\-
ECOLOGICAL PATTERNS IN THE DEVELOPMENT, SETTLEMENT
AND RECRUITMENT OF ARCHAEOGASTROPODS
FROM THE OREGON COAST
by
MATTHEW CURRY KAY
A THESIS
Presented to the Department of Biology
and the Graduate School of the University of Oregon
in partial fulfillment of the requirements
for the degree of
Master of Science
March 2001
11
"Ecological Patterns in the Development, Settlement and Recruitment of
Archaeogastropods from the Oregon Coast", a thesis prepared by Matthew C. Kay in
partial fulfillment of the requirements for the Master of Science degree in the Department
of Biology. This thesis has been approved and accepted by:
Dr. Richard B. Emlet, Chair of the Examining Committee
Date
Committee in charge:
Accepted by:
Dr. Richard B. Emiet
Dr. Jan Hodder
Dr. Alan L. Shanks
Dr. Stephen S. Rumrill
Vice Provost and Dean of the Graduate School
III
An Abstract of the Thesis of
Matthew Curry Kay
In the Department of Biology
for the degree of
to be taken
Master of Sceince
March 2001
Title: ECOLOGICAL PATTERNS IN THE DEVELOPMENT, SETTLEMENT
AND RECRUITMENT OF ARCHAEOGASTROPODS FROM THE
OREGON COAST
Approved: _
Dr. Richard B. Emlet
Larvae of the limpets Lottia digitalis and LOffia asmi, as well as larvae of the flat
abalone Halioits walallensis. develop into lecithotrophic veliger larvae and settle into
benthic habitat after a breif planktonic stage. Larvae of 1. digitalis settled and
metamorphosed upon rocky substrata, as well as the barnacle P. polymerus, collected
from an adult habitat high in the rocky intertidal. In contrast, substrata from mid and low
intertidal zones failed to induce settlement and metamorphosis in larvae of L. digitalis.
These results suggest that recruitment into high intertidal habitat is driven by settlement
rather than post settlement processes. New recruits within an adult habitat were most
abundant low within the adult range and upon north-facing slopes of rocks. Larvae of H
walallensis that experienced a five-day extension of their competence period exhibited
accelerated rates of metamorphosis, as well as accelerated juvenile growth rates, relative
to larvae that were presented with settlement surfaces at initial competence.
IV
CURRICULUM VITA
NAME OF AUTHOR: Matthew Curry Kay
PLACE OF BIRTH: Santa Barbara, California
DATE OF BIRTH: January 29, 1973
GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED:
University of Oregon
Universidad San Francisco de Quito (Quito, Ecuador)
DEGREES AWARDED:
Master of Science in Biology, 2001, University of Oregon
Bachelor of Science in Biology, 1996, University of Oregon
AREAS OF SPECIAL INTEREST:
Natural History of Coastal Marine Invertebrates
Reproductive Biology and Ecology of Marine Gastropods
PROFESSIONAL EXPERIENCE:
Graduate Teaching Fellow, Oregon Institute of Marine Biology (University of
Oregon), Charleston, 1998-2000
Undergraduate Teaching Assistant, Department of Biology, University of Oregon,
Eugene: Fall 1993, Fall 1995, Winter 1995
AWARDS AND HONORS:
Neil Richmond Memorial Fellow, 2000
Conchologists of America Student Research Award, 1999
GRANTS:
Western Society of Malacologists Student Research Grant, 1999
PADI: Project AWARE Research Grant, 1999
v
VI
ACKNOWLEDGEMENTS
I am profoundly appreciative to Dr. Jan Hodder. This thesis would not exist, nor
would it have been as enjoyable to complete, if not for her good will and generosity. I
am equally appreciative to my advisor, Dr. Richard Emlet, for his expert advice and
criticisms, as well as his personal interest in my professional development. Dr. Emlet has
been an enthusiastic and supportive mentor, and his lessons will not soon be forgotten. I
was honored to have Dr. Alan Shanks and Dr. Steve Rumrill serve on my committee.
I have enjoyed the company of 1. Schaeffer and the late N. Richmond, whether
diving for flat abalone or discussing natural history. I am grateful to M. Berger, who
offered assistance during a diving collection and (with equal willingness!) statistical
analysis of data. A. PuIs made trips to the intertidal more insightful and enjoyable. B.
Butler made forays into the OIMB library both productive and humorous.
The facuIty and staff of the Oregon Institute of Marine Biology were unwavering
in their support of research objectives. Financial support essential to this project was
provided by NSF grant UFE-203221 to J. Hodder, by NSF grant OCE-9911682 to R.
Emlet, and the following awards to M. Kay: Conchologists of America Student Research
Award, Western Society of Malacologists Student Research Grant, and a grant from
PADI: Project AWARE.
Finally, I thank my parents Dr. 1.T. and G.c. Kay, whose waves of affection and
support have shaped oceans of opportunity.
\-
Vll
DEDICATION
This thesis is dedicated, however inadequately, to the memory of the late Neil
Richmond. Neil's premature death paradoxically served as a reiteration of life's most
fundamental lesson: how to live. I am reminded of Neil, and his love for life, when I read
the following passage from John Steinbeck's "The Log From the Sea of Cortez". The
following reflections were inspired as the Crew of the Western Flyer steamed south
through rolling seas along the California coastline - a scene in which, no doubt, Neil
would have glowed.
