W&M ScholarWorks
Dissertations, Theses, and Masters Projects Theses, Dissertations, & Master Projects
1983
Distribution of macrobenthic invertebrates on the North Carolina continental shelf with consideration of sediment, hydrography and biogeography
Donald Paul Weston College of William and Mary - Virginia Institute of Marine Science
Follow this and additional works at: https://scholarworks.wm.edu/etd
Part of the Marine Biology Commons
Recommended Citation Weston, Donald Paul, "Distribution of macrobenthic invertebrates on the North Carolina continental shelf with consideration of sediment, hydrography and biogeography" (1983). Dissertations, Theses, and Masters Projects. Paper 1539616903. https://dx.doi.org/doi:10.25773/v5-k0af-nv08
This Dissertation 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]. INFORMATION TO USERS
This reproduction was made from a copy of a document sent to us for microfilming. While the most advanced technology has been used to photograph and reproduce this document, the quality of the reproduction is heavily dependent upon the quality of the material submitted.
The following explanation of techniques is provided to help clarify markings or notations which may appear on this reproduction.
1.The sign or “target” for pages apparently lacking from the document photographed is “Missing Page(s)”. If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure complete continuity.
2. When an image on the film is obliterated with a round black mark, it is an indication of either blurred copy because of movement during exposure, duplicate copy, or copyrighted materials that should not have been filmed. For blurred pages, a good image of the page can be found in the adjacent frame. If copyrighted materials were deleted, a target note will appear listing the pages in the adjacent frame.
3. When a map, drawing or chart, etc., is part of the material being photographed, a definite method of “sectioning” the material has been followed. It is customary to begin filming at the upper left hand comer of a large sheet and to continue from left to right in equal sections with small overlaps. If necessary, sectioning is continued again—beginning below the first row and continuing on until complete.
4. For illustrations that cannot be satisfactorily reproduced by xerographic means, photographic prints can be purchased at additional cost and inserted into your xerographic copy. These prints are available upon request from the Dissertations Customer Services Department.
5. Some pages in any document may have indistinct print. In all cases the best available copy has been filmed.
University Micrrifilms International 300 N. Zeeb Road Ann Arbor, Ml 48106 r; 8407031
Weston, Donald Paul
DISTRIBUTION OF MACROBENTHIC INVERTEBRATES ON THE NORTH CAROLINA CONTINENTAL SHELF WITH CONSIDERATION OF SEDIMENT, HYDROGRAPHY AND BIOGEOGRAPHY
The College of William and Mary in Virginia Ph.D. 1983
University Microfilms International 300 N. Zeeb Road, Ann Arbor, Ml 48106
PLEASE NOTE:
In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark V .
1. Glossy photographs or pages ______
2. Colored illustrations, paper or print ______
3. Photographs with dark background ______
4. Illustrations are poor copy ______
5. Pages with black marks, not original copy ______
6. Print shows through as there is text on both sides of page ______
7. Indistinct, broken or small print on several pages ^
8. Print exceeds margin requirem ents ______
9. Tightly bound copy with print lost in spine ______
10. Computer printout pages with indistinct print ______
11. Page(s)______lacking when material received, and not available from school or author.
12. Page(s)______seem to be missing in numbering only as text follows.
13. Two pages numbered ______. Text follows.
14. Curling and wrinkled pages ______
15. Other ______
University Microfilms International
DISTRIBUTION OF MACROBENTHIC INVERTEBRATES
ON THE NORTH CAROLINA CONTINENTAL SHELF
With Consideration of Sediment, Hydrography and Biogeography
A Dissertation
Presented to
The Faculty of the School of Marine Science
The College of William and Mary 1n Virginia
In Partial Fulfillment
Of the Requirements for the Degree of
Doctor of Philosophy
by
Donald P. Weston
1983 APPROVAL SHEET
This dissertation is submitted in partial fulfillment of
the requirements for the degree of
Doctor of Philosophy
^ Author
Approved, July 1983
Donald F. Boesch, Ph.D
QyL*, 9 . (yy^ — Jo h /J. Magnflson, Ph.D. (j Un/versity of Wisconsin-madison TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS...... v
LIST OF TABLES...... vii
LIST OF...... FIGURES...... ix
ABSTRACT...... xii
INTRODUCTION...... 2
CHAPTER 1. Macrobenthos-sediment relationships off Cape Hatteras, North Carolina: application of multivariate analytical techniques.
Abstract ...... 7 Introduction ...... 8 Methods ...... 10 Statistical analyses ...... 12 Results Temperature and sa lin ity ...... 14 Sediments ...... 16 Numerical classificatio n ...... 19 Discriminant analysis...... 26 Ordination ...... 29 Discussion Evaluation of multivariate techniques ...... 35 Significance of sediment parameters ...... 38 Literature cited...... 44 Appendix ...... 51
CHAPTER 2. The effect of an oceanic front on the macrobenthos of the North Carolina continental shelf.
Abstract ...... Introduction ...... Methods ...... Results Sediments ...... Temperature and sa lin ity ...... Biota ...... Zoogeography ...... Relative importance of physical variables ...... Temporal s ta b ility ...... Page
Discussion ...... 97 Literature cited...... 102
CHAPTER 3. Comparative biogeography of the North Carolina continental shelf macrobenthos.
Abstract ...... 106 Introduction ...... 107 Methods ...... 110 Results Physical environment ...... 113 Biological environment ...... 115 Blogeographic a ffin itie s ...... 117 Rate of faunal change with latitu d e ...... 125 Discussion ...... 129 Biogeography and larval dispersal ...... -132 Pal eobi ogeography ...... 135 Conclusions ...... 141 Literature cited...... 142 Appendices...... 149
VITA ...... 154
iv ACKNOWLEDGMENTS
This work could not have been accomplished without the support of my co-major professors, Donald Boesch and Robert Diaz. Don's support and guidance, both during and after his tenure at VIMS, was invaluable and very much appreciated. I am grateful for Bob's advice and assistance, particularly in regard to the numerous details associated with graduation while I was completing my degree work in abstenia. I am grateful to all my committee members for their review of this dissertation.
Numerous students and staff of the University of Wisconsin assisted with the field sampling. The contributions of John Magnuson,
Cynthia Harrington, Frank Rahel, Donald Stewart and Stephen Brandt are especially appreciated. I am particularly grateful to John Magnuson for allowing the use of sediment and physical data.
The computer analyses would not have been possible without the time and patience of Pamela Langley. Many thanks go to Terri Kirby for the patience with which she suffered through typing of the major portion of the manuscript. Acknowledgments (continued)
Numerous people deserve recognition for their assistance in species identifications: Nancy Maciolek Blake, Marian Pettibone, Peter
Kinner, Mary Petersen (Polychaeta); Christen Erseus (Oligochaeta);
Louis Kornicker (Ostracoda); Edward Bousfield (Amphipoda).
Finally a special thank you is due to my wife Deborah, whose constant support and encouragement made completion of this dissertation possible. The original manuscript benefitted greatly from her skill in editing and our discussion of Ideas.
This research was supported by the National Science Foundation under Grant No. 0CE8002062 to the University of Wisconsin. ■L
LIST OF TABLES
Chapter 1
Table Page
1. Temperature (°C) and salin ity (°/oo) observations (mean and range) of bottom water 1n the three strata during the four sampling periods (from physical-chemical data base, J. Magnuson, University of Wlsconsin-Madison) ...... 15
2. Temporal variation In grain size percentages at a site within each of the three sampling strata. Successive samples at each site were taken within 1 km of each o th e r...... 18
3. Untransformed mean values of sediment parameters used In the discriminant analysis ...... 27
4. Standardized discriminant function coefficients and total structure coefficients showing relative Importance of all variables on the f ir s t two discriminant functions (DF). Those considered most important are underlined ...... 28
5. Spearman rank correlation coefficients between sedi ment parameters and sample scores on the first two DCA axes. (*=o<<0.05, **=®*<0.01) ...... 33
Chapter 2
Table Page
1. Grain size percentages and selected sediment para meters for all stations samples ...... 73
2. Temperature and salinity observations of bottom water within each of eight equidistant strata during each of the four sampling periods (from physical- chemical data base, J. Magnuson, University of Uisconsin-Madison) ...... 74
3. Untransformed mean values of all variables used in the discriminant analysis ...... 89
v ii Chapter 2 (continued)
Table Page
4. Total structure coefficients for all variables on the three discriminant functions. Those variables considered most Important on each function are underlined ...... 91
5. The three numerically dominant species at four selected sampling sites during each of the four sampling periods. The number 1n parentheses in dicates the total number of Individuals collect ed 1n the three replicate samples ...... 93
Chapter 3
Table Page
1. Seasonal and annual means and ranges of bottom water temperatures (°C)’throughout the study area (from physical-chemical data base, 0. Magnuson, Uni versity of Wlsconsin-Madison) ...... 114
2. Macrofaunal groups collected. Numbers 1n parentheses are total number of species collected during all four sampling periods and number collected during June alone ...... 116
3. Biogeographic affinities and distributions of species in the four dominant macrobenthic groups, as In dicated by percentages of total species con sidered (n). Collections from all four sampling periods are Incorporated ...... 120 LIST OF FIGURES
Chapter 1
Figure Page
1. Location of the three sampling strata with the speci fic collection sites indicated. Depth contours 1n meters ...... 11
2. Grain size parameters of surficial sediments 1n the study area: (a) median grain size, (b) percentage of silt and clay ...... 17
3. Temporal stability of mean grain size of surficial sediments in the H a t t d n is :M1d stratum as evidenced by the comparability of samples taken seven years apart: (a) samples taken in 1970 and 1971 (re drawn from Hunt et a l., 1971; (b) samples taken in 1977 and 1978 during the present study ...... 20
4. Numerical classification of all collections of macro benthos. Strata~H:N=Hatteras:North; H:M=Hatte- ras: M1d; H:S*Hatteras:South. Season—(Sp)=Spr1ng; (Su)=Suitmer; (FJ^Fal 1; (W)=W1nter ...... 21
5. Separation of station groups defined by numerical classification on the basis of median grain size and sediment sorting coefficient ......
6. Species characteristic of Group 1 stations and their abundance as a function of median grain size. Median grain size (x axes) given in phi units, density (y axes) given in number of individuals/ ......
7. Species characteristic of station Groups 2, 3 and 4 and th eir abundance as a function of median grain size. Median grain size (x axes) given in phi units, density (y axes) given in number of indivi duals/m^......
8. Centroids (indicated by asterisks) and limits of station groups in two-factor discriminant space. Vectors are determined from the total structure coefficients and indicate the relative orientation of the environmental variables. SRT=sediment sorting coefficient, SHL=percentage shell and gravel, VCS=percentage very coarse sand, CS=per-
ix Page centage coarse sand, MS-percentage medium sand, FS=percentage fine sand, VFS*percentage very fine sand, SLT=percentage s iltand ctay ...... 30
9. Location of station groups 1n ordination space. Length of axes are scaled 1n proportion to the eigenvalues ...... 32
Chapter 2
Figure Page
1. The location of sampling stations on the North Carolina continental shelf. Stations are designated by minutes north of latitude 35° 00.5'N...... 70
2. Isotherms observed during each of the four sampling periods 1n along-shelf transects across the frontal area. Solid triangles denote the position of eachbathythermograph cast ...... 76
3. Continuous record of bottom water temperature as measured by a stationary chart recorder anchored 0.5 m above the bottom at 35°35.5'N, 75°13.0'W: (a) temperature during 8-14 August 1977; (b) temperature during 23-29 October 1977... 79
4. Selected community parameters at each station 1n the June 1977 sampling transect: (a) density of total macrofauna; (b) total number of species collected 1n three replicate grab samples ...... 80
5. Percentage of species in each of four distributional catagories at all stations along the June 1977 sampling transect ...... 82
6. Changes in density (no. 1ndiv./m2) across the study area 1n June 1977 for selected macrobenthic species. Note use of log scale for densities of Harmothoe extenuata and Spisula solldissima simi1 i s . Shaded area represents zone of most frequent frontal occurrence ...... 84
7. Changes in density (no. indiv./m2) across the study area in June 1977 for the polychaete Glycera oxycephala. Shaded area represents zone of most frequent frontal occurrence ...... 86
x t .
Page 8. Numerical classification of all stations sampled in June 1977. Station 39 was sampled on three separate occasions ...... 88
9. Centroids (Indicated by asterisks) and limits of station groups 1n two-factor discriminant space. Vectors indicate the relative orien tation of the most Important environmental variables ...... 92
10. Temporal stability, as determined by the Peterson index, for five sampling site s. — ) = stations 35, 39 and 49 for all sampling periods; ( — ) * stations 23, 25, 39 and 49 for three sampling periods excluding August; (— ) = stations 31, 35, 39 and 49 for three sampling periods excluding October ...... 95
Chapter 3
Figure Page
1. Location of the ten sampling areas on the North Carolina continental shelf ...... 111 2. Geographic extent of the eight distributional categories into which all macrofaunal species considered were grouped. Arrows Indicate species distributions may extend beyond area shown ...... 118
3. Proportional distribution of species within the eight distributional categories as grouped by major taxon and sampling area. Number 1n parentheses indicates number of species of each taxon collected ...... 122
4. Rate of faunal change with latitude for the four major macrofaunal groups. Mean latitudinal distance and standard deviation between all station pairs within the prescribed similarity categories 1s shown ...... 126
5. Proportional distribution of station pairs within each of the prescribed similarity categories for the four major macrofaunal groups ...... 127
xi Ml
ABSTRACT
The macrobenthic invertebrates of the North Carolina continental shelf in the vicinity of Cape Hatteras were sampled during four seasonal cruises, June 1977 to January 1978. The diverse assemblage of macrobenthic organisms collected was comprised of over 600 species distributed among 14 phyla. A variety of factors found to structure species distributions in the study area are treated: 1) macrobenthos-sediment relationships, 2) species distributions within an area affected by an oceanic front, and 3) comparative biogeography and paleobiography.
Macrobenthos-sediment relationships were investigated in the area surrounding Diamond Shoals, a submarine extension of Cape Hatteras, where a broad range of sediment textures was found in a very localized area. Four benthic assemblages, each representative of specific sediment regimes, were recognized in the Diamond Shoals region. The results of multivariate analysis were interpreted as indicating that the percentages of very fine sand and silt and clay were of greatest biological significance. It is suggested that the importance of the finer particles is due to their influence on sediment permeability and organic content. The degree of particle sorting was also important in accounting for some faunal differences with fossorial species predominating in the most well-sorted sediments.
Thermal factors were found to be the dominating factor in species distributions on the shelf north of Cape Hatteras, an area occupied by a sharp thermal front between Gulf Stream and Virginia Shelf Water. This front was characterized by a temperature gradient of 12°C and a salinity gradient of 5°/oo over a distance of Z7 km or less. The benthic community exposed to the greatest thermal variability within the front was more speciose than the benthos of more thermally stable areas to the north and south, but otherwise demonstrated no unique characteristics. For several species the front represented a zoogeographic barrier with cold-water species restricted to the northern portion of the frontal area and warm-water species restricted to the southern portion.
The North Carolina region in general was confirmed to represent a zoogeographic barrier to several major macrobenthic taxa. At mid-shelf depths the Cape Hatteras region was far more effective in limiting the northward distribution of southern species than in limiting the southward distribution of northern species. Biogeographic affinities, extent of geographic range, and ability to traverse the Cape Hatteras area were compared among the four most speciose macrofaunal groups, the Polychaeta, Amphipoda, Bivalvia and Gastropoda. The polychaetes and amphipods generally exhibited the broadest geographic distributions while the molluscs, particularly the gastropods, were the most narrowly distributed. These differences are related to the dispersal capabilities and comparative degree of eurytopy among the four macrofaunal groups considered. The present-day distributions and biogeographic affinities of the North Carolina macrofauna are also a function of the geologic history of the northwestern Atlantic and the evolutionary origin of the fauna.
xii DISTRIBUTION OF MACROBENTHIC INVERTEBRATES
ON THE NORTH CAROLINA CONTINENTAL SHELF
With Consideration of Sediment, Hydrography and Biogeography INTRODUCTION
The continental shelf in the vicinity of Cape Hatteras, North
Carolina has a number of geological, physical and biological characteristics which make it unique among all other shelf areas along the North American Atlantic seaboard. Much of its uniqueness can be attributed, either directly or indirectly, to the bathymetry of the region. Along most of the American Atlantic coast north of Cape
Canaveral, Florida the continental shelf exceeds 100 km in width.
However off Cape Hatteras the shelf narrows to only 23 km. Much of this distance is occupied by Diamond Shoals, a submarine extension of
Cape Hatteras characterized by dynamic, shifting sands and water depths of generally less than 10 m.
Because of the constriction of the continental shelf off Cape
Hatteras and the shallow depths of Diamond Shoals, there is little opportunity for exchange of shelf water around the Cape. Consequently the hydrographic regimes on either side of the Cape are very different. South of Cape Hatteras the shelf is occupied by the
Carolina Shelf Water. Excluding inshore waters, water temperatures of
Carolina Shelf Water are generally high (>15°C) throughout the year.
The shelf north of Cape Hatteras is occupied by the Virginia Shelf
Water. This water mass is characterized by pronounced seasonal variability. While high temperatures north of Cape Hatteras are nearly as great as those south of the Cape, low temperatures can reach
2 3
4°C. The seaward boundary of shelf waters 1n the Cape Hatteras region
1s formed by the Inshore margin of the Gulf Stream. The Gulf Stream flows northward along the outer shelf throughout the South Atlantic
Bight but turns eastward near Cape Hatteras, moving away from the coast.
As a consequence of the hydrographic characteristics of the Cape
Hatteras area, the region 1s of great biological significance as a zoogeographic barrier. Nearly every major group of marine flora and fauna examined has been found to exhibit rapid latitudinal change In species composition within the Cape Hatteras area. A large number of tropical and subtropical species reach their northern limit In the vicinity of Cape Hatteras, while many boreal species are similarly limited 1n th eir southward distribution.
In June 1977 an Intensive, m ulti-disciplinary study was launched by the University of Wisconsin to explore the relationship between many of the unique features of the Cape Hatteras region and the biota.
A sharp Gulf Stream-Vlrginia Shelf Water front located north of Cape
Hatteras was of particular Interest, but a broad portion of the North
Carolina shelf was Investigated, extending from off Cape Lookout northwards to an area 55 km south of the Virginia border. Four sampling cruises, spaced to as permit evaluation of seasonal variability, were made between June 1977 and January 1978. Three major components of the biota were intensively sampled: 1) demersal fishes, 2) megabenthic invertebrates, particularly decapods and echinoderms, collected by o tter trawl, and 3) macrobenthic 4
Invertebrates collected by Smith-Mclntyre grab. A variety of physical and chemical parameters were also measured concurrently with the blologial sampling.
This dissertation examines the macrobenthic invertebrates of the
Cape Hatteras region and attempts to Identify the factors responsible for the distribution patterns observed. Physical determinants of distribution (substrate, temperature and salinity) and Intraspecific biological factors (e.g. dispersal ability) are given primary consideration. Interspecific biological factors (e.g. predation, competition) which may influence distribution could not be explicitly evaluated using this sampling scheme. Physical and biological factors which structure faunal distribution are addressed in three Independent chapters, each of which examines factors operating on successively broader spatial scales.
Chapter 1 considers faunal changes occurring over relatively short distances of a few kilometers as a result of substrate differences. Only samples collected in the 1mned1ate vicinity of
Diamond Shoals are considered 1n order to Include a broad range of sediment types yet minimize differences in temperature and salinity which might occur over a greater latitudinal distance. The relationhip between grain size and faunal distribution is examined by several multivariate techniques including discriminant analysis and detrended correspondence analysis, a recently developed ordination procedure. Chapter 2 discusses changes In faunal composition occurring over a 50 km distance extending from Oregon Inlet to slightly north of Cape
Hatteras. This area 1s occupied by a sharp oceanic front characterized by strong gradients 1n temperature and salinity. The
Importance of the front In structuring the benthic community 1s evaluated, as well as Its role as a blogeographlc barrier 1n limiting the latitudinal range of benthic organisms.
Distributions of species on a global scale are considered 1n
Chapter 3 as the relative effectiveness of the Cape Hatteras area as a zoogeographic barrier to a barrier to a variety of major benthic taxa
1s evaluated. A taxon's ability to traverse the Cape Hatteras area Is considered In light of each taxon's mode of reproduction and dispersal capabilities. Global patterns of distribution are interpreted in relation to both present day dispersal mechanisms and evolutionary patterns of dispersal throughout the Tertiary and Quaternary. Chapter 1
MACROBENTHOS-SEDIMENT RELATIONSHIPS OFF
CAPE HATTERAS, NORTH CAROLINA:
APPLICATION OF MULTIVARIATE ANALYTICAL TECHNIQUES
D. P. Weston
6 ABSTRACT
The complex current regime associated with Cape Hatteras, North Carolina results 1n an unusually broad range of sediment textures for open shelf areas of comparable size. The median grain size ranged from coarse to very fine sand, while the percentage of si It and clay ranged from 0 to 27%. Four assemblages of macrobenthic species were recognized, separable on the basis of sediment type: 1) a muddy, very fine sand assemblage dominated by Lumbrinerls 1mpat1ens, 2) a fine to medium sand assemblage domnlated by Polyqordlus sp., 3) a well-sorted, fine sand assemblage dominated by Protonaustorlus cf. delchmannae, and 4) a medium to coarse sand assemblage characterized b.y Hemlpodus roseus and Hesionura elonqata. Multiple discriminant analysis and detrended corespondence analysis, a linear ordination technique, were used to identify which of 8 sediment parameters were most useful 1n Interpreting faunal patterns. Sediment sorting, as reflective of sediment mobility, was Important 1n determining the dominance of fossorlal species. The percentages of very fine sand and silt and clay were found to be of biological significance because of th eir effects on sediment permeability and the type and availability of food resources. INTRODUCTION
The Importance of sediment in structuring macroinvertebrate communities has been often demonstrated, but mainly 1n estuaries
(McNulty et a l., 1962; Nichols, 1970; Bloom et a l., 1972) or shallow marine embayments (Sanders, 1958; Young and Rhoads, 1971; Franz, 1976;
Blernbaum, 1979). This emphasis 1s largely a result of both the accessibility of these environments and the wide diversity of sediment types present in relatively close proximity. Animal-sediment relationships on the open continental shelf have received considerably less attention and several of the studies of this environment have found l i t t l e influence of sediment type on animal distribution
(Buchanan, 1963; Day et a l ., 1971).
