% Location(s) - Penetrated body posterior to left pectoral fin, head lying near heart. Host Reaction - Entrance lesion in side. Dis placed tissue internally, with large necrotic area around head of copepod. Anisakid nematode larvae (Nematoda, Ascaridoidea, Anisakidae) Incidence - 1/12, 8.3% Location(s) - Encysted in mesentery. Host Reaction - Worm was heavily melanized. Tetraphyllidean plerocercoid (Cestoda, Tetraphyllidea) Incidence - 2/12, 16.7% Location(s) - Encysted in mesentery outside stomach, encysted in mesentery outside intestine. Host Reaction — none apparent Previous Studies None. HOST 16 Ceratoscopelus maderensis [18/10/55.6%] No. of parasite species recovered = 4 Parasites Trypanorhynch larva (Cestoda, Trypanorhyncha) Incidence - 1/18, 5.6% Location(s) - Encysted in mesentery. Host Reaction - Encapsulated with fibrous material. Tetraphyllidean plerocercoid (Cestoda, Tetraphyllidea) Incidence - 5/18, 27.8% Location(s) - Encysted in wall of pyloric caeca, encysted in mesentery outside stomach. 223 Host Reaction - none apparent Pseudophyllidean plerocercoid (Cestoda, Pseudophyllidea) Incidence - 3/18, 16.7% Location(s) - Encysted in walls of pyloric caeca. Host Reaction - none apparent Anisakis type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 2/18, 11.1% Location(s) - Free under pyloric caeca, encysted in mesentery. Host Reaction - none apparent Cysts [cause(s) and type(s) unknown] Incidence - 2/18, 11.1% Location(s) - Encysted outside intestine, encysted in mesentery. Host Reaction - Encapsulation in fibrous material. Cyst in mesentery melanized. Previous Studies Collard (1968) examined C_» townsendi, the Pacific form of Ceratoscopelus and reported a 33.3% infection level consisting of nematodes (anisakine larvae), cestodes (tetra- and pseudophyllidean plerocercoids) trematodes (hemiurids and fellodistomes) and a copepod. HOST 17 Lampanyctus macdonaldi [14/4/28.6%] No. of parasite species recovered = 2 Parasites Nybelinia sp. postlarva (Cestoda, Trypanorhyncha, Tentaculariidae) Incidence - 1/14, 7.1% Location(s) - Encysted in mesentery outside stomach. Host Reaction - Parasite cyst infiltrated by granular material. Ichthyophonus type fungal spores (Fungi, Zygomycetes, Entomophthorales) Incidence - 2/14, 14.3% Location(s) - Encysted in mesentery, inside pancreatic tissue. Host Reaction - Fungi in pancreatic tissue are spores. Hyphae in mesentery are surrounded by fibrous tissue. Cysts [cause(s) and type(s) unknown] Incidence - 1/14, 7.1% Location(s) - Encysted in liver. Pathology/Host Reaction - Displacement of liver tissues. 224 Order - Beryciformes Suborder - Stephanoberycoidei Family - Melamphaidae HOST 18 Scopelogadus beanii [69/33/47.8%] No. of parasite species recovered = 5 Parasites Steringophorus sp. adult (Trematoda, Digenea, Fellodistomidae) Incidence - 1/69, 1.4% Location(s) - Free inside intestine. Host Reaction - none apparent Lecithophyllum sp. (?) adult (Trematoda, Digenea, Hemiuridae) Incidence - 1/69, 1.4% Location - Free inside stomach. Host Reaction - none apparent Tetraphyllidean plerocercoid (Cestoda, Tetraphyllidea) Incidence - 3/69, 4.3% Location(s) - Encysted in mesentery. Host Reaction - none apparent Contracaecum type larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 1/69, 1.4% Location(s) - Free outside intestine. Host Reaction - none apparent Anisakid nematode larva (Nematoda, Ascaridoidea, Anisakidae) Incidence - 1/69, 1.4% Location(s) - Encysted in mesentery. Host Reaction - Fibrous cyst around worm, infil tration with granular material. Ichthyophonus type fungal spores and hyphae (Fungi, Zygo mycetes, Entomophthorales) Incidence - 22/69, 31.9% Location(s) - Encysted in liver, encysted in mesentery. Host Reaction - Fungi are very pathogenic in liver tissue. Infections range from 1-5 cysts approximately 2-10 mm in diameter. Infections with greater number of cysts have little liver tissue remaining. In worst case of infection, entire visceral cavity was filled with connective tissue and cysts; visceral organs pressed up against spine, marked reduction in body weight of fish for size class. Host response appears to be of the granulomatous inflammatory type. Cysts [cause(s) and type(s) unknown] Incidence - 1/69, 1.4% Location(s) - mesentery Host Reaction - Encapsulation in fibrous material, infiltration with granular material. 225 Previous Studies Collard (1968) examined 10 specimens of mizolepis bispinosus and noted 3 as being parasitized. Two of the specimens possessed hemiurid trematodes, and two had greater than 100 unknown cysts. These cysts were located in the liver (2), heart (1), mesentery (many), peritoneum (many), caeca (many), and wall of colon/rectum (many) of one fish, and the liver (3 cysts) of the other. It is quite possible that these were fungal in nature. Section B. Discussion of Major Parasite Groups Part B of the results is a brief overview of the major parasite groups found within these hosts. Cestoda The cestodes were the most abundant parasites recorded, and were found in all but one family, the Malacosteidae (which may have reflected sample size rather than actual absence). Worm burden ranged from one to over one hundred individuals, and the major infection loci were inside the intestine or pyloric caeca. Three orders were represented, the Trypanorhyncha (= Tetrarhynchidea), Tetraphyllidea, and the Pseudophyllidea in decreasing order of abundance. The presence of trypanorhynchs was surprising as Collard (1968) and Noble and Collard (1970) stated that "...mesopelagic fishes do not serve as either the second intermediate or final hosts of trypanorhynch cestodes.” This is not the case in the present study, since five species of Nybelinia and two species of Tentacularia (species differentiated by tentacular hook morphology) were found in 226 nine species of hosts; infection intensities ranged from one to four individuals per host. All trypanorhynchs found were postlarvae as described by Stunkard (1977), and indicated that the hosts probably served as second intermediate hosts. The Tetraphyllidea were the most abundant order of cestodes (found in the same number of host species as trypanorhynchs, but many more individuals) and frequently formed heavy infestations, especially in the piscivores Chauliodus sloani and Stomias boa ferox. Much of the material was found with the scolex inverted, presumably in response to fixation. Those with the scolex everted appeared to closely resemble the types referred to by Dollfus (1964) as Scolex pleuronectis bilocularis and Scolex polymorphous unilocularis. The plerocercoids with biloculate bothridia appeared identical to the illustrations and descriptions presented by Stunkard (1977) of Ceratobothrium xanthocephalum. Adult Ceratobothrium, parasitic in elasmobranchs, possess a pair of hornlike projections on the posterior edge of each upper loculus, but no such projections were noticed by this author nor mentioned by Stunkard for the plerocercoids he examined. Large tetraphyllids of the family Phyllobothriidae were found in Nemichthys scolopaceus. The structure of the scolex in these specimens is similar to that of Phormobothrium, a parasite of skates (family Rajidae) described by Alexander (1963). The large, flaplike bothridia with numerous ridges, and short muscular myzorhynchus were clearly visible in several specimens. 227 The Pseudophyllidea were the least abundant of the cestode orders found (four host species), but were the only group in which adult forms were recovered. Three adult specimens of a new genus and species of cestode belonging to the family Triaenophoridae were found in the upper intestine of two individuals of Eurypharynx pelecanoides, indicating that this species, at least, is a definitive host for a cestode species (Campbell and Gartner, in prep.). Pseudophyllidean infection intensities were low, usually only one worm per host. Noble and Collard (1970) attributed this to the larger size of pseudophyllids and resulting possible intraspecific competition for space in the host. Trematodes Digenetic trematodes were found in over 50% of the potential host families examined. Although rare in the individuals from Norfolk Canyon (4.9% overall infection), digenes seem to occur more frequently in meso- and bathypelagic fishes than was originally thought, an idea that has gained support in other recent studies (Noble and Orias, 1970, 1975; Orias et al. , 1978). Infrequent occurrence in host individuals may be related to dispersal of the parasite in intermediate hosts throughout the vast three-dimensional midwater environment. The trematodes found belonged to the families Opocoelidae, Hemiuridae, Fellodistomidae, and Hirudinellidae. With the exception of the latter, these families are all primarily parasites found in 228 cold, shallow waters. Geographic submergence of cold waters may be one mechanism by which they have penetrated the deep-sea (Munroe, pers. comm.). Pseudopoecelus sp. was the most common digene found in two host species (both eels), comprising a 4.2% infection of all host individuals examined. Pseudopoecelus species have been reported in deep-sea fishes (benthic forms) by Manter (1934) and Armstrong (1974). The hemiurids, probably Lecithophyllum sp. were found in two bathypelagic hosts, Scopelogadus beanii and Gonostoma bathyphilum (one specimen of each infected). Lecithophyllum irelandeum was described from an unidentified melamphaid host by Orias et al. (1978). These worms were extremely small, and some may have been passed over during dissection of the hosts. A specimen of the fellodistome S teringophorus sp. was also recovered from S^. beanii. (another worn was found in a Scopelogadus specimen which is not included in this study). Infection by this trematode related to the feeding habits of the host. Members of the Hirudinellidae generally are found as adults in large oceanic predators such as tunas, dolphin-fishes, and billfishes (Yamaguti, 1958). The presence of a single immature specimen (possibly Botulus microporus , Overstreet, pers comm.) in Chauliodus sloani may also be related to feeding and will be discussed later. 