"We sat on a crate oloranges and thought what good men most biologists are. the
tenors olthe scientific world - temperamental ... loud-laughing, and healthy. Once in a
while one comes across the other kind - what used in the university to be called a "dry-
ball" - but such men are not really biologists. They are the embalmers ofthe field, the
picklers who see only the preservedform oflife without any ofits principle. The true
biologist deals with life, with teeming boisterous life, and learns something from it, learns
that thefirst rule oflife is living ... Your true biologist will sing you a song as loud and
off-key as will a blacbmith ... and meanwhile he is very good company ... "
For Neil: September 24, 1954 - June 8, 1999
VIII
TABLE OF CONTENTS
Chapter Page
1. GENERAL INTRODUCTION 1
II. SPAWNING, LARVAL DEVELOPMENT AND METAMORPHOSIS
OF THE LIMPETS Iottia digitalis AND Iottia asmi
(PATELLOGSTROPODA: LOTTIDAE) 5
Introduction " , " " " '" 5
Materials and Methods 7
Spawning and Gamete Collection 7
Larval Culture 8
Light and Scanning Electron Microscopy 9
Results 11
Spawning and Gamete Collection 11
Ioffia digitalis 11
Ioffia asmi 12
Larval Development 13
Discussion '" , , " 16
Spawning 16
Larval Development. 20
III. LABORATORY EVIDENCE THAT NATURAL RECRUITMENT IN
THE LIMPET Iottia digitalis (PATELLOGASTROPODA: LOTTIDAE)
IS DRIVEN BY SETTLEMENT RATHER THAN POST-SETTLEMENT
PROCESSES 33
Introduction 33
Materials and Methods 36
Data Analysis 42
IX
Results 43
Discussion 46
IV. ANOTHER ANGLE ON THE RECRUITMENT OF HIGH INTERTIDAL
LIMPETS - THE CASE OF Lottia digitalis REVISITED 63
Introduction 63
Materials and Methods 65
Site Description 65
Observations 66
Identification and Enumeration of Small Individuals 67
Data Analysis 68
Results 69
Discussion 72
V. SPAWNING, LARVAL DEVELOPMENT AND THE EFFECT OF AN
EXTENDED METAMORPHIC COMPETENCE PERIOD IN
LARVAE OF THE FLAT ABALONE, Haliotis walallensis STEARNS 81
Introduction 81
Materials and Methods 84
Collection and Spawning 84
Fertilization and Larval Culture 85
Extension of Competence Period Experiment. '" 86
Data Analysis 88
Results 89
Collection and Spawning 89
Early Development and Pigment Partitioning 90
Later Development. 92
Extension of Competence Period Experiment. 93
xDiscussion 95
Spawning and Larval Development. 95
Extension of Competence Period Experiment. 97
VI. CONCLUDING SUMMARY 112
BIBLIOGRAPHy 114
Xl
LIST OF FIGURES
Figure Page
1. The Total Number of Eggs Released, Relative to Shell Length,
For Indivdual Females of Lottia digitalis and Lottia asmi That Were
Induced to Spawn in the Laboratory 24
2. Light Micrographs of Developmental Stages in Lottia asmi 25
3. Light Micrographs of Developmental Stages in Lottia digitalis 26
4. Scanning electron micrographs of: (A) Lottia digitalis larval shell in lateral view,
rotated slightly to expose the shell aperature 27
5. Scanning electron micrographs of postlarval (A) Lottia digitalis
and (B) Lottia asmi 28
6. Fates of larvae of Lottia digitalis 3 days (A) and 6 days (B) after introduction
to field-collected settlement surfaces. These larvae are progeny of adults
spawned on February 12, and were introduced to settlement surfaces on
February 18 55
7. Fates of larvae of Lottia digitalis 3 days (A) and 6 days (B) after introduction
to field-collected settlement surfaces. These larvae are progeny of adults
spawned on March 2, and were introduced to settlement surfaces on
March 8 56
8. Growth of sibling Lottia digitalis postlarvae introduced to High/Adults and
Mid/Adults substrata in the laboratory (see materials and methods for
substrata descriptions) 57
9. Fates oflarvae of Lottia digitalis, from two different sets of parents, 6 days
after introduction to macroalgal settlement surfaces on March 8 (A) and
March 12 (B) 58
10. Fates of Lottia digitalis larvae introduced to isolates of the High rock
from adult habitat (see materials and methods for description of
isolates) 59
Xll
11. A postlarval LOffia digitalis that metamorphosed in the presence of, and is
actively feeding upon, the filamentous green alga (FGA) isolated
from High rock collected in adult L. digitalis adult habitat 60
12. Fates of larvae of Lottia digitalis introduced to rock collected from adult
habitat high in the intertidal (High Rock) and the stalked barnacle
Pollicipes polymerus 61
13. Composite results for censuses of LoUia digitalis from 25 different 0.25m2
quadrats within an adult habitat. 76
14. Least squares linear regressions for various features of the 25
quadrats sampled in July 2000 77
15. Least squares linear regressions for various features of the 25
quadrats sampled in October 2000 ,. '" 78
16. Least squares linear regressions for various features of the 25
quadrats sampled in January 2001 79
17. (A) Uncleaved egg (e) of Haliotis walallensis, surrounded by its egg
membrane (m) 100
18. The percentage of Haliotis walallensis larvae having shed their
vela at given sample intervals (x-axis) 101
19. The percentage of Haliotis walallensis larvae having initiated adult
shell growth at given sample intervals (x-axis) 102
20. Size of postlarval Haliotis walallensis after introduction to settlement
dishes 103
21. Size of postlarval Haliotis walallensis after initial competence '" 104
Xlll
LIST OF TABLES
Table Page
1. Laboratory spawning summary for LoUia digitalis and LoUia asmi 29
2. Summary of development in LoUia asmi at 13.00 C, and Lotfia digitalis
at 8.5 & 13.00 C 30
3. Egg sizes of free spawning northeastern Pacific lottiids 31
4. Summary of laboratory spawning events in Haliotis walallensis 105
5. Summary of development in Haliotis walallensis at 12.5-13.00 C and
8.0- 9.0oC 106
6. Repeated measures ANOVA on the effect of an extended competence
period on time to velar shedding in larvae of Haliotis walallensis
(with introduction to settlement surfaces as t = 0) 107
7. Repeated measures ANOVA on the effect of an extended competence