The continental shelf off Cape Hatteras, North Carolina has a number of unique features which make 1t especially suitable for investigations of animal-sediment relationships. Diamond Shoals, extending seaward across two-thirds of the shelf off Cape Hatteras, is characterized by water depths of less than 10 m. The complex current regime associated with Diamond Shoals results 1n an unusually broad range of sediment textures for open shelf areas of comparable size.
For example, in the Middle Atlantic Bight, the percentage of s ilt and clay of shelf sediments is generally less than 10% (Boesch and Bowen, in prep.). In contrast, shelf sediments in the Cape Hatteras region may range from near 0 to in excess of 27% silt and clay.
8 Additionally, the sedimentary environment of the Cape Hatteras area may have important zoogeographlc implications. The Cape Hatteras area acts as an effective zoogeographlc barrier for a wide variety of faunal groups (Briggs, 1974; Cerame-V1vas and Gray, 1966). Though temperature lim itations are undoubtedly of major Importance in the termination of many species distributions 1n this area, factors other than temperature appear to be Involved 1n some cases (Vernberg and
Vernberg, 1970). Harrington (1981) suggested that the unstable sands of the Diamond Shoals area limit the distribution of those echinoderm species not having a sufficiently long pelagic larval life to survive passage over the Shoals.
This study had three objectives:
1. to describe the sedimentary environment of the Cape Hatteras
area and its relationship to the macrobenthic communities,
2. to examine the utility of several multivariate statistical
techniques 1n Investigation of animal-sediment relationships,
3. to identify the sediment grain size parameters most
responsible for community patterns. METHODS
Collections were made during four cruises of the R/V Eastward spring (M^y 31-June 14, 1977), sumner (August 3-16, 1977), fall
(October 17-31, 1977) and winter (January 4-16, 1978). The sampling area (Figure 1) was subdivided into three strata, Hatteras:North,
Hatteras:Mid and Hatteras:South, located to the north, east and south of Diamond Shoals, respectively. The sampling sites were aligned along the 30 m depth contour, but because of the steepness of the bottom topography, sampling depths varied from 23-54 m. On each cruise, sampling locations were chosen randomly within each stratum with equalized sampling effort between strata. A total of 76 samples was collected for biological and sediment analyses over the four cruises.
At each s ite , bottom water temperature and salinity were measured using a shallow-water mechanical bathythermograph (BT) and Beckman induction salinometer. Measurements were also taken periodically with a reversing thermometer and Hytech salinometer, and the BT and Beckman salinometer readings corrected accordingly.
Macrofauna was collected using a 0.1 m2 Smith-McIntyre grab.
Most organims were separated from the sediment by elutriation through a 0.5 mm mesh Nitex screen, with the material remaining in the container washed through a 1.0 mm mesh screen. Both the 0.5 and 1.0
10 1 . .
Figure 1 Location of the three sampling strata with the
specific collection sites indicated. Depth contours
in meters.
11 E.
• N
\ r - '
T. Of •••-V
•• 75 75 40' 75^30' 75* 20'
IOO urn fractions were placed 1n MgCl2 t0 anesthetize the organisms, and then transferred to 10% buffered formalin stained with Rose Bengal.
Sediment samples collected from each grab were dried at 100°C and
100 g aliquots obtained by use of a sediment cutter. The material was sieved through a series of Standard Sieves with silts and clays combined as pan weight. Sediment parameters were determined using modified Inman measures (Inman, 1952).
Statistical analyses
In order to delimit zones of faunal similarity cluster analysis was performed, deleting those species which occurred In 3 or less samples. Algorithms used Included log-transformat1on of species abundances, the Bray-Curt1s sim ilarity measure (Bray and Curtis,
1957), and flexible sorting wlthyfl established at -0.25 (Clifford and
Stephenson, 1975).
Ordination was performed using detrended correspondence analysis
(DCA, Hill and Gauch, 1980). DCA 1s a modification of reciprocal averaging ordination which alleviates many of the problems Inherent in the la tte r technique (H111 and Gauch, 1980). Specifically, DCA eliminates the 'arch' or 'horseshoe' effect of reciprocal averaging resulting from the quadratic dependency of the second axis on the first. Additionally, DCA rescales the axes to eliminate the compression of ordination distances at th eir ends.
Multiple discriminant analysis (N1e et al., 1975) was used to evaluate the significance of the group separation derived by cluster analysis, and to interpret this separation 1n light of a variety of sedimentary variables. Sediment percentages used 1n the analysis were transformed by arcsinVp'to induce normality. The sorting coefficient was left untransformed as it was already in logarithmic ‘phi1 units. RESULTS
Temperature and salinity
The Cape Hatteras region 1s marked by abrupt temperature and
salinity fronts between widely divergent water masses (Herbst et a l.,
1979). However the three Hatteras sampling strata were all exposed to
similar conditions of temperature and salinity owing to their
relatively small geographic extent (Table 1). The greatest
differences observed 1n mean temperature and salinity between strata
during any cruise were 2.5°C and 1.1 °/oo during the fall sampling
period, a period in which the Hatteras:North strata was influenced by
a thermal front (Harrington, 1981).
The temperature of bottom water varied seasonally from 16°C
during the winter cruise to 27°C during the summer sampling period.
Bottom salinity ranged from 32 °/oo, indicative of southerly-flowing
Virginia Shelf Water, to 36 °/oo, representative of the
northeasterly-flowing Gulf Stream. It is noteworthy that the lowest
salinities measured were during the fall and winter cruises,
Indicating intrusion of Virginia Shelf Water in the study area during this period. This intrusion has been attributed to prolonged periods
of strong northeasterly winds which may force Virginia Shelf Water
over Diamond Shoals and into Raleigh Bay (Wells and Gray, 1960; Gray
and Cerame-Vivas, 1963; Hunt et a l., 1977).
14 Table 1. Temperature (°C) and salinity (%>o) observations (mean and range) of bottom water 1n the three strata during the four sampling periods (from physical-chemical data base, J. Magnuson, Univ. of Wisconsin-Madison).
Strata
Season Hatteras:North Hatteras:Mid Hatteras:South
SPRING (June 1977) Temperature 24.4 23.9 24.1 (23.8 - 25.2) (22.5 - 24.9) (23.2 - 24.7)
Salinity 35.03 34.70 34.82 (34.32 - 35.94) (34.20 - 35.00) (33.91 - 35.60)
SUMMER (Aug. 1977) Temperature 26.9 25.5 ?•> 4 (26.3 - 27.5) (23.5 - 26.2) (25.1 -*25.9)
Salinity 35.09 35.22 35.35 (34.65 - 35.57) (34.69 - 35.58) (34.74 - 36.25)
FALL (Oct. 1977) Temperature 20.3 22.8 21.3 (16.4 - 26.7) (19.4 - 26.3) (19.6 -*25.6)
Salinity 33.00 34.08 34.11 (31.68 - 35.71) (32.92 - 35.72) (33.23 - 35.50)
WINTER (Jan. 1978) Temperature 17.3 17.3 15.8 (15.4 - 19.2) (15.8 - 18.6) (14.8 - 16.7)
Salinity 35.68 36.00 35.57 (32.64 - 36.29) (35.10 - 36.30) (33.28 - 35.98)
15 Sediments
The surficlal sediments of the study area showed considerable spatial variation as a result of both the bathymetry of the Diamond
Shoals area and the complex current regime. Median grain size (Figure
2a) ranged from 0.11 0 (coarse sand) to 3.41 0 (very fine sand). The coarsest sand occurred 1n portions of the Hatteras:North stratum. The finest sand was 1n an area of the Hatteras:South stratum which probably receives fine sediments winnowed from Diamond Shoals (Hunt et a l., 1977).
The percentage of silt and day (Figure 2b) was generally low.
The majority of the surflcial sediments of the study area had less than 5% s ilt and clay though a restricted region of the Hatteras:Mid and Hatteras:South strata had as much as 27%.
In order to determine the comparability of sediment data over the four cruises, the question of temporal change In substrate composition needs to be addressed. However since no attempt was made to return to previously sampled stations on successive cruises seasonal changes are difficult to assess. The best available data, provided by seasonal samples taken within less than 1 km of each other (Table 2), indicated l i t t l e temporal change. The only variation evident was shift towards the finer particle size with time in Hatteras:South. Since this variation could also reflect substrate patchiness, these apparent seasonal differences may be a misrepresentation of the true temporal changes. Figure 2 Grain size parameters of surficial sediments in the
study area: (a) median grain size, (b) percentage
of silt and clay.
17 ( 0 )
MEOIAN GRAIN SIZE ■ iV E R Y FINE SAND (3-4 0) ESfl FINE 8AND (2-3 0) E H MEDIUM SAND (1-20) E 3 COARSE SAND (0-1 0 )
PERCENT SILT AND CLAY □ < 5 % Eh) 5 - 1 0 % H i 10-15% ■ I 15-20% ■ I >20% . m
Table 2. Teiporal variation in grain size percentages at a site within each of the three sanpling strata. Successive sanples at each site were taken within 1 km of each other. 10 2 2 ft •r I - 00 *52 *8 00 Li- •r- ■§ £ s * 2 • © • • Fd fe fe Fd O - O 10 0 *1-0 *-« O O O B 8 IB 8 8 I 10 «o co CO S 8 8 8 £ o o o o o o o o o o 0 H H 00 H H K H 8 R 3 W &J fe a * i < & © 18 e a a fe S 1 8SK ? r4 «? I CO S © 3 a 3 • • fe fe n * 2 * o t H v-l & © rH a a e s ^ is fe a & a 8 88 fe 5t R & e ? 3 5? fe a g Ol(O e e e s a s a fe fe fe o o o o a a a a a a a s f 8 fe 3 • • • • • • • I S M i S - p 9 • • 83 Ir 3 r- -I £ 3 - • w © •• 1- -0 0 1-10 co u> Fd s a s a s a a a s O r-»VO© • » • i • • • • • • • 19 Even on a long-term basis substrate composition is very stable as evidenced by the comparability of two independent studies. In 1970 and 1971, Hunt et al. (1977) sampled sediments around Diamond Shoals in an area that partially overlapped the present study area. The mean grain size (Figure 3) showed remarkable consistency considering the fact that the two sets of samples were taken seven years apart. Numerical classification Cluster analysis was performed to group the stations according to faunal sim ilarity. Four discrete groups resulted (Figure 4) with the second group consisting of three smaller groups. It is important to note that temporal changes were of no significance at the sim ilarity levels at which groups were defined. Seasonal faunal differences were not responsible for the major groupings but were expressed only as sub-groups within each major cluster. It is clear that sedimentary parameters were primarily responsible for group separation with group differences largely explained on the basis of median grain size (Figure 5). The faunal dissimilarity of groups 2 and 4 is surprising since their median size was similar, however the sediments of group 4 were generally better sorted. Group 1 was comprised of the fine and very fine sands of Hatteras:South and the southern portion of Hatteras:Mid. The percentage of silt and clay was greatest in this group, ranging from 10 to 27% and averaging 16.5%. Species with a high fidelity to Group 1 stations are illustrated in Figure 6. Lumbrineris impatiens was the numerically dominant species in the fine and very fine sands with Figure 3. Temporal stability of mean grain size of surficial sediments in the Hatteras:Mid stratum as evidenced by the comparability of samples taken seven years apart: (a) samples taken in 1970 and 1971 (redrawn from Hunt et a l., 1971); (b) samples taken in 1977 and 1978 during the present study. 20 Z/W///4" 'W / S 'V *• • ' o — w w A 'A '/ A 1 Figure 4. Numerical classification of all collections of macrobenthos. Strata—H:N=Hatteras:North; H:M=Hatteras:Mid; H:S=Hatteras:South. Season— (Sp)=Spring; (Su)=Summer; (F)=Fall; (W)=Winter. 21 -0 .5 SIMILARITY S' Figure 5 Separation of station groups defined by numerical classification on the basis of median grain size and sediment sorting coefficient. 22 6 . MEDIAN MEDIAN GRAIN SIZE (0) SORTING (0) Figure 6. Species characteristic of Group 1 stations and their abundance as a function of median grain size. Median grain size (x axes) given in phi units, density (y axes) given in number of individuals/m^. 23 100 o+------~------.~~.. ~~ 100 1000 -II .!i! ~ .:::o- C!l:: ISO 100 o+------+------~~~~·~11r-.~ ~::l I ~ 100 .-.+ aoo 100 100 o+------+------.~-11-- IOO 100 10 IS o+------+------.~-11r~~--~ 5 10 ~::l0+------+------~---~~--~·~----~ a ..~~~~~ et 100 aoo S' 24 densities reaching over 100 per m^. Other species not illustrated but with a similar affinity to Group 1 stations included Mediomastus californiensis and Prlonospio cristata (Appendix 1). Group 2 was composed of medium and fine sand stations located throughout the study area. Groups 2A and 2B included stations with coarser and finer substrates respectively. Characteristic species (Figure 7) included the polychaetous annelids Magelona cf. pettiboneae, Glycera oxycephala and Polygordius sp. Stations of Group 3 were the coarsest grained 1n the study area, with the percentage of silt and clay less than 1%. This group was comprised of stations primarily 1n the Hatteras:North area but included some sites in both Hatteras:M1d and Hatteras:South. Spisula solidissima sim ilis was the only species abundant in Group 3 stations, although it was not faithful to stations of this group. Many species fidel to Group 3 stations were small, interstitial burrowers such as Hemipodus roseus, Hesionura elonqata and Pionosyllis gesae (Figure 7). These were all collected in low numbers, possibly as a result of loss through the 0.5 mm sieve used. Group 4 stations occupied a relatively small geographic area in Hatteras:North and Hatteras:Mid. Median grain size at these stations ranged from 1.9 to 2.4 0 (fine sand) with less than 1% s i l t and clay. With the exception of one outlier, sediments of Group 4 statons were the best sorted of any in the study area with a standard deviation around the mean grain size about 0.5 0. Three fossorial amphipods, Figure 7. Species characteristic of station Groups 2,3 and 4 and their abundance as a function of median grain size. Median grain size (x axes) given in phi units, density (y axes) given in number of individuals/m2. 25 100 •0 0 i l f • 0 100 soo “ .3 too 100 ^ I * 0 too ° I too S 5 too too B0 o I 0 OJ? $o o BO too 80 tB 0 8B BO ro BO tB a. E o o o» 0 o I I K I * BB O 80 BO 10 S'S 0 g 3 O 9 ♦ ♦ » £ 10 BO I i B 000 5 s BBOO 3 6 0 *r I s BBOO a. B 0 00 3 CL o O 300 tr. BOO 0 100 II 0 100 BOO 1 SOO SI 26 Protohaustorius c f. deichmannae, Rhepoxynius epistomus and Bathyporeia parkeri were characteristic of Group 4 stations (Figure 7). Discriminant analysis Multiple discriminant analysis was performed to identify the sediment variable(s) which would best account for the station groupings created by numerical classification. These groups were discriminated on the basis of eight sediment variables including the sediment sorting coefficient and the percentages of shell and gravel, very coarse sand, coarse sand, medium sand, fine sand, very fine sand and silt and clay (Table 3). Phi size intervals were as defined by Wentworth (1922). Skewness, kurtosis or measures of central tendency were considered redundant to information better conveyed by the grain size percentages and were not included. The overall chi-squared test of among-group differences was highly significant (pc? = 422.94, d.f. = 40). The first and second functions accounted for 70.40% and 26.47% of the among-group variance respectively. Only these functions were considered further in the interpretation. The standardized discriminant function coefficients (SDFC) (Table 4) indicate the relative contribution of the variables in calculating the discriminant scores on each function, and are commonly used as a measure of the relative importance of the variables in discriminating between the groups (Green 1971; Green and Vascotto, 1978). On discriminant function I the sediment sorting coefficient had, by far, the greatest standardized discriminant function coefficient (1.011). L- ^ &I £8 ss a fc a a eg• o * vo • r-i • o • o • cl oo g 3 g 53 Ul S • ••••• tJj CM $ 00 O * - 8 S 8 & S S S si &s «® £ a a o . iA 52, a s a 8 s 5 'id • ••••• s w s ^ s ? 01 +-> s c ‘i 9 a a s s d -«*!••• 00 d - cvj jy cvj I 8• a••••• a ffi at a r-l 00 C\J rH 00 O • ’S ' ® ffi » 8 fc 8 5 1 !!= ^ O• *H ••••• o O h** o • fe j j r 'id > 5? 8 a R 8 ffl O• O«•••• O O •—I o 1te •fee•r* § en vo w *d- t s K3 a PO 00 § CD 27 Table 4. Standardized discriminant function coefficients and total struc ture coefficients showing relative importance of all variables on the first two discriminant functions (DF). Those considered most important are underlined. Standardized discriminant Total structure function coefficients coefficients Variable DF I DF II DF I DF II Sorting 1.011 0.569 0.312 -0.540 % Shell and Gravel -0.529 -0.003 -0.339 -0.611 % Very Coarse Sand -0.346 -0.640 -0.208 -0.732 % Coarse Sand 0.020 -0.026 -0.251 -0.822 % Medium Sand -0.240 -0.426 -0.574 -0.738 % Fine Sand -0.354 0.698 -0.541 0.832 % Very Fine Sand 0.491 0.188 0.970 0.179 % S ilt and Clay 0.356 -0.135 0.984 0.049 28 The sorting coefficient was also important on discriminant function II (SDFC = 0.569) as were the percentages of very coarse (SDFC = -0.640) and fine (SDFC = 0.698) sand. An alternative method of identifying those values of greatest importance in discriminating between groups is the total structure coefficients (TSC) (Klecka, 1980). These coefficients (Table 4) are calculated as the Pearson product-moment correlations between the variables and the discriminant scores on each function. They can be represented graphically, following the procedure of Johnson (1977), as vectors in two-factor discriminant space (Figure 8). Using this procedure discriminant function I is largely representative of the percentages of silt and clay and very fine sand (TSC = 0.984 and 0.970 respectively). The function clearly discriminates between the finer sediments of groups 1 and 2B and the coarser sediments in the remainder of the study area. Discriminant function II, with about one-third the discriminating power of the first function, is more d ifficu lt to interpret. Several of the variables have high total structure coefficients on this function although the percentages of fine sand and coarse sand are the most highly correlated (TSC = 0.832 and -0.822 respectively). The function generally serves to differentiate groups 2A and 3, the coarsest-grained station groups in the study area. Ordination Detrended correspondence analysis was performed in which the groups defined by numerical classification were plotted in ordination E. Figure 8. Centroids (indicated by asterisks) and limits of station groups in two-factor discriminant space. Vectors are determined from the total structure coefficients and indicate the relative orientation of the environmental variables. SRT=sediment sorting coefficent, SHL=percentage of shell and gravel, VCS=percentage of very coarse sand, CS= percentage coarse sand, MS=percentage medium sand, FS=percentage fine sand, VFS=percentage very fine sand, SLT=percentage s i l t and clay. 30 DF.n *4 * z c —DF I space (Figure 9). Only the first (eigenvalue 0.642) and the second (eigenvalue 0.364) axes are presented. Neither the third or fourth axes (eigenvalues 0.279 and 0.217, respectively) are considered further in the analysis. The groups defined by numerical classification generally showed clear spatial separation in the ordination, largely on axis 1. The very fin e sand sta tio n s of H atteras:South (Group 1) showed a high degree of faunal similarity forming a tightly spaced group scoring low on axis 1. The coarser sands of Hatteras:North (Group 3) formed a diffuse grouping scattered over much of axis 2. The segregation of Group 2C in the numerical classification was not reflected in the ordination, with the positions of stations in that group overlapping those of Groups 2A, 3 and 4. Though the ordination axes are not necessarily a reflection of any specific environmental parameter, the axes can frequently be co rrelated with one or more physical v a ria b les. Spearman rank correlation coefficients (Siegel, 1956) were calculated between the sta tio n scores on the f i r s t two ordination axes and the same sediment variables used in the discriminant analysis (Table 5). Axis 1 was most highly correlated with the percentage of silt and clay and secondarily with the percentage of very fine sand. Sorting and the percentage of medium and fine sands had smaller but statistically significant r$ values. Axis 2 showed the greatest correlation with the percentage of fine sand. Sorting and the percentages of very 51: Figure 9 Location of station groups in ordination space. Length of axes are scaled in proportion to the eigenvalues. 