229 Nematoda Nematodes formed a surprisingly small percentage of the parasite faunas recovered in this study (6 .1 % of individuals examined were infected), when compared to the results of Collard (1968), in his study on Pacific midwater fishes. His data indicated a 19.6% overall level of infection. However, closer inspection of the host species examined revealed that of the 1,122 specimens examined by Collard, 944 were myctophids, most of which were from species that commonly migrate into surface waters at night, and in so doing, may pick up a higher incidence of nematodes than deeper dwelling non-migratory species. Of the 221 specimens infected with nematodes, only 12 were non-myctophid species. In comparison, only 18 of the 6 6 8 individuals examined for this study were migratory myctophids; 3 of these had larval nematodes. Therefore, the 7.0% overall infection level may be a more realistic value for midwater fishes in general. Anisakine nematodes were most abundant forming 4.9% of the total number of infected individuals in 11 species of hosts. The remaining 1.2% (3 host species) was contributed by spiruridan worms of the family Rhabdochonidae. The anisakine worms were of genera commonly reported from marine fishes (Cannon, 1977; Norris and Overstreet, 1975), namely, Anisakis type, Thynnascaris type, Contracaecum type, and Terranova type. Because of the difficulties involved in confirming identification, the worms were only identified to generic group. For a description of these generic groups, see Myers, 1975. Spiruridan adults were found in Nemichthys scolopaceus (6 hosts) 230 Photostomias guernei (1 host), and Serrivomer beanii (1 host). The worms in Nemichthys and Serrivomer were identified as Johnstonmawsonia o p 0 Machida (1978) described Johnstonmawsonoides nemichthyos, a closely related nematode, from Nemichthys scolopaceus. However, several morphological differences, most notably in egg structure and spicule length, indicated that the worms examined for this study did not conform to the description of J. nemichthyos. Fungi Fungi have rarely been reported from deep-sea pelagic fishes, a situation that is more probably due to a lack of intensive investigation rather than actual absence of the organisms. Collard (1969) reported a Lagenidium sp. from two hosts, and Agius (1968) examined Ichthyophonus occurrence in a trichiurid species. In this study, fungal spores and hyphae were recovered from 5 species of fishes, 3 of which were bathypelagic (Lampanyctus macdonaldi, Gonostoma bathyphilum, and Scopelogadus beanii), and 2 mesopelagic (Derichthys serpentinus and Gonostoma elongatum) species. The hosts L. macdonaldi, (3. elongatum, and G_. bathyphilum had low infections (1-2 individuals). However, almost one-third of the Scopelogadus, and nearly 95% of the Derichthys examined possessed fungi in some stage of development. Because of a lack of fruiting bodies in the material examined, or fresh samples for growing in culture, a specific identification of the 231 fungus could not be made. From morphological appearances of the spores (Fig. 1 a-d), hyphae (Fig. 2 a-f) and the nature of the infection and host response (formation of complex granulomas, Fig. 3 a-f), the fungi found in this study were tentatively labelled as Ichthyophonus type. Ichthyophonus, which has also been mistakenly called Ichthyosporidium (a protozoan, see Sprague, 1965 for review of the taxonomic problem), has been found in numerous species of marine and freshwater fishes. Agius (1978) reported the first occurrence of it in a deep-sea fish, the pelagic trichiurid Aphanopus carbo. Ichthyophonus is frequently found in, and often causes high mortality in, a number of commercially important fish species such as the sea herring (Sindermann and Scattergood, 1954), the cod (Hendricks, 1972; Mo He r , 1974), the yellowtail flounder (Ruggieri et al., 1970), and the rainbow trout (Miyazaki and Kubata, 1977a, 1977b, 1977c). The tentative type identification was made from the figures in the above-mentioned studies, as well as general descriptions of the morphology of Ichthyophonus (Ryebilin and Migaki, 1975). The spores pictured in Fig. 1 a-d, although variable in size and shape, closely resemble the description given and figure shown by Ribelin and Migaki (1975) , "range in size from 10 to 250 microns and possess a double wall. The outer wall thickness appears related to spore size, while the inner wall is consistently quite narrow." 232 Figure 1. Ichthyophonus type fungal spores from: (a) Gonostoma bathyphilum- 155mm S.L., CI-73-10, Sta. 95 Spores in stomach, Semichon's Acetocarmine stain, Whole Mount (126x) (b) Scopelogadus beanii- 100mm S.L., GI-74-04, Sta. 85 Spores in liver tissue, Harris* Hematoxylin-Eosin stain, 5y section (63x) (c) Derichthys serpentinus- 266mm T.L., 69-8(Alb. IV), Sta.03 Spores outside stomach, Semichon*s Acetocarmine stain, Whole Mount (63x) (d) Derichthys serpentinus- 239mm T.L., GI-76-01,Sta. 32 Spores outside intestine (pancreas?), Harris' Hematoxylin-Eosin stain, 5p section (lOOx) 234 Figure 2. Ichthyophonus type fungal hyphae from: (a) Scopelogadus beanii- 111mm S.L., GI-76-01, Sta. 57 Hyphae in liver cyst, McManus* PAS stain, 5 section (126x) (b) S^. beanii— 111mm S.L., GI-76-01, Sta. 57 Hyphae in liver cyst, McManus' PAS stain, 5 section (126x) (c) S^. beanii- 104mm S.L., GI-74-04, Sta. 74 Hyphae in liver cyst, McManus' PAS stain 5 section (126x) 236 Figure 2. (cont’d) Ichthyophonus type fungal hyphae from: (d) Scopelogadus beanii- 96mm S.L., G 1-76-01, Sta. 57 Hyphae from cyst in mesentery outside intestine, Semichon's Acetocarmine stain, Whole Mount (25x) (e) Gonostoma elongatum- 105mm S.L., GI-76-01, Sta. 34 Hyphae from cyst in mesentery, Semichon’s Acetocarmine stain, WhoJLe Mount (63x) (f) Scopelogadus beanii- 96mm S.L., GI-76-01, Sta. 57 Closeup of Fig. 2(d), (lOOx) 238 Figure 3. Ichthyophonus type fungal infections from Scopelogadus beanii (All photographs are of liver cysts, 5y sections) (a) 100mm S.L., GI—74-04, Sta. 85 Harris1 Hematoxylin-Eosin stain, (63x) (b) Same as Fig. 3(a) (c) Same as Fig. 3(a) 240 Figure 3. (cont'd) Ichthyophonus type fungal infections from Scopelogadus beanii (d) 102mm S.L., GI-74-04, Sta. 84 McManus PAS stain, (63x) (e) 111mm S.L., GI-76-01, Sta. 57 McManus PAS stain, (16.25x) (f) 104mm S.L., GI-74-04, Sta. 96 McManus PAS stain, (16.25x) M 9.kmM(WW?- tipi >c. > • ,, <* '?* ‘ ,^/tH9hH! 242 Hyphal structure also agreed well with published information. Sinderman and Scattergood (1954) stated "Germination of spores and hyphal bodies...produces characteristic pseudopodium-like hyphae usually with a length not in excess of the parent spore, and never more than 3 or 4 times this diameter." This is clearly seen in the hyphae shown in Fig. 2 a-c, while Fig. 2 d-f show the large number of hyphae present in whole mounts of the cyst material. The granulomatous nature of the infection and host response is indicated if Fig. 3 a-f. Fig. 3a shows a closeup of a section through the liver of Scopelogadus beanii. In the figure, "L" denotes normal liver tissue seen in the left of the photograph. Centrally, "F" indicates a layer of fibrotic tissue which forms the cyst wall. Epithelioid cells ("E") are seen densely concentrated and encapsulated within the wall. This host reaction, termed a focalized granulomatous response (Sypek, pers. comm.) is seen more clearly in Fig. 3b, which shows the foreign material to be a fungal spore. Leukocytic infiltration to the area of the cyst also occurred, in addition to the laying down of fibrotic tissue and infiltration by epithelioid cells. This is shown in Fig. 3c where the leukocytes ("LK") lie adjacent to the fibrotic tissue. Within the cyst where hyphae are present, reduced numbers of epithelioid cells occur (Fig. 3 c-f). This may be due to one of several factors: a more diffuse granulomatous response (Sypek, pers. comm.); successful destruction of the foreign material by the host, 243 and consequent lessening of epithelioid cell infiltration; destruction of the epithelioid cell infiltrate by the hyphae within the cyst. This last possibility may be the case, as in Fig. 3f, hyphae are noted (arrows) penetrating the cyst wall. In addition, as the fungus seems to be a slowly spreading pathogen (see Discussion), the presence of very large cysts (10-20 mm diameter) in severe infection cases indicates inability of the host to successfully combat infection. Apparently, In Scopelogadus, although culture work would need to be done in order to prove it, the hyphae are capable of destroying the epithelioid cells, as these advanced cysts appear highly necrotic (Fig. 3 e-f). In the larger cysts, the hyphae were clustered around a granular substance, which, upon treatment with hydrochloric acid, proved to be calcium carbonate. Calcification occurs in many vertebrates when necrotic tissue is unable to heal (Walter and Israel, 1970) and can be a terminal response to focal necrosis (Zwerner, pers. comm.). In Derichthys, host response appears somewhat different, and more effective. A thick, opaque substance appeared to infiltrate the cysts, and in many cases obliterate the hyphae. Until more material is obtained and sectioned, this cannot be explained further. The pathogenic nature of Ichthyophonus, especially in relation to Scopelogadus beanii will be dealt with more fully in the discussion. 244 External morphology of the cysts also had the typical appearance indicated in previously cited wprks. Cysts were generally white or brownish, and were variable in both number and size (probably a factor of severity and longevity of infection). In Scopelogadus, cysts ranged from about 3-20 mm in diameter, and usually were clearly defined with respect to one another, while in Derichthys, clusters of small (1-3 mm) nodules were found. This size discrepancy may be due to the amount of tissue available to the fungus. The primary target organ in Scopelogadus was the liver, while in Derichthys apparently it was the much smaller pancreatic tissues. Figure 4 a-d shows the various stages of infection and cyst size in Scopelogadus, while Fig. 4e shows the mycelium, or hyphal masses, excised from two cysts (within each of these masses, but indistinguishable in the photograph are calcium carbonate granules) . Moller (1974) stated that different host species had different organs that were the major target sites for Ichthyophonus. As indicated previously, the target site in Scopelogadus was the liver, while in Derichthys, it may have been the pancreas, although more work needs to be done to confirm this. Copepoda Parasitic copepods were found on two hosts, Benthosema glaciale, and Phofcosiomias guernei. Both were adult females of the genus Sarcotretes. 245 Figure 4 Ichthyophonus type fungal infections from Scopelogadus beanii (All photographs are of preserved liver tissue and encysted fungi, or of excised hyphal mass) (a) 107mm S.L. R/V Knorr 71, Sta. 65, Net 1 (MOCNESS net) Van Cleave's Hematoxylin stain applied to cysts to provide contrast (30x) (b) 108mm S.L., GI-75-08, Sta. 35, Van Cleave's Hematoxylin stained applied to cysts to provide contrast (30x) 247 Figure 4 (cont’d). Ichthyophonus type fungal infections from Scopelogadus beanii (c) 97mm S.L., CI-73-10, Sta. 56 No stain (30x) (d) 100mm S.L., GI-74-04, Sta. 83 No stain (30x) (e) Same as above. Hyphal masses excised from two cysts pictured in Fig. 4(d) 249 Gopepods appeared to be a minor constituent of the parasite faunas, probably because of the extreme difficulties in location of the widely dispersed potential hosts. A canthocephala A single adult female, Echinorhynchus sp. was located in the intestine of Nemichthys scolopaceus. This group also constituted a very minor percentage of the parasites found. Unidentified Cysts These cysts probably comprise a variety of things, such as infective larvae which entered the wrong host, strong host response which overcame and destroyed an invading organism, scar tissue from earlier infections, etc. Although a number of these were found in various hosts, they probably had little effect on the hosts. DISCUSSION Comparison of Meso- vs Bathypelagic Levels of Parasitism The parasites collected from the fish specimens in this study were primarily those which required intermediate hosts in their life cycles. Thus, an analysis of the level of parasitism in meso- and bathypelagic fishes must take the their diet into account, i.e. the. presence and availabity of parasite-carrying prey (Noble, 1973). It has been well documented that biomass decreases with depth in the pelagic environment (see Marshall, 1954, 1971 for reviews of literature). As a result, prey availability becomes lower as one descends from the epipelagic to the bathypelagic zone. This decrease in food becomes especially marked in areas of low surface productivity such as the central ocean gyres. The mesopelagial, as with most of its parameters, is intermediate in biomass (prey availability levels), between the epi- and bathypelagic zones. However, a number of mesopelagic fishes possess several additional sources of energy that are not available to bathypelagic predators. These include diel