period on time to initiation of adult shell growth in larvae of Haliotis
walallensis (with introduction to settlement surfaces as t = 0) 108
8. Repeated measures, nested ANOVA on the effect of an extended competence
period on the size of Haliotis walallensis postlarvae 5, 10, 15 days after
introduction to settlement dishes 109
9. Repeated measures, nested ANOVA on the effect of an extended competence
period on the size of Haliotis walallensis postlarvae 5, 10, 15 and 20 days
after initial competence 110
10. Developmental periods for haliotids from a wide geographic range 111
CHAPTER I
GENERAL INTRODUCTION
Benthic marine invertebrates with planktonic larval stages impose an apparently
critical challenge on their progeny. Not only must offspring survive various perils of the
plankton, they must also return to suitable adult habitat if they are to recruit into
reproductive populations. This thesis describes experiments and observations that were
conducted in an effort to better understand how marine gastropod larvae achieve the
transition from plankton to benthos. The organisms that I employed in my studies were
the ribbed limpet, Loffia digitalis (Patellogastropoda: Lottiidae) (Rathke 1833) and the
flat abalone, Halitois vvalallensis (Vetigastropoda: Haliotidae) (Steams).
Knowledge of reproduction and larval development in a given organism is
fundamental to an understanding of its settlement and recruitment biology.
Appropriately, I provide a thorough description of the reproduction and larval
development for both of the organisms under study in this thesis. Chapter II is focused
entirely on the spawning and development of L. digitalis, as well as its congener Lottia
asmi, whereas larval development of H. walallensis is described in Chapter V. These
descriptions are of value not only in the context of this thesis, but have broader
application since there is no published, complete description of larval development for L.
digitalis, L. asmi or H. walallensis.
2Chapter III of this thesis describes settlement experiments in which larvae of L.
digitalis were exposed to various substrata from an adult habitat. The purpose of these
experiments was to identify likely natural settlement surfaces for this species. L. digitalis
is found high in the intertidal where individuals occupy one of two habitat types: rocky
substrata and the stalked barnacle Pollicipes polymerus (Morris et al. 1980). Whereas the
recruitment of L. digitalis onto P. polymerus has not been investigated, distributional
studies based on length-frequency data suggest that L. digitalis settles and recruits
directly into rocky, high-intertidal adult habitat (Frank 1965; Breen 1972). Two factors,
however, suggest that L. digitalis larvae might settle lower in the rocky intertidal, and
then undertake an upward postlarval migration. First, the youngest individuals observed
by Frank (1965) and Breen (1972) were consistently seen at lower elevations within the
adult habitat range, but were many months old by the time they were visible. Second, a
similar indirect recruitment pattern has been hypothesized for other species of limpets
that live high in the rocky intertidal (Corpuz 1981; Delany et al. 1998). Previous
descriptions of L. digitalis recruitment do not address the possibility of indirect
recruitment in this species. Information regarding settlement preferences presented in
this report, when considered in tandem with size distributions previously observed by
Frank (1965) and Breen (1972), addresses the relative importance of initial settlement
versus post-settlement processes during the recruitment of L. digitalis into rocky high
intertidal habitat. Additionally, I exposed larvae of L. digitalis to live specimens of P.
polymerus to determine if the former settles (and possibly recruits) directly onto this host.
..,
-'
Chapter IV of this thesis reports on observations of the abundance and distribution
of1. digitalis recruits (0-1 year-olds) in an high intertidal, boulder field habitat. This
habitat allowed me to observe recruit abundance relative to a) height on the shore, and b)
the directional orientation (i.e. north vs. south facing) of substrata in the habitat.
Chapters III and IV provide complementary perspectives of recruitment of1. digitalis
into high intertidal habitat.
Chapter V of this thesis, in addition to describing larval development of H.
walallensis, reports on the effect of a five-day extension of metamorphic competence (the
stage at which larvae are morphologically and metabolically able to metamorphose and
inhabit benthic habitat) in larvae of H. walallensis. The significance of this research, and
relevance to this thesis, is as follows. Since larvae of benthic marine invertebrates are
weak swimmers, and readily become entrained in currents, they have limited spatial
control over their return trip to adult habitat (Shanks 1995). However, they might exert
some temporal control by extending their competence periods, thereby increasing the
probability that they will encounter suitable settlement surfaces (Morgan 1995). The
ecological benefit of an extended competence period, however, could be negated if the
extension were to impose energetic or physiological stress on a larva (Wendt 1998). This
is especially true for non-feeding larvae with a finite supply of endogenous nutrition
(Pechenik et al. 1998). The ways in which an extended competency period can
compromise postlarval performance are manifold. The most severe repercussion is the
inability of a larva to successfully undergo metamorphosis once an appropriate surface is
encountered (Woollacott et al. 1989; Pechenik and Cerulli 1991). When larvae that have
4experienced prolonged larval periods do successfully undergo metamorphosis, they often
experience suppressed juvenile growth rates (Pechenik et al. 1993; Woollacott et al.