32 GL o o o o o o o o Z SIXV Table 5. Spearman rank co rrelatio n c o effic ien ts between sediment parameters and sample scores on the first two OCA axes. (* = «*< 0.05, ** = <*< 0.01) V ariable Axis I Axis II Sorting -0.428** 0.633** % Shell and Gravel 0.221 0.326* % V e r y Coarse Sand 0.008 0.506** % Coarse Sand -0.080 0.483** % Medium Sand 0.362** 0.130 % Fine Sand 0.541** -0.755** % V e r y Fine Sand -0.678** 0.070 % S ilt and Clay -0.775** 0.211 33 coarse sand and shell and gravel showed progressively lesser but statistically significant correlations. DISCUSSION Evaluation of multivariate techniques It is often the goal of ecologists to relate some measure of species importance, usually abundance, to one or more environmental variables. While conceptually straightforward, this goal frequently becomes d if f ic u lt to a tta in when applied to real d ata. Even simple correlation between species abundance and the value for some environmental variable requires assumptions to be made which are often too restrictive. Assumptions of linearity and monotonicity between species abundance and an environmental fa c to r often can not be met unless the measured range of the environmental factor encompasses only a small portion of its total range throughout the species habitat. In the presence of a strong environmental gradient, most species distribution curves become bell-shaped (Austin, 1976) and simple linear correlations are not longer applicable. Ordination techniques have frequently been applied not only to show separation of the entitites in ordination space but to identify those environmental factors responsible for this separation (Nichols 1970; Biernbaum 1979; Wenner and Boesch 1979). This is accomplished by correlating the scores on one or more ordination axes with the measured environmental factors. The major problem with this procedure is that most indirect ordination techniques assume a linear relationship between the axes and species abundances (Noy-Meir and Whittaker, 1977). With even a moderate degree of environmental 35 36 variation among the samples (beta diversity), species distribution curves are curvilinear and nonmonotonic, and any ordinations based on assumptions of linearity can produce misleading results (Gauch and Whittaker, 1972; Gauch et a l., 1977). Ordination techniques unconstrained by lin e a rity (caten atio n s, Noy-Me1r, 1974) are av ailab le but all have serious drawbacks (Noy-Meir and Whittaker, 1977). DCA is a linear ordination technique, and therefore constrained by the same theoretical limitations Inherent in these techniques, yet it does make corrections for the effects of curvllinearity. Tests with simulated coenoclines have shown DCA Induces little of distortion even with high beta-diversity (H111 and Gauch, 1980). Most notably, the distortion created by the quadratic dependency of the second axis on the first in reciprocal averaging ("horseshoe" or "arch" effect) has been eliminated. Interpretation of the second axis in terms of an environmental fa c to r, which often has not been possible (Wenner and Boesch, 1979), becomes more fe a sib le . DCA was used in the present study to id e n tify sedimentary facto rs responsible fo r community differences by correlation of station ordination scores with measured grain size parameters. The ecological credibility of the results obtained in this study, and the general agreement of these results with those of discriminant analysis, supports the validity of this approach. Multiple discriminant analysis (Rao, 1952; Cooley and Lohnes, 1962) has been extensively used in other fields such as the social sciences, but has received only limited attention in ecological 37 investigations. In marine studies discriminant analysis has been used to Identify mechanisms of resource partitioning among benthlc polychaetes (Flint and Rabalais, 1980) and amphlpods (Schaffner and Boesch, 1982) and to relate station groups to physical variables (Shin, 1982; Vecchione and Grant, 1983). The second approach is used herein, in an attempt to Identify environmental factors (or unmeasured correlates) responsible for faunal differences among stations. The approach utilized (Green and Vascotto, 1978) involves initial segregation of stations into faunally-homogeneous groups by use of numerical classification. Multiple discriminant analysis is then employed to identify those environmental variables on which the biotically-derived station groups show the greatest separation. Standardized discriminant function coefficients are usually considered measures of each variable's importance 1n discriminating among groups. Use of this approach in the present study led to the conclusion that the sediment sorting coefficient was considerably more important than any other variable in discriminating between groups. However this does not seem intuitively correct and it is likely that this finding results from the high intercorrelation among the other variables. Grain size percentages of adjacent phi size intervals (e.g. coarse sand and very coarse sand) may be highly correlated, thus the discriminating information they carry may be similar. Their standardized discriminant function coefficients may be small even though their joint contribution is very important. Similar difficulties arise with grain size percentages that are widely separated on the phi size spectrum (e.g. coarse sand and very fine 38 sand) for these are likely to have a high Inverse correlation. The sorting coefficient had a high standardized discriminant function coefficient probably because 1t was the only variable used which was not likely to be inherently correlated with any other variable. Total structure coefficients have been used successfully (Dueser and Shugart, 1978; 1979; Flint and Rabalais, 1980) as alternative measures of a variable's importance. They may be more accurate than standardized discriminant function coefficients in cases of high variable colinearity (Johnson, 1977; Klecka, 1980). They are calculated only as correlations between independent variables and the discriminant scores and are therefore unaffected by correlations among the variables. Total structure coefficients can also be used to present the variables as vectors in a two-function discriminant plot (Overall and Klett, 1972; Johnson, 1977). Interpretation of the discriminant functions in terms of total structure coefficients leads to the more credible conclusion that the percentages of silt and clay and very fine sand are the most important sedimentary variables in discriminating between groups. Many of the coarser sand grades are of importance in the second discriminant function. The sorting coefficient is only of moderate importance in both of the first two discriminant functions. Significance of sediment parameters Benthic communities and their member organisms are distributed in a continuum along an environmental gradient (Mills, 1969). Nonetheless it remains common practice to recognize discrete faunal 39 assemblages with the realization that their junctures are often not as discrete as might be Implied. Assemblages of benthlc species have often been designated on the basis of substrate association (Franz, 1976 and numerous others) and the same appears possible on the North Carolina continental shelf. Recognizable assemblages Include: 1) a muddy, very fin e sand assemblage (Group 1) dominated by Lumbrineris impatiens, 2) a fine to medium sand assemblage (Group 2) dominated by Polygordius s p ., 3) a w e ll-so rte d , fin e sand assemblage (Group 4) dominated by Protohaustorius cf. deichmannae. and 4) a medium to coarse sand assemblage (Group 3) characterized by Hemlpodus roseus and Hesionura elonqata. A goal of this study 1s to identify the grain size parameters most responsible fo r community p a tte rn s. The faunal assemblages recognized have been designated by reference to the median grain size of the sediments in which they are found, however, median grain size per se is probably not the factor to which the organisms are responding. Median grain size alone is not responsive to differences in the grain size distribution which could have profound biological consequences yet be unexpressed by any measure of central tendency. Efforts to determine what physical properties are of greatest importance in structuring communities have frequently been frustrated by a high correlation among the environmental variables. Community changes have often been related to depth (Nichols, 1970; Lie, 1978; Flint and Rabalais, 1980), but in shallow marine environments sedimentary or other physical variables which vary with depth are more likely to be the controlling factors rather than depth itself. In most benthic studies the degree of sediment sorting has been of minimal use in in te rp re tin g community d ifferen ces, however th ere are some notable exceptions. Nichols (1970) found sim ilarity among polychaete assemblages to be best correlated with sorting, although he suggested that sorting was only a manifestation of the clay content. Gray (1974) argued that sorting, as a measure of structural complexity, may be a determinant of species diversity. Greater species diversity should be found on those sediments within are more poorly sorted, since a wider variety of particle types would be available for utilization by the benthos. A similar argument has been advanced relating deposit-feeding species diversity to food particle and total particle diversity (Whitlatch, 1981). In the Cape Hatteras area no relationship was found between sorting and number of species (Spearman rank correlation,®* > .05) but it should be noted that number of species is not necessarily correlated with species diversity which contains an evenness component as w ell. In the present study the degree of particle sorting was of greatest importance in differentiating Group 4 stations from the remainder of the station groups. Although the median particle sizes of Group 4 sediments were sim ilar to those of Group 2, th e Group 4 sediments were better sorted and characterized by dramatically different fauna. The importance of particle sorting was probably due to its correlation with sediment mobility, a property more difficult to quantify. The well-sorted sands of Group 4 are indicative of a dynamic hydraulic regime with a high degree of sediment mobility. Tubiculous and non-fossorial burrowers would be rapidly washed out of 41 sediment in such an environment. Only fossorial species, such as the haustoriid and phoxocephalid amphipods, are able to burrow rapidly enough in order to maintain purchase. The percentage of gravel has been identified as an important factor in differentiating between benthic assemblages and, in some instances, has been found to be the most important environmental factor in determining the distribution of species, feeding types and life styles of benthic amphipods (Blernbaum, 1979). The percentage of shell and gravel was of no detectable biological significance in the Cape Hatteras area but this finding is believed to be a result of the small range of shell and gravel content encountered (up to 23% but generally less than 1%). The percentage of fine sand was found to be the most important variable on the second axis of the ordination and the second discriminant function. Differences in the fine sand component were most important in distinguishing the interstitial burrowers of Group 3. The importance of fine sand has been noted for both certain amphipod species (Schaffner and Boesch, 1982) and the whole macrobenthic community (Boesch and Bowen, in p re p .; Boesch, in prep.) in the Middle Atlantic Bight, though it should be noted that in these studies fine sand was considered to be the aggregate of both the fine and very fine sand size classes used here. Boesch (1n prep.) speculated that the exclusion of interstitial species with increasing proportions of fine sand is due to the occulusion of the interstitial spaces. The fact that many species of Group 3 stations were 42 interstitial burrowers 1s probably reflective of the low proportion of fine sand particles in stations in this group. The sediment variables found most useful for interpreting faunal distributions were found to be the percentages of very fine sand and silt and clay. Both scored highly in the discriminant anlaysis and o rd in atio n . The percentage of s i l t and clay appeared s lig h tly more important in both analyses but because both percentages were highly correlated it is difficult to determine their independent contributions. The importance of the very fine sand and silt/clay components is, in part, a reflection of their role in determining sediment permeability. Sediment permeability is diminished as the proportion of finer particles in the sediment increases (Webb, 1958). Dissolved oxygen is unable to penetrate as deeply in sediments containing a high proportion of fine particles, and thus deep-burrowing species are restricted to those which can maintain contact with the sediment-water interface (e.g. by elongate siphons or burrow irrigation) or have physiological adaptations to anaerobic conditions. A second reason that finer particles were found to be so important in structuring benthic communities is their crucial role in determining the type and abundance of food resources. Sanders (1958) found the distribution of many deposit-feeding macrobenthos to be correlated with the percentage of silt and clay. He speculated that the amount of clay was the more important v ariab le since clay particles have the greatest surface:volume ratio, allowing adherence of relatively more organic matter. The greater surface area also provides more attachment sites for microorganisms which serve as a food resource for deposit-feeders. Levinton (1977) suggested that the size of bacterial populations, which in turn determines the abundance of deposit-feeders, is limited by the space available for attachment. The proportion of clay or silt/clay has been found to be the most biologically important sediment parameter in several studies (Sanders, 1958; Nichols 1970; Ki^rboe, 1979). Although these studies found percentages of silt and clay ranging as high as 50 to 90%, it appears that silt and clay is of equal biological significance in the considerably lower concentrations found on the North Carolina shelf. Even modest changes in silt and clay content in the range observed of 0 to 27% can have profound biological consequences. Both dynamic and static properties of sediments are of importance in structuring benthic communities on the North Carolina continental shelf. Dynamic properties, as reflected 1n the degree of particle sorting, are of importance in determining the dominance of fossorlal species. The importance of static properties of sediment are best reflected in the proportion of cle^s, silts and finer grades of sand. The primary mechanisms by which the finer particles structure the community are twofold: 1) th e ir e ffe c t on sediment perm eability, and 2) their influence on the type and availability of food resources. The abundance of coarser particles may be important for select species, but on a community level, coarser particles are only of importance if they become large enough to permit the establishment of a hard-substrate community. LITERATURE CITED Austin, M. P. 1976. On non-linear species response models in ordination. Vegetatio 33:33-41. Biernbaum, C. K. 1979. Influence of sedimentary factors on the distribution of benthic amphipods of Fishers Island Sound, Connecticut. J . Exp. Mar. B iol. Ecol. 38:201-223. Bloom, S. A., J . L. Simon and V. D. Hunter. 1972. Animal-sediment re la tio n s and community analysis of a F lorida estu ary . Mar. Biol. 13:43-56. Bray, J . R. and J . T. C u rtis. 1957. An ordination of th* upland fo re st communities of southern Wisconsin. Ecol. Monogr. 27:320-349. Buchanan, J. B. 1963. The bottom fauna communities and their sediment relationships off the coast of Northumberland. Oikos 14:154-175. Boesch, D. F. and M. A. Bowen. In prep. Bathymetric d is trib u tio n of assemblages of macrobenthos in the Middle Atlantic Bight, U.S.A. Boesch, D. F. In prep. Distribution and ecology of macrobenthos in relation to mesoscale topography on the continental shelf of the Middle Atlantic Bight. 44 Briggs, J. C. 1974. Marine Zoogeography. McGraw-Hill, New York. 475 pp. Cerame-Vivas, M. 0. and I. E. Gray. 1966. The distribution pattern of benthic invertebrates of the continental shelf off North Carolina. Ecology 47:260-270. Clifford, H. T. and W. Stephenson. 1975. An Introduction to Numerical C la ss ific a tio n . Academic P ress, New York. 229 pp. Cooley, W. W. and P. R. Lohnes. 1962. Multivariate Procedures for the Behavioral Sciences. Wiley, New York. Day, J . H., J . G. Field and M. P. Montgomery. 1971. The use of numerical methods to determine the distribution of the benthic fauna across the continental shelf of North Carolina. J. An. Ecol. 40:93-125. Dueser, R. D. and H. H. Shugart, Jr. 1978. Microhabitats in a forest-floor small mammal fauna. Ecology 59:89-98. Dueser, R. D. and H. H. Shugart, Jr. 1979. Niche pattern in a forest-floor small mammal fauna. Ecology 60:108-118. Flint, R. W. and N. N. Rabalais. 1980. Polychaete ecology and niche patterns: Texas continental shelf. Mar. Ecol. Prog. Ser. 3:193-202. Franz, D. 1976. Benthic molluscan assemblages in relation to sediment gradients in northeastern Long Island Sound, Connecticut. Malacologia 15:377-399. 46 Gauch, H. G. and R. H. Whittaker. 1972. Comparison of ordination techniques. Ecology 53:868-875. Gauch, H. G., Jr., R. J. Whittaker and T. R. Wentworth. 1977. A comparative study of reciprocal averaging and other ordination techniques. J. Ecol. 65:157-172. Gray, I. E. and M. J. Cerame-Vivas. 1963. The circulation of surface waters in Raleigh Bay, North Carolina. Limnol. Oceanogr. 8:330-337. Gray, J. S. 1974. Animal-sediment relationships. Oceanogr. Mar. Biol. Ann. Rev. 12:223-261. Green, R. H. 1971. A multivariate statistical approach to the Hutchinsonian niche: bivalve molluscs of central Canada. Ecology 52:543-556. Green, R. H. and G. L. Vascotto. 1978. A method for the analysis of environmental factors controlling patterns of species composition in aquatic communities. Water Res. 12:583-590. Harrington, C. L. 1981. Responses of benthic macrofauna to the dynamics of a Gulf Stream front on the continental shelf near Cape Hatteras, North Carolina. Ph.D. dissertation. Univ. of Wisconsin-Madison. H erbst, G. N., D. P. Weston and J . G. Lorman. 1979. The d is trib u tional response of amphipod and decapod crustaceans to a sharp 47 thermal front north of Cape Hatteras, North Carolina. Bull. B1ol. Soc. Mash. 3:188-213. H ill, M. 0. and H. G. Gauch, Jr. 1980. Detrended correspondence analysis: an Improved ordination technique. Vegetatlo 42:47-58. Hunt, R. E., D. J. P. Swift and H. Palmer. 1977. Constructional shelf topography, Diamond Shoals, North Carolina. Geol. Soc. America Bull. 88:299-311. Inman, D. I. 1952. Measures for describing the size distribution of sediments. J. Sed. Pet. 22:125-145. Johnson, R. M. 1977. Multiple discriminant analysis: marketing research applications, p. 65-79. hi: J. N. Sheth (ed.) Multivariate Methods for Market and Survey Research. American Marketing Assoc., Chicago. Kitfrboe, T. 1979. The distribution of benthic invertebrates in Holbaek Fjord (Denmark) in re la tio n to environmental fa c to rs. Ophelia 18:61-81. Klecka, W. R. 1980. Discriminant Analysis. Sage University Paper series Quantiative Applications 1n the Social Sciences, series no. 07-019. Sage Publications, Beverly Hills and London. Levinton, J. S. 1977. Ecology of shallow water deposit-feeding communities Quisset Harbor, Massachusetts, p. 191-227. hl: B. C. Coull (ed.) Ecology of Marine Benthos. Univ. of South Carolina Press, Columbia. 48 Lie, U. 1978. The quantitative distribution of benthic macrofauna in Fanafjorden, western Norway. Sarsla 63:305-316. McNulty, J. K., R. C. Work and H. B. Moore. 1962. Some relationships between the infauna of the level bottom and the sediment 1n south Florida. Bull. Mar. Sci. Gulf Caribbean 12:322-332. Mills, E. I. 1969. The community concept in marine zoology, with comments on continua and instability in some marine communities: a review. J . Fish. Res. Bd. Canada 26:1415-1428. Nichols, F. H. 1970. Benthic polychaete assemblages and their relationship to the sediment in Port Madison, Wisconsin. Mar. Biol. 6:48-57. Nie, N. H., C. H. Hull, J. G. Jenkins, K. Steinbrenner and D. H. Bent. 1975. SPSS: Statistical Package for the Social Sciences. McGraw-Hill, New York. Noy-Meir, I. 1974. Catenation: quantitative methods for the definition of coenoclines. Vegetatio 29:89-99. Noy-Meir, I. and R. H. Whittaker. 1977. Continuous multivariate methods in community an aly sis: some problems and developments. Vegetatio 33:79-98. Overall, J. E. and C. J. Klett. 1972. Applied Multivariate Analysis. McGraw-Hill, New York. 49 Rao, C. R. 1952. Advanced S ta tis tic a l Methods in Biometrical Research. Wiley, New York. Sanders, H. L. 1958. Benthic studies in Buzzards Bay. I. Animal - sediment relationships. Limnol. Oceanogr. 3:245-258. S chaffner, L. C. and D. F. Boesch. 1982. Spatial and temporal resource use by dominant benthic Amphipoda (Ampeliscidae and Corophiidae) on the Middle Atlantic Bight outer continental shelf. Mar. Ecol. Prog. Ser. 9:231-243. Shin, P. K. S. 1982. Multiple discriminant analysis of macrobenthic infaunal assemblages. J . Exp. Mar. B iol. Ecol. 59:39-50. Siegel, S. 1956. Nonparametric Statistics for the Behavioral Sciences. McGraw-Hill, New York. Vecchione, M. and G. C. Grant. 1983. A multivariate analysis of planktonic moluscan distribution in the Middle Atlantic Bight. Cont. Shelf Res. 1(4):405-424. Vernberg, F. J . and W. B. Vernberg. 1970. Lethal lim its and the zoogeography of the faunal assemblages of coastal Carolina w aters. Mar. B iol. 6:26-32. Webb, J. E. 1958. The ecology of Lagos Lagoon. V. Some physical properties of lagoon deposits. Phil. Trans. Soc., Ser. B, Biol. Sci., no. 683, 241:393-419. 50 Wells, H. W. and I. E. Gray. 1960. The seasonal occurrence of Mytilus edulis on the Carolina coast as a result of transport around Cape Hatteras. Biol. Bull. 119:550-559. Wenner, E. L. and D. F. Boesch. 1979. Distribution patterns of epibenthic decapod Crustacea along the shelf-slope coenocline, Middle Atlantic Bight, U.S.A. Bull. Biol. Soc. Washington 3:106-133. Wentworth, C. K. 1922. A scale of grade and class terms for clastic sediments. J. Geol. 30:377-392. Whitlatch, R. B. 1981. Animal-sediment relationships in intertidal marine benthic habitats: some determinants of deposit-feeding species d iv e rs ity . J. Exp. Mar. B iol. Ecol. 53:31-45. Young, D. K. and D. C. Rhoads. 