vertical migrations into food-rich epipelagic or near-epipelagic waters for the purpose of utilizing higher prey concentrations (Hopkins and Baird, 1973, Merrett and Roe, 1974; Kinzer, 1977; Ozawa et al., 1977; Clarke, 1978; Worner, 1979, and others). In addition, schooling of many mesopelagic species, especially myctophids (e.g. Ceratoscopelus maderensis, Backus et al., 1968) provide relatively large concentrations of prey for 250 251 mesopelagic (and other) piscivores such as Chauliodus or Stomias (see Feeding Habits section of this study). A third source may possibly be found around the density layer formed by the permanent thermocline. The "rain" of organic material, primarily dead plankton from the surface layers, tends to be stopped or slowed in its descent as it reaches the density gradient formed by the rapid decrease in temperature. Numerous animals are thought to congregate in waters around and below this food-rich layer. Another organically rich source that may draw mesopelagic fishes is the "nepheloid layer,” a layer that occurs near the bottom over the continental slope in the Western North Atlantic. Presence of zooplankters in this layer may draw midwater fishes whose habitat range encompasses depths where the layer may be encountered. The bathypelagic zone, in contrast, has no additional energy sources, and its inhabitants are mostly non-migratory and lethargic (Baird, 1971; Marshall, 1971), and appear to be non-schooling. The environment is uniformly dark and cold (Marshall, 1954) and presents no density barriers to trap material. Many of the types of prey items found in the deep ocean have been found to act as intermediate hosts for various parasites. Copepods, chaetognaths (Noble, 1973; Noble and Orias, 1970; Reimer et al., 1975), cephalopods (Dollfus, 1964; Overstreet and Hochberg, 1975; Stunkard, 1977), jelly plankters (Dollfus, 1964; Reimer, 1976), and euphausiids (Smith, 1971) have all been cited as first intermediate hosts for major parasite groups. 252 Kennedy (1975) stated that "...distribution of the parasite population throughout that of its host is not random but...majority of parasites are found in relatively few host individuals." Therefore, the less abundant a possible intermediate host is, the lower the probability of a predator feeding on one of the infected intermediate hosts. This is especially true of the sparsely populated bathypelagic environment. Thus, levels of parasitism should decrease with depth, a trend which is indicated in this study and others (Collard, 1970; Noble and Collard, 1970; Noble, 1973). Table 3 shows infection levels both in individual host species and overall between meso- and bathypelagic host groups. For the purposes of the following discussion, the "upper," "upper/lower," and "lower" mesopelagic species in that table are grouped together. The two "transitional" species, are excluded, because of small sample sizes and a lack of knowledge regarding their habits. Bathypelagic species exhibited higher than expected parasite infections (1 out of 3 specimens infected), however mesopelagic fishes showed lower percentages than the studies previously mentioned. Upon examination of individual host species, anomalously low infection levels were noted in the mesopelagic species Nemichthys scolopaceus and Argyropelecus aculeatus, whereas the bathypelagic Scopelogadus beanii exhibited a conspicuously high percentage of infected individuals. Possible reasons for these anomalies are discussed later. PERCENTAGE OF INFECTION IN HOST SPECIES BY DEPTH ZONES. to O P CM C3 O P Ph W w W PSJ X o P3 PC W h h P OQ <3 <3 h D C <3 M o X P O M <3 O pp >- P O M w O o o H O w S P C O 3 P H i-3 C/0 h-f H W S CP § P W 525 p P H 4 h h 2 h h i —' '— - S'S B-S '— ' y- —' '— '— r-> CNI LO S-S ■K i \ ✓— O L —V /— /"N CO S'S t o l S'S T3 LO LO CN o o •H "0 O O O O o —1 r— H * H • « ol w s 525 1 — -H - H • 4-J d - P P 03 CN *H LO 4-1 P N P P B p a CL 253 254 Removal of these species and recalculation of overall infections alter the percentages to values that closely agree with other results (Collard, 1970), with 63.7% parasitism in mesopelagic species vs 15.0% bathypelagic fishes parasitized (Table 4). A comparison of the incidence of infection within meso- and bathypelagic host groups by parasite group (Table 4) indicated results quite different, however, than the findings of Collard (1968, 1970) and suggestions made by Noble and Orias (1975) and Orias et al. (1978). Collard (1970) examined depth-related differences in the parasite-mix of 953 mesopelagic and 134 bathypelagic fishes. He indicated that parasite group comparisons showed higher incidences of larval cestodes and digenetic trematodes in bathypelagic fishes, while nematode incidence was greater in the mesopelagic specimens. He theorized that this was due to the largely piscivorous nature of bathypelagic fishes, which fed on mesopelagic fishes, thereby acquiring their infections. Nematodes, however, were thought not to occur the prey-to-predator host transfer and so were more abundant in mesopelagic species. The disagreement in findings between this study and Collard's results may be explained in several ways. The simplest explanation is that geographic location of the host affects the parasite types that infect it (Noble, 1973). Another explanation may lie in an examination of diet. Of the 6 species of bathypelagic predators (149 255 TABLE 4 PARASITE GROUP INFECTION LEVELS IN MESO- AND BATHYPELAGIC SPECIES (TRANSITIONAL SPECIES EXCLUDED) PARASITE MESOPELAGIC BATHYPELAGIC GROUP NO. % NO. % Cestoda 143 28.4 8 5.4 Trematoda, Digenea 30 6.0 3 1.3 Nematoda 36 7.1 4 2.7 Fungi ■19 3.8 25 16.8 Copepoda 1 0.2 0 0.0 Acanthocephala 1 0.2 0 0.0 Unidentified Cysts 19 3.8 3 2.0 Unidentified Tumors 3 0.6 0 0.0 Total No. Examined 504 149 50.0 28 .9 Total No. Parasitized 252 43 Totals minus anomalous species (N. sco lopaceus , A. aculeatus- Meso-; Scop . beanii- Bathypelagic) Total No. Examined 237 80 63.7 15 .0 Total No. Parasitized 151 12 256 specimens) examined in the feeding habits section of this study, only 8 individuals in 3 species (5.4%) had fish or fish remains in their stomachs. This figure might have been somewhat higher, but Gonostoma bathyphilum appeared to actively regurgitate its stomach contents. In at least two species, Lampanyctus macdonaldi and Scopelogadus beanii, crustaceans (amphipods) apparently were the preferred food item. Serrivomer beanii had fish in 4 out of 14 stomachs with food, and whereas only crustacean remains were found in Eurypharynx pelecanoides and S^. brevidentatus, other workers have reported fishes in the stomachs of these species (Bohlke, 1966; Beebe and Crane, 1936). Thus, Collard*s claim that bathypelagic species are generally piscivorous would appear to be incorrect. In fact, some of his data also showed this. In earlier work (1968) Collard examined stomach contents of 42 specimens of what he termed bathypelagic fishes, which were incorporated into the 1970 study. Of the 38 fishes which had food in the stomach, 28 had unidentifiable debris, 7.5 had crustacean remains, and only 2.5 (8.9%) had fish material. From this data, it is difficult to see how he generalized bathypelagic fishes as piscivorous. In the present study only one parasitic group, the fungi, showed a markedly higher percentage of infection in the bathypelagic vs mesopelagic fishes. The three major parasite groups (cestodes, digenetic trematodes, and nematodes) all occurred in much higher percentages in mesopelagic fishes. Copepods and acanthocephalans were of sufficient rarity as to be discounted. The incidence of fungi was 257 high (16.1%) in the bathypelagic species because of a high level of infection in one species, Scopelogadus beanii, which will be discussed later. Effects of Parasitism on Hosts Noble and Collard (1970) have indicated that parasites in deep-sea fishes are rarely pathogenic on a macropcopically observable scale. Noble and Collard's examination of various midwater fish species and their parasite-mix found only two protozoan species which caused observable damage to their hosts. McCormick, Laurs, and McCauley (1967) described epizoic hydroids on several species of myctophids taken off Oregon, and noted that some damage was inflicted on the host epidermis by the parasites. Agius (1978) noted the presence of Ichthyophonus type fungal infections in a species of trichiurid and suggested a possible link between its presence and the absence of small specimens of the fish in catches of recent years. Results obtained in this study show one definite and one probable case of pathogenicity. In addition, several possibly detrimental host- parasite interactions were observed. The Ichthyophonus type fungal infection in Scopelogadus beanii was found to be pathogenic. Fishes with infected livers were found to have less functional liver tissue than those with apparently normal livers. Table 5 shows this phenomenon quite clearly. As previously 258 mentioned, the fungus apparently actively destroys the liver in Scopelogadus by necrosis of the tissue, a process which has been observed previously with respect to this disease (Ruggieri et al., 1970). The liver in fishes "can be considered to occupy a central position in the metabolism of the organism" (Harder, 1975). It is the central organ in a wide variety of functions. Emulsification of fat, detoxification of materials into excretable forms, and metabolism and storage of carbohydrate materials and fats are all primary metabolic operations carried out by the liver. Damage of the liver or liver tissue must have severe consequences on the affected host(s), especially when the host is an inhabitant of an energy-depauperate environment, where unwanted drain on energy stores or prevention of maximal energy uptake must have a particularly pronounced effect. From these inferences, it was thought that debilitating effects of the pathogen might be manifested in a reduction of host weight. Accordingly, a comparison of mean weights of specimens with no apparent liver parasite with those of specimens with infected livers was made (Table 6). A total of 43 specimen weights was available, of which 15 weights were from infected individuals. In all size range categories, the infected hosts showed lower body weights. A non-parametric statistical test, the Smirnov test (or Kolmogorov-Smirnov two sample test) was performed to determine if the 259 o oo m lo v© o oo LO O O rHO on O I I I I I I I I I I I EH EH CO SC 00 v© m in - O OO O O O r"» inoiotocntnroin'Jtnoo O O n O O O O n ooooooooooo pH rH i—I 1— I r—I l—f I—t i—) I—! rH rH 260 TABLE 6 COMPARISON OF WEIGHTS OF SCOPELOGADUS BEANII SPECIMENS NORMAL VS. INFECTED (ICHTHYOPHONUS TYPE FUNGUS) INDIVIDUALS. SIZE NO~! n o ; X NORMAL x INFECTED RANGE, (nun)______EXAMINED PARASITIZED WEIGHTS (g) WEIGHTS (g) 81- 90 4 1 9.9 8.6 91-100 10 3 15.9 14.9 101-110 26 10 20.9 17.7 111-120 3 1 21.4 19.5 261 weight differences were statistically significant. In all categories, the differences were not significant (except in the 101-110 mm category at p = 0.80 level). This lack of statistically significant differences can be attributed to variability in the data; different intensities and degrees of infection, whether stomachs had food or were empty, external damage to host (lack of skin, scales, etc.) resulting from capture. Thus, the results noted, although not statistically significant , may well show a biologically Important trend. In vitro studies of Ichthyophonus show that growth occurs over a range of temperatures (3°-20°C) with an optimum at 10°C (Sindermann and Scattergood, 1954). Since the habitat temperature of J3. beanii around Norfolk Canyon is ca. 4°C and below, the fungus is in the lowest range of its growth temperature and so would probably grow rather slowly. An examination of Table 7, which presents the incidence of infected individuals over the size range of Scopelogadus, shows that there may be some correlation between the habitat temperature of S. beanii and fungal growth rate. Scopelogadus livers appear to remain normal until the fishes reach a length of about 85 mm, which is approximately the size at sexual maturity for the species (Ebeling and Weed, 1963), at which point macroscopic manifestations of the disease appear. Sectioning of the smaller, apparently uninfected livers may reveal quiescent spores, but this was not done in the present study. 