1989). Retarded growth, no matter what the cause, has ostensibly grave implications for
organisms that must secure resources (i.e. space) in highly competitive environments. A
more subtle physiological consequence of extended larval period, which has yet to be
thoroughly researched, is the ability of newly-settled juveniles to endure physical and
biological stresses (Pechenik et al. 1998). The extension of competence periods in
marine invertebrate larvae, although ecologically adaptive, is not always an
inconsequential event. Chapter V explores the postlarval affects of a five-day extension
of the metamorphic competence period in larvae of H walallensis.
In essence, this thesis explores the patterns in marine gastropod larval ecology
that facilitate successful recruitment into adult habitats. As evidenced by the results,
these patterns manifest on both spatial and temporal scales. The mere existence of these
patterns suggests that settlement and recruitment is not a random process in the taxa I
studied. Furthermore, as the field of larval ecology develops and grows, the insights of
its practitioners must be integrated into knowledge of adult biology and ecology in order
to achieve a more complete understanding of the natural history of marine invertebrates.
\-
5
CHAPTER II
SPAWNING, LARVAL DEVELOPMENT AND METAMORPHOSIS
OF THE LIMPETS LoUia digitalis AND LOUia asmi
(PATELLOGASTROPODA: LOTTIDAE)
Introduction
The patellogastropod limpets LaUia digitalis (Rathke, 1833) and LoUia asmi
(Middendorff, 1847), are members of the Lottiidae Gray 1840 (superfamily
Acmaeoidea), which is one of three families formed from the recently fragmented
Acmaeidae (then superfamily Patellacea) (Lindberg, 1986). The 25 lottiid species that
inhabit the northeastern Pacific (Lindberg, 1981) serve as an impressive example of
ecological radiation, as some species have formed strict associations with particular algal
species (Test, 1945; Yonge, 1962; Proctor, 1968), gastropods (Test, 1945; Eickenberry &
Wickizer, 1964), and vertical habitat zones (Rickets & Calvin, 1968; Underwood, 1979;
Morris et aI., 1980).
LaUia digitalis and 1. asmi occupy very different microhabitats in the intertidal
zone. 1. digitalis is common in high intertidal rocky habitats from the Aleutian Islands
(Alaska) to southern Mexico (Morris et aI., 1980). In contrast, L. asmi is restricted
almost exclusively to the shells of its gastropod hosts, Tegulafunebralis and Tegula
gallina, throughout its range from British Columbia to Isla Socorro, Mexico (Morris et
al., 1980). The adult biology and ecology of L. digitalis have been studied extensively
(Castenholz, 1961; Frank, 1965; Hardin, 1968; Breen 1971; Wootton, 1992, 1993;
Hobday 1995), and the same is true for L. asmi (Test, 1945; Eickenberry & Wickizer,
1964; Alleman, 1968; Evans, 1992; Lindberg, 1990; Lindberg & Pearse, 1990).
Despite the scrutiny placed upon the adult biology of 1. digitalis and 1. asmi,
larval development for L. digitalis has been described only to the veliger stage (Koppen
et al., 1996; Holyoak et al., 1999), and there is no published description of development
for L. asmi. Aside from work with L. digitalis, and Lindberg's (1983) description of
reproductive anatomy in the brooders Erginus spp., detailed developmental study of the
25 northeastern Pacific lottiids has been restricted to Tectura scutum (Karp, 1970) and
Acmaea incessa (Proctor, 1968). Developmental studies are important because
developmental mode influences the dispersal ability, genetic structure, ecological
resilience, and evolutionary persistence of marine invertebrate taxa (Strathmann, 1985;
Pechenik, 1999). Accounts of species development also contribute to our understanding
of life history evolution (Grahame & Branch, 1985; Havenhand, 1993; Goddard, 1996)
and phylogenetics (Emlet, 1988; Hickman, 1992; Ponder & Lindberg, 1996).
This study describes and compares the laboratory spawning, larval development,
and metamorphosis of L. digitalis and L. asmi. My observations reveal that these
ecologically unique species share nearly identical patterns of early development. Both
species freely spawn their gametes, which develop into lecithotrophic veliger larvae that
reach competence at 5.25-5.5 days after fertilization (13oC). L. asmi shed smaller eggs
than L .digitalis. Small egg size may be an adaptation to increase fecundity in the small-
6
\-
7
bodied L. asmi, but this conclusion is hampered by a lottiid developmental literature that:
a) represents relatively few species, and b) suggests that intraspecific variation in egg size
may not be uncommon in the family. Because free spawning lottiids do not need to
secure egg masses in safe habitats for development, this reproductive mode may have
been conducive to their ecological radiation into novel benthic habitats.