1971. Animal-sediment relations in Cape Cod Bay, Massachusetss I. A transect study. Mar. Biol. 11:242-254. M l Appendix I. Macrofaunal species collected and average density (indiv./to?) in each station grotp. Station gnxps arranged in order of increasing median grain size (decreasing phi) in order to illustrate substrate affinity for each species. Total Individuals Collected 1 2B 4 2C 2A 3 CNIDARIA Anthozoa Renillidae Renilla reniformis PLAmELMINTHES Tirbellaria Discocelidae Coronadena mutabilis RHYNCHOCOELA Nemertea sp. A 29 10 3 1 2 1 Nemertea sp. B a 6 3 6 Nemertea sp. D 1 1 Anopla Lineidae Cerebratulus lacteus 20 6 4 1 Micrura leidyi 8 2 3 Zygetpolla rufens 8 3 1 Enopla Anphiporidae Anphiporus bioculatus ANNELIDA Polychaeta Polynoidae Hanrothoe extenuata 10 3 2 1 Harmotfoe sp. A 1 1 Sigalionidae Eipanthalis kinbergi 1 1 Grubeulepis sp. A 9 2 Sthenelais boa 2 1 Sthenelais linricola 10 2 Kigalion arenlcola 1 51 Appendix I . continued Total Individuals Collected 1 2B 4 2C 2A 3 Chrysopetalidae Bhawania goodei 4 1 2 1 Paleanotus heteroseta 15 1 17 Anphinomidae Chloeia viridis 10 1 2 7 1 Pseudeurythoe paucibranchiata 3 1 1 Phyllodocidae Anaitides groenlandica 1 1 Anaitides arenae 10 4 2 Anaitides longipes 10 3 3 Anaitides mucosa 1 1 Anaitides cf. panamensis 1 1 Anaitides sp. B 3 4 GenetylTTs castanea 2 1 1 Eteone lactea 1 2 Eunida sanfuinea 26 9 5 Eulalia macroceros 1 1 Paranaites speciosa 3 1 2 Paranaites polynoides 1 1 7 Hesionuri"ongata 10 I Hesionidae Gyptis brevipalpa 3 2 1 1 mcrcphthalmus sp. A 1 1 Microphthalmus sp. B 2 1 Microphthalmiis sp. C 2 1 Podarke obscura 2 1 • Heteropodarke heteranorpha 4 4 1 Pilargiidae Siqantra tentaculata 50 18 7 Siganbra bassi 2 3 Pilargis" sp. A 2 1 Cabira incerta 1 2 Synelmis albini 1 2 Syllidae Brania wellfleetensis 3 1 1 1 Pionosyllis gesae 4 2 2 Syllis (Tvposyllis) arnica 2 1 1 Syllis (Ehlersia) comuta 1 1 Exogone lourei 1 1 Exogone verugera 1 1 Odontosyl1 is fulgurans 3 4 53 Appendix I . continued Total Individuals Collected 1 28 4 2C 2A Streptosyllis arenae 5 1 2 1 1 Parapionosyllis sp. A 1 1 Eurysyllis tiEeinculata 2 1 1 Nereididae Ceratonereis irritabilis 5 1 3 Ceratonereis longicirrata 12 17 Nereis acumnata 3 1 Nereis qrayi 2 1 Nereis succinea 11 4 1 Nereis riisei 15 2 2 10 1 Platynereis dunerilii 14 5 1 Ceratocephale oculata 299 112 20 6 6 1 Nqihtyidae Nephtys pieta 157 12 22 27 30 50 11 Aglaophantis verrilli 337 126 35 Inermonyhtys inermis 3 1 3 Sphaerodoridae Sphaerodoropsis sp. A 2 1 1 Glyosridae Glycera americana 8 3 1 Glycera dibranchiata 13 2 10 4 Glycera oxycephala 44 1 6 44 6 Hemipodus roseus 8 6 Goniadidae Goniada littorea 207 60 22 8 24 10 Goniada teres 8 3 2 Onuphidae Oniphis eremita 11 1 3 9 1 Mooreoniphis pallidula 14 4 7 Mooreontphi? nebulosa- 46 12 3 20 Diopatra ciprea 25 9 3 Diopatra tridentata 1 2 Eunicidae Etnice vittata 9 1 2 4 1 Eunice websteri 1 1 Lintrineridae Lutfcrineris latreilli 1 1 Lurbrineris tenuis 3 1 Lintrineris' inpatiens 6282 2504 23 2 6 1 Linbrineris' cruzensis 6 1 2 1 Lurbrineris coccinea 4 1 2 1 Appendix I . continued Total Individuals Collected 1 2B 4 2C 2A 3 Lurbrineris cf. januarii 5 2 Lurbrinerides dayi 2 3 Arabellidae Drilonereis longa 1 1 Dorvilleidai Dorvillea sp. A 14 6 Protodorvillea gaspeensis 1 l ProtodorvilTia kefersteini 6 7 1 Schistomeringos caeca 2 1 1 Orbiniidae Leitoscoloplos foliosus 4 1 2 1 Naineris sp. A 2 1 Scoloplos acmeceps 39 14 5 1 2 Scoloplos rubra 1 1 Paraonidae Aricidea suecica 30 12 Aricidea wassi 53 38 18 Aricidea catherinae 123 48 3 1 Aricidea sinplex 1 1 Aricidea cerrutii 2 4 Aricidea fragilis 4 1 3 Aricidea sp. A 33 11 7 1 Levinsenia gracilis 6 2 2 Paraonis fulgens 5 1 7 Paraonis pygoeniqnatica 5 4 Cimyhorus americanus 6 2 1 Paradoneis lyra 3 1 Spionidae Polydora social is 6 3 6 Polydora websteri 1 1 Polydora concharun 3 2 3 Laornce cirrata 3 1 Prionospio cirrifera 21 5 1 Prionospio pyqnaea 47 16 3 Prionospio cirrobranchiata 2 4 Prionospio cristata 646 224 84 4 49 1 Prionospio dayi 12 4 2 1 Prionospio fallax 2 1 Spio cf. setosa 4 1 2 2 Spio pettiboneae 48 6 7 5 7 10 Spio piqnentata 19 8 Appendix I . continued Total Individuals Collected 1 2B 4 2C 2A 3 Spiophanes bortoyx 89 26 22 2 7 1 Spiophanes wiqleyi 3 1 Spiophanes missionensis 33 12 2 1 Nblacoceros indicus 5 2 Paraprionospio pinnata 1085 416 57 1 12 6 Dispio uncinata 9 3 2 1 1 Scolelepis texana 12 4 3 Scolelepis squamata 2 1 Scolelepis sp. B 2 3 Aonides paucibranchiata 3 4 Magelonidae Maqelona cf. rosea 20 2 12 3 4 Maqelona cf. pettiboneae 52 10 5 1 2 37 1 Maqelona cf. phyllisae 231 87 13 1 2 6 Poecilochaetidae Poecilochaetus johnsoni 84 28 15 1 4 Chaetopteridae Spiochaetopterus oculatus 25 7 7 2 3 Cirratulidae Caulleriella sp. A 15 6 7 1 Caulleriella sp. B 3 1 1 Caulleriella sp. C 11 4 16 Tharyx annulosus 3 3 1 Chaetozone cf. setosa 38 12 8 2 1 Chaetozone sp. B 6 2 3 Cossuridae Cossura lonqocirrata 16 4 Ctenodrilidae Ctenodrilidae sp. A 21 5 Flabelligeridae Pherusa sp. B 2 3 Opheliidae Armandia maculata 69 18 8 4 2 9 4 Armandia aqilis 11 3 2 2 Ophelia denticulata 5 1 6 1 Travisia forbesii 1 1 Polyophthalmus pictus 1 1 Capitellidae Notomastus latericeus 277 107 27 2 4 Notcmastus americanus 7 3 Mediomastus califomiensis 646 224 17 1 7 Appendix I . continued Total Individuals Collected 1 2B 4 2C 2A 3 Amastiqos caperatus 1 1 Maldanidae Asychis carolinae 3 1 1 Macroclynsne zonal is 5 1 1 Owenidae Owenia fusiformis 190 67 20 2 10 Myriochele cf. oculata 20 8 1 Sabellaridae Sabellaria vulgaris 219 86 5 Anphictenidae Cistena gouldii 1 1 Cistena regalis 2 1 Arrpharetidae Anpharete cf. acutifrons 12 2 1 1 6 Anpharete americana 2 1 Anphicteis ginnerT" 3 1 Terebellidae Pista cristate 646 224 83 4 49 Polycirrus eximius 4 6 Polycurus carofinensis 1 1 Loiima medusa 1 1 Trichobranchidae Trichobranchus glacial is 2 3 Sabellidae Chone sp. A 1 1 Fuchohe incolor 3 1 4 Megalomna bioculatun 11 4 3 Potarnilla renlformis 1 1 Serpulidae Hydroides protulicola 3 1 Pomatoceros americanus 2 1 Polygordi idae Polygordius sp. 221 23 52 11 102 96 Oligochaeta Tubificidae TUbifocoides maureri 7 1 7 1 Tubificoides dukei 21 4 5 1 7 Tubificoides sp. A 6 4 Tibificoides sp. B 2 1 2 Liimodriloides monothecus 1 1 Bathydrilus longus 1 1 57 Appendix I . continued Total Individuals Collected 1 28 4 2C 2A 3 MXLUSCA Gastropoda Phasianellidae Tricolia sp. 2 1 Rissoidae Alvania pelaqica 2 1 1 Vitrinellidae Vitrinella sp. 2 1 7 Caecidae Caecun pulchellun 153 58 3 2 7 Caecun cooperi 7 1 7 Caecun carol ini anun 6 4 Caecun cibitatun 6 4 Melanellidae Strcntjiformis auricinctus 1 1 Niso aeqlees 4 1 3 Crepidulidae Cal^ptraea centralis 2 3 Crepidula fomicata 2 1 Naticidae Pol inices diplicatus 5 1 Sinun perspectivun 4 1 2 Siqatica carolinensis 2 1 1 Muricidae Eipleura caudata 1 1 Murex leviculus 1 1 Col urfaellidae Astyris lunata 7. 4 Anachis cf. pretrii 3 1 Nassaridae Nassarius albus 1 1 Olividae Olivella sp. 3 1 1 Oliva sayana 1 1 Turridae Kurtziella rubella 1 1 Kurtziella limonitella 2 1 Terebridae Terebra dislocata 2 1 Terebra concava 4 1 1 3 Pyramidellidae Odostomia sp. 52 19 3 Appendix I . continued Total Individuals Collected 1 2B 4 2C 2A 3 Acteonidae Acteon pinctostriatus 51 18 3 4 Acteocinidae Acteocina candei 9 4 Cylichnidae Cylichnella bidentata 26 9 1 Philinidae Phi line sagra 5 1 3 Bullidae Bulla striata 2 1 Haminoeidae Haminoea solitaria 1 1 Atys sandersoni 1 1 Retusidae Volvulella persinrilis 8 Volvulella recta 1 1 Pleurobranchidae Pleurobranchaea hedgpethi 2 1 1 Bivalvia Nuculidae Nucula proxima 7 1 9 Nuculanidae Nuculana acuta 60 23 3 Solariyacidae Solemya velun 3 1 1 Aricidae Anadara transversa 3 1 ^ ilid a e Crenella divaricata 2 3 Musculus lateralis 6 1 Modiolus modiolus squamosus 3 1 1 Pinnidae Atrina seminuda 2 1 Pectinidae Cyclopecten nanus 3 4 Aequipecten mucosus 1 1 Placopecten maqellanicus 2 1 1 Argopecten gibbus 1 1 Anomiidae Anemia sinplex 1 1 Limidae Lima pellucida 1 1 Appendix I . continued Total Individuals Collected 1 2B 4 2C 2A 3 Lucinidae Parvilucina multilineata 111 42 3 1 2 Lucina nassula 2 1 Lucina pectinata 3 1 2 Divaricella quadrisulcata 4 1 1 Codakia costata 1 1 Thyasiridae Thyasira trisinuata 14 3 3 Ungulidae Diplodonta pmctata 4 1 1 Leptonidae Pythinella cuneata 1 2 Cardiidae Laevicardium pictun 1 1 Mactridae Spisula solidissima similis 751 11 2 48 2 696 Solenidae Ensis directus 1 1 Tellinidae Macoma tenta 1 1 Tellina versicolor 313 88 48 4 8 71 Tellina aequistriata 3 1 Tellina nitens 2 1 Strigilla cf. mirabilis 9 1 2 1 6 Semelidae Semele bellastriata 1 1 Semele nuculoides 16 9 Abra aequalis 163 60 17 Veneridae Mercenaria mercenaria 2 1 Chione cancellata 4 6 Chione qrus 1 1 Chione latilirata 1 1 Crassinella linulata 10 1 11 Cooperellidae Cooperella atlantica 1 1 Chamidae Arcinella comuta 1 1 Corbulidae Corbula barrattiana 310 81 175 1 2 Corbula swiftiana 20 7 3 1 Varicorbula operculata 566 188 2 134 60 Appendix I . continued Total Individuals Collected 1 28 4 2C 2A 3 Pandoridae Pandora bushiana 14 4 6 1 Pandora arenosa 1 1 Lyansiidae Uyonsia hyalina 187 65 33 1 Cuspidariidae Cardiomya costellata 1 1 Scaphopoda Dentaliidae Dentaliun calamus 1 1 Siphodentaliidae Cadulus carolinensis 17 5 6 Cadulus tetraschistus 5 6 1 Cephalopoda Sepiolidae Rossia tenera 1 1 Loligo pealeii 5 1 2 2 ARTHROPODA Py:nogonida Phoxichiliidae Anoplodactylus petiolatus 3 1 1 Ostracoda Asteropidae Antoleberis americana 6 2 1 Parasterope cf. hulinqsi 6 2 1 2 SarsielIidae Sarsiella texana 1 1 Sarsiella spinosa 5 2 Sarsiella ozotothrix 3 1 Sarsiella disparilis 2 1 1 Sarsiella tubipora 1 1 Halocyprididae Euconchoecia chierchiae 11 3 2 1 1 Philomedidai Harhansus paucichelatus 14 2 2 9 Pseudophilomedes sp. A 2 3 Cytheridae Pterygocythereis americana 5 1 2 Appendix I . continued Total Individuals Collected 1 2B 4 2C 2A 3 tyysidacea Mysidae Bownaniella portoricensis 7 1 1 2 Anchialina typica 1 1 Cumacea Diastylidae Oxyurostylis smithi 8 1 2 2 3 Bodotriidae Cvclaspis varians 30 5 7 1 2 Isopoda Anthuridae Qyathura burbanki 1 1 Apanthura magnificat 5 1 3 1 Xenanthura breyiteTson 149 57 12 Ptilanthura tricarina 1 1 Sphaeromatidae Ancinus depressus 1 1 Serolidae Serqlis mgrayi 1 1 Idoteidae” Edotea triloba 3 1 1 1 Anphipoda Anpeliscidae Anpelisca verrilli 77 31 Anpelisca agassizi 5 6 1 Anpelisca cristoides 6 9 Anphilochidae Gitanopsis sp. A 1 1 Aoridae Lgrbos smithi 2 1 Lgrbos websteri 8 1 1 7 Microdeutopus rnyersi 3 1 Miorodeutopus sp. A 1 1 Acuninodeutopus naglei 5 1 3 Pseudunciola obliquua 1 1 Rildardanus laminosa 1 1 Unciola irrorata 1 1 Argissidae Argissa hamatipes 6 1 1 3 Appendix I . continued Total Individuals Collected 1 2B 4 2C 2A 3 Bateidae Batea catharinensis 1 1 Corophndae Cerapus tubularis 3 4 Corophiun acutun 1 2 Erichthonius brasiliensis 19 2 20 1 Siphonoecetes smithianus 2 1 1 Gammaridae Elasropus sp. A 12 2 10 Melita appendiculata 1 1 Jerbarnia amencana 3 4 Haustoriidae Acanthohaustorius intermedius 6 1 8 Acanthohaustorius nrillsi 4 2 2 Bathyporeia parkerT 128 2 64 6 3 1 Protonaustorius cf. deidmarma 3088 1 3 1610 74 6 Haustoriidae (nr. Platyischnopus) sp. A 12 6 1 Isaeidae Photis puyiator 20 4 7 10 Gamreropsis Sutherland!' 1 1 Liljeborgiidae Listriella bamardi 9 3 3 Oedicerotidae Synchelidiim atnericanun 114 34 7 6 19 1 Phoxocephalidae Rhepoxynius epistanus 378 1 12 179 46 7 2 Metharpinia floridanus 59 1 15 2 6 17 Stenothoidae Parametopella cypris 1 1 Parametopella inquilinus 2 1 Synopudae Tiron tropakis 12 1 11 CaprelIidae Caprella equilibra 1 1 Luconacia incerta 5 2 Decapodef Penaeidae Trachypenaeus constrictus 1 1 Sicyonia typica 1 1 Solenocera atlantidis 7 4 2 Appendix I . continued Total Individuals Collected 1 2B 4 2C 2A 3 Sergestidae Lucifer faxoni 27 4 5 1 2 3 2 Lucifer typus 1 1 Pasiphaeidae Leptochela serratorbita 1 1 Leptochela bermudensis 6 1 2 2 1 Palaemonidae Periclimenes americanus 1 Alpheidae Automate evermanni 3 Ogyrididae Ogyrides alphaerostris 2 Processidae Processa henphilli 8 2 1 Crangonidae Crangon septemspinosa 1 Paguridae Pagurus annulipes 1 Galatneidae Munida irrasa 1 Porcellanidae Euceramus praelongus 2 Albuneidae Albunea panetii 2 Albunea qibbesii 5 Dramiidae Drcmidia antillensis Calappidae Hepatus epheliticus Leucosiidaie Persephone mediterranea Raninidae Raninoides laevis Majidae Conodes trispinosus Batrachonotus fragosus Portunidae Ovalipes stephensoni Goneplacidae Frevillea hirsuta Appendix I . continued Total Individuals Collected 1 2B 4 2C 2A 3 Pinnotheridae Pimixa retinens 3 1 Pinnixa sayana 1 1 Stomatopoda SquilIidae Odontodactylus brevirostris 1 1 SIPUNCULA Sipunculus nudus 1 1 Golfingia trichocephala 231 90 5 1 1 Golfingia elongate 1 1 Phascolion strorrfci 3 1 Aspidosiphon albus 18 3 8 9 Aspidosiphon qosnoldi 15 1 3 17 ECHIURA Thalassema philostracun 1 2 BRACHIOPODA Glottidea pyramidata 27 8 2 2 7 1 ECHINOGERMATA Asteroidea Astropectinidae Astropecten diplicatus 1 1 Ophiuroidea Anphiuridae Micropholis atra 1 1 Anphiodia trychna 3 3 1 Echinoidea Schizasteridae Moira atropos 8 3 CEPHALOCHORDATA Branchiostoma caribaeun 21 1 1 16 5 Total nuriber of species 368 275 92 72 49 172 80 Chapter 2 THE EFFECT OF AN OCEANIC FRONT ON THE MACROBENTHOS OF THE NORTH CAROLINA CONTINENTAL SHELF D. P. Weston 65 ABSTRACT A sharp front between Gulf Stream and Virginia Shelf Water masses was observed on the North Carolina continental shelf northeast of Cape Hatteras. This front was characterized by a temperature gradient as great as 12°C and a salinity gradient of about 5°/oo. These changes occurred over a latitudinal distance of 27 km during the study period but had earlier been observed to occur within 1 km. The front was found to undergo frequent, rapid and unpredictable movements across the shelf, subjecting the benthos to temperature changes as great as 8°C over a 3 hr period. The front was highly effective as a zoogeographic barrier with many warm-water and cold-water macrobenthic species reaching the limit of their distributions at the front. In addition warm-water species were restricted to the area south of Cape Hatteras by seasonal migrations of the front which permit cold Virginia Shelf Water to reach the Cape. The role of the front as an ecotone was evaluated, and though there was an increased number of species in the area of the front, there was no evidence of a higher macrofaunal density or the presence of ecotonal species. The lack of a strong ecotonal response in the benthos is attributed to their immobility and the spatial instability of the front. 66 INTRODUCTION Scientific investigations of oceanic fronts have focused almost exclusively on the water column and its biota. The distribution of nutrients, phytoplankton, zooplankton and nekton associated with fronts have all received much attention (Savidge, 1976; Fournier et al., 1977; Iverson et al., 1979; Mills, 1980). Information on the benthic communities in the vicinity of fronts is scarce. This lack of attention to the benthos of fronts 1s, in part, a result of the fact that fronts are water column phenomena, and therefore initial interest has concentrated on the pelagic component of the biota. Intensive benthic sampling has also been complicated by the fact that fronts typically occur in deep water (200 m or more). With the discovery of a shallow-water (ca. 30 m) front on the North Carolina continental shelf, intensive sampling of the benthic community associated with a major oceanic front became fe a sib le . During the spring of 1975 investigators from the University of Wisconsin noted the presence of a Gulf Stream-Virginia Shelf Water front located 35 km northeast of Cape Hatteras. Bottom water temperatures across the front changed up to 12°C over less than 1 km (J. Magnuson, Univ. of Wisconsin, unpub. data). Coincident with the thermal gradient was a change in salinity of 4-5°/oo over the same distance. The front was dynamic and found to migrate across the shelf at speeds of up to 6 cm/sec. 67 68 Because of the dynamic nature of the fro n t, the benthic community in the area is exposed to frequent, rapid and unpredictable fluctuations in temperature and salinity as the area is alternately exposed to Gulf Stream water and Virginia Shelf Water masses. In the relatively stable environment of the continental shelf, the front represents a unique habitat, characterized by physical inconstancy of temperature and salinity. The physical environment might therefore be expected to impart a unique structure to the macrobenthic community of the front, distinguishing it from the biotic assemblages of the hydrographically stable areas to the north and south. It is an objective of this study to evaluate the role of the front in determining species d is trib u tio n s , community stru c tu re and temporal variability of the macrobenthic populations. Investigation of the frontal area is of even greater significance because of the biogeographic importance of the Cape Hatteras region. Since the work of Dana (1853), the North Carolina area in general, and Cape Hatteras in particular, has been often recognized as a zoogeographic barrier to a diverse array of faunal groups (Briggs, 1974; Cerame-Vivas and Gray, 1966). The front is a juncture between the warm Gulf Stream water and the cooler Virginia Shelf Water, so it is reasonable to expect that it may also serve as a transition zone between the benthic fauna associated with these two water masses. This study is designed to assess the effectiveness of the front as a zoogeographic barrier, and to identify the physical factor(s) responsible for delimiting species distributions. K. METHODS Sampling stations were established along a north-south transect (75°13.0'W) extending from 35°23.5‘N to 35°51.5'N (Figure 1), and were designated by minutes north of latitude 35°00.5'N (e.g. station 36 located at 35°36.5'N). Stations were spaced two nautical miles apart at each end of the transect, and one nautical mile apart in the middle of the transect. Water depth varied from 33-42 m among the stations. The entire transect was sampled in June 1977 and a few select stations occupied in August and October 1977 and January 1978. Only the more comprehensive June data are included in these analyses unless otherwise specified. Macrofauna were collected using a 0.1 m^ Smith-Mclntyre grab with three replicate samples taken at each site. Most organisms were separated from the sediment by elutriation through a 0.5 mm mesh screen. The material remaining after elutriation was washed through a 1.0 mm mesh screen. All animals in both the 0.5 and 1.0 mm fractions were preserved in formalin, sorted to major taxa and identified. Sediment samples collected from each grab were dried at 100°C and 100 g aliquots of each sample were obtained by use of a sediment cutter. The material was sieved through a series of Standard Sieves with siltis and clays combined as pan weight. 69 K’ Figure 1 The location of sampling stations on the North Carolina continental shelf. Stations are designated by minutes north of latitude 35°00.5'N. 70 Chataptaka — \B ay - '35*50' North Carol! m Oragon Inltt 'Ralaigh Bay • * 5 * 4 0 ' ATLANTIC OCEAN Pamlico Sound ■ -35*30' ■ -35*20' 9 * CAPE HATTERAS -■35*1 O' —I------1------1------— ^ ------h 75°40' 75°30' 75°20' 75° I 71 Water temperature and salinity were measured at each station using a shallow-water mechanical bathythermograph (BT) and Beckman salinometer. Measurements taken periodically with a reversing thermometer and Hytech salinometer were used to correct the BT and Beckman salinometer meter readings. On some cruises a continuous record of bottom temperature was obtained by use of a waterproof thermograph (Ryan Model J) positioned 0.5 rn above the bottom at 35°35.5'N, 75°13.0'W. Several multivariate techniques were employed in analysis of the biological and geological data. In order to delimit zones of faunal sim ilarity, cluster analysis was performed in which species occurring in two or less samples were deleted. Algorithms included log-transformation of species abundances, the Bray-Curtis similarity measure (Bray and Curtis, 1957), and flexible sorting with established at -0.25 (Clifford and Stephenson, 1975). Multiple discriminant analysis (Nie et a l., 1975) was used to evaluate the significance of the group separation derived by cluster analysis, and to interpret this separation in light of a variety of physical variables. Sediment percentages used in the analysis were transformed by arcsin/p to induce normality. RESULTS Sediments The sediments 1n the study area were fairly well-sorted fine sands with a median grain size ranging form 1.12-2.97 0 (Table 1). The percentages of s i l t and clay were generally small (< 7%), with a tendency for those stations in the southern portion of the front to have a slightly greater proportion of silt and clay than those in the north. Little change in sediment parameters was noted in seasonal sampling, indicatin g minimal temporal and sm all-scale sp atial v a ria tio n . Temperature and salinity An abrupt front between two distinctly different water masses was observed during all four cruises. Though referred to herein as a Gulf Stream-Virginia Shelf Water front, salinity measurements (Table 2) indicate that some dilution of Gulf Stream water had occurred. Salinity of Gulf Stream water is greater than 36°/oo (Stefansson and Atkinson, 1967) while salinity of bottom water on the southern edge of the front was generally 35.2-35.8°/oo. Presumably some mixing of Gulf Stream water with shelf water from the South Atlantic Bight had occurred. The magnitude of the thermal change across the front and the sharpness of the gradient varied seasonally (Figure 2). The greatest 72 Table 1. Grain size percentages and selected sediment parameters for all stations sanpled. Median Percent Percent Grain Sorting Shell and Percent Silt and Station Size(0) ______(0)______Gravel______Sand______Clay June 1977 51 2.24 0.55 0.02 99.54 0.44 49 2.40 0.65 0.18 98.95 0.87 47 2.39 0.64 0.06 99.CH 0.94 45 1.12 1.17 4.67 95.85 0.68 43 1.13 1.05 9.15 96.27 0.68 42 1.89 1.79 8.15 89.51 2.34 41 2.16 0.99 0.87 96.12 3.GL 40 2.02 1.35 3.74 92.45 4.14 39 2.42 1.20 1.71 95.24 3.05 38 1.60 0.85 1.10 98.23 0.67 37 1.66 0.76 0.65 97.54 1.81 36 1.53 0.70 0.43 98.61 1.30 35 1.93 0.59 0.10 99.33 0.56 34 2.51 0.77 0.09 96.85 3.06 33 2.57 0.74 0.06 96.37 4.59 31 2.73 0.75 0.04 93.45 6.51 29 2.80 0.81 0.11 96.17 3.66 27 2.96 0.60 0.04 95.60 4.35 25 2.97 0.61 0.09 95.19 4.72 23 2.94 0.50 0.03 96.52 3.66 August 1977 49 2.40 0.95 1.69 97.44 0.87 39 2.05 0.94 1.02 95.82 3.16 35 1.67 0.58 0.30 99.03 0.67 31 2.80 0.66 0.06 93.64 6.30 October 1977 49 2.45 0.58 0.14 98.87 0.99 39 2.08 1.58 5.95 91.72 2.32 35 2.16 0.79 0.99 98.29 0.71 34 2.51 0.90 0.62 95.95 3.43 23 2.90 0.50 0.03 97.05 2.92 January 1978 49 2.40 0.97 1.81 96.39 1.80 39 2.10 0.63 0.40 99.39 0.21 35 2.13 0.76 0.83 96.61 0.56 31 2.59 0.56 0.09 97.55 2.36 23 2.96 0.48 0.03 96.03 3.94 73 pi. Table 2. Terrperature and salinity observations of bottom water within each of eight equidistant strata during each of the four sanpling periods (from physical-chemical data base, J. Magnuscn, Univ. of Wisconsin-Madison). I & M £ z § *b s 6 S •6 i f b S3 •n o o o o o g 8888 g g * o o o o w*? S 8 9 3F8 89E 0O * 19 10 * » 74 s 9 1 h t t o o o o o o o o h 3 8 8 8 S 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 9 ^ $ 8 8 8 8 3 ro 8 $ C8 O m C p ro ro C O C O 3 3 O 8 O 8 O 8 | 8 O | 8 O 8 | O O | O j j oj> (y O H iO o c a N o s ^ i in a i a a i a i w f i r m M SM ^-O CO N O 8 8 8 8 ? S « a 8 9 3 3 3 3 8 8 8 8 OLfiNOiOlNinN *»■to & & & & & & & ro ro ro& ro ro roro ro o o o o o o o o • ••••••• ...... • ••••••• ro ro co ro ro 0 o o o o o o o 0O 8 ^ 3 8 8 8 8 in 00 S 3 3 S 8 8BS830' 3 8 S B 8 8 h »— v h W& Table 2. Continued 3* M z ffi s c f s a is f jf & B S — OO???