262 TABLE 7 DISTRIBUTION OF FUNGUS FOR SIZE RANGE OF SCOPELOGADUS BEANII EXAMINED SPECIMEN LENGTH NUMBER NUMBER % (STANDARD LENGTH IN mm) EXAMINED INFECTED INFECTED 41- 50 1 0 0.0 51- 60 3 0 0.0 61- 70 4 0 0.0 71- 80 4 0 0.0 81- 90 6 2 33.3 91-100 15 6 40.0 101-110 32 12 37.5 111-120 _4 2 50.0 Totals 69 22 31.2 263 Although the appearance of the disease appears concomitant with attainment of sexual maturity, Ruggieri et al. (1970) noted that the reproductive system is rarely attacked directly by the fungus, and so reproductive capability may be unaffected. Hormonal changes or a lowering of host resistance because of the energy demanded for gonadal development may cause an outbreak of the disease when maturity is attained. The largest infected Scopelogadus specimens were still ripe, indicating that the infection did not affect maturation of gonads. Despite this, the fish may die before spawning if the disease has progressed far enough due to the dual energy drain of liver damage and egg or sperm production. A probable case of pathogenicity can be deduced from the Ichthyophonus-type fungal infection in Derichthys serpentinus. About 95% of the specimens examined were infected in the region believed to be pancreatic tissue. The pancreas is an important organ, producing digestive enzymes and insulin. A lack of production due to pancreas impairment or destruction would also debilitate affected fish. Pancreatic tissue is diffusely distributed along the mesenteries (Harder, 1975) which would make it more difficult for a slowly spreading pathogen to impair pancreatic function significantly. Host reaction was marked and vigorous with many of the hyphae apparently destroyed by infiltration of material (epithelioid cells?). Although hyphae were destroyed, so too was pancreatic tissue destroyed. This suggests the infection to be pathogenic in Derichthys. More sectioned material is required before pathogenicity is verified. 264 Other possible detrimental effects of parasitic infections were suggested. Tissue deformation and damage by larval trypanorhynchs may lead to secondary infection by bacteria or fungi. Occlusion of the lumen of the intestine by the bodies of the Phormobothrium plerocercoids and Pseudophyllidean adults may have occurred. In both cases the large bodies of the worms were coiled such that they almost completely blocked the intestine. This could have resulted from contraction during fixation. If it occurred in nature, then passage of food through the intestine was probably severely impeded. In addition, the presence of the adult cestodes in the intestine of Eurypharynx caused distension of the intestine to several times its normal diameter. Mechanical damage by the heads and necks of the parasitic copepods was quite obvious. Blood drain on the hosts (63 mm Benthosema glaciale, 97 mm Photostomias guernei) may have had detrimental effects on various functions of the hosts (e.g. vertical migration in Benthosema). In addition, secondary infection at the wound site also was possible. Damage to the intestine of Nemichthys by Echinorhynchus sp. was severe in that a large diameter (>1 mm) hole was bored through the intestinal wall. Infection by leakage of toxic fecal materials and/or secondary bacterial invasion are possible results of such a wound. These additional parasitic effects were, mainly those observed in a few individuals of a species rather than any large percentage of the population. Nevertheless, these effects may act to remove a viable 265 member of the population, which, in a large three dimensional, energy impoverished environment, may be an important reproductive element of a population. Diet, Feeding Habits and Parasitism Nemichthys scolopaceus, Argyropelecus aculeatus, and Scopelogadus beanii showed levels of parasitism that were anomalous for their respective host group habitats. These discrepancies may be linked to feeding habits. Nemichthys scolopaceus possessed a lower percentage (23.4%) of individuals infected than did most other mesopelagic fish species examined (Table 3). This could be the result of its stenophagic habits. Of 114 stomachs with food, 10 had euphausiid prey remains, while the rest contained either Sergestes arcticus, sergestid shrimp, or unidentifiable crustacean remains (possibly sergestid material). Because of the skewed population distribution of parasites, i.e., large numbers of parasites infecting few hosts (Kennedy, 1975), fish that prey on a single food item or narrow range of prey types may ingest fewer parasites (Collard, 1970) than those that prey on a wide variety of prey items, as is the case with most of the other mesopelagic hosts sampled. This type of trend may be true for invertebrate predators. Piscivores, however, may acquire a higher incidence of parasites in the manner suggested by Collard (1970), i.e. transfer of parasites 266 from fish prey to the predator. Chauliodus sloani and Stomias boa ferox both appear to be totally piscivorous in the region of Norfolk Canyon. Both species also show high percentages of infection (Table 3), especially with larval cestodes. Argyropelecus aculeatus presents a perplexing situation in that the species apparently feeds on a wide variety of prey (Merrett and Roe, 1974, Feeding Habits section, this study), and is a vertical migrator, entering the surface waters at night (Schultz, 1964; Baird, 1971; personal observation). As a result, if assumptions on parasite incidence are valid, the parasite burden in this species should be relatively high; comparable to other mesopelagic species. Instead, the infection level was very low (13.3%). Two explanations may account for this variation from expected levels of infection. The first is that the sample size was small (15 specimens examined). However, a number of other mesopelagic species had comparable numbers of specimens examined, yet still showed typically high percentages. The other is that Argyropelecus, because its feeding and migratory habits possibly expose it to a high probability of infection, may have evolved an effective host response or immunological system. Collard (1968) examined 18 Argyropelecus specimens from 3 species and found only 2 specimens from 1 species to be infected, which may lend some support to this theory, although extensive studies would have to be carried out before a definite statement could be made. A much higher percentage of Scopelogadus beanii were infected with parasites in comparison with other bathypelagic species. The 267 majority of these individuals were infected with an Ichthyophonus type fungus (see Results section on Fungi). This type of fungus apparently infects the host via ingestion of spores or infected material (Sindermann and Scattergood, 1954; Agius, 1978) although whether the spores are free in the environment or within prey items is unclear. Transmission of the fungus via the prey of the host does not seem to be a viable probability. Scopelogadus primarily feeds on hyperiid amphipods of the genera Vibilia, Phronima, and Parathemisto (see section on Feeding Habits). Two other host species (Gonostoma elongatum, Lampanyctus macdonaldi) besides S. beanii, of the five found to be infected with the fungus, also feed on hyperiids of the genus Parathemisto. A problem with this explanation is that _L. macdonaldi fed more heavily on Parathemisto, yet had a relatively light infection. In addition, an extremely high proportion of the Derichthys serpentinus sampled were afflicted with the fungus, but did not appear to feed on any hyperiids. A possible way in which fungal material may be ingested is near-bottom interaction, through which pelagic fishes swimming near the bottom (in this case, the continental slope) might come in contact with a higher percentage of infected or infectious material than is present in the pelagial, and so acquire a greater level of infection. A large number of the reported hosts for Ichthyophonus are benthic or demersal. This may be the result of dead or moribund carrier organisms falling to the bottom and being consumed. Spores may perhaps be liberated during decomposition of infected material and 268 concentrate in the "nepheloid" layer and be available for consumption by the swallowing action of fishes swimming near the bottom. Of the 5 species of fishes found to be parasitized with the fungus, two, L^. macdonaldi (Craddock, pers. comm.) and Gonostoma bathyphilum (Musick, Sulak, pers. comm.), have been observed quite near to, and even In contact with the bottom. In G. bathyphilum a cluster of Ichthyophanus type spores was collected from inside the stomach. Derichthys serpentinus is a heavy bodied eel, similar to the closely related species Nessorhamphus ingolfianus, which was implicated by Sedberry (1975) as probably occurring near the bottom. I), serpentinus and N_. ingolf ianus were often captured together by the otter trawl, which may indicate that Derichthys also may spend time near the bottom. Marshall and Merrett (1975), noted that the vertical migrations of midwater fishes often bring them near or into contact with the bottom, as a result, a strong migrator such as Gonostoma elongatum could at times approach the bottom and be in position to pick up fungal spores. Evidence for near-bottom activities of Scopelogadus is somewhat better documented. The presence of benthic food items (see section on Feeding Habits) in several stomachs indicates contact with the bottom. In addition, several individuals were found to be infected with Steringophorus sp., a fellodistomid digenetic trematode (although only 269 one individual of S^. beanii examined for this study was infected) . Known fellodistomid life histories usually have clam and ophiuroid intermediate hosts. Markle and Wenner (1979) used the presence of fellodistomes to indicate periodic near-bottom aggregations of the mesopelagic zoarcid Melanostigma atlanticum. The presence of Steringophorus sp. in Scopelogadus may also point to near-bottom presence of individuals. Incidence of parasitism in fish hosts may depend upon prey consumed and/or parasite host-specificity. This can be seen by examination of the presence of the different trypanorhynch cestodes in different hosts. Five species of Nybelinia and two of Tentacularia were recovered from eight species of hosts. Several species showed host overlap which possibly can be traced to host diet. Bathypelagic predators had no trypanorhynchs• Because several species fed heavily on hyperiid amphipods and jelly plankton, this may indicate that hyperiids and jelly plankters do not serve as carriers of trypanorhynch larvae. Nybelinia sp. 1 showed the greatest overlap among hosts, being found in 4 species; Gonostoma elongatum, Derichthys serpentinus, Nessorhamphus ingolfianus, and Nemichthys scolopaceus. In all four species, euphausiids were found as prey items, being common in Nessorhamphus and Gonostoma, in about 