Materials and Methods
Spawning and Gamete Collection
Adult Lattia digitalis were collected from South Cove of Cape Arago, OR. (430
18' N; 124° 24' W), during winter and spring of 2000 (Table 1). During the same time
period, adult Lattia asmi (on their hosts, Tegulafunebralis) were collected from the Cape
Arago site, and also from Sunset Bay State Park, OR. (43° 20' N; 1240 22' W). The
limpets were held in a l30C incubator at the Oregon Institute of Marine Biology (OIMB),
Charleston, OR., where they were spawned within one day of collection.
To induce spawning, groups of 6-8 adult L. digitalis and L. asmi (removed from
their hosts) were separately placed in seawater in 2L glass finger bowls (l8cm diameter,
7cm tall) and exposed to a weak solution of hydrogen peroxide (HzOz), thermal shocking,
and/or vigorous bubbling. Hydrogen peroxide treatments were prepared according to
Hahn (1989). Specifically, each liter of seawater was buffered with 6.6ml of 2M
Tris(hydroxymethyl)aminomethane (Tris-base), allowed to equilibrate for 15 minutes,
then spiked with 3ml of a freshly prepared 6% HzOz solution. Adult limpets were soaked
in this solution for two hours. In the thermal shocking treatments, limpets were held in
Swater while the temperature was increased from 130 C to 16-1SoC over a two-hour
period. Seawater in the vigorously bubbled treatments was agitated for two hours using
standard aquarium airstones. All three spawning treatment stimuli were terminated by
the addition of0.45llm filtered seawater (FSW) at nOc, which thoroughly rinsed and
refilled each vessel. Each vessel of treated adults was then placed in an incubator at
nOc with a photoperiod representative of field conditions, and monitored continuously
for gamete release.
Immediately upon release, eggs were reverse-siphoned through a 300llm nitex
mesh sieve and pipetted into 500ml beakers of isothermal 0.451lm FSW. After the eggs
had settled to the bottom of these beakers (~1O minutes), the vessels were decanted and
refilled twice for a total of three rinsings. Sperm were collected from spawning males
when possible, but in some cases they were stripped with a fine bore hypodermic syringe
and activated in seawater. A stock solution of sperm in seawater, typically ~ 5 xl06
sperm per ml FSW, was prepared using a hemocytometer to calculate cell density.
Larval Culture
Larval culture took place in 0.451lm FSW at nOc under a photoperiod
representative of field conditions. Rinsed, unfertilized eggs were settled into uncrowded
monolayers (~ 500 eggs/cm2) upon the bottom of lL glass beakers. Sperm were then
added from the stock solution to yield a final concentration of 200,000 sperm per ml of
0.451lm FSW. This concentration was maintained for five minutes, after which time the
cultures were decanted and refilled with fresh 0.451lm FSW. Control batches of eggs
\-
9
(which were not exposed to sperm solution) were maintained along with the fertilization
vessels to ensure that fertilization occurred at a known time. Cultures were periodically
agitated during early development.
Soon after trochophores began to swim, the upper half of each culture was
decanted into a 2.5L culture vessel and diluted to a density of 1 larva per 2-4 ml FSW. In
this manner, waste and debris accumulated in the fertilization vessels were eliminated
and healthy, actively swimming trochophores were isolated for final culture. From this
point in time until completion of the protoconch, cultures were unrinsed and periodically
agitated.
When larvae first displayed the ability to withdraw soft tissues into their shells
(i.e. protoconch and operculum fully formed), they were rinsed in a partially submerged
100 ~lln nitex sieve. From this point in time until settlement, larvae were rinsed and
transferred to fresh 0.45/lm FSW once daily, and the cultures were agitated with paddles
that swung at 12 beats min. -1 (described in Strathmann, 1971). To observe
metamorphosis, larvae of both species were settled onto fragments of rock that were
collected from the adult habitat of each species. To measure the effect of temperature on
development of L. digitalis, the above protocol was followed at 8.50 C (in addition to
l3oC) for adults collected on February 12,2000.
Light and Scanning Electron Microscopy
Larval development was monitored with a compound microscope, and major
stages were recorded with photomicrographs. Ciliated blastulae, trochophores, and early
10
stage veligers were fixed for photography in a buffered solution of 4% formaldahyde (1 %
sodium borate). Late-stage veligers were exposed to low levels of MgCh (2-4 drops of
7.5% MgCh per ~20ml of FSW) for 2-12 hours in order to prevent velar retraction.
Developmental features were measured to the nearest 4.4 J-lm with an ocular micrometer.
Measurements of larval shells were taken: a) across the shell opening, perpendicular to
the velum's bilateral symmetry, and b) from the tip of the shell opening to posterior end
of the protoconch. These two measures were termed length and width, respectively, and
are diagrammed by Hadfield & Strathmann (1990).
The veliger shells were also observed with scanning electron microscopy (SEM).
To clear tissues from larval and postlarval shells, anaesthetized individuals were soaked
for six hours in a 3: 10 solution of bleach and distilled water. To dissolve residual lipids,
shells were held in distilled water and gradually introduced to 100% methanol, exposed
to a 1: 1 mixture of methanol and chloroform (1 hour), then rehydrated through the
methanol series to distilled water (Pederson & Page, 2000). Once the shells were
cleaned, they were placed in absolute acetone and stored until examination. Specimens
were subsequently air-dried, stub mounted, platinum coated, and examined with an AMR
1000A scanning electron microscope. All larval shell SEM work employed shells from
competent larvae.