■ ? ? ? > ? P I F P P d d d S R d H S GSe S iG I S s K s s S S B s S S w w w w 00 roro com £2&£2£ift&&2&2 ro ro co ro J> & & & & & & & C& o O o 8 o C o 8?$S O o o C o o O co C O C O C0)0> O o> Ol Ol COl O^ O Ot O C n H s n i H CO N CO CO *H o CO CNJ H m O 44 ococNjuocNjmr^o llllllll • ••••••• • ••«•••• • ••••••• vq H p r**! N O H i FdE55 a a 4 00 W cA $ $ $ & & vo a t—H H H oo oo 444 * cj .-H j> c C*> VO <*■ r^. ^ r j n c o co co f «l CO «tl- Ift 75 «SJ-CO H r H a J) N O s S 8 S ? i & 8 8 2 OCOCNJCOCO^frCOCO 3 8 3 E 8 & S 8 8 co coS3&&S3£2£2QS5 co CO co CO CO O l —» —li i H r S O i 3 O S O 8 O R ? O a R i 8 OOOOOOOO ? ? ? ? ? ! W S d d a d d d d s s s s i-H^xd-coCTvcomr^ s s s a a N CJ N CJ — N CJ CNJ CVJ CNJ »—I CNJ CNJ CNJ CNJ 3 3 S 8 O O 8 O 8 O 3 O 8 O O O ON ON ONON ON ON ON ON O O H O C O H O i O 3 3 3 8 8 8 3 3 & & & & & & & CO& COCO COCO CO COCO CO CO CO CO CO y ( CO nj) ior-*U3r-%^cQr^r-* • ••••••• • ••••••• • ••••••• • ••••••• • ••••••• oo «a-i to o» oo O Ol«t ^ C Q Ol O CO Q ^ t « l O IO ct > ov Figure 2. Isotherms observed during each of the four sampling periods in along-shelf transects across the frontal area. Solid triangles denote the position of each bathythermograph cast. 76 OCTOBER JANUARY K4 20 2 30 77 temperature differences across the front were observed in June and August. During those months bottom temperatures on the warm side of the front were 24-26°C while temperatures on the cold side were about 14°C. In August this temperature gradient was observed over a north-south distance of 27 km, indicating an average temperature change of 0.4°C/km. However, since the sampling transect was not necessarily perpendicular to the front the temperature gradient was probably sharper than that observed. In October seasonal overturning of the Virginia Shelf Water increased bottom water temperatures on the cold side of the front to 18°C. In January the front was poorly defined on the basis of temperature alone, but salinity measurements clearly showed the influence of both Virginia Shelf Water and the Gulf Stream. The dynamic nature of the front is indicated by the range and standard deviation of physical observations made successively in several latitudinal strata (Table 2). During June and August most of the movements of the fro n t (in d icated by a high standard deviation) were confined to the area extending from 35°28'N to 35°40'N. In October the front shifted southward and occupied the southern portion of the study area (35°20' - 35°36'N) for much of the sampling period. On some occasions during October the front was observed to move out of the study area, permitting intrusion of Virginia Shelf Water as far south as Cape H atteras (35°08'N). An appreciation for the actual thermal conditions to which benthic organisms in the front were exposed can be best obtained from 78 a continuous temperature record. In August (Figure 3a) bottom tem peratures continuously flu ctu ated over a range of 11°C. The ra te of thermal change was found to be as great as 8°C within a 3 hr. period (2200 hrs., August 10 - 0100 hrs., August 11). In October (Figure 3b) fro n tal movements were much less e r ra tic than in August, and bottom temperatures remained constant over relatively long periods of time. During much of the October monitoring period the front was located away from the instrument site and the area was under the influence of Virginia Shelf Water. Only for two brief periods did the front move northward, permitting the intrusion of warmer Gulf Stream water over the monitoring site. Biota During the four sampling periods a total of 372 species and 21,177 individuals were obtained. Sixty-nine percent of these taxa and seventy percent of the total individuals were collected during the June sampling period alone. Not only were a greater number of samples taken at that time, but recent heavy recruitment by several species also contributed to the large numbers. The total macrofaunal density at each station sampled during the June cruise was highly variable and showed no consistent trend across the study area (Figure 4). The number of species collected was greatest in the central portion of the study area, with the exception of the anomalously low number of species at station 36. Continued declines in the number of species was evident south of station 31 and north of station 42. Figure 3. Continuous record of bottom water temperature as measured by a stationary chart recorder anchored 0.5 m above the bottom at 35°35.5'N, 75°13.0'W: (a) temperature during 8-14 August 1977; (b) temperature during 23-29 October 1977. 79 UJ UJ I- H < < a Q in ro CM ro CM (Do) 3H n±Va3dW 3± Figure 4. Selected community parameters at each station in the June 1977 sampling transect: (a) density of to ta l macrofauna; (b) to ta l number of species collected in three replicate grab samples. 80 NO. OF SPECIES DENSITY ( INDIV./m8 x 10* ) 70 90 40 50 80 20 0 3 60 25 iut nrh f 5'N) .5 0 0 * 5 3 of north (minutts 30 STATION 5045 Zoogeography: The role of the front as a zoogeographic barrier was investigated by examining the zoogeographic affinities of the fauna throughout the study area. Using data only from the June cruise, all species collected were categorized into one of four distribution types based on records from the scientific literature: 1. Southern - distributed only to the south of North Carolina, 2. Northern - distributed only to the north of North Carolina, 3. Widespread - distributed both to the north and south of North Carolina, 4. Other - all east coast records from North Carolina only. The proportion of species 1n each distribution type was calculated for each station (Figure 5). The majority of species in the sampling area were in the widespread category, with distributions generally extending at least from Florida to Massachusetts. The percentage of widespread species remained f a ir ly constant throughout the study area (ca. 65%), and showed no co n sisten t trend with la titu d e (Spearman rank c o rrelatio n (rs) = 0.14, ©<> 0.05). There was evidence of a transition across the front from species of warm-water affinities at southern stations, to species of cold-water affinities at the northern stations. The percentage of southern species ranged from 11.9 to 28.6% and showed a significant decrease with increasing latitude (rs = -0.54, e< < 0.01). Figure 5 Percentage of species in each of four distributional categories at all stations along the June 1977 sampling transect. 82 iN30«3d 3Aiivnnwno 001 IS- Northern species comprised a progressively smaller proportion of the total fauna with decreasing latitude (rs = 0.65,®* < .01). A number of species showed abrupt decreases in density at the front and often a termination of their distribution. Harmothoe extenuata. Lumbrineris fra g ilis. Polycirrus eximius and Cerastoderma pinnulatum were all abundant in the northern portion of the study area, but were absent from the southern portion which was frequently inundated by warmer water (Figure 6). In the case of P. eximius, this .observation may be coincidental, since the species is known to range as far south as the Gulf of Mexico. However, the front does appear to represent a true distributional Umiit for L. fragilis and C. pinnulatum. No individuals of either species were found at mid-shelf depths to the south of the front despite extensive sampling off Cape Hatteras and Cape Lookout (Weston, unpub. data). Day et a l., (1971) reported L. frag ilis off Cape Lookout at depths of 80-200 m. Apparently this species is able to extend its range southward beyond the front but only by submergence. Abrupt decreases in density of southern species at the front were even more common than were such decreases for northern species. Some notable examples of species abundant only to the south of the front include Spisula solidissima sim ilis, Glottidea p.yramidata, Aglaophamus verri11i and Cylichnella bidentata (Figure 6). G. pyramidata and C. bidentata range only as far north as North Carolina, and those individuals found at the southern edge of the front represent the northernmost records for the species. A. verriHi is also primarily a Figure 6. Changes in density (no. indiv./m2) across the study area in June 1977 for selected macrobenthic species. Note use of log scale for densities of Harmothoe extenuata and Spisula solidissima similis. Shaded area represents zone of most frequent frontal occurence. 84 BL frgqllla Polyclrryt gxlmlm pinnulatum Splaulo Qlottidaa xolldlMimg pyromidata Aflloophomm Cyiichntllq verri Hi bidantoto 85 warm-water species but there are sporadic records as far north as Virginia. The dramatic decrease in abundance of S. solidissima similis at the front and at all stations to the north 1s more likely a result of sediment or factors other than temperature, since the species is distributed as far north as Massachusetts. Many species were found exclusively or were collected 1n disproportionate numbers on only one side of the front, but few species showed a high fidelity to the front Itself. The polychaete Glycera oxycephala, was the only species which demonstrated fidelity to stations with the greatest physical variability though at low densities (Figure 7). Since this species is common across the North Carolina continental shelf (Day et a l., 1971; Weston, Chap. 1, this dissertation) its occurrence in only a restricted region of the sampling transect is probably unrelated to the physical variability of the region. Relative importance of physical variables: While the thermal change across the front is perhaps the most dramatic, changes in salinity or substrate may also be partially responsible for faunal differences. Identification of the physical factors responsible for biotic patterns of distribution was attempted through the use of discriminant analysis, following the procedure of Green and Vascotto (1978). Stations were first segregated into groups using numerical classification, and the physical variables which best accounted for group separation were identified using multiple discriminant analysis. Only the June data were included in these 2 Figure 7. Changes in density (no. indiv./m ) across the study area in June 1977 for the polychaete Glycera oxycephala. Shaded area represents zone of most frequent frontal occurence. 86 WL DENSITY (N 0 ./m 2 ) 0 2 0 3 10------oxvceohala Glycera Glycera 3 ' 0 4 ° 5 3 ‘ 0 3 ° 5 3 1 ...... OT LATITUDE NORTH 1 - - 1 Wafflm W 1 / «««*« “ l l l l i f J 11111111 ' 1 ' ' ‘ 0 5 ° 5 3 analyses in order to eliminate seasonal variation in the physical variables. Four station groups were identified by numerical classification (Figure 8). Group 1 included the 4 southernmost stations of the transect, an area predominantly under the influence of Gulf Stream water. Group 2 stations were located in the center of the transect, in the area most frequently occupied by the front. Group 3 stations occupied a broad area in the central portion of the transect, overlapping Group 2 stations but extending farther both to the north and south. Group 4 was comprised of the 5 northernmost stations. Discriminant analysis was performed using the sediment variables listed in Table 1 and hydrographic variables 1n Table 2. The variables used (Table 3) included median grain size, percentage of shell and gravel, percentage of sand, percentage of silt and clay, sediment sorting coefficient (standard deviation of particle sizes about the mean), mean temperature, thermal variability (standard deviation of repeated temperature observations), mean salinity and salinity variability. The importance of each variable on each discriminant function was determined by use of total structure coefficients (Klecka, 1980). These coefficients were calculated as the Pearson product-moment correlation between the discriminant function scores and the measured environmental variable. When some of the variables may be correlated use of total structure coefficients is preferable to the more often seen standardized discriminant function coefficients (Klecka, 1980). Figure 8 Numerical classification of all stations sampled in June 1977. Station 39 was sampled on three separate occasions. 88 0.0 0.2 0 . 4 0.6 0.8 I i.... » ■t... SIMILARITY 23 25 r C 27 29___ 35 38 39(2) 36 ___ 40 •42 •41 ■39(1) -39(3) -31 33 34 3 7___ 43 45 ■47 ■49 GROUP GROUP 4 GROUP 3 GROUP 2 GROUP I -51 ML •P» 8 R £ S® © o d o 9 R J? R d 00 o 83 8 8 « 8R S & 8 83 E3 o d r H CNJ &i d a 8 CNJ r H CNJ r H r H £ sd r H R o r H r H o & !3 W Rj S {J? R 8 8 1C 8 9 d CNJ CNJ r H s 8 8 S3 a 8 a 3 *d- a> LO 89 90 The overall ch1-squared test of among-group differences was highly significant (x2 = 94.17, d.f. = 27). Each of the three axes held a significant portion of the discriminating Information (Table 4). The first discriminant function was largely representative of differences 1n thermal variability between station groups. Thus, among-group faunal differences are best interpreted on the basis of the frequency of exposure to the thermal front. Sediment variables (sorting, percent sand) were only of moderate importance. The highest total structure coefficients on discriminant function II were those of mean temperature and mean salinity which were highly correlated to one another 1n the study area. The second function, therefore, represented a north-south gradient of faunal change. The southern stations of Group 1 scored highest on function II and the northern stations Of Group 4 scored lowest (Figure 9). The percentage of silt and clay (and its influence on the median grain size) were also of Importance on this function. Discriminant function III, like function I, was largely reflective of thermal variability. Several sediment parameters (median grain size, percent sand, percent silt and clay) were of slightly less importance. Temporal stab ility : Seasonal collections were made at a few selected stations in order to permit assessment of temporal variability. A high degree of seasonal turnover of the numerically dominant species was evident (Table 5). Only the archiannelid Pol.yqordius sp., and the polychaete Table 4. Total structure coefficients for all variables on the three discriminant functions. Those variables considered most important on each function are underlined. Discriminant Finction I II III Percentage of variance 67.90 24.82 7.29 Mean tenperature -0.13 0.90 0.31 Thermal variability -0.70 -0.22 0.56 Mean salinity -0.09 0.89 0.22 Salinity variability -0.46 -0.11 0.35 Median grain size 0.13 0.67 -0.44 Percentage of shell and gravel -0.08 -0.21 -0.13 Percentage of sand 0.43 -0.44 0.48 Percentage of silt and clay -0.37 0.76 -0.48 Sediment sorting coefficient -0.48 -0.18 -0.21 91 Figure 9. Centroids (indicated by asterisks) and limits of station groups in two-factor discriminant space. Vectors indicate the relative orientation of the most important environmental variabdes. 92 DFH GROUP I GROUP 3 DF I GROUP 2 ) GROUP 4 R. a a \ t/l "8 . 8 ■e « cs in • r - s § CO 5 8 1 I a-'s $1 g g g ""S •I— s i I!ft »p» I? OLI S3 8 8 9 S 5 5 6 6 K /> si 1 B B •i—' s !i l. i § .(5 a £ > r H § I c$w - CO fe ,2-»w 0 0 — r“ s i m § • & s •r— 8 •n«o a 5 ‘ LT) (O % 93 Nephtys pi eta , were consistently among the dominant species during all four sampling periods. The oUgochaete Tub1f1co1des dukei was abundant throughout most of the year but only at the northernmost sampling site . More typically, dominant species exhibited high abundances during one sampling period and were to tally absent or 1n low densities during the other periods. Examples of such species Include the molluscs Splsula solidissima slmilis and Abra aequalis which showed heavy larval recruitment 1n the southern front area prior to June. Further examples of such species Include Notomastus latericeus, Spiophanes bombyx and Ampharete amerlcana. The relative temporal stab ility for each of the sampling sites can be calculated by use of the Peterson Index (Peterson, 1975) which uses the mean Bray-Curtls similarity between all possible pairs of collections at a site as a measure of temporal stab ility of the community. Because of the data gaps created by the lack of samples from station 23 in August and station 31 1n October, three alternative procedures wre used in calculating this Index. One set of calculations was done using stations 35, 39 and 49 over all four sampling seasons. A second calculation was performed using stations 23, 35, 39 and 49 but for only three seasons, excluding August. Finally, the index was calculated for stations 31, 35, 39 and 49 for three seasons excluding October. Regardless of which procedure was used in calculating the index, the results were nearly the same (Figure 10). The community at the southernmost s ite , station 23, showed the greatest temporal stab ility , Figure 10. Temporal stability, as determined by the Peterson index, for five sampling sites. ( ------) = stations 35,39 and 49 for all sampling periods; ( ----- ) = stations 23,25, 39 and 49 for three sampling periods excluding August; (------) = stations 31,35,39 and 49 for three sampling periods excluding October. 95 INDEX OF TEMPORAL STABILITY 0.3- 23 mnts ot o 35° 5i) i .5 0 °0 5 3 of north (minutes STATION 53 49 39 35 96 followed closely by the northernmost site, station 49. Stations 35 and 39 in the center of frontal area consistently demonstrated the lowest temporal stab ility . DISCUSSION The Gulf Stream-Virginia Shelf Water front examined is part of a larger Gulf Stream-shelf water frontal system extending from Florida to North Carolina. Farther to the north as the Gulf Stream turns eastward, slope water intrudes along the western margin of the Gulf Stream, preventing any interaction of shelf water with the main body of the Gulf Stream. South of North Carolina the Gulf Stream-shelf water front is typically located near the shelf break. However, because of the narrowness of the continental shelf and the steepness of the bottom topography off Cape Hatteras, this front is found in shallower water and closer to the coast than at any point north of Cape Canaveral. Meanders in the Gulf Stream result in movements of this front which are frequent, rapid and variable in extent. During the study period the bottom water temperatures across the front differed as much as 12°C within a 27 km distance. Additionally there was a 5°C difference in mean annual bottom temperatures over a latitudinal distance of only 60 km. Consequently the front represents a formidable zoogeographic barrier to both northern and southern species alike. In the study area, the front is responsible for many of the numerous faunal range terminations generally ascribed to Cape Hatteras. Many of the species collected demonstrated an abrupt decrease in density in the region most frequently affected by the front. For some species the frontal zone represents the known terminus of their distribution along the east coast of North America. 97 98 The zoogeographic influence of the front is even more striking if one considers the annual frontal migrations which restrict many immobile warm-water stenothermal species to the shelf south of Cape Hatteras. In October 1977 the front was observed to migrate out of the study area, allowing cold Virginia Shelf Water to reach Cape Hatteras. The seasonal intrustion of Virginia Shelf Water around the Cape and Into the northern and western portions of Raleigh Bay is wel1-documented (Bumpus and Pierce, 1955; Wells and Gray, 1960; Gray and Cerame-Vivas, 1963; Hunt et a l., 1977). As a result of movements of the front, and the intrusion of cooler waters, many warm-water species (e.g. Nereis riisei, Calyptraea centralis, Tellina aequistriata) common off Cape Lookout do not range as far north as the study area, but instead reach their distributional limits in Raleigh Bay or at Cape Hatteras (Weston, 1979; unpub. data). The majority (ca. 65%) of the species collected were widely distributed and therefore presumably eurytoplc. Their distributions encompassed both the warm waters of the South Atlantic Bight and the cooler waters of the Middle Atlantic Bight. It would be expected that these species would be the most prevalent in a physically unstable area such as the front. The occurrence of southern species (ca. 20% of the total fauna) in the frontal area is more intriguing. These species are distributed throughout the South Atlantic Bight, and in most cases, the Caribbean as well. On the basis of their distribution, 1t is presumed that their tolerance of cold water is very limited. In the frontal zone Individuals of these southern species can be exposed to daily temperature changes as great as the annual temperature changes encountered throughout the remainder of their distribution. It is possible that the occurrence of southern species in the frontal area results from the fact that the thermal requirements for successful reproduction are often more stringent than for survival alone. In the frontal area, populations of southern species may be able to survive the rigors of the physical environment, but are unable to reproduce successfully. These species may be maintained in the frontal area only by continued recruitment from populations in warmer waters to the south. The physical variability of the front appeared to exert some influence on the macrobenthic community structure. Temporally, faunal stability was minimized in the zone of greatest physical variability. Spatially, the boundaries of faunal assemblages were determined largely on the basis of frequency of exposure to the front. Substrate differences among the stations were of some biological significance, even in the relatively small range of sediment types encountered (0 to 7% silt and clay). However, the influence of substrate on faunal distribution was overshadowed by the importance of hydrographic factors, particularly temperature. While salinity varied concurrently with temperature in the frontal region, the thermal variability appeared more important in the interpretation of faunal distribution patterns. Based on interpretation of the discriminant analysis, it appeared that the macrobenthos is better able to cope with salinity changes 1n the range of 30 to 36°/oo than temperature variation between 13° and 24°C. In light of the abrupt physical discontinuity at the front, it is tempting to view the front as an ecotone, a transitional zone created by the juxtaposition of two or more diverse habitats. Ecotones are characterized by sharp truncation of species distribution curves, and often, an increased number of species, an increased population density or biomass and unique species which are absent in the adjoining habitats (Odum, 1971). The fishes of the frontal region have been clearly shown to respond to the front as an ecotone with increased species richness and density in the region with the sharpest thermal gradient (Magnuson et a l., 1981). Ecotonal species were also observed, through these ecotonal species were not seasonally persistent. During certain months species such as Micropogonias undulatus, Lei ostomus xanthurus and Cynosci on regal is were col 1ected in extremely high densities at the front but were absent in the adjoining, thermally stable areas to the north and south. In contrast, the benthos showed only slight evidence of an ecotonal response to the frontal area. The number of species in the frontal area was slightly elevated relative to those stations outside the zone of greatest physical variability. However, the total macrofaunal density showed no consistent pattern and there was l i t t l e evidence of ecotonal species. The contrast in the response of the benthic community to that of the fishes is believed to be a result of differential mobility (Herbst et al., 1979). The fishes are able to 101 migrate with the front as it moves across the shelf, remaining in a zone of preferred temperature and salinity. In contrast, the benthos is essentially sessile and therefore unable to follow frontal movements. The high variability 1n the position of the front prevents the development of an ecotonal benthic community. A front that attains a high spatial permanence, perhaps because of coastal morphology or bathymetric irregularities (Atkinson and Targett, 1983), may evoke a more profound ecotonal response 1n the macrobenthic community than was observed for the Carolinian front. LITERATURE CITED Atkinson, L.P. and T.E. Targett. 