10% of Nemichthys examined, and rare in Derichthys, a situation which was reflected in parasite incidence. The cestode occurred in 13 (50.0%) specimens of Nessorhamphus, 5 (10.4%) of Gonostoma, 3 (1.2%) of Nemichthys, and 1 270 (5.6%) of Derichthys. Apparently, the preferred intermediate hosts may be pelagic eels, because 3 of the 5 cestodes in Gonostoma were in various stages of destruction from host reaction. Sternoptyx diaphana was infected with 2 species of trypanorhynchs, Nybelinia sp. 2 and Tentacularia sp. 1. No link between their presence and Sternoptyx diet has been discerned. Sternoptyx primarily fed on hyperiid amphipods, which have been discounted as intermediate hosts. Ostracods may be the carrier, but until more specimens are examined, no definite statement may be made. Nybelinia sp. 3, 4, and 5 were each found in separate hosts, Argyropelecus aculeatus, Chauliodus sloani, and Stomias boa ferox, respectively. Nybelinia sp. 3 was found in a specimen of Argyropelecus that had ingested a cephalopod. Cephalopods are common intermediate hosts for Nybelinia larvae (Dollfus, 1964, Stunkard, 1977). The presence of two different species of Nybelinia in Chauliodus and Stomias is perplexing, especially because both fish fed on virtually the same prey (see Feeding Habits section). The only explanation is that each species of cestode was specific for a host type. The two predators had the same species of Tentacularia (sp. 2) infecting them. Nybelinia sp. 4 and 5, and Tentacularia sp. 2 both apparently pass through fishes as larvae and are picked by the piscivorous hosts. All the trypanorhynchs belong to a single family, the Tentaculariidae. Tentaculariids are adult in elasmobranchs 211 (Schmidt, 1970) which may indicate the importance of midwater fishes in the diet of elasmobranch fishes. Sedberry (1975) noted the presence of a myctophid in the stomach of the deep-sea squalid CentroscyIlium fabricii, while examination of the stomachs of the deep-sea shark Deania calceus (Crabtree, pers. comm.; personal observation) revealed several with midwater fish remains. Another possible indication of midwater fish importance as intermediate hosts of parasites in feeding chains may be indicated by the presence of a large hirudinellid trematode, possibly Botulus microporus (Overstreet, pers. comm.), in the intestine of Chauliodus sloani. These worms grow to maturity in such fishes as tuna, (Yamaguti, 1958), which indicates that midwater fishes, especially the vertically migrating forms may serve as prey items and also as carriers of parasites for predators in the epipelagic zone. CONCLUSIONS Certain components of the dominant midwater fish community living near Norfolk Canyon apparently occur regularly near bottom. This hypothesis is supported by trawl data food habits, and internal parasite data and by published accounts. In an oceanic environment, the water column overlies many submerged topographic features such as seamounts, or as in the present study, the Continental Slope and Rise. A pelagic organism occurring near this submerged land mass, whether it migrates vertically or maintains a restricted habitat range (i.e. non-migratory) may approach this bottom interface (Marshall and Merrett, 1977), either at the lower limits of the animals1 vertical migratory range or by approaching the point of intersection between the bottom and the water column of a non-migrators’ depth range. One suggested biological implication of the near-bottom presence of pelagic animals is that an increased food source is available to benthic or demersal predators. Pereyra et al. (1969) found that, off Oregon, mesopelagic animals vertically migrating to the surface at night were transported shorewards by current action. In the shoaler areas, the downward vertical migration during the day brought the midwater animals near the bottom where they were preyed upon extensively by demersal predators. Sedberry and Musick (1978) found that several species of demersal fishes of the upper, middle, and lower slope around Norfolk Canyon primarily consumed pelagic prey. 272 273 They felt that this was due to increased presence of the prey near the bottom. In predators of the lower slope and rise, pelagic items were of less importance in the diet, because the depths that the predators occurred in were below the deepest depths of most prey migrations. However, it was postulated that near-bottom zooplankton and micronekton concentrations occurred at these lower depths, because suspension of organic particulates in the near-bottom nepheloid layer was believed to provide rich nutrient sources for these plankters. Near-bottom plankton concentrations may also represent an enhanced prey source for midwater fishes. At the deepest levels of downward migrations in species that vertically migrate, and especially for non-migratory species (e.g. Gonostoma bathyphilum, Derichthys serpentinus) this concentration of potential prey near bottom may provide a far richer food supply than open midwaters. Submersible observers have commented on the frequent presence of midwater fishes and various zooplankton near bottom, often within a distance of several meters (Musick, Sulak, Craddock, pers. comms.). Isaacs and Schwartzlose (1965) noted from sonic traces the interaction of pelagic zooplankton of the Deep Scattering Layer (DSL) with the sea floor where the bottom was shallower than the deepest depth of migration. Interactions of the DSL with bottom features were noted on Precision Echo Sounding Recorder (PESR) traces during early sampling cruises (R/V Eastward) in the Norfolk Canyon area (pers. o b s .). 274 Feeding habits analysis of 18 dominant species of midwater fishes examined in the present study showed that only a few prey species were consistently ingested. One species of fish (Ceratoscopelus), one decapod (Sergestes), three euphausiid species (Nematoscelis, Meganyctiphanes, and Euphausia), three hyperiid amphipods (Vibilia, Paratemisto, and Phronima), and unidentified gelatinous zooplankton remains (primarily hydromedusae) constituted about 80% of all the stomach contents found. Most of these prey items have been found to often occur near bottom in appreciable numbers. Sedberry and Musick (1978) found Ceratoscopelus maderensis, Meganyctiphanes norvegica , Sergestes arcticus, and hyperiid amphipods in the stomachs of several species of demersal predators captured around Norfolk Canyon. Crabtree (pers. comm.) found S_. arcticus in a number of stomachs of the benthopelagic macrourid Nezumia bairdii from Norfolk Canyon collections. McEachran (1976) identified the only species of euphausiid in the stomachs of four species of skates (family Rajidae) as Meganyctiphanes norvegica. Marshall and Merrett (1977) noted that Eupahusia krohnii and Nematoscelis megalops extended close to the bottom off northwest Africa. Golovan and Pakhorukov (197 5) indicated that the benthopelagic alepocephalid, Alepocephalus bairdi, fed mostly on gelatinous zooplankton which "forms a false bottom at a height of 28-30 m from the bottom... formed by food organisms: Coelenterata, Pyrosoma and bathypelagic fishes." The near-bottom presence of 275 gelatinous plankters was observed in the Norfolk Canyon area by J. A. Musick (pers. comm.) during a dive aboard the D.S.R.V. Alvin. Prey items previously documented to occur near bottom may have been consistently present in the stomachs of the dominant Norfolk Canyon midwater fish species because the otter trawl captured fishes which were near the bottom when taken, where they fed on near-bottom prey species. The dominant midwater species captured mainly comprised adult fishes, although the net proved capable of capturing diminutive species (e.g. Cyclothone braueri) and juveniles (e.g. 18 mm Benthosema glaciale). A number of accounts indicate that adult forms are often found at greater depths than younger members of a species, and may migrate less (Gibbs and Roper, 1970; Clarke, 1974; Nafpaktitis et al., 1977). Therefore larger fishes may be near-bottom more often, and more susceptible to otter trawl capture. The internal parasite faunas of the dominant species were examined in order to see if parasite compositions supported the hypothesis of regular near-bottom interaction by these fishes. It has been established that midwater fishes possess fewer kinds of internal parasites, and that adult forms are rare (Noble and Collard, 1970; Noble, 1973). A midwater fish remaining near the more heavily parasitized demersal or benthic habitats could acquire heavier incidences of infection. Markle and Wenner (1979), following this theory, proposed that Melanostigma atlanticum (one of the dominant 276 species found in this study) and Xenodermichthys copei occurred near the bottom at certain times of the year based on incidences of infection of adult trematodes and larval nematodes. They hypothesized that this seasonal near-bottom interaction enhanced reproductive shoaling along a two-dimensional plane. Results from the present study on 18 species of dominant midwater fishes showed several distinct differences from the only other comprehensive study of midwater fish parasites (Collard, 1968). Unfortunately, Collard1s study utilized only mesopelagic species (primarily myctophids) from the Pacific. Collard found fewer numbers and kinds of parasites in mesopelagic fishes than in epipelagic or benthic fishes. In addition, he noted the extremely rare occurrence of adult parasites, suggesting that mesopelagic fishes merely serve as intermediate or paratenic hosts. In 1087 specimens of fishes examined, he found overall infection rates of 7.1%, 0.8%, 19.7%, and 1.9% for cestodes, trematodes, nematodes, and fungi, respectively. Comparing these values to the present study results, from 668 specimens examined, overall infections were 22.6%, 4.9%, 6.4%, and 6.1% for the respective groups. Cestodes, trematodes, and fungi occurred at 3.2, 6.1, and 3.4 times the frequency in Norfolk Canyon fishes, although the fungi mainly infected two species, Scopelogadus beanii and Derichthys serpentinus. The threefold lower percentage of nematode infection in Norfolk Canyon fishes was perplexing until it was realized that only 143 of Collards ' study specimens were from non-myctophid species, while 624 of the specimens examined for the 277 present study were non-myctophids. If myctophid infections of nematodes are removed from analysis, Collards' results indicate a 6.2% incidence of infection, as compared to 5.7% in the Norfolk Canyon material. It may be that the generalized diet of many myctophids (Gjosaeter, 1973; Worner, 1979, the present study) causes high incidences of nematodes in this family. Collards' overall incidence of adult parasites was 1.1% if externally attached copepods were excluded. Frequency of occurrence of adult parasites in Norfolk Canyon collection specimens was 6.1%. All but one of the trematodes recovered were adult. Eight of 41 nematode hosts harbored adult worms. A new genus and species of adult pseudophyllidean cestode was found in Eurypharynx pelecanoides, indicating that at least one midwater fish species serves as definitive host for cestodes (Campbell and Gartner, in prep.). Because of the difficulty in assessing life history strategies of midwater parasites, and the lack of comprehensive data for parasites in midwater fish populations in the North Atlantic, firm conclusions supporting near-bottom interactions based on parasite data cannot be made. However, the higher incidence of most major parasite groups, and increased presence of adult forms suggested that the parasite-mix in the dominant midwater fishes examined in the present study was different than that found in previous studies based on collections made in the water column with midwater trawls. APPENDIX A STATIONS AT WHICH MIDWATER FISHES WERE CAPTURED BY 45T OTTER TRAWL ( * -denotes water haul) CRUISE CI-73-10 : ATION DATELAT.