I I
Results
Spawning and Gamete Collection
Lottia digitalis
Of the animals collected on February 12, 2 of 30 spawned after immersion in
H20 2, whereas 7 of 30 adults spawned after exposure to vigorous bubbling (Table 1).
Similarly, of the animals collected on March 2,3 of24 spawned after immersion in H20 2,
whereas 8 of 24 adults spawned after exposure to vigorous bubbling. In light of these
results, vigorous bubbling was employed throughout the season to induce gamete release
in L. digitalis. (In some instances, thermal shocking and/or H20 2 were also applied, but
never in the absence of bubbling.)
Physical contact between/among spawning males and females was not typical.
Eggs and sperm were extruded from beneath the right anterior portion of the adult shells.
Eggs were released singly, while sperm were extruded as a viscous milk-like fluid. Male
and female gametes were negatively buoyant. The eggs were "rusty" brown in color
upon release, but turned a pale green color after ~5 minutes.
The spherical eggs released by 5 different females had mean diameters of 150.9
~lm (SD = 0.69, n = 25), 151.0 ~m (SD = 1.5, n= 25), 151.1 ~m (SD = 1.5, n = 25), 152.2
~m (SD = 4.1, n = 25), and 167.4 ~m (SD = 3.1, n = 25). Each egg was surrounded by a
transparent jelly coat, (262.3 ~m in diameter, SD = 4.6, n = 25 eggs from a single
female), which was visible when eggs were suspended in "Sumi" ink particles
12
(Strathmann, 1987). The smallest female to release eggs was 16mrn long, and the largest
number of eggs released was 42,000 (Fig. 1). As evident in Figure 1, there was a trend of
increased fecundity with increased adult size in L. digitalis. The lanceolate sperm from
one male had mean head and tail lengths of 8.0 /-1m (SO = 0.8) and 70.0 /-1m (SD = 3.7),
respectively (n = 16).
LoWa asmi
Ten of twenty adults that were collected on April 19 spontaneously released
gametes on April 20 (Table 1). Of the animals collected on April 25, 8 of 20 spawned
after immersion in H20 2, while 3 of 20 spawned after exposure to rigorous bubbling.
Efforts to induce spawning after this date were met with limited success. Eggs and sperm
were extruded in the same manner as in L. digitalis.
The spherical eggs released by 5 different females had mean diameters of 135.9
/-1m (SO = 2.8, n = 25), 130.5 /-1m (SO = 4.4, n = 25), 135.0 /-1m (SD = 3.1, n = 25), 134.6
/-1m (SO = 2.8, n = 25), and133.2 /-1m (SO = 3.4, n = 25). Eggs were rusty brown in color
upon release, but turned pale green within 5 minutes. The smallest female to release eggs
measured 7 mm, and the largest number of eggs released was 13,000. The lanceolate
sperm from one male had a mean head and tail length of 7.0 /-1m (SD = 1.6) and 63.5 /-1m
(SD = 2.4), respectively.
\-
13
Larval Development
The sequence and timing of developmental stages were nearly identical in L. asmi
and L. digitalis. Due to this similarity, some early stage photographs for L. asmi and L.
digitalis are presented in alternating order in Figures 2 and 3. For example, Figure 2B
illustrates the two-cell stage for L. asmi, whereas Figure 3B illustrates an 8-cell embryo
of L. digitalis. The timing of developmental events at I3.00 C was synchronous within
each species, but differed slightly between the two (Table 2). L. digitalis larvae reared at
8.SoC took nearly twice as long to reach metamorphic competence as larvae reared at
13°C (see Table 2 for the specific timing of developmental events).
In both species, the first reliable evidence of fertilization was the formation of a
polar body (Figs. 2A, 3A). In contrast, separation of the egg membrane from the egg was
imperceptible or extremely slight. The first two cleavages were complete, equal,
meridional and always initiated near the polar body (Fig. 2B). The third cleavage was
unequal, equatorial, and dexiotropic, such that the resulting 8-celled embryos were
comprised of four macromeres and four micromeres surmounted by the polar body (Fig.
3B). The fourth and fifth divisions were synchronous, but the 6th division was
asynchronous.
At approximately 9 hrs (13oC) post fertilization, the gastrulae bore three tiers of
cilia which beat feebly and rotated the embryos at rates no faster than I revolution 20
seconds· I (Fig. 2C). At this stage, the polar body persisted atop the animal pole. At
approximately 10 hours, however, the polar body was resorbed and apical cilia appeared
14
in its place. Unfortunately, I was unable to determine the process by which gastrulation
occurred.
Trochophores were well formed and swam actively 14-16 hours after fertilization.
Soon after each species began active swimming, the telotroch (anal tuft) and refractive
bodies were formed (Fig. 2D). In early trochophores of both species, steady swimming
was often punctuated with abrupt bursts of speed. During these short bursts, which lasted
less than 0.25 sec., larvae swam distances of> I0 body lengths. This sprint behavior was
consistently induced when larvae were touched with an eyelash probe, and they often
responded in this manner when a probe passed them in close proximity.