1983. Upwelling along the 60-m iso bath from Cape Canaveral to Cape Hatteras and its relationship to fish distribution. Deep-Sea Res. 30:221-226. Bray, J.R. and J.T. Curtis. 1957. An ordination of the upland forest comnunities of southern Wisconsin. Ecol. Monogr. 27:320-349. Briggs, J.C. 1974. Marine Zoogeography. McGraw-Hill, New York. 475 pp. Bumpus, D.F. and E.L. Pierce. 1955. The hydrography and the d is tr i bution of chaetognaths over the continental shelf off North Carolina. Pap. Mar. Biol, and Oceanogr., Deep-Sea Res., Suppl. to 3:92-109. Cerame-Vivas, M.J. and I.E. Gray. 1966. The distribution pattern of benthic invertebrates of the continental shelf off North Carolina. Ecology 47:260-270. Clifford, H.T. and W. Stephenson. 1975. An Introduction to Numerical Classification. Academic Press, New York. 229 pp. Dana, J.D. 1853. On an isothermal oceanic chart, illustrating the geographical distribution of marine animals. Am. J. Sci. 16:153-167, 314-327. Day, J.W., J.G. Field and M.P. Montgomery. 1971. The use of numeri cal methods to determine the distribution of the benthic fauna across the continental shelf of North Carolina. 0. Anim. Ecol. 40:93-125. Fournier, R.O., J. Marra, R. Bohrer and M. Van Det. 1977. Plankton dynamics and nutrient enrichment of the Scotian Shelf. J. Fish. Res. Board Can. 34:1004-1018. Gray, I.E. and M.J. Cerame-Vivas. 1963. The circulation of surface waters in Raleigh Bay, North Carolina. Limnol. Oceanogr. 8:330-337. Green, R.H. and G.L. Vascotto. 1978. A method for the analysis of environmental factors controlling patterns of species composition in aquatic communities. Water Res. 12:583-590. Herbst, G.N., D.P. Weston and J.G. Lorman. 1979. The distributional response of amphipod and decapod crustaceans to a sharp thermal 102 front north of Cape Hatteras, North Carolina. Bull. B1ol. Soc. Wash. 3:188-213. Hunt, R.E., D.J.P. Swift and H. Palmer. 1977. Constructional shelf topography, Diamond Shoals, North Carolina. Geol. Soc. America Bull. 88:299-311. Iverson, R.L., L.K. Coachman, R.T. Cooney, T.S. English, J .J . Goering, G.L. Hunt, J r ., M.C. Macauley, C.P. McRoy, W.S. Reeburg and T.E. Whitledge. 1979. Ecological significance of fronts in the southeastern Bering Sea. Iru R.J. Livingston (ed.), Ecological Processes 1n Coastal and Marine Systems, Marine Science Vol. 10. Plenum Press, New York. pp. 437-466. Klecka, W.R. 1980. Discriminant Analysis. Sage Univ. Paper Series Quantitative Applications in the Social Sciences, series no. 07-019. Sage Publications, Beverly Hills and London. Magnuson, J .J ., C.L. Harrington, D.J. Steward and G.N. Herbst. 1981. Responses of macrofauna to short-term dynamics of a Gulf Stream front on the continental shelf. In: F.A. Richards (ed.) Coastal Upwelling. Coastal and Estuarlne Sciences, Vol. I. American Geophys. Union, Washington, D.C. pp. 441-448. Mills, E.L. 1980. The structure and dynamics of shelf and slope ecosystems off the north east coast of North America. Itk K.R. Tenore and B.C. Coull (eds.), Marine Benthic Dynamics, Belle Baruch Library in Marine Science No. 11. Univ. South Carolina Press, Columbia, pp. 25-47. Nie, N.H., C.H. Hull, J.E. Jenkins, K. Stelnbrenner and D.H. Bent. 1975. SPSS: Statistical Package for the Social Sciences. McGraw-Hill, New York. Odum, E.P. 1971. Fundamentals of Ecology. W.B. Saunders Comp., Philadelphia. 574 pp. Peterson, C.H. 1975. Stability of species and of community for the benthos of two lagoons. Ecology 56:958-965. Savidge, G. 1976. A preliminary study of the distribution of chloro phyll a . in the vicinity of fronts 1n the Celtic and Western Irish Seas. Estuarine and Coastal Mar. Sci. 4:617-625. Stefansson, V. and L.P. Atkinson. 1967. Physical and chemical pro perties of the shelf and slope waters off North Carolina. Tech. Rep., Duke Univ. Marine Lab., Beaufort, North Carolina. 230 pp. Wells, H.W. and I.E. Gray. 1960. The seasonal occurrence of Mytilus edulis on the Carolina coast as a result of transport around Cape Hatteras. Biol. Bull. 119:550-559. 104 Weston, D.P. 1979. Distribution of macrobenthic Mollusca and Amphi- poda in relation to a sharp thermal front: Cape Hatteras region, North Carolina. M.A. Thesis, College of William and Mary, Williamsburg, Virginia. 120 pp. Chapter 3 COMPARATIVE BIOGEOGRAPHY OF THE NORTH CAROLINA CONTINENTAL SHELF MACROBENTHOS D. P. Weston 105 ABSTRACT The benthic macroinvertebrates of the central continental shelf off North Carolina were sampled 1n a broad zone extending from Cape Lookout to 85 km north of Cape Hatteras. Most species were of warm-water affinity with distributions extending southward into the Caribbean. For 37% of the species collected the North Carolina area represented the northward terminus of th eir distributions. In contrast only 4.5% of the species collected were at their southern limit within the study area, with most northern species continuing th eir distribution southward to Florida. The Polychaeta, Amphlpoda, Bivalvia and Gastropoda were contrasted with regard to biogeographlc affinities, extent of geographic range, and ability to traverse the Cape Hatteras area. The polychaetes and amphipods generally exhibited the broadest geographic distributions while the bivalves, and particularly the gastropods, were the most narrowly distributed. These differences in geographic distribution were not Interpretable solely on the basis of the relative dispersal capabilities among the four groups considered, but also suggested a comparatively lower degree of eurytopy in the molluscs. The present-day distributions and biogeographic a ffin ities of the macrofauna were also a function of the geologic history of the northwestern Atlantic and the evolutionary origin of the fauna. 106 INTRODUCTION Biogeographers have long attempted to divide the marine environment along the continental margins Into discrete faunal provinces. Unfortunately they have rarely been able to agree on how to do so, with the result that the literature contains a bewildering variety of alternative proposals and terminologies for marine provinces and climatic zones. For example, Hazel (1970) presented nine alternative classifications proposed by various workers for the biogeographic subdivision of the Atlantic coast of North America. Even these nine schemes are far from all-inclusive. There are many other less widely recognized proposals for the delineation of faunal provinces along the North American Atlantic coast. Two primary reasons for this difficulty in establishing a commonly accepted biogeographic scheme are apparent. Biogeographers working with shallow-water or intertidal fauna may not arrive at the same provincial boundaries as those working with fauna in deeper waters. Barriers limiting the distribution of species on the continental shelf may not be equally as effective in limiting the distribution of intertidal species (Jackson, 1974). Biogeographic provinces stra tifie d in relation to depth have been proposed to deal with this problem (Bowen et a l., 1979; Cerame-Vivas and Gray, 1966; Franz and M errill, 1980a; Watling, 1979). 107 The differences in biogeographic patterns between the major faunal groups are an additional confounding factor in biogeography. Numerous biogeographic studies have been done on the molluscan fauna (Coomans, 1962; Franz and M errill, 1980a, 1980b; Hall, 1964; Johnson, 1934) but all faunal groups do not behave sim ilarly. Temperature is generally regarded as the single most important factor in determining faunal distributions (Gunter, 1957). However, if this were the only factor involved, all faunal groups would be expected to have nearly identical biogeographic patterns, and this is clearly not the case. This study is designed to contrast the biogeographic patterns of several major inacrofaunal groups 1n the vicinity of Cape Hatteras, North Carolina, an area widely recognized as a region of great biogeographic significance (Briggs, 1974). Jackson (1974) and others have suggested that species with a long pelagic larval life should generally have broader geographic ranges than those without a pelagic life phase or one of only short duration. This study will examine the hypothesis that macrofaunal groups with a high proportion of species having planktotrophic larvae (e.g. Bivalvia) should be better able to traverse Cape Hatteras and have broader geographic ranges than groups without planktotrophic larvae (e.g. Amphlpoda). Biogeographic patterns are often best understood by consideration of geologic history and past opportunities for faunal migrations (Franz and M errill, 1980b). It is also the objective of this study to review the geologic history of the North Carolina marine environment K, _...... 109 throughout the Tertiary and Quaternary in an attempt to understand the evolutionary origin of the present-day continental shelf macrofauna. METHODS Ten sampling areas on the North Carolina continental shelf (Figure 1) were located within a region extending 85 km to the north and 69 km to south of Cape Hatteras in a depth range of 23-54 m. The region between Cape Hatteras and Oregon Inlet was occupied by a dynamic Gulf Stream-Virginian Shelf Water front. This area was subdivided into five strata with each stratum designated by the minutes of latitude between its midpoint and latitude 35°00.5'N (e.g. the frontal stratum extending from 35°23.5'N to 35°27.5'N was designated Front:25). Sampling areas were occupied in June, August and October 1977 and January 1978. Macrofauna was collected using a 0.1 m2 Smith-Mclntyre grab. Most organisms were separated from the sediment by elutriation through a 0.5 mm mesh screen. The material remaining after elutriation was washed through a 1.0 mm mesh screen. All animals in both the 0.5 and 1.0 mm fractions were preserved in formalin, sorted to major taxon and identified to species. During the June sampling period nine samples were taken in each frontal stratum and eight samples taken in each of the remaining areas. During the three other sampling periods combined, a total of 17 samples was taken in each area with the exception of the frontal strata where sampling intensity was variable. Sediment samples collected from each biological grab were dried at 100°C and 100 g aliquots of each sample were obtained by use of a 110 Figure 1. Location of the ten sampling areas on the North Carolina continental shelf. Ill NORTH '« OREGON QFRONT-*49 J INLET _ _ I LJFRONT:43 \ []FR0NT:37 [ _ _ i [jFR0NT:3l ' i [~] Fr o n t *25w/ HATTERAS: NORTH CAPE HATTERAS HATTERAS:MID / / HATTERAS sSOUTH SOUTH CAPE LOOKOUT sediment cutter. The material was sieved through a series of Standard Sieves with silts and clays combined as pan weight. Water temperatures was measured at each station using a shallow-water mechanical bathythermograph (BT). Measurements taken periodically with a reversing thermometer were used to correct the BT reading. RESULTS Physical environment In spite of the relatively narrow latitudinal range encompassed in this study, there were large differences in the thermal regimes among the sampling sites (Table 1). The mean annual temperature in the South area was 8°C greater than that in the North area even though they are separated by a latitudinal distance of only 154 km. Seasonal changes in the bottom water temperature of the South area closely paralleled atmospheric thermal conditions with the warmest bottom water temperatures (26.0°C) observed 1n August and the coldest temperatures (17.3°C) observed in January. Bottom water of the North area was warmest in October (17.7°C). This is typical of the stratified shelf waters of the Middle Atlantic Bight, because of seasonal overturning of the water column which occurs each fall. The influence of both warm and cold bottom waters was evident in the Front areas, particularly 1n June and August. This region has been found to be frequently occupied by a sharp Gulf Stream-Virginian Shelf Water front (Herbst et a l., 1979; Magnuson et a l., 1981). The front has been found to move across the shelf exposing this region to rapid changes in temperature and salinity. The Hatteras sampling areas were predominantly under the influence of warm water with bottom water temperatures sometimes 113 & Table 1. Seasonal and annual means and ranges of bottom water tenperatures (°C) throughout the study area (from physical-chemical data base, J. Magnuson, University of Wisconsin-Madison). 3 • ^ H3 SH H < H 3 CO i£j r^. n co 3 cn3 • ci 9 S w H 3 9 & S 5? Km* w cvj 1 s • 3 3 co 3 SH 4 3 4 9 3 o ^ > - - O9CO CO 9 9 «-3 0-3 o>3 ,^3 3 o t rH *~K 9 9 3 3 3 3 o O H 4 3 4 3 4 3 C T i F dCO K 3 4 3 co 3 CO Fd O' 9 *—N 9 r-| ***N .-“V 3 3 CM* r-s. • H £3 4 d 4 9 4 9 4 9 &54 4 d 4 9 cm So co 9 ocrf 114 d> 3 s? w 3 & St 3 CO P3 Km* 3 3 CO CO •94 9 • H 9 - o? cvi 3** 4 9 00 22.7 26.0 22.7 17.3 22.2 (22.2-23.3) (25.2-26.7) (20.2-26.7) (16.2-18.1) (16.2-26.7) exceeding those of the South area. Cold water may periodically intrude into the area as evidenced by the low temperature (16.4°C) observed in Hatteras:North during October. Sediments throughout the area were generally fine to medium sands with less than five percent silt and clay. Around Cape Hatteras, the modification of currrent patterns by Diamond Shoals resulted in the establishment of a variety of sedimentary regimes. In portions of the Hatteras .’South and Hatteras:M1d areas, the protection afforded by the Shoals allowed deposition of finer materials. The proportion of s ilt and clay exceeded 20% in these areas. In much of the Hatteras:North area, current velocities were apparently greater since sediments were extremely well-sorted sands (standard deviation about the mean 0.4-0.6 0) with negligible amounts of s ilt and clay (<0.1%). Biological environment Diverse macrobenthic assemblages were present on the North Carolina continental shelf (Table 2), with over 624 species collected, 422 of which were collected during the June cruise alone during which sampling was most intensive. Most groups were represented by too few species to allow analysis of biogeographic patterns, therefore only the Polychaeta, Bivalvia, Gastropoda and Amphipoda will be considered in d e ta il. . A large number of undescribed species was collected (ca. 40), particularly in the South area, reflecting our inadequate knowledge of the fauna of the South Atlantic Bight. Description of some new taxa Table 2. Macnofaunal groups collected. Nutters in parentheses are total nutter of species collected during all four sanpling periods and nuiter collected during Jute alone. CNIDARIA Crustacea (continued) Anthozoa (1+/1+) - partially analyzed Mysidacea (4/2) Hydrozoa - unanalyzed Cunacea (B*-/&*-) - partially analyzed Tanaidacea - utanalyzed PLATYHELMINTHES (2/1) Isopoda (12/10) Anphipoda (56/47) RHYN0000ELA (12/5) Decapoda (52/20) ANNELIDA SIPUNCULA (8/7) Poly:haeta (258/188) Oligochaeta (13+/13+) - partially analyzed ECHIURA (1/1) MXLUSCA PHORONIDA - unanalyzed Polyplacophora (3/2) Scaphopoda (5/2) ECTOPRQCTA - utanalyzed Gastropoda (68/41) Bivalvia (77/47) BRACHIOPOQA (1/1) Cephalopoda (2/0) CHAET0GNA7HA - utanalyzed ARTl-ROPOQA Pytnogonida (4/3) ECHINOOERMATA Crustacea Asteroidea (2/0) Cephalocarida (1/1) Ophiuroidea (4/1) Ostracoda (24/19) Holothuroidea (1/0) Ccpepoda - utanalyzed Echinoidea (2/0) Cirripedia - unanalyzed Stomatopoda (2/1) CEPHALOCHORDATA (1/1) 116 E. 117 collected in this study have already appeared in the literature (Bertelsen and Weston, 1980; Erseus, 1982) and others are in preparation. Twenty-two range extensions have been established by these collections (Appendix 1) and are included in biogeographic analyses below. Biogeographic affinities The biogeographic affinities of species collected in each of four dominant groups were evaluated on the basis of species distributions given in the literature. Each species collected was assigned to one of eight distributional categories (Figure 2) using the provincial boundaries for the United States east coast as delineated by Hall (1964), with the exception that the Caribbean/Carolinian boundary was placed at Cape Canaveral rather than near St. Augustine, Florida as Hall suggested. The distributional cateogries used and their abbreviations are: Caribbean-Carolinian (Cb-C) - south of Cape Canaveral to south of Cape Hatteras, Carolinian (C) - north of Cape Canaveral to south of Cape Hatteras, including the northern Gulf of Mexico, Carribbean-Virginian (Cb-V) - south of Cape Canaveral to south of Cape Cod, Carolinian-Virginian (C-V) - north of Cape Canaveral to south of Cape Cod, Caribbean-Nova Scotian (Cb-NS) - south of Cape Canaveral to north of Cape Cod, Carol inian-Nova Scotian (C-NS) - north of Cape Canaveral to north of Cape Cod, Figure 2. Geographic extent of the eight distributional categories into which all macrofaiinal species considered were grouped. Arrows indicate species distributions may extend beyond area shown. 118 (A O CAPE CANAVERAL Virginian (V) - north of Cape Hatteras to south of Cape Cod, Virginian-Nova Scotian (V-NS) - north of Cape Hatteras to north of Cape Cod. Not included in these analyses were undescribed or undetermined species (e.g. Grubeulepis sp. A, Magelona sp.), species with uncertain distributions (e.g. Mooreonuphis nebulosa, Schistomeringos rudolphi), species endemic to North Carolina (e.g. Polydora hartmanae, Polycera chilluna, Cryoturris elata and Maera carol ini ana), and species with all East Coast records from North Carolina (e.g. Heteropodarke hetermorpha, Lumbrineriopsis paradoxa). Species of southern affinity were far better represented than those of northern affinity (Table 3). In the Mollusca, species primarily distributed south of Cape Hatteras outnumbered those primarily distributed north of Hatteras by a six to one ratio. A greater northern component was evident in the Polychaeta and Amphipoda, but southern species s till outnumbered northern species by a ratio of two to one. The relative proportions of species with 1) distributions traversing Cape Hatteras, 2) broad distributions in the western Atlantic, and 3) distributions extending beyond the western Atlantic were used to determine which groups were most broadly distributed. By all these criteria the Polychaeta and Amphipoda had the broadest distributions. Approximately 70% of the species in both groups were distributed on both sides of Cape Hatteras while only about 40% of the Bivalvia and Gastropoda had distributions extending around Hatteras. Though most of the amphipod species (70.5%) had distributions Table 3. Biogeographic affinities and distributions of species in the four dominant macrobenthic groups, as indicated by percentages of total species considered (n). Collections from all four sampling periods are incorporated. S- ‘S-3 - *A-q3) A-3 ‘ SN-q3 ‘ (SN-3 uuapaM uuapaM anusp apads d a p sa paqnquqsip . i:o pb qqjiou pub i|:pos j.0 S-3 o Rub UV (SN-q3) apuBLTV UU34S3M aas u^aqqnos sapads uuaq^aoN sapads (A-q3 ‘3 ‘3-q3) ‘3-q3) ‘3 (A-q3 SN-3) ‘ A ‘ (SN-A /Clsp.lm sapads ul UBOuauiBLqduiv DLiueueLndiuy eaB adBQ seua^BH paqnquqsip paqnquqsip ouuapua ouuapua 120 rH - c >- vn oo CM LT) in OII UJ VO a. o —J CM o 10 (0 10 □C r-» < CO CM h- < to o l-H 00 co in 00 co r%. 00 CO > O c rH 00 CM oo 0i to CD c c :c oo h- c ^ o ex2 MCM CM § 10 a: LO o < • h i n ii CM o io in o\ & . s f o LO LO C\J o c < 2 a. >3-° in in CM CM (0 10 CM 10 CO o c < 00 in LO —I II co CM 0» extending on both sides of Cape Hatteras, very few (4.5$) had distributions extending from the Caribbean province to the Nova Scotian province. Only two species, Rhepoxynius epistomus and Ampelisca agassizi were distributed along the entire east coast of the United States. Of the 176 species of polychaetes collected, only 41.5% were endemic to the western A tlantic. In contrast, 65.9% of the amphipods and 90.8% of the molluscs were endemic. There were more species with amphiamerican distributions than there were with amphiatlantic distributions. Latitudinal changes in the biogeographic affin ities of the fauna were evaluated using the June data only (Figure 3). In the Polychaeta the most common distributional types were the Caribbean-Virginian and Carolinian-Virginian. The stenothermal warm-water species (Cb-C) comprised an appreciable percentage of the fauna only south of Cape Hatteras and particularly in the South sampling area. Though Caribbean-Carolinian species comprise 18% of the polychaete fauna of the South area, they are generally found in very low densities. Examples included Aphrogenia alba, Mexieulepis weberi, Phyllodoce castenae and Glycera tesselata. Carolinian species represented 9 to 20% of the polychaete fauna throughout the study area, even in the sites north of Cape Hatteras where they would not be expected to occur since their northern boundary is generally considered to be Cape Hatteras. Figure 3. Proportional distribution of species within the eight distributional categories as grouped by major taxon and sampling area. Number in parentheses indicates number of species of each taxon collected. 122 SOUTH HATTERAS* HATTERAS* HATTERAS* FRONT* FRONT* FRONT* FRONT* FRONT* NORTH SOUTH MIO NORTH 25 31 37 43 49 tl] i L j i l l u l l l u L l c f l » i » l “ L b ] J « 1 ; (0 1 (50) 1 0 8 40-] 40-] 4 0 -i -i 0 4 i 40-1 01 (86) 50-1 ) T » 50-, M] 0, (17) 50-, (25) 1 0 5 : i l ; OYHEA BIVALVIA FOLYCHAETA L h <«•> . - J O O O A O O A • z s •z •z • ) , 4 , . O A 1 a a 0 > > »•> »•> « » f" f I 1 : -I 1 -1 i h b l 1 L 1 L i 1 . 1 L 1 l l a l l 3 3 ( ) 3 (2 ■ o > 5 > « > o o I L 1 s I I ' > i i .0 1 (13) (16) (13) (13) (II) >m . r f f f I 1 (14) I L i U l l I OA8TROPOOA l J L 8 l u l l IIOO I 8 M aw.. w la I I I ? 7 O?r o j > (IS) ( (14) 3) (3 ( 12 (A) (I) ( 8 t 9 ) ) ) 1 U 6) (6 l l ] 1 1 - U L a J a ( 1 l l l l 1 a l l l l 1 1 U I I O A A O > O O a AMAHIAODA I'1 * I '*1 1 r f J f H , , l i . J h O I I l l l L I ) (IS (24) (12) 6) (6 10 (•) 6) (6 (9) 1 ) Polychaete species of northern affinity were represented almost entirely by those of Carolinian-Nova Scotian distributions. Virginian and Virginian-Nova Scotian components, species limited in th eir southward distribution at Cape Hatteras, were rare. Only two Virginian polychaete species were collected (Goniadella gracilis, Cabira incerta) and both of these were found north of Cape Hatteras. Four Virginian-Nova Scotian species were collected (Travisia forbesi, FIabel1i gera a ffin is, Cossura longocirrata, Protodorvillea gaspeensis). Of these £. affinis was collected south of Cape Hatteras and its occurrence there had been tentatively suggested by Day (1973). The bivalves displayed biogeographic affinities very different from those of the polychaetes. Carolinian and Caribbean-Nova Scotian species were absent with the exception of the Caribbean-Nova Scotian species Thyasira trisinuata in the South area. The dominant bivalve component was the subtropical fauna distributed throughout the Caribbean and Carolinian provinces. The percentage of Caribbean-Carolinian bivalves was greatest in the South area (52%) but remained high throughout the study area. Even in the North area Caribbean-Carolinian species (Diplodonta punctata, Pi tar fulminatus) comprised 18% of the fauna. Bivalves in sampling sites north of Cape Hatteras demonstrated a strong cold water affinity with comparatively high percentages of Carolinian-Nova Scotian species (e.g. Solemya velum, Ensis directus) and Virginian-Nova Scotian species (PIacopecten magellanicus, Cerastoderma pinnulatum). South of Front:43 a greater percentage of warm-water Caribbean-Virginian species were collected (e.g. Varicorbula operculata, Parvilucina m ultilineata, Argopecten gibbus). The gastropods, like the bivalves, were primarily of subtropical affinity with a high proportion (25 to 100%) of Caribbean-Carolinian species (e.g. Phi line sagra, Cyclichnella bidentata, Niso aegless). The Caribbean-Virginian component (e.g. Acteon punctostriatus, Caecum carolinianum) became increasingly important in the southern areas. Species of northern affinity were rarely collected. Unlike all other groups examined, the gastropods had few Carolinian-Nova Scotian species throughout the study area. A Virginian component was absent and the only Virginian-Nova Scotian species collected was Lunatia heros. The dominant distributional type in the amphipods throughout most of the study area was the Carolinian-Virginian component. Some of the most abundant species of this type were Ampelisca v e r r illi, Lembos websteri, Bathyporeia parkeri, Pnotis pugnator and Listriella barnardi. Caribbean-Carolinian species were collected at most sites although they were best represented at Front:43 (Ampelisca cristoides, Metharpinia floridanus, Phtisca marina) and the South area (A. cristo id es, M. floridanus, Lembos unicornis, Microdeutopus myersi, Podocerus b rasilien sis). As in most of the other faunal groups, Carolinian-Nova Scotian species were most prevalent in the northern portion of the study area, and the Virginian-Nova Scotian species were poorly represented. Rate of faunal change with latitude A comparison of the species composition of the North and South areas is perhaps the simplest method to evaluate the rate of faunal change with latitude and the effectiveness of Cape Hatteras as a zoogeographic barrier. Using combined data from all four cruises, the gastropods show the greatest faunal change with only 11% of the species common to both the South and North areas. Amphipods show the least faunal change with 42% of the species common to both the South and North areas. Bivalves and polychaetes have 32% and 35% shared species respectively. A modification of a method originally proposed by Sanders (1977) was also used to evaluate the relative rate of faunal change with latitude for the four dominant macrobenthic taxa. The Bray-Curtis sim ilarity measure (Bray and Curtis, 1957), with log-transformation of abundances, was used to calculate the faunal similarity between all possible station pairs sampled during the June cruise. Similarity values obtained from the 45 possible station pairs were grouped into categories and the mean latitudinal distance and standard deviation calculated for all station pairs in each similarity category (Figure 4). The percentage of the total station pairs in each similarity category was also calculated (Figure 5). Gastropods demonstrated a tendency to have relatively short latitudinal distances between station pairs with a high sim ilarity, indicative of rapid faunal turnover between sampling areas. The gastropods also had a comparatively larger percentage of station pairs ■ S k , Figure 4 Rate of faunal change with latitude for the four major macrofaunal groups. Mean latitudinal distance and standard deviation between all station pairs within the prescribed similarity categories is shown. 126 E JtC UJ o z < ►“ CO .o < z Q 3 .o H 5 .o oc UJ o z < .o H CO .o O 3 H I- < I I ’I I I O 0> 0> 0> 0 ) m o o sioio^roN — o <£> IO (O , . dcioddooo oooooooo <0 iq * K) CO - A <0 U) <; K? «sl — Aimvimis Figure 5 Proportional distribution of station pairs within each of the prescribed similarity categories for the four majo,r macrofaunal groups. 127 >0.70 0.60-0.69 SIMILARITY 0.50-0.59 0.40-0.49 GASTROPODA POLYCHAETA o o o oo o £ O CI O — o CVI o rO o o — 0 CVJ ro 0 I A I I 1 128 in low similarity categories (53% of station pairs with similarity <0.30). Thus, among the four dominant taxa the gastropods had the narrowest ranges and showed the greatest dissimilarity among the sampling areas. The polychaetes, bivalves and amphipods were sim ilar to one another and distinct from the gastropods in their rate of faunal change. For the polychaetes, the proportional distribution of station pairs was skewed slightly more toward higher similarity categories than for the other faunal groups. However the amphipods tended to have a greater latitudinal distance between station pairs at any given similarity level than either the polychaetes or bivalves. Overall the polychaetes and amphipods showed the least faunal change throughout the study area, followed closely by the bivalves. The gastropods showed the greatest faunal change. 15.... DISCUSSION Most macrobenthic species of the central continental shelf in the vicinity of Cape Hatteras exhibited warm-water affinities. This pattern was clearly seen in all four major macrofaunal groups examined, the Polychaeta, Bivalvla, Gastropoda and Amphipoda, with species of warm-water affinity comprising about one-half to three-fourths of the total species. The dominant faunal component was the stenothermal, subtropical group of species distributed throughout the Caribbean and along the east coast of the United States as far north as North Carolina. Species of the second largest faunal group were also of Caribbean a ffin ities but reached th eir northern boundary between North Carolina and Massachusetts. The Cape Hatteras region was clearly effective as a faunal boundary to the northward range extension of southern species. Thirty-seven percent of the species collected reached th eir northern limit within the study area, although the percentage varied greatly among the four macrofaunal groups (20 to 57%). Numerous Carolinian species for which Cape Hatteras is generally considered to be the northern limit were found up to 85 km north of the Cape near the latitude of Oregon Inlet (Appendix 2). The collection of southern species on the inner and central shelf between Cape Hatteras and Oregon Inlet has previously been reported for molluscs (Merrill et a l., 1978) and podocopid ostracods (Hazel, 1975). 129 The occurrence of Carolinian species within the southern portion of the Virginian province does not Indicate that Cape Hatteras 1s Ineffective as a faunal barrier. Rather 1t reflects the gradual nature of faunal change between provinces. Four some southern species, occurrences north of Cape Hatteras may represent extrallmital, non-reproducing populations ('pseudopopulatlons' of Milelkovsky, 1971) resulting from the northward transport of larvae around Cape Hatteras by the Gulf Stream. This situation would not apply to amphipods collected north of the Cape since they have direct development. However some of the molluscs collected north of Cape Hatteras may be incapable of reproduction and dependent upon continued influx of larvae from the south for maintenance of their populations. The occurrence of populations incapable of reproduction or with a high rate of reproductive failure is probably typical of junctures between biogeographic provinces. Species of cold-water affinity were poorly represented on the North Carolina continental shelf comprising only 17% of the total fauna. The majority of these species were distributed on both sides of Cape Hatteras, and reached their southern limit in northern Florida. Only 4.5% (15 species) of the species collected reached their southern limit of distribution in the Cape Hatteras area. The weak boreal influence is further illustrated by the absence of many typical cold-water genera in the collections such as Brada (Polychaeta), Astarte, Propebela (Mollusca), Harpinia and Hippomedon (Amphipoda). 131 Two factors are probably responsible for the poor representation of boreal fauna in the collections. Many boreal species which are able to penetrate southward beyond Cape Cod are unable to reach Cape Hatteras, but instead reach their southern limit within the Middle Atlantic Bight. This pattern has been demonstrated for many species of polychaetes (Kinner, 1978), opisthobranch molluscs (Franz, 1970), amphipods (Bousfield, 1973) and benthic algae (Searles and Schneider, 1980). The absence of shal1ow-water boreal species south of New Jersey has been attributed to warm summer temperatures (Franz, 1970). A lternatively, many boreal species exhibit equatorward submergence in the Middle Atlantic Bight (Dickinson et a l., 1980; Franz and M errill, 1980a). Boreal species which are able to reach North Carolina may occur in water depths greater than were sampled in this study (54 m). While some boreal species may be distributed in this manner (Gray et a l., 1968), most of the megafauna of the outer shelf between Cape Hatteras and Oregon Inlet 1s of tropical rather than boreal affinity (Cerame-Vivas and Gray, 1966). Fourteen species were collected that were either endemic to North Carolina (4 species) or that have been recorded in the northwestern Atlantic only from North Carolina (10 species). It is doubtful that these species will be found unique to North Carolina as additional collections are made on the continental shelf. The distributions of most of these species will probably be found to extend southward from North Carolina as the benthos of the shelf throughout the South Atlantic Bight becomes as well known as that of the Middle Atlantic Bight. Biogeography and larval dispersal The four macrofaunal groups considered are representative of a wide variety of reproductive strategies and consequently have differing capacities for larval dispersal. At one end of the spectrum are the Bivalvia, with the highest proportion of species having a pelagic larval phase. Direct, lecithotrophic and planktotrophic development have all been reported in the Bivalvia, but the vast majority of temperate and subtropical species, the dominant groups 1n this study, have plankotrophic larvae (Mileikovsky, 1971; Ockelmann, 1965). The Gastropoda and Polychaeta are Intermediate 1n larval dispersal capabilities since some species may have a long pelagic larval phase while other species may forego the pelagic larval phase entirely (Fauchald, 1983; Mileikovsky, 1971; Pettibone, 1982; Robertson, 1976; Thorson, 1950). At the other extreme are the Amphipoda, with no pelagic larval stage. Like all peracaridan crustaceans, amphipods brood th e ir eggs and development is direct. It has been suggested that the greater the length of pelagic larval life the greater should be the animal's ability to bridge geomorphological barriers and the broader should be its geographic range (Hansen, 1980; Jackson, 1974; Scheltema, 1977; Shuto, 1974). Even a relatively minor barrier may limit the distribution of a species without a pelagic life phase since it would have to survive within a hostile environment in order to traverse the barrier. 133 Because of the sedimentary regime of Diamond Shoals, the gemorphology of Cape Hatteras has been suggested to lim it the northward distribution of several echinoderm species lacking a pelagic larval phase of sufficient duration to enable passage around the Cape (Harrington, 1981). If the proportion of species having planktotrophic larval stage was the factor controlling the breadth of distribution in the four macrofaunal groups examined, one would expect the bivalves to be the most broadly distributed, the polychaetes and gastropods intermediate, and the amphipods most narrowly distributed. However based on both the faunal changes within the Cape Hatteras area and the extent of geographic range throughout the Atlantic Ocean in general, the amphipods and polychaetes were the most broadly distributed and the molluscs, particularly the gastropods, were the most narrowly distributed. The amphipods were more broadly distributed than would be expected on the basis of larval dispersability alone, while the bivalves and, particularly, the gastropods were surprisingly narrowly distributed. Three mechanisms probably account for the ease with which the amphipods have bridged the Cape Hatteras barrier. For a minority of the species, most notably the haustoriid amphipods, the sedimentary regime of Diamond Shoals is probably not an inhospitable environment. These species are characteristic inhabitants of the dynamic, well-sorted sands which are typical of the Shoals and continuous along much of the American Atlantic seaboard. For other species, 134 Individuals may periodically be transported around the Cape and even across the Atlantic while clinging to flotsam. The Importance of planktonlc dispersal of adult amphlpods by rafting has become Increasingly evident 1n investigations of remote oceanic Islands (Barnard, 1965; 1970; 1976). Species collected such as Caprella equilibra, Lembos websterl, Melita appendiculata and Jassa falcata are all associated with material such as hydrolds, algae or seagrasses which would increase their potential for dispersal by rafting. The ability of Infaunal amphipods, such as the ampellscids, to traverse Cape Hatteras is more problematic. These amphlpods are not apt to be found in the dynamic sands of Diamond Shoals nor are they associated with raftable material. Their opportunity to traverse Cape Hatteras may have occurred in the geologic past, probably during Pleistocene interglacial periods. This possibility is discussed at length below. Despite the existence of a planktotrophic larval stage, distributions of the molluscs were narrower than those of the amphipods and polychaetes. Within the study areea this was reflected in the dissimilarity of the North and South areas and the rate of latitudinal faunal change. On a broader scale, bivalves and gastropods had fewer species with amphiatlantic or amphiamerican distributions. The correlation between duration of planktonic larval life and extent of geographic range presupposes a range of physiological tolerance sufficient to allow survival in divergent climatic regimes. The comparatively narrow distributions of the molluscs and particularly the gastropods, 1n spite of an efficient dispersal mechanism, may be the result of relatively limited environmental tolerance (stenotopy). However it is almost Impossible to directly determine the relative degree of eurytopy among macrofaunal groups. Instead the degree of eurytopy is generally inferred from geographic distribution (Hansen, 1980; Jackson, 1974; Sanders, 1977). Some experimental evidence is available suggesting greater eurytopy among the Crustacea in comparison to the Mollusca, at least in regard to salinity tolerance (Pearse, 1936; Topping and Fuller, 1942; Wells, 1961). Vermeij (1972) suggested that Crustacea of high-intertidal areas are generally more eurytopic than gastropods in the same habitat. Paleobiogeography The Cape Hatteras area was of varying effectiveness in limiting species distributions among the four macrofaunal groups examined, but the fact that the distributions of a large number of taxa do terminate in this area is indisputable. However throughout the history of the North A tlantic, Cape Hatteras has not always been as effective a zoogeographic barrier as it is today. The Tertiary history of the North Atlantic in reference to Cape Hatteras was discussed by Franz and Merrill (1980b). From the opening of the North Atlantic Basin in the late Triassic (ca. 200 my ago) to the mid-Pliocene (ca. 3 my ago) the Gulf Stream flowed along the e n tire Atlantic coast from Florida to Newfoundland (Berggren and H o llister, 1977). Investigations of Miocene deposits show that species of warm-water affin ity were able to 136 penetrate northwards along the coast beyond Cape Hatteras (Hazel, 1971). Molluscs show no evidence of a faunal barrier at Cape Hatteras at that time (Hecht, 1969; Hecht and Agan, 1972). Two relevant events occurred during the Pliocene. First, elevation of the Isthmus of Panama (ca. 4 my ago) created a continuous land bridge between North and South America, severing the connection between the Atlantic and Pacific and ending exchange of warm-water marine fauna. Secondly, the development of glaciation in the Northern Hemisphere (ca. 3 my ago) resulted in the establishment of the Labrador Current System in the northwestern A tlantic. The Labrador Current System deflected the Gulf Stream southward and established the biogeographic provinces along the American Atlantic coast as we know them today (Berggren and Hollister, 1977). It was at this time, during the mid to late Pliocene, that the Cape Hatteras area first became effective as a faunal barrier. With continued climatic deterioration during the remainder of the Pliocene and into the Pleistocene, the gradient of faunal change in the Cape Hatteras area became increasingly sharper. The transition between cold and warm-water fauna at Cape Hatteras was probably most abrupt during Pleistocene glacial periods. North of the Cape temperatures were lower than at present (McIntyre et a l., 1976) permitting the intrusion into the Middle Atlantic Bight of many cold-water species which today are limited by Cape Cod (Hazel, 1968). South of Cape Hatteras sea temperatures were much as they are today. Supplementing the thermal barrier at Cape Hatteras was the physical barrier imposed by the Cape itself. At present the shelf off Cape Hatteras is narrower (23 km) than the shelf at any other point along the eastern seaboard north of Cape Canaveral. With the lower sea levels accompanying glaciation the shelf area available for passage of shallow-water benthic species around Cape Hatteras would have been virtually eliminated. During Pleistocene Interglacial periods the Cape Hatteras area became less effective in limiting the distribution of shallow-water benthic species. Pleistocene Interglacial deposits in Virginia contain many warm-water species which today are restricted to waters south of Cape Hatteras (Richards, 1936; 1962; Valentine, 1971; Oaks et a l., 1974). A 15 m rise above the present sea level would have increased the extent of the continental shelf off Cape Hatteras from the 23 km of today to 135 km. Shelf waters north and south of Cape Hatteras would have intermingled and diffused over a broad expanse of shallow shelf (Valentine, 1971). The marked convergence of isotherms which exists today would not have existed, and shallow-water benthic organisms would have had unimpeded passage around the Cape. Hansen (1980) has demonstrated that species without planktonic larvae are able to transverse geographic barriers during periods of transgression which would be impassable during regressive phases. The marine fauna of the northwest Atlantic has been shown to have a strong similarity to that of the northeast Atlantic, with the percentage of amphiatlantic species decreasing southward (Franz, 1970; Franz and M errill, 1980b; Hazel, 1970; Kinner, 1978). The species collected off North Carolina showed a closer relationship to the Pacific Ocean than to the eastern Atlantic (Table 3) in spite of the 138 fact that there 1s no warm-water connection between the Pacific and Atlantic at present. It is believed that the affinity of the North Carolinian fauna with that of the Pacific can be explained only with regard to the geologic history of the area. The origin of much of the present North Carolinian fauna can be traced to eastern Pacific progenitors which migrated into the western Atlantic prior to the Pliocene establishment of the Central American land bridge. For example, of the four Ampelisea (Amphipoda) species collected in this study (A. vadorum, A. verrilli. A. agassizi, A. cristoides), three are found 1n the eastern Pacific today or are derived from eastern Pacific stocks (Mills, 1965; 1967). The Atlantic 'A. cristoides1 may be distinct from the Pacific species (R. Heard, pers. comm.) but its morphological similarity suggests a common origin. Only A. vadorum has little affinity with eastern Pacific species and is probably derived from eastern Atlantic stocks of the late Mesozoic or early Tertiary (M ills, 1965). The phoxocephalid amphipods collected (Metharpinia floridanus, Rhepoxynius epistomus) are found in only the western Atlantic and eastern Pacific and probably migrated into the Atlantic via the Central American region as well (Barnard, 1960). Of the five glycerld polychaetes collected (Hemipodus roseus, Glycera tesselata, ja. americana, G. dibranchiata and G. oxycephala) all have eastern Pacific populations while only one (£. tesselata) is found in the eastern Atlantic. Nearly all the molluscs collected are endemic to the western Atlantic but their east Pacific origin is evident in the number of shared genera and cognate species in the eastern Pacific (Vermeij, 1978). The fact that so few mollusc species are common to the Atlantic and Pacific, suggests a more rapid evolutionary divergence in the molluscs since the Pliocene than in either the polychaetes or amphipods. Throughout much of the Tertiary period species immigrating into the western Atlantic via Central America were able to extend their range throughout the Caribbean, Gulf of Mexico and along the North American Atlantic coast, aided in their northward dispersal by the Gulf Stream. By the Miocene epoch a continuous warm-water fauna existed along the North American coast from Florida to Newfoundland. With the formation of the Labrador Current and the southward displacement of the Gulf Stream a cooler, seasonally-variable marine climate developed on the shelf north of Cape Hatteras. Stenothermal species became restricted to the waters south of Cape Hatteras while eurythermal species were able to adapt to the climatic deterioration and maintain populations on both sides of the Cape. It is from these later groups of eurytopic species of Miocene ancestry that the 'transhatteran' fauna of Franz and Merrill (1980a; 1980b) is derived. The transhatteran fauna has been defined as predominantly estuarine and shallow-shelf species which are endemic to the American Atlantic coast (Franz and M errill, 1980a; 1980b). Most are distributed north to Cape Cod though they may extend as far north as Newfoundland. Their southern boundary may be anywhere from Florida to Brazil. The vast majority of species are of southern affinity but some cold-water species of North Pacific ancestry may also exhibit transhatteran distributions (Franz et a l., 1981). Franz and Merrill (1980a) coined the term 'transhatteran' to describe a species type rather than a geographical zone. The 'Transhatteran Province' of Watling (1979), although restricted to estuarine and shelf habitats, is not sufficiently different from the 'Transatlantic Province' of Woodward (1851-1856) and Johnson (1934) to warrant distinction. Species with transhatteran distributions were well represented in the collections comprising 57% of the total species collected. Forty percent of the molluscs which Franz and Merrill (1980) designated as transhatteran were collected in the present study. A difficulty arises however regarding the question of endemicity. Franz and Merrill (1980a; 1980b) regarded endemicity to the American Atlantic coast as an important distinguishing criterion for transhatteran fauna although two of the transhatteran species they listed (Franz and Merrill, 1980a, App. 3, 4) are found in the eastern Pacific (Thyasira trisinuata, Terebra dislocata). Of the species collected in the present study which have transhatteran distributions, approximately half are endemic to the American Atlantic. In the Mollusca, the group for which the concept of endemic transhatteran species was initially devised, few of the transhatteran species are distributed beyond the American A tlantic. However 35% of the Amphipoda and 63% of the Polychaeta which have distributions traversing Cape Hatteras are not endemic. For groups such as these which typically contain many wide-ranging species, the inclusion of endemicity in the definition of transhatteran species seems too restrictive. CONCLUSIONS 1. The macroinvertebrates of the continental shelf in the vicinity of Cape Hatteras were predominantly of warm-water affinity. The dominant faunal groups were those distributed from the Caribbean northward to North Carolina or, 1n some cases, as far north as Massachusetts. 2. A large number of southern species generally considered to be limited in th eir northward distribution at Cape Hatteras were, in fact, collected up to 85 km north of the Cape. Their occurrence north of the Cape is considered to be reflective of a gradual rather than an abrupt faunal change between biogeographic provinces. 3. Species of northern affinity were poorly represented at mid-shelf depths on the North Carolina shelf. Most northern species either reach th eir southern limit 1n the Middle Atlantic Bight well to the north of Cape Hatteras, or are found only in the deeper waters of the outer shelf off North Carolina. 4. For fauna of the central shelf, the North Carolina area is far more effective in limiting the northward distribution of southern species than i t is in limiting the southward distribution of northern species. 5. The Cape Hatteras reigon was least effective in limiting the distribution of amphipods and polychaetes and most effective in limiting the distribution of gastropods. The ability of the amphipods to traverse Cape Hatteras may have been achieved by rafting of adults or passage around the Cape during Pleistocene interglacial periods. 6. The ability to traverse Cape Hatteras and the extent of geographic distribution in general is not dependent merely upon efficiency of larval dispersal but is also determined by the degree of eurytopy. Molluscs, in general, are presumed to be more stenotopic than either the polychaetes or amphipods. 7. The effectiveness of the Cape Hatteras area as a biogeographic barrier has varied greatly during the geologic history. It was probably least effective prior to the m1d-Pliocene and during Pleistocene interglacial periods. 8. The 'transhatteran1 fauna of Franz and Merrill (1980a) comprised a significant component of the North Carolinian fauna. The transhatteran fauna, as originally conceived, was comprised primarily of species endemic to the American Atlantic, but this criterion is considered too restrictiv e for most macrofaunal groups. 141 LITERATURE CITED Abbott, R. T. 1974. American Seashells. 2nd ed. Van Nostrand Relnhold Co., N.Y. 663 pp. Barnard, J. L. 1960. The amphipod family Phoxocephalidae in the eastern Pacific Ocean, with analysis of other species and notes for a revision of the family. Allan Hancock Pacific Expeditions 18(3):1-375. Barnard, J. L. 1965. Marine Amphipoda of Atolls in Micronesia. Proc. U.S. Nat. Mus. 117:459-551. Barnard, J. L. 1970. Sublittoral Gammaridea (Amphipoda) of the Hawaiian Islands. Smithsonian Contr. Zool. 34. 286 pp. Barnard, J. L. 1976. Amphipoda (Crustacea) from the Indo-Pacific tropics: A review. Micronesia 12(1):169-181. Berggren, W. A. and C. D. Hollister. 1974. Paleogeography, paleobiogeography and the history of circulation in the Atlantic Ocean. Jn: W. W. Hay (ed.), Studies in Paleo-oceanography. Society of Economic Paleontologists and Mineralogists, Spec. Publ. No. 20. pp. 126-186. Bertel sen, R. D. and D. P. Weston. 1980. A new species of Sclerobregma (Polychaeta: Scalibregmatidae) from off the southeastern United States. Proc. Biol. Soc. Wash. 93(3):708-713. Boesch, D. F., J. N. Kraeuter and D. K. Serafy 1977. Distribution and structure of communities of macrobenthos on th outer continental shelf of the Middle Atlantic Bight: 1975-1976 Investigations. Chap. 6 jji: Middle Atlantic Outer Continental Shelf Environmental Studies, Vol. II. Chemical and Biological Benchmark Studies. Prepared by Virginia Institute of Marine Science, Gloucester Point, Va. under Contract No. 08550-CT-5-42 with the Bureau of Land Management, U.S. Dept, of Interior, 111 pp. Blake, J. A. 1979. Four new species of Carazziella (Polychaeta: Spionidae) from North and South America, with a redescription of two previously described forms. Proc. Biol. Soc. Wash. 92(3):466-481. Bousfield, E. L. 1973. Shallow Water Gammaridean Amphipoda of New England. Comstock Publ. Assoc., Cornell Univ. Press, Ithaca, N.Y. 312 pp. 142 Bowen, M. A., P. 0. Smyth, D. F. Boesch and J. van Montfrans. 1979. Comparative biogeography of benthic macrocrustaceans of the middle Atlantic (U.S.A.) continental shelf. Bull. Biol. Soc. Wash. 8:214-255. Bray, J. R. and J. T. Curtis. 1957. An ordination of the upland forest communities of southern Wisconsin. Ecol. Monogr. 27:325-349. Briggs, J. C. 1974. Marine Zoogeography. McGraw H ill, N.Y. 475 pp. Cerame-Vivas, M. J. and I. E. Gray. 1966. The distribution pattern of benthic invertebrates of the continental shelf off North Carolina. Ecology 47:260-270. Coomans, H. E. 1962. The marine mollusk fauna of the Virginian area as a basis for defining zoogeographical provinces. Beaufortia 9:83-104. Day, J. H. 1973. New Polychaeta from Beaufort, with a key to all species recorded from North Carolina. NOAA Tech. Rep. NMFS CIRC-375. 140 pp. Uickinson, J. J ., R. L. Wigley, R. D. Brodeur and S. Brown-Leger. 1980. Distribution of gammaridean Amphipoda (Crustacea) in the Middle Atlantic Bight region. NOAA Tech. Rep. NMFS SSRF-741. 46 pp. Dorsey, J. H. 1977. A new species of Rhynchospio (Polychaeta: Spionidae) from San Clemente Island, California. Bull. So. Calif. Acad. Sci. 76(1):1-4. Erseus, C. 1982. Revision of the marine genus Smithsonidrilus Brinkhurst (Oligochaeta, Tubificidae). Sarsia 67:47-54. Fauchald, K. 1983. Life diagram patterns in benthic polychaetes. Proc. Biol. Soc. Wash. 96(1):160-177. Foster, N. M. 1971. Spionidae (Polychaeta) of the Gulf of Mexico and the Caribbean Sea. Studies on the Fauna of Curacao and other Caribbean Islands, No. 129. 183 pp. Franz, D. R. 1970. Zoogeography of Northwest Atlantic opisthobranch molluscs. Mar. Biol. 7:171-180. Franz, D. R. and A. S. M errill. 1980a. Molluscan distribution patterns on the continental shelf of the Middle Atlantic Bight (Northwest A tlantic). Malacologia 19(2):209-225. 144 Franz, D. R. and A. S. M errill. 1980b. The origins and determinants of distribution of molluscan faunal groups on the shallow continental shelf of the northwest Atlantic. Malacologia 19(2):227-248. Franz, D. R., E. K. Worley and A. S. M errill. 1981. Distribution patterns of common seastars of the middle Atlantic continental shelf of the northwest Atlantic (Gulf of Maine to Cape Hatteras). Biol. Bull. 160:394-418. Gaston, G. R. and D. P. Weston. 1983. Benthos. Chap. 6 In: L. R. DeRouen et al. (eds.), West Hackberry Brine Disposal Project, Post-discharge Studies. Prepared by McNeese State University, Lake Charles, La. for the Dept, of Energy under DOE Contract No. DE-AC96-80P010288. Gray, I. E., M. E. Downey and M. J. Cerame-Vivas. 1968. Sea-stars of North Carolina. Fish. Bull. 67(1).-127-163. Gunter, G. 1957. Temperature. _In: J. W. Hedgpeth (ed.), Treatise on Marine Ecology and Paleoecology, Vol. I. Geol. Soc. Am., Mem. 67. pp. 159-184. Hall, C. A., J r. 1964. Shallow-water marine climates and molluscan provinces. Ecology 45(2):226-234. Hansen, J. A. 1980. Influence of larval dispersal and geographic distribution oh species longevity in neogastropods. Paleobiology 6(2):193-207. Harrington, C. L. 1981. Responses of benthic macrofauna to the dynamics of a Gulf Stream front on the continental shelf near Cape Hatteras, North Carolina. Ph.D. dissertation. University of Wisconsin - Madison. 142 pp. Hazel, J. E. 1968. Pleistocene ostracode zoogeography in Atlantic coast submarine canyons. J. Paleontology 42(5):1264-1271. Hazel, J. E. 1970. Atlantic continental shelf and slope of the United States — Ostracode zoogeography in the southern Nova Scotian and northern Virginian faunal provinces. U.S. Geol. Surv. Prof. Paper 529-E, 21 pp. Hazel, J» E. 1971. Paleoclimatology of the Yorktown Formation (upper Miocene and lower Pliocene) of Virginia and North Carolina. In: H. J. Oertli (ed.), Paleoecologie ostracodes. Centre Rech., Pau, Bull. 5 supp. 1, p. 361-375. Hazel, J. E. 1975. Ostracode biofacies in the Cape Hatteras, North Carolina, area. Bull. Am. Paleont. 65:463-487. 145 Hecht, A. D. 1969. Miocene distribution of molluscan provinces along the east coast of the United States. Geol. Soc. Am. Bull. 80:1617-1620. Hecht, A. D. and B. Agan. 1972. Diversity and age relationships in Recent and Miocene bivalves. Sys. Zool. 21:308-312. Herbst, G. N., D. P. Weston and J. G. Lorman. 1979. The distributional response of amphipod and decapod crustaceans to a sharp thermal front north of Cape Hatteras, North Carolina. Bull. Biol. Soc. Wash. 3:188-213. Jackson, J. B. C. 1974. Biogeographic consequences of eurytopy and stenotopy among marine bivalves and their evolutionary significance. Am. Naturalist 108:541-560. Johnson, C. W. 1934. List of marine Mollusca of the Atlantic coast from Labrador to Texas. Proc. Boston Soc. Nat. Hist. 40:1-204. Kinberg, J. G. H. 1855. Nya slagter och arter of Annelider. Ofv. Svenska Vetensk. Akad. Forh. 12:381-388. Kinner, P. C. 1978. The distribution and ecology of errantiate Polychaetes on the continental shelf from Cape Cod to Cape Hatteras. M.S. thesis, Univ. Delaware. 159 pp. Magnuson, J. J ., C. L. Harrington, D. J. Stewart and G. N. Herbst. 1981. Responses of macrofauna to short-term dynamics of a Gulf Stream front on the continental shelf. _In: A. Richards (ed.), Coastal Upwelling. Coastal and Estuarine Sciences, Vol. I. Am. Geophys. Union, Washington, D.C. pp. 441-448. McIntyre, A. et a l. (35 other authors). 1976. Glacial North America 18,000 years ago: a CLIMAP reconstruction. Geol. Soc. Am. 145:43-75. M errill, A. S., R. C. Bullock and D. R. Franz. 1978. Range extensions of mollusks from the Middle Atlantic Bight. Nautilus 92:34-40. Myers, A. A. 1981. Amphipod Crustacea. I. Family Aoridae. Mem. Hourglass Cruises, Fla. Dept. Nat. Res. Vol. V, Pt. V. 75 pp. Mileikovsky, S. A. 1971. Types of larval development in marine bottom invertebrates, their distribution and ecological significance: a re-evaluation. Mar. Biol. 10:193-213. Mills, E. L. 1965. The zoogeography of North Atlantic and North Pacific ampeliscid amphipod crustaceans. Sys. Zool. 14(2):119-130. Mills, E. L. 1967. A reexamination of some species of Ampelisca (Crustacea: Amphipoda) from the east coast of North America. Canadian J. Zool. 45:635-652. Oaks, R. Q., J r ., N. K. Coch, J. E. Sanders and R. F. Flint. 1974. Post-Miocene shorelines and sea levels, southeastern Virginia. In: R. Q. Oaks, Jr. and J. R. DuBar (eds.), Post-Miocene Stratigraphy -- Central and Southern Atlantic Coastal Plain. Utah State Univ. Press, Logan, Utah. 275 pp. Ockelmann, W. K. 1965. Developmental types in marine bivalves and th eir distribution along the Atlantic coast of Europe. In: Proc. First European Malacological Congr, 1962. pp. 25-35’. O'Gower, A. K. and J. M. Wacasey. 1967. Animal communities associated with Thalassia, Diplanthera and sand beds in Biscayne Bay. I. Analysis of communities in relation to water movements Bull. Mar. Sci. 17(1):175-210. Pearse, A. S. 1936. Estuarine animals at Beaufort, North Carolina. J. Elisha Mitchell Sci. Soc. 52:174-222. Perkins, T. H. 1980. Review of species previously referred to Ceratonereis mirabilis, and descriptions of new species of Ceratonereis, Nephtys, and Goniada (Polychaeta). Proc. Biol. Soc. Wash. 93(1):1-49. Perkins, T. H. 1981. Syllidae (Polychaeta), principally from Florida, with descriptions of a new genus and twenty-one new species. Proc. Biol. Soc. Wash. 93(4):1080-1172. Pettibone, M. H. 1966. Revision of the Pilargidae (Annelida: Polychaeta), including descriptions of new species, and redescription of the pelagic Podarmus ploa Chamberlain (Polynoidae). Proc. USNM 118:15-208. Pettibone, M. H. 1969. Revision of the aphroditoid polychaetes of the family Eulepethidae Chamberlin (=Eulepid1nae Darboux; = Pareulepidae Hartman). Smithsonian Contr. Zool. 41. 44 pp. Pettibone, M. H. 1982. Annelida. Iji: McGraw Hill Synopsis and Classification of Living Organisms. McGraw-Hill, N.Y. pp. 1-43 Richards, H. G. 1936. Fauna of the Pleistocene Pamlico Formation of the southern Atlantic Coastal Plain. Bull. Geol. Soc. Am. 47:1611-1656. Richards, H. G. 1962. Studies on the marine Pleistocene; parts 1, 2 Am. Phil. Soc. Trans., n.s. 52(3):1-141. 147 Rioja, E. 1946. Estudies anelldologicos. XIV. Observaciones sobre algunos poliquetos de las costas del Golfo de Mexico. An. Inst. Biol. Mex. 17:193-203. Robertson, R. 1976. Marine prosobranch gastropods: larval studies and systematics. Thalassia Jugoslavica 10:213-238. Sanders, H. L. 1977. Evolutionary ecology and the deep-sea benthos. In: C. E. Goulden (ed.), The Changing Scenes in Natural Sciences, 1776-1976. Spec. Publ. 12, Acad. Nat. Sci., Phila. pp. 223-243. Scheltema, R. S. 1977. Dispersal of marine invertebrate organisms: Paleobiogeographic and biostrati graphic implications. In; E. G. Kaufman and J. E. Hazel (eds.), Concepts and Methods of Biostratigraphy, Dowden, Hutchinson and Ross, Stroudsburg, Pa. pp. 73-108. Searles, R. B. and C. W. Schneider. 1980. Biogeographic affinities of the shallow and deep water benthic marine algae of North Carolina. Bull. Mar. Sci. 30(3):732-736. Shuto, T. 1974. Larval ecology of prosobranch gastropods and its bearing on biogeography and paleontology. Lethaia 7:239-256. Tenore, K. R. 1979. Macroinfaunal benthos of South Atlantic/Georgia Bight. Chap. 7 In; South Atlantic Benchmark Program, Vol. 3. Results of Studies of Georgia Bight of South Atlantic Ocean. Prepared by Texas Instruments Inc. under Contract No. AA550-CT7-2 with the Bureau of Land Management, U.S. Dept of Interior. Thorson, G. 1950. Reproductive and larval ecology of marine bottom invertebrates. Biol. Rev. 25(1):1-45. Topping, F. L. and J. L. Fuller. 1942. The accomodation of some marine invertebrates to reduced osmotic pressures. Biol. Bull. 82:372-384. Uebelacker, J. M. (in prep.) Syllidae, Chap. 30 hi; B. A. Vittor and P. G. Johnson (eds.), Standardization of Identifications of Benthic Polychaetous Annelids from the Gulf of Mexico Outer Continental Shelf. Dept, of Interior, Bureau of Land Management Tech. Rep. Valentine, P. C. 1971. Climatic implications of a late Pleistocene ostracode assemblage from southeastern Virginia. Geol. Surv. Prof. Pap. 683-D. 28 pp. Vermeij, G. J. 1972. Endemism and environment: some shore molluscs of the tropical A tlantic. Am. Nat. 106:89-101. Vermeij, G. 0. 1978. Biogeography and Adaptation. Harvard Univ. Press, Cambridge. 332 pp. Watling, L. 1979. Zoogeographic affin ities of northeastern North American gammaridean Amphipoda. Bull. Biol. Soc. Wash. 3:256-282. Wells, H. W. 1961. The fauna of oyster beds, with special reference to the salinity factor. Ecol. Monogr. 31:239-266. Woodward, S. P. 1851-1856. Manual of the Mollusca, or, a rudimentary tre a tise of recent and fossil shells. Weale, London. 486 pp. . m Appendix 1. Range extensions of benthic polychaetes, anphipods and molluscs collected. d • id I r= 00 ■ 'rd £ ff a * a f jf S - 8 . 5 - f 8 * 8. .S>8 8. *5 8- -ff 8. .5 03 i ? H 8 •r— .a c e A I f i -S S? P ? % ■SPBi o f rt cofo uZ« L 1 •r 1 as O rt) CO £'43 ]§ I- •r— i I 1 V) » id CO jjilu il I *os I $ $ to ■ ? ! * o> OO^f^wO - ^ ? 5 « n f M i •r- £ fi S3. (/) O I l J S 5 •5 3= i 'Jd O) id •r* 3 8 W 111 o f f I •r— fep.E CO & ? * C I fl ffrS E* id CO I . ? ' , id ^ CVJ A a o> co S f A Appendix 1. Continued H jij 8 f—i i—CT> ■r- H.§ Og) £3 § 1 ~ (d c o\c f p t e S - f I foI >3 *K-f - K * i>£3 1 1 . 5 £ “ 3? t £= M •r— •r— ■3.5 - a J"? >£>~ .r . * g f - I CM (/) & $ $ S3 $ . g CO I . i s rH s •r* . Q IS t b *r.s IS O) * O I •r" . a IS c cn A 1981 •r* s CD — r 4-a. - 8 te * 8 jS U vO * j *5 - 4 r—• ul^ u —Z J 1 P C 3 8 1^ ! Z i O s ( 5 I 3 8 I n> ? r (/) •r™ o - £ »w» s fc c <0 *r— 2 , 3 - I « o 3 i fo j3 & £ p * ? * n z u «P «-H O E □ g & g<55 *r— O c 3 18 8 a-R.g CM 8 » I * $ H r • < ? 5 CM £< : | a 5 . 8 - IS P 6 « « Appendix 1. Continued ’i I 8 | i—i CO p % s * S -Ptb 8 5? fo j3 o f ? 5 i £ ? I I p ‘fegSE E ^ — CO 51 r— *o a) s a w a l 4-> O 5 I £ •r— TJ E ' p e f ' i s l Wy ro S f 4-? h Z . 3S?fctfK3 S3 . |3 3S?fctfK3 . t 8 i 8 O L M C I I I I CM H I I I O C g (75g "5 ■o • • f> L • • • • O) 151 . m Appendix 1. Continued 152 WL. Appendix 2. Caribbean-Carolinian (Cb-C) and Carolinian (C) species collected 17 to 85 km north of Cape Hatteras (Front:25 to North sampling areas). I—II E5 CQ S •P“ o ■o I i (/> Donald Paul Weston Born in Plainfield, New Jersey, 28 August 1953. Graduated from Watchung Hills Regional High School, Warren, New Jersey, June 1971. Received B.S. in Biology from Juniata College, Huntingdon, Pennsylvania, June 1975. Entered the College of William and Mary, School of Marine Science, September 1975. Held research assistantship in Bureau of Land Management - Outer Continental Shelf Studies, September 1975 through May 1977. Held research assistantship in North Carolina Thermal Front Study, June 1977 through October 1980. Awarded M.A. in Marine Science from College of William and Mary, May 1979. Served as Research Scientist and Assistant Professor of Biology (Adjunct) at McNeese State University, Lake Charles, Louisiana, December 1980 through June 1983. 154