(N) LO N G . (W) LOCAL TIME DEPTH(m) 38 7-VI-73 36°38.71 74*44.0* 2250 123 44 8-VI-73 36®40.4' 74*40.0* 2032 335 45 9-VI-73 36*42.61 74*38.2* 0001 390 47 9-VI-73 36*36.6 * 74*41.9' 0630 316 48 9-VI-73 36.34.8f 74*38.9* 1125 751 49 9-VI-73 36®41.5’ 74*37.4* 1405 742 50 9-VI-73 36*44.2 * 74*35.4* 2045 786 51 10-VI-7 3 36® 38.5 * 74*39.4* 0140 680 52 10-VI-73 36®33.51 74*42.8* 0515 501 53 10—VI-73 36°35.5 * 74*42.0* 0736 753 55 10-VI-73 36°38.0' 74*42.8* 1326 118 56* 10-VI-73 36*35.6 * 74*37.5* 1808 1190 57 10-VI-73 36°43.6’ 74*32.8’ 2158 1194 75 ll-VI-73 37®04.6' 74*36.6* 2230 403 76 12-VI-73 37®09.O ’ 74*34.4* 0259 270 78 12-VI-73 37*00.7* 74*36.7* 0748 292 79 12-VI-73 37®03.7’ 74*38.2* 1059 2.52 81 12-VI-73 36*56.2' 74*36.6 2030 729 82 13-VI-73 37°04.4’ 74*32.0’ 0108 716 83 13-VI-73 37°03.5’ 74*33.3’ 0545 986 84 13-VI-73 37°03.2 1 74*34.1* 0923 636 85 13-VI-73 37*03.9' 74*31.5* 2125 776 89 14-VI-73 37°01.7' 74*37.7* 1007 367 90 14-VI-73 36° 54.6’ 74*27.5* 1425 1488 91 14-VI-73 37°03.0* 74*20.6* 1924 1719 92* 14-VI-73 37*03.2 * 74*28.0* 2342 1876 93* 15-VI-73 37®04.61 74*12.9* 0509 1664 94* 15-VI-73 37°09.21 74*19.3* 0855 1753 95 15-VI-73 37*01.6' 74*22.5* 1245 1591 97 16-VI-73 36°50.2* 74*32.5* 0028 1350 279 APPENDIX A(cont ’d) CRUISE GI-74-04 DATE LAT.(N) L O N G .(W) LOCAL TIME DEPTH(m) 68 16-XI-74 36°42.8f 74°38.2’ 1500 269 69 16-XI-74 36*42.8' 74° 36.61 1730 642 71 17-XI-74 36°37.6 ’ 74°28.0r 0221 1701 72 17-XI-74 36® 36.3’ 74°24.81 0636 1730 73 17-XI-74 36*41.51 74°18.51 1337 1910 74 17-XI-74 36*36.9’ 74°16.2’ 1732 1985 75 17-XI-74 36*33.2' 74*01.7’ 2125 2400 79 18-XI-74 36*43.2’ 74° 38.0’ 1151 260 81 18-XI-74 36°38.9' 74°38.6’ 1703 578 82 18-X1-74 36° 36.81 74°38.7 ’ 2056 699 83 19^X1-74 36*40.7 f 74*30.5’ 0057 1408 84 19-XI-74 36°36.0' 74°24.4’ 0543 1763 85 19-XI-74 36*37.0’ 73*21.8’ 1057 1823 86 19-XI-74 36°41.6' 73°47.0’ 1702 2642 87 19-XI-74 36*44.5’ 73°45.9’ 2340 2624 88 20-XI-74 36°44.5 T 74°35.4 ’ 0812 809 89 20-XI-74 36* 32.5’ 74°40.1 ’ 1312 912 90 20-XI-74 36° 39.5 T 74°40.4’ 1531 686 91 20-XI-74 36°43.7’ 74°37.6’ 2003 376 92 20-XI-74 36* 39.6’ 74° 37.8’ 2230 200 95 22-XI-74 37*05.0’ 74°05.7 ’ 1515 2115 96 22-XI-74 37*08.6’ 73°54.7 ’ 2211 2170 97 23-XI-74 37*00.3’ 74°15.0’ 0435 1441 98 23-XI-74 36°59.6 ’ 74*33.5’ 1107 750 00 23-XI-74 37° 03.3’ 74*31.8’ 1915 614 02 24-XI-74 37*05.1’ 74*39.7’ 0631 315 03 24-XI-74 37°03.0' 74*30.9’ 1013 694 04 24-XI-74 37*03.6’ 74*34.1’ 1248 312 05 24-XI-74 37°05.41 74*23.6’ 1840 1525 06 24-XI-74 37°06.11 74*23.7’ 2158 1403 07 25-XI-74 37*03.5’ 74*30.9’ 0533 655 CRUISE GI-75-08 10 9-IX-75 37°05.3’ 74*40.8’ 1707 306 14 10-IX-75 37°08.4’ 74*30.5’ 0657 330 15 10-IX-75 37°07.7’ 74*31.9’ 1027 277 17 10-IX-75 37°04.5 f 74*30.6’ 1854 662 19 ll-IX-75 36° 57.4’ 74*37.5’ 0446 221 22 ll-IX-75 36°58.6 ’ 74*33.8’ 1630 906 27 12-IX-75 37°09.9’ 74*24.8’ 1152 1287 28 12-IX-75 37° 08.5 ’ 74*21.9’ 1713 1663 29 12-IX-75 37°00.0 ’ 74*19.0’ 2312 1698 280 APPENDIX A(cont’d) DATE LAT.(N) LONG.(W) LOCAL TIME DEPTH(m) 13-IX-75 36°57.6 f 74° 22.2’ 0527 1943 31 13-IX-75 36°57.7 f 74°32.6' 0929 1109 33 13-IX-75 37°01.7 T 73°59.11 2239 2257 34 14-IX-75 36°57.31 73° 39.5’ 0748 2767 35 14-IX-75 36°57.9' 73°21.5 ? 1445 2933 36 15-1X-75 36°59.0I 72°58.0’ 0055 3083 82 17-IX-75 36°37.0’ 74°41.4' 1904 198 84 17-IX-75 36°44.0' 74°38.31 2340 304 85 18-IX-75 36°43.51 74°34.5' 0207 647 86 18-IX-75 36°36.81 74°39.6' 0515 619 87 18-IX-75 36°36.21 74°38.9 0832 898 88 18-IX-75 36° 42,5’ 74°32.1’ 1314 1189 89 18-1X-75 36°40.1 ’ 74°25.41 1737 1644 90 19-IX-75 36°38.7' 74° 23.0' 0130 1743 91 19-IX-75 36°39.0’ 74°21.6 T 0712 1842 93 19-IX-75 36°37.11 74°25.0' 1730 1803 95 20-IX-75 36°41.2 T 74°41.5’ 0215 118 98 20-IX-75 36°40.2 ’ 74° 37.6’ 0932 932 99 20-IX-75 36° 36.5 74*39.01 1526 918 CRUISE GI-76-01 24 23- 1-76 36°59.2' 73°43.4' 0901 2679 25 23- 1-76 37°07.9 r 73°53.2' 1555 2250 26 23- 1-76 37°06.5 T 73° 05.8' 2115 2242 27 24- 1-76 36°56.0? 74° 01.51 0102 2427 28 24- 1-76 37°01.31 74°13.61 1020 1916 29 24- 1-76 37°04.5' 74°23.0* 1758 1458 31 25- 1-76 37°01.5' 74° 34.0f 0439 712 32 25- 1-76 37°01.5’ 74° 32.8' 0750 818 34 25- 1-76 37°04.5 1 74° 32.01 1335 452 35 25- 1-76 37°08.5’ 74° 32.5’ 1620 221 45 26- 1-76 36° 37.5’ 74°45.01 1820 109 47 26- 1-76 36°40.5 ’ 74°40.5' 2233 384 48 27- 1-76 36° 37,5' 74°42.5 f 0140 313 49 27- 1-76 36° 38.5 ? 74°40.0r 0435 616 50 27- 1-76 36° 42. 5 ’ 74°37.3 r 0728 884 51 27- 1-76 36° 43.8’ 74° 37.0’ 1022 760 52 27- 1-76 36° 37.5' 74° 33.6 f 1430 1470 54 28- 1-76 36° 33.0 T 74°25.0' 0824 2000 55 28- 1-76 36° 36.5' 73o58.0, 2110 2575 56 29- 1-76 36°40.0’ 73°36.5' 0631 2920 57 29- 1-76 36° 39. 5 ’ 74° 27.0’ 1713 1635 58 29- 1-76 36°39.8’ 74°28.5' 2122 1578 59 30- 1-76 36°39.5 T 74°35.7 T 0152 1108 APPENDIX A(cont’d) STATION DATE LAT.(N) LON G .(W) LOCAL TIME DEPTH(m) 60 30- 1-76 36°40.5 1 74° 37.5 ? 0643 792 61 30- 1-76 36°41.2 1 74°38.6' 1015 486 63 30- 1-76 37900.0 1 74°37.0f 1720 340 64 30- 1-76 37°04.01 74° 36.0T 2010 226 67 31- 1-76 37°07.5' 74*32.0? 0305 534 68 31- 1-76 37*01.5' 74®39.0 ? 0543 150 53* 27- 1-76 36°32.4' 74°27.1' 2103 1910 282 APPENDIX B CAPTURE DATA FOR NON-DOMINANT MIDWATER FISHES ( ^-indicates species that may be year-round members of Norfolk Canyon midwater community which were underrepresented in trawl collections) [NO.OF STATIONS] DEPTH NO.OF SPECIMENS STRATUM(A)(m) SPECIES______SEASON (S) CAPTURED OF CAPTURE (S) Saccopharynx flagellum SP [1 1 1501-2000 Bathylagus compsusC?)** SU [1 2 1501-2000 FL [3 3 >1000 WR [1 1 1501-2000 B. longirostris* sp [1 1 1001-1500 SU [1 1 1501-2000 WR [1 1 1501-2000 Xenodermichthys copei* SU [1 1 201- 500 FL [3; 3 201-1000 WR [2 2 201-1000 Cyclothone braueri SP [1 1 1501-2000 C. pallida SP [1 2 1501-2000 Vinciguerria s p . FL [1 1 1501-2000 Argyropelecus gigas* SP [1 1 501-1000 SU [2 2 501-1500 FL [1 1 501-1000 Chauliodus danae SU [1 1 1501-2000 Stomias affinis SP [1 1 501-1000 Borostomias antarcticus SP [1 1 1001-1500 Rhadinesthes decimus WR [1 1 1001-1500 Melanostomias spilorhynchus SU [2 2 1001-2000 WR [2 2 >751 Echiostoma barbatum WR [1 1 1501-2000 Eustomias bigelowi SU [1 1 1001-1500 Trigonolampa miriceps SU [1 1 1001-1500 Grammatostomias flagellibarba SU [1 1 1501-2000 Lobianchia gemellari SP [1 1 1001-1500 SU [2; 2 1501-2000 FL [1 1 1001-1500 L. dofleini SU [1 5 1501-2000 FL [3 5 751-2000 WR [1 1 1001-1500 Diaphus dumerilii* SP [3 11 75- 750 WR [1 1 201- 500 Lampanyctus crocodilus* SP [1 1 501- 750 SU [1 2 > 2000 WR [1 1 1501-2000 2G3 APPENDIX B(cont’d) [NO.OF STATIONS] DEPTH NO.OF SPECIMENS STRATUM(A)(m) SPECIES SEASON(S) CAPTURED OF CAPTURE(S) Lampanyctus ater SP [1 1 501- 750 SU [1 1 1501-2000 L. cuprarius FL [1 1 1501-2000 Lepidophanes guentheri SU [1 1 1501-2000 Gonichthys coccoi FL [1 1 751-1000 Myctophum taf f ine FL [2 2 501-2000 M. punctatum WR [1 1 501- 750 Hygophum hygomii SU [4 14 751-1000 Notoscopelus resplendens SU [1 1 1501-2.000 WR [1 1 1501-2000 Symbolophorus veranyi WR [1 1 1501-2000 Neoscopelus macrolepidotus SU [1 1 1501-2000 FL [1 4 201- 500 Polyipnus asteroides* SP [3 7 201-1000 SU [1 1 751-1000 FL [4 5 201- 750 Mentodus rostratus WR [1 1 1001-1500 Evermannella balbo FL [1 1 501- 750 Scopelarchus analis FL [1 1 1501-2000 Melanocetus johnsonii WR [1 1 751-1000 Ceratias holbolli WR [1 1 1001-1500 Cryptopsaras couesi WR [2 2 501-1000 Poromitra capito SP [1 1 1001-1500 P. crassiceps* SP [1 1 1501-2000 SU [1 1 1501-2000 FL [1 1 >2000 Scopeloberyx opisthopterus* SP [4 10 1001-2000 SU [1 2 1501-2000 FL [2 10 1501-2000 S. robustus* SP [1 1 1501-2000 FL [2 2 1501-2000 WR [1 1 1501-2000 Melampbaes suborbitalis SU [1 1 >2000 Anoplogaster cornuta SU [1 1 1501-2000 Rondeletia loricata SU [1 1 1501-2000 FL [1 1 1001-1500 Cetomimus gilli WR [1 1 >2000 Chiasmodon sp. A FL [1 1 >2000 WR [1 1 >2000 C. niger SP [2 3 751-1500 Promethichthys prometheus SU [2 2 75- 200 Nesiarchus nasutus WR [1 1 201- 500 Benthodesmus simoyi SP [2 2 501-2000 B. atlanticus FL [1 1 501-750 284 APPENDIX B(cont’d) [NO.OF STATIONS] DEPTH NO.OF SPECIMENS STRATUM(A)(m) SPECIES SEASON(S) CAPTURED OF CAPTURE(S) Lampadena speculigera SP II] 1 1501-2000 FL II] 1 1501-2000 SP- Cruise CI-73-10, Spring SU- Cruise GI-75’-08, Summer FL- Cruise GI-74-04, Fall WR- Cruise GI-76-01, Winter 285 a a 0) a) c p a, co co 4-> rH rH rH rH rH rH I—1 P-. Q TO . CM CO b~~ rH X OO i—H o CO • i—i d a) S O 38°27.3f , 73°03.5’ ; 2068 ; 1852 1 ; o 4-1 CP •rH CU •H • CO t-H CM c- CM CO CO U o- CO CO r H c- M0 CO CP a 38: <: 4-1 s rH rH r H r—1 0) O O i—1 r H ~d" rH 03 cd X M0 CP rH CU v_r d CO d m o Pi o o *w» Pm #* d CO •H r-^ •H d 4-J d 4-1 v—✓ 4-) 03 C o Cu d cd cd CU H •H cd cd Ml CO 4-) TO <1 4-1 ,p •H CO d CO •H Q Cd CM d O 4A: 38°51.0' 72°16.0’ , ; station 34: 36°58.1'N, 74°11.0'W; 2200 m; 0911 hrs.; 1 spec 4-J CO r—1 r •* •rH n #k d a CO PS o m m m CM 4-> o M CO XI CO O •rH u Sh Cruise X> •rH s CO Pi P Pi X d CU a> d O d a Q w 286 Vi O u CM CU cu *G CO (U CN 'C}' CN CN r—I C — »— I i— I i—H LITERATURE CITED Agius, C. 1978. Infection by an Ichthyophonus-like fungus in the deep-sea scabbard fish Aphanopus carbo (Lowe) (Trichiuridae) in the North East Atlantic. Jour. Fish Diseases 1:191-193. Alexander, C. G. 1963. Tetraphyllidean and Diphyllidean Cestodes of New Zealand Selachians. Trans. Royal Soc. New Zealand. 3(12 ) :117-142. Armstrong, H. W. 1974. A study of the Helminth Parasites of the family Macrouridae from the Gulf of Mexico and Caribbean Sea: Their systematics, ecology and zoogeographical implications. Ph.D. Dissertation Texas A&M University. 329 pp. Backus, R. H., G. W. Mead, R. L. Haedrich, and A. W. Ebeling. 1965. The mesopelagic fishes collected during Cruise 17 of the R/V Chain, with a method for analyzing faunal transects. Bull. Mus. Comp. Zool. 134(5):139-158. Backus, R. H., J. E. Craddock, R. L. Haedrich, D. L. Shores, J. M. Teal, A. S. Wing, G. W. Mead, and W. D. Clarke. 1968. Ceratoscopelus maderensis: Peculiar sound-scattering layer identified with this Myctophid fish. Science 169(3831):991-993. Backus, R, H., J. E. Craddock, R. L. Haedrich, and D. L. Shores. 1969. Mesopelagic fishes and thermal fronts in the Western Sargasso Sea. Mar. Biol. 3:87-106. Backus, R. H., J. E. Craddock, R. L. Haedrich, and D. L. Shores. 1970. The distribution of mesopelagic fishes in the equatorial and western North Atlantic Ocean. In: Proceedings of an International Symposium on Biological Sound-Scattering in the Ocean, G. B. Farquhar (Ed.), U.S. Government Printing Office, Washington, D.C. 20-40. Backus, R. H., J. E. Craddock, R. L. Haedrich, and B. H. Robison. 1977. Atlantic mesopelagic zoogeography. In: Fishes of the Western North Atlantic. Sears Found. Mar. Res. 1(7):266-287. Badcock, J. 1970. The vertical distribution of mesopelagic fishes collected on the Sond Cruise. Jour. Mar. Biol. Assn. U.K. 50:1001-1044. 288 Badcock, J., and N. R. Merrett. 1977. On the distribution of midwater fishes in the Eastern North Atlantic. In: Oceanic Sound Scattering Prediction, N. R. Andersen and B. J. Zahuranec (Eds.). Marine Science 5:249-282. Plenum Press. Baird, R. C. 1971. The systematics, distribution, and zoogeography of the marine hatchetfishes (family Sternoptychidae). Bull. M u s . Comp. Zool. 142(1):1-128. Baird, R. C., T. L. Hopkins, and D. F. Wilson. 1975. Diet and feeding chronology of Diaphus taaningi (Myctophidae) in the Cariaco Trench. Copeia. 1975:356-365. Barnes, R. D. 1974. Invertebrate Zoology. W. B. Saunders Co., Philadelphia. 870 pp. Be, A. W. H . , J. M. Forns, and 0. A. Roels. 1971. Plankton abundance in the North Atlantic Ocean. In: Fertility of the Sea, J. D. Costlow, Jr. (Ed.) 1:17-50. Gordon and Breach Science Publishers. Beebe, W. 1935a. Deep-sea fishes of the Bermuda oceanographic expeditions. Family Derichthyidae. Zoologica. 20(1):1-24. Beebe, W. 1935b. Deep-sea fishes of the Bermuda oceanographic expeditions. Family Nessorhamphidae. Zoologica. 29(2):25-51. Beebe, W. 1937. Preliminary list of Bermuda deep-sea fish. Zoologica. 22(14):197-208. Beebe, W . , and J. Crane. 1936. Deep-sea fishes of the Bermuda oceanographic expeditions. Family Serrivomeridae. Zoologica. 20(3):53-102 . Beebe, W . , and J. Crane. 1944. Deep-sea fishes of the Bermuda oceanographic expeditions. Family Nemichthyidae. Zoologica. 22(27):349-383 . Beebe, W . , and J. Crane. 1947. Eastern Pacific expeditions of the New York Zoological Society XXXVII Deep-sea Ceratioid fishes. Zoologica. 31(11):151-182 . Belman, B. W., and M. E. Anderson. 1979. Aquarium observations on feeding by Melanostigma pammelas (Pisces: Zoarcidae). Copeia. 366-369. Bengston, D. 1973. Vertical distribution and life history of the fish family Nemichthyidae in the Bermuda Ocean acre. M.S. Thesis, Univ. of Rhode Island. 70 pp. 289 Bertelsen, E. 1951. The Ceratioid fishes. Ontogeny, taxonomy, distribution and biology. Dana-Report No. 39. 1-276. Bigelow, H. B., and M. Sears. 1939. Studies of the waters on the continental shelf, Cape Cod to Chesapeake Bay. III. A volumetric study of the zooplankton. M us. Comp. Zool. Harvard Mem. 54(4):189-373. Bohlke, J. E. 1966. Order Lyomeri. In: Fishes of the Western North Atlantic. Sears Found. Mar. Res. 1(5):603-628. Borodulina, 0. D. 1972. The feeding of mesopelagic predatory fish in the open ocean. Jour. Ichthy. 12(4):692-702. Bowman, T. E. 1960. The pelagic amphipod genus Parathemisto (Hyperiidea: Hyperiidae) in the North Pacific and adjacent Arctic Ocean. Proc. U. S. Nat. Mus. : 112(3439) 343-392. Bowman, T. E., and H. E. Gruner. 1973. The families and genera of Hyperiidea (Crustacea: Amphipoda) Smithsonian Contr. Zool. 146:1-64. Bullock, W. L. 1969. Morphological features as tools and as pitfalls in Acanthocephalan systematics. In: Problems in Systematics of Parasites, G. C. Schmidt (Ed.). University Park Press 49-76. Cailliet, G. M. 1972. The study of feeding habits of two marine fishes in relation to plankton ecology. Trans. Amer. Micros. Soc. 91(1):88-89 . Cannon, L. R. G. 1977. Some larval Ascaridoids from southeastern Queensland marine fishes. Int. Jour. Parasit. 7:233-243. Castle, P. J. 1970. Distribution, larval growth, and metamorphosis of the eel Derichthys serpentinus Gill, 1884 (Pisces, Derichthyidae). Copeia. 1970:444-452. Childress, J. J., and R. P. Meek. 1973. Observations on the feeding behavior of a mesopelagic fish (Anoplogaster cornuta: Beryciformes). Copeia. 1973:602-603. Clarke, G. L. 1940. Comparative richness of zooplankton in coastal and offshore areas of the Atlantic. Biol. Bull. 78:226-255. Clarke, G. L. 1970. Light conditions in the sea in relation to the diurnal vertical migrations of animals. In: Proceedings of an International Symposium on Biological Sound-Scattering in the Ocean, G. B. Farquhar (Ed.) U. S. Government Printing Office, Washington, D.C. 41-50. 290 Clarke, T. A. 1973. Some aspects of the ecology of Lanternfishes (Myctophidae) in the Pacific Ocean near Hawaii. Fish. Bull. 71(2):401-433. Clarke, T. A. 1974. Some aspects of the ecology of Stomiatoid fishes in the Pacific Ocean near Hawaii. Fish. Bull. 72(2):337-351. Clarke, T. A., and P. J. Wagner. 1976. Vertical distribution and other aspects of the ecology of certain mesopelagic fishes taken near Hawaii. Fish. Bull. 74(3):635-645. Cohen, D. M. 1964. Suborder Argentinoidea. In: Fishes of the Western North Atlantic. Sears. Found. Mar. Res. 1(4):1-70. Collard, S. B. 1968. A study of parasitism in mesopelagic fishes. Ph.D.Dissertation, Univ. of California, Santa Barbara. 251 pp. Collard, S. B. 1970a. Forage of some Eastern Pacific midwater fishes. Copeia. 1970:348-354. Collard, S. B. 1970b. Some aspects of host-parasite relationships in mesopelagic fishes. In: A Symposium on Diseases of Fishes and Shellfishes, S. F. Snieszko (Ed.). Spec. Publ. No. 5, American Fisheries Society 41-56. Devany, T. 1969. Ecologic interpretation of distribution of the Lanternfishes (Myctophidae) in the Straits of Florida. Ph.D. Dissertation, Univ. of Miami. 431 pp. DeWitt, F. A., Jr., and G. M. Cailliet. 1972. Feeding habits of two bristlemouth fishes, Cyclothone acclinidens and C^. signata (Gonostomatidae). Copeia. 1972:868-871. Dollfus, R. PI. 1964. Enumeration des Cestodes du plancton et des invertebres marins (6 e contribution). Ann. Parasit. (Paris). 39(3):329-379. Ebeling, A. W. 1962. Melamphaidae I. Systematics and zoogeography of the species in the bathypelagic fish genus Melamphaes Gunther. Dana-Report N o . 58. Ebeling, A. W . , and W. H. Weed, III. 1963. Melamphaidae III. Systematics and distribution of the species in the bathypelagic fish genus Scopelogadus Vaillant. Dana-Report No. 60:1-58. Ebeling, A. W., R. M. Ibara, R. J. Lavenberg, and F. J. Rohlf. 1970. Ecological groups of deep-sea animals off southern California. Bull. Los Angeles Cnty. Mus. Nat. Hist. Science. 6:1-43. 291 Ebeling, A. W., and G. M. Cailliet. 1974. Mouth size and predator strategy of midwater fishes. Deep-Sea Res. 21:959-968. E g e , V. 1934. The genus Stomias Cuv., taxonomy and biogeography. Dana-Report No. 5:1-58. Ege, V. 1940. Chauliodus Schn., bathypelagic genus of fishes. A Systematic, Phylogenetic and Geographical Study. Dana-Report no. 31:1-148. Gibbs, R. H., Jr. 1969. Taxonomy, sexual dimorphism, vertical distribution, and evolutionary zoogeography of the bathypelagic fish genus Stomias (Stomiatidae). Smithsonian Contr. Zool. 31:1-25. Gibbs, R. H., Jr., R. J. Goodyear, M. J. Keene, and D. W. Brown. 1970. Biological studies of the Bermuda Ocean acre. II. Vertical distribution and ecology of the lanternfishes (family Myctophidae) Report to the U. S. Navy Underwater Systems Center. Contract No. N00140-70-C-0307:1-141. Gjosaeter, J. 1973. The food of the Myctophid fish, Benthosema glaciale (Reinhardt), from Western Norway. Sarsia. 52:53-58. Gorelova, T. A. 1978. The feeding of lanternfishes, Ceratoscopelus warmingi and Bolinichthys longipes, of the family Myctophidae in the western equatorial part of the Pacific ocean. Jour. Ichthy. 18(4):588-598 . Gosner, K. L. 1971. Guide to identification of marine and estuarine invertebrates. Wiley-Interscience. 693 pp. Grant, G. C. 1979. Zooplankton of the water column and neuston. In: Middle Atlantic Outer Continental Shelf Environment Studies. Vol. II-A. Chemical and Biological Benchmark Studies to the Bureau of Land Management, Contract No. AA550-CT-6-62. Grey, M. 1956. The distribution of fishes found below a depth of 2000 meters. Fieldiana: Zoology 36(2):1-337. Grey, M. 1964. Family Gonostomatidae. In: Fishes of the Western North Atlantic. Sears Found. Mar. Res. l(4):78-240. Grice, G. D., and A. D. Hart. 1962. The abundance, seasonal occurrence and distribution of the epizooplankton between New York and Bermuda. Ecol. Monogr. 32 :287-309. Haedrich, R. L. 1972. Midwater fishes from a warm-core eddy. Deep-Sea Res. 19:903-906. 292 Halliday, R. G. 1970. Growth and vertical distribution of the glacier lanternfish, Benthosema glaciale, in the Northwestern Atlantic. Jour. Fish. Res. Bd. Canada. 27(1):105-116 . Harbison, G. R., D. C. Biggs, and L. P. Madin. 1977. The associations of Amphipoda Hyperiidea with gelatinous zooplankton - II. Associations with Cnidaria, Ctenophora and Radiolaria. Deep-Sea Res. 24:465-488. Harder, W. 1975. Anatomy of fishes. Part I: Text. E. Schweizerbart'sche Verlagsbuchhandlung Stuttgart. 612 pp. Harrison, C. M. H. 1967. On methods for sampling mesopelagic fishes. Symp. Zool. So c . Lond.19:71-126. Hendricks, J. D. 1972. Two new host species for the parasitic fungus Ichthyophonus hoferi in the Northwest Atlantic. Jour. Fish. Res. Bd. Canada. 29(12):1776-1777. Holton, A. A. 1969. Feeding behavior of a vertically migrating lanternfish. Pac. Sci. 23:325-331. Hopkins, T. L., and R. C. Baird. 1973. Diet of the hatchetfish Sternoptyx diaphana. Mar. Biol. 21:34-46. Hopkins, T. L. and R. C. Baird. 1977. Aspects of the feeding ecology of oceanic midwater fishes. In: Oceanic Sound Scattering Prediction, N. R. Andersen and B. J. Zahuranec (Eds.). Mar. Sci. 5:325-360. Plenum Press. Idyll, C. P. 1964. Abyss. The deep-sea and the creatures that live in it. Thomas Y. Crowell Co., N. Y. 396 pp. Jahn, A. E. 1976. On the midwater fish faunas of Gulf Stream rings, with respect to habitat differences between Slope Water and northern Sargasso Sea. Ph.D. Dissertation, Woods Hole Oceanographic Institution. 173 pp. Kampa, E. M. 1970. Photoenvironment and sonic sensing. In: Proceedings of an International Symposium on Biological Sound Scattering in the Ocean, G. B. Farquhar (Ed.). U.S. Government Printing Office, Washington, D.C. 51-59. Kennedy, C. R. 1975. Ecological animal parasitology. Halsted Press, J. Wiley & Sons, N.Y. 163 pp. Kinzer, J. 1977. Observations on feeding habits of the mesopelagic fish Benthosema glaciale (Myctophidae) off Northwest Africa. In: Oceanic Sound Scattering Prediction, N. R. Andersen and B. J. Zahuranec (Eds.). Marine Science 5:381-392. Plenum Press. 293 Kobayashi, B. N. 1973. Systematics, zoogeography, and aspects of the biology of the bathypelagic fish genus Cyclothone in the Pacific Ocean. Ph.D. Dissertation, Univ. of California, San Diego. 487 pp. Krefft, G. 1976. Distribution patterns of oceanic fishes in the Atlantic Ocean. Selected Problems. Rev. Trav. Inst. Peches Marit. 40(3+4):439-460 . Krueger, W. H . , and G. W. Bond. 1972. Biological studies of the Bermuda Ocean Acre. III. Vertical Distribution and Ecology of the Bristlemouth Fishes (family Gonostomatidae) Report to the U.S. Navy Underwater Systems Center Contract No. N00140-72-C-0315:1-50. Krueger, W. H . , M. J. Keene, and A. A. Keller. 1975. Systematic analysis of midwater fishes obtained at Deepwater Dumpsite 106 May 1974. In: NOAA Dumpsite Evaluation Report 75-1. May 1974 Baseline Investigation of Deepwater Dumpsite 106:359-388. Krueger, W. H., R. H. Gibbs, Jr., R. C. Kleckner, A. A. Keller, and M. J. Keene. 1977. Distribution and abundance of mesopelagic fishes on Cruises 2 and 3 at Deepwater Dumpsite 106. In: NOAA Dumpsite Evaluation 77-1:377-422. Lancraft, T. M., and B. H. Robison. 1980. Evidence of postcapture ingestion by midwater fishes in trawl nets. Fish. Bull. 77(3):713-714. Leavitt, B. B. 1935. A quantitative study of the vertical distribution of the larger zooplankton in deepwater. Biol. Bull. 68(1):115-130. Leavitt, B. B. 1938. The quantitative vertical distribution of macrozooplankton in the Atlantic Ocean Basin. Biol. Bull. 74(3):376-394. Legand, M., and J. Rivaton. 1969 . Cycles biologiques des poissons mesopelagiques de l ’est de l'ocean Indien. Troisieme note: Action predatrice des poissons micronectoniques. Cah. 0 .R.S.T.O.M., ser. Oceanogr. 7(3):29-45. Machida, M. 1975. Nematodes from the deep-sea fishes of Suruga Bay. I. A New Rhabdochonid Nematode, Johnstonmawsonoides nemichthyos from the Nemichthyid Fish. Bull. Natn. Sci. Mus. Ser. A^ 1 (1 ):1-4. Madin, L. P., and G. R. Harbison. 1977. The associations of Amphipoda Hyperiidea with gelatinous zooplankton - I. Associations with Salpidae. Deep-Sea Res. 24:449-463. 294 Manter, H. W. 1934. Some digenetic trematodes from deep-water fish of Tortugas. Florida papers from the Tortugas Laboratory, Carnegie Inst. Washington, Pub. 435:261-345. Markle, D. F., and C. A. Wenner. 1979. Evidence of demersal spawning in the mesopelagic Zoarcid fish Melanostigma atlanticum with comments on demersal spawning in the Alepocephalid fish Xenodermichthys copei. Copeia 1979:363-365. Marshall, N. B. 1954. Aspects of deep-sea biology. Hutchinson, London. 380 pp. Marshall, N. B. 1960. Swimbladder structure of deep-sea fishes in relation to their systematics and biology. Discovery Report 31 :1-122. Marshall, N. B. 1971. Explorations in the life of fishes. Harvard Univ. Press, Mass. 204 pp. Marshall, N. B. 1975. The life of fishes. Universe Books, N. Y. 402 pp. Marshall, N. B., and N. R. Merrett. 1977. The existence of a benthopelagic fauna in the deep-sea. In: A Voyage of Discovery: George Deacon 70th Anniversary Volume, M. Angel (Ed.), 483-497. Pergamon Press. Mauchline, J., and L. R. Fisher. 1969. The biology of Euphausiids. Adv. Mar. Biol. 7:1-454. McCormick, J. M . , R. M. Laurs, and J. E. McCauley. 1967. A hydroid epizoic on Myctophid fishes. Jour. Fish. Res. Bd. Canada. 24(9):1985-1989. McCosker, J. E., and M. E. Anderson. 1976. Aquarium maintenance of mesopelagic animals: A Progress Report. Bull. South. Cal. Acad. Sci. 75(2):211-219. McEachran, J. D., D. F. Boesch, and J. A. Musick 1976. Food Division within two sympatric species-pairs of skates (Pisces: Rajidae) Mar. Biol. 35:301-317. McNaughton, S. L., and L. L. Wolf. 1979. General Ecology. Holt, Rinehart, and Winston, New York. Mead, G. W., and I. Rubinoff. 1966. Avocettinops yanoi, a new nemichthyid eel from the southern Indian Ocean. Breviora, Mus. Comp. Zool. 241:1-6. Mead, G. W . , and S. A. Earle. 1970. Notes on the natural history of snipe eels. Proc. Cal. Acad. Sci. 38(5):99-103. 295 Merrett, N. R., and H. S. J. Roe. 1974. Patterns and selectivity in the feeding of certain mesopelagic fishes. Mar. Biol. 28:115-126. Miyazaki, T., and S. S. Kubota. 1977a. Studies on Ichthyophonus diseases of fishes - I. Rainbow Trout Fry. Bull. Fac. Fish. Mie Univ. 4:45-56. Miyazaki, T., and S. S. Kubota. 1977b. Studies on Ichthyophonus diseases of fishes - II. YearlingRainbow Trout - Clonic Infection. Bull. Fac. Fish. Mie Univ. 4:57-65. Miyazaki, T., and S. S. Kubota. 1977c. Studies on Ichthyophonus diseases of fishes - III. Life Cycle of Ichthyophonus Affected Rainbow Trout. Bull. Fac. Fish. Mie Univ. 4:67-80. Moller, H. 1974. Ichthyosporidium hoferi (Plehn et Mulsow) (Fungi) as parasite in the Baltic Cod (Gadus morhua L.) Kieler Meersforsch. 3091):37-41. Morrow, J. E., Jr. 1964. Suborder Stomiatoidea. Families Chauliodontidae, Stomiatidae, and Malacosteidae. In: Fishes of the Western North Atlantic, Sears Found. Mar. Res. 1(4):274-310; 523-549. Mukhacheva, V. A. 1964. The composition of species of the genus Cyclothone (Pisces, Gonostomatidae) in the Pacific Ocean. Translation from Trudy Instituta Oceanologii Akad. Nauk. U.S.S.R. 73:93-138. Musick, J. A. 1973. Mesopelagic fishes from the Gulf of Maine and the adjacent continental slope. Jour. Fish. Res. Bd. Canada. 30(1) :134-137 . Musick, J. A. 1979. Community structure of fishes on the continental slope and rise off the middle Atlantic Coast of the United States. Special Scientific Report No. 96 of the Virginia Institute of Marine Science, School of Marine Science, College of William and Mary. 125 pp. Myers, B. J. 1975. The nematodes that cause anisakiasis. J. Milk Food Technol. 38(12):744-782. Nafpaktitis, B. G., R. H. Backus, J. E. Craddock, R. L. Haedrich, B. H. Robison, and C. Karnella. 1977. Family Myctophidae. In: Fishes of the Western North Atlantic, Sears Found. Mar. Res. 1(7):13-265. Nelson, J. S. 1976. Fishes of the world. Wiley-Interscience. 416 pp. 296 Nielsen, J. G., and D. G. Smith. 1978. The eel family Nemichthyidae. Dana-Report No. 88:1-71. Noble, E. R. 1960. Fishes and their parasite-mix as objects for ecological studies. Ecology. 41(3):593-596. Noble, E. R. 1973. Parasites and fishes in a deep-sea environment. Adv. Mar. Biol. 11:121-195. Noble, E. R., and S. B. Collard. 1970. The parasites of midwater fishes. In: A Symposium on Diseases of Fishes and Shellfishes, S. F. Snieszko (Ed.). Spec. Publ. No. 5, American Fisheries Society: 57-68. Noble, E. R . , and J. D. Orias. 1970. Aponurus californicus: new Trematode from the deep-sea smelt Leuroglossus stilbius. Trans. Amer.Micros. Soc. 89(3):413-417. Noble, E. R., and J. D. Orias. 1975. Parasitism in the bathypelagic fish, Melanostigma pammelas. Int. Jour. Parasit. 5:89-93. Norris, D. E., and R. M. Overstreet. 1975. Thynnascaris reliquens sp. n. and TN habena (Linton, 1900) (Nematoda: Ascaridoidea) from fishes in the Northern Gulf of Mexico and Eastern U.S. Seaboard. Jour. Parasit. 61(2):330-336. Omori, M. 1974. The biology of pelagic shrimps in the ocean. Adv. Mar. Biol. 12:233-324. Orias, J. D., E. R. Noble, and G. D. Alderson. 1978. Parasitism in some East Atlantic bathypelagic fishes with a description of Lecithophyllum irelandeum sp. n. (Trematoda). Jour. Parasit. 64(1):49-51. Overstreet, R. M . , and F. G. Hochberg, Jr. 1975. Digenetic trematodes in Cephalopods. Jour. Mar. Biol. Assn. U. K. 55:893-910. Ozawa, T . , K. Fujii, and K. Kawaguchi. 1977. Feeding chronology of the vertically migrating Gonostomatid fish, Vinciguerria nimbaria (Jordan and Williams), off Southern Japan. Jour. Ocean. Society Japan 33:320-327. Paxton, J. R. 1967. Biological notes on Southern California lanternfishes (family Myctophidae). Calif. Fish Game 37(2):111-120. Pearcy, W. G. 1964. Some distributional features of mesopelagic fishes off Oregon. Jour. Mar. Res. 22(1):83-102. 297 Pearcy, W. G., and R. M. Laurs. 1966. Vertical migration and distribution of mesopelagic fishes off Oregon. Deep-Sea Res. 13:153-165. Pearcy, W. G., H. V. Lorz, and W. Petersen. 1979. Comparison of the feeding habits of migratory and non-migratory Stenobrachius leucopsarus (Myctophidae). Mar. Biol. 51:1-8. Pinkas, L., M. S. Oliphant, and I. L. K. Iverson. 1971. Food habits of albacore, bluefin tuna, and bonito in California waters. Calif. Dept. Fish. Game Fish Bull. 152:105 pp. Raymont, J. E. G. 1963. Plankton and productivity in the Oceans. Pergamon Press. 660 pp. Reimer, L. W. 1976. Metazerkarien in Wirbellosen der Kuste von Madras. Angew. Parasitol. 17(1):33-43. Reimer, L. W., S. Hnatiuk, and J. Rochner. 1975. Metacercarien in planktontieren des mittleren Atlantik. Wiss. Zeitschrift der Padogogischen Hochschule "Lisellote Herrmann" Gustrow. 2/75:239-258. Richardson, P. 1976. Gulf stream rings. Oceanus. 19(3):65-68. Robison, B. H. 1972. Distribution of the midwater fishes of the Gulf of California. Copeia. 1972:448-461. Robertson, D. A., P. E. Roberts, and J. B. Wilson. 1978. Mesopelagic faunal transition across the subtropical convergence east of New Zealand. N. Z. Jour. Mar. F. W. Res. 12(4):295-312. Rofen, R. R. 1966. Family Paralepididae. In: Fishes of the Western North Atlantic. Sears Found. Mar. Res. 1(5):205-461. Roule, L., and L. Bertin. 1929. Les poissons Apodes appartenant au sous-ordre des Nemichthydiformes. Danish "Dana" Expeditions 1920-22. No. 4:1-113. Ruggieri, G. D., R. F. Nigrelli, P. M. Powles, and D. G. Garnett. 1970. Epizootics in yellowtail flounder, Limanda ferruginea Storer, in the Western North Atlantic caused by Ichthyophonus, an ubiquitous parasitic fungus. Zoological. 55(3):57-62. Ruzecki, E. P. 1979. On the water masses of Norfolk Canyon. Ph.D. Dissertation, College of William and Mary. 293 pp. Ribelin, W. E., and G. Migaki. 1975. Pathology of fishes. The Univ. of Wisconsin Press. 1004 pp. 298 Samyshev, E. Z., and S. V. Schetinkin. 1971. Feeding patterns of some species of Myctophidae and Maurolicus muelleri: caught in the sound-dispersing layers in the Northwestern African area. Annales Biol. 1971:212-215. Schultz, L. P. 1964. Family Sternoptychidae. In: Fishes of the Western North Atlantic Sears Found. Mar. Res. 1(4):241-273. Sedberry, G. R., III. 1975. Food habits of some demersal fishes of the continental slope and rise. M. A. Thesis, College of William and Mary. 66 pp. Sedberry, G. R., and J. A. Musick. 1978. Feeding strategies of some demersal fishes of the continental slope and rise off the mid-Atlantic coast of the U.S.A. Mar. Biol. 44:357-375. Sindermann, C. J., and L. W. Scattergood. 1954. Diseases of the fishes of the Western North Atlantic. II. Ichthyosporidium disease of the sea herring (Clupea harengus). Dept. Sea Shore Fish. Res. Bull. 19:40 pp. Smith, J. W. 1971. Thysanoessa inermis and T.- longicaudata (Euphausiidae) as first intermediate hosts of Anisakis sp. (Nematoda: Ascaridata) in the Northern North Sea, to the North of Scotland and at Faroe. Nature. 234(5330) :478. Sokal, R. R., and F. J. Rohlf. 1969. Biometry. W. H. Freeman and Co. 776 pp. Sprague, V. 1965. Ichthyosporidium Caullery and Mesnil 1905, the name of a genus of fungi or a genus of sporozoans? Syst. Zool. 14:110-114. St. John, P. A. 1958. A volumetric study of zooplankton distribution in the Cape Hatteras area. Limnol. Oceanogr. 3, 4:387-397. Stunkard, H. W. 1977. Studies on Tetraphyllidean and Tetrarhynchidean Metacestodes from squids taken on the New England Coast. Biol. Bull. 153(2):387-412. Tchernavin, V. V. 1953. The feeding mechanisms of a deep-sea fish Chauliodus sloanei Schneider. Brit. Mus. (Nat. Hist.). 101 pp. Tighe, K. A. 1975. Systematics, vertical distribution, and life histories of the Serrivomerid eels in the Bermuda Ocean acre. M.S. Thesis, Univ. of Rhode Island. 299 Vinogradov, M. E. 1970. Vertical distribution of the oceanic zooplankton. Translation from Russian by National Technical Information Service, Israel Program for Scientific Translations. 339 p p . Walter, J. B., and M. S. Israel. 1970. General pathology. J. and A. Churchill. 1116 pp. Warsh, C. E. 1975. Physical oceanography historical data for Deepwater Dumpsite 106. In: NOAA Dumpsite Evaluation Report 75-1. May 1974 Baseline Investigation of Deepwater Dumpsite 106:105-140. Weitzman, S. 1974. Osteology and evolutionary relationships of the Stenoptychidae, with a new classification of Stomiatoid families. Bull. A m e r . Mus. Nat. Hist. 153(3):331-478. Wenner, C. A. 1978. Making a living on the continental shelf slope and in the deep-sea: Life history of some dominant fishes of the Norfolk Canyon area. Ph.D. Dissertation, College of William and Mary. 294 pp. Wiebe, P. H., K. H. Burt, S. H. Boyd, and A. W. Morton. 1976. A multiple opening/closing net and environmental sensing system for sampling zooplankton. Jour. Mar. Res. 34(3): 313-326. Windell, J. T., and S. H. Bowen. 1978. Methods for study of fish diets based on analysis of stomach contents. In: Methods for Assessment of Fish Production in Fresh Waters, T. Bagenal (Ed.). IBP Handbook No. 3, Blackwell Scientific Publications 219-226. Worner, F. G. 1979. Nahrungsbiologie dreier Myctophidenarten, Benthosema glaciale (Reinhardt, 1837), Ceratoscopelus maderensis (Lowe, 1839) und Myctophum (M.) punctatum Rafinesque, 1810, Nordwestafrikanischen Auftriebsgebietes. "Meteor" Forsch.-Ergebnisse Reihe D, No. 30:41-61. Yamaguti, S. 1959. Cestodes of vertebrates. Systema Helminthum. Vol. 2:1-860. 300 VITA John Vianney Gartner, Jr. Born in Rockville Centre, Long Island, N.Y., August 9, 1954. Graduated from Albertus Magnus High School, Bardonia, N.Y., June 1972. Received A.A.S. from Rockland Community College, Suffern, N.Y., January 1974. Entered Southampton College of Long Island University, September 1974, graduated June 1976, magna cum laude, B.S. in Marine Biology. Entered the School of Marine Science, College of William and Mary (Virginia Institute of Marine Science) September 1976, as a graduate research assistant, Division of Fisheries, Department of Ichthyology.