At 16 hours, the stomodaeum was visible immediately beneath the prototrochal
girdle (Fig. 3C). The prototrochal girdles of each species were comprised of 21
individual cells. Girdle cells in L. digitalis and L. asmi were approximately 35-40 and
30-35 /lm long, respectively. L. digitalis trochophores were 177 /lm long (SD = 8.5, n=
30) and 164 /lm wide at the prototrochal girdle (SD = 7.6, n = same 30), while L. asmi
trochophores were 155 /lm long (SD = 4.4, n= 30) with a 145 /lm girdle width (SD = 8.4,
n = same 30).
During protoconch formation, the telotroch migrated anteriorly along the ventral
surface of the larva, until it was located immediately posterior to the foot rudiment (Figs.
2E, 3E). Torsion was not a punctuated event in either species, but rather, it took place
over a 2-4 hour period. Soon after torsion was completed, the operculum, the first
epipodial tentacle, and both eyespots were formed. At this stage, larval behavior changed
markedly. Rather than resorting to flight as a defensive strategy (as in trochophores),
15
veligers withdrew into their shells at the slightest provocation. Both species swam
upwards, as trochophores and veligers, but the roles of photo- and geotaxis were not
determined.
The veliger larvae of L. digitalis and L. asmi (reared at 13.00 C) reached
metamorphic competence at 5.5 and 5.25 days, respectively. Larvae were deemed
competent when they crawled on/adhered to glassware, and when they bore 3+ tubercles
on their cephalic tentacles (as described in Seki & Kan-no, 1981). Upon settling, larvae
reared forward on their shells and swept laterally across the substratum with their
cephalic region. Velar loss, the first evidence of metamorphosis, and the presence of
food (plant matter) in the gut were detectable 1 day after settlement in both species. It
was not determined whether velar cells were shed, resorbed, or ingested by the newly
formed juvenile.
Development of the radular complex was rapid in postlarvae of both species, and
was particularly visible in L. digitalis (Figs. 2H and 3H). By 2 days after settlement, the
ribbon-like radulae of postlarvae were clearly visible. By 10 days after settlement, L.
digitalis postlarvae developed well-vascularized radular muscles that were red in color
when oxygenated, but quickly turned blue-purple when postlarvae were placed between a
microscope slide and cover slip for observation (and ostensibly deprived of oxygen). In
contrast, L. asmi postlarvae did not develop obvious vascularisation of their radular
muscles until 20-25 days after settlement, and even then the muscles were not as
intensely red as in L. digitalis. Postlarvae of both species actively rasped the substratum
with their radulae, but they were sedentary and typically remained in small (9-16mm2)
16
areas for consecutive days. In many instances, larvae settled in clusters of 4-10
individuals. It is unclear, however, whether this was a display of gregarious settlement or
a pattern caused by patchiness of substratum quality. At 13oC, L. digitalis grew to a size
of 1mm in 35-40 days, whereas L. asmi took 45-50 days to reach that size (Table 2).
Larval shells of1. asmi averaged 206.2 Ilm wide and 144.5 Ilm long (SD's = 2.5
and 1.6, respectively; n = 10 individuals from one female with average egg diameter =
133.2 Ilm). See Hatfield and Strathmann (1990) for a description of dimensions. Larval
shells of L. digitalis averaged 224.2 Ilm wide and 154.9 Ilm long (SD's = 2.33 and 3.2,
respectively; n = 10 individuals from a female with average egg diameter = 151.0 Ilfn).
There was no difference between L. asmi and L. digitalis with regard to protoconch
sculpturing (Fig. 4). Both species initiated adult shell growth 2-4 days after settlement.
The adult shell was initially secreted symmetrically, as illustrated in Figs. 4 C&F. In
slightly older postlarvae, however, the protoconch appears as to be displaced clockwise
relative to the adult shell's long axis (Fig. 5).
Discussion
Spawning
Broadcast spawning is a common reproductive mode among littoral
archaeogastropods of the northeastern Pacific (Karp, 1970; Strathmann, 1987; Holyoak,
1988; Hickman, 1992; Moran, 1997). One potential benefit of broadcast
17
spawning is that, unlike eggs tethered in benthic egg masses, free-spawned gametes are
delivered from potential stresses encountered in the adult habitat (Underwood, 1979;
reviewed by Pechenik, 1999). This advantage has been linked to the prevalence of
broadcast spawning among high-intertidallittorinids (Mileikovsky, 1975), although Reid
(1990) has proposed an alternative phylogenetic hypothesis for this trend. Among
northeastern Pacific lottiids, almost all of which free spawn their gametes, it is unclear if
broadcast spawning has been selected for ecologically or is constrained phylogenetically.
The occurrence of nonpelagic development in the northeastern Pacific lottiids Erginus
apicina, Erginus maskalevi, and Erginus sybaritica (Lindberg, 1983), however, suggests
that developmental mode is not under strict phylogenetic constraint in the Lottiidae.
Since free-spawning adults are not obligated to locate safe microhabitats for benthic egg
development, this reproductive mode may have been permissive in the ecological
radiation of the Lottiidae.
An additional hurdle to egg mass formation might also exist for Lattia asmi. It is
estimated that adult L. asmi change hosts once every 13 to 24 hours (Test, 1945;
Eickenberry and Wickizer, 1964), and that in a given night 40% of animals transfer hosts
(Lindberg, 1990). Given these movement regimes, deposition of an egg mass upon a T.
funebralis host shell would be tantamount to abandonment. Since T. funebralis are
highly mobile, and egg masses seldom confer complete protection from abiotic
environmental stresses (Pechenik 1999), a female L. asmi who has abandoned her egg
mass would have little control over the microclimate in which it develops. If a female
were to brood her egg mass on a T. funebralis shell, and thereby buffer the microclimate,
18
she would be forced to fast since adult L. asmi consume up to 7.0% of the algae on a
typical host every hour (Eikenberry & Wickizer, 1964).
Since at least the 1960's, it has been observed that certain patellaogastropods
spawn coincidentally with heavy seas (Fretter and Graham, 1962). More recently,
Shanks (1998) observed that field spawning in Laffia digitalis is apparently associated
with winter storm events. The effectiveness of vigorous bubbling as a means of inducing
gamete release, during the course of this study, may be related to this life history trait in
L. digitalis. Indeed, winter spawning (when storms are most frequent) is common among
eastern Pacific patellogastropods (Fritchman 1961 a, b, c, 1962; Phillips 1981; Koppen et
aI., 1996). In light of this seasonal timing, the high sperm viscosity of L. asmi and L.
digitalis may be an adaptation to resist dilution in vigorously agitated water (e.g. Thomas,
1994).
Koppen et al. (1996) and Holyoak et al. (1999) report mean egg sizes of 197.5 and
146.3 ).lm, respectively, for L. digitalis. The 150.9-167.4).lm egg sizes that I measured
approach the observations of Holyoak et aI., but are much smaller than those of Koppen
et al. While it is challenging to account for such a wide variation in intraspecific egg
size, and Holyoak et al. suggest that an erroneous micrometer calibration may be
culpable, such patterns are not unprecedented in the lottiid literature. For example, Karp
(1970) and Nicotri (1974) report egg sizes of 150 and 190 ).lm, respectively, in Tectura
scutum. In three individual Laffia pelta females, collected simultaneously from the San
Juan Islands, egg sizes ranged from 128-144 ).lm (Hadfield & Strathmann, 1996). Due to
-------~--
19
the relatively sparse literature on lottiid development, it is uncertain whether intraspecific
egg size variation is the rule or the exception in this family.
Egg sizes in L. asmi are smaller than egg sizes described for most other eastern
Pacific lottiids (Table 3). The magnitude of the differences is best appreciated when one
considers the volumetric differences calculated from the measured egg diameters. For
example, egg diameters of 130 ~m for L. asmi and 150 ~m for L. digitalis indicate that
eggs of the latter are more than 1.5 times the volume of the former.
Reduced egg size, relative to other lottiids, might be a distinct result of the
ecological constraints imposed by habitat choice and reproductive mode in L. asmi. For
marine invertebrates with planktonic development, reproductive success is dependant
upon high fecundity (Grahame & Branch, 1985; Emlet et aI., 1987; Rumrill, 1990).
Fecundity in small-bodied invertebrates, however, is acutely limited by the adults' ability
to allocate energy and space for gonad development (Menge, 1975; Underwood, 1979;
Gallardo & Perron, 1982; Strathmann & Strathmann, 1982) Body size in L. asmi is
constrained by the shape and size of its host, Tegula funebralis (Lindberg & Pearse,
1990), but the association provides L. asmi with an escape from high interspecific
competition among intertidal limpets (Murphy, 1976). Thus, the ecological benefits of
this commensalism impose limitations on adult size and, by extension, gonadal bulk in L.
asmi. Reduced egg size, however, permits higher fecundity in a gonad of fixed size
(Strathmann, 1986). Therefore, reduced egg size in L. asmi might be an adaptation to
increase fecundity in this small-bodied broadcast spawning limpet.
20
Unfortunately, the strength of this argument is handicapped by the paucity of
published egg sizes and fecundities for eastern Pacific lottiids. Furthermore, intraspecific
variations in egg size (as described above for L. digitalis, L. pelta and T. scutum)
confound efforts to determine the adaptive significance of this life history trait within the
Lottiidae. Finally, recruitment successes attributable to higher fecundity must
compensate for any liabilities associated with production of smaller offspring (e.g. Emlet
& Hoegh-Guldenberg, 1997).
Larval Development
Early cleavage patterns in L. digitalis and L. asmi were typical of development in
lottiids and other patellogastropods (Smith, 1935; Kessel, 1964; Karp, 1969; Van Den
Biggelaar & Haszprunar, 1996). The absence of a 24-cell resting stage during 5th
cleavage (synchronous 5th cleavage) was observed in both species and is presumed to be
ancestral among the gastropods (Van Den Biggelaar, 1993). This feature has been used
to phylogenetically distinguish the patellogastropoda from members of the closely related
vetigastropoda (Van Den Biggelaar & Haszprunar, 1996). No cytoplasmic markers or
pigments, as have been observed in some gastropods (Collin 1997; Moran, 1997; Kay,
pers. obsv.), were present or segregated during early cleavage.
The timing of development in L. digitalis and L. asmi was similar to other lottiids
and patellogastropods (Smith, 1935; Proctor, 1968; Strathmann, 1987; Hadfield &
Strathmann, 1996). The fact that L. asmi embryos developed slightly faster than those of
L. digitalis, while egg size in the former was smaller than in the latter, is consistent with
21
the "egg-to-juvenile" period model posited by Havenhand (1993). The effect that
temperature had on developmental time in L. digitalis is consistant with observations of
other gastropod larvae (Leighton, 1972; Hahn, 1989; Goddard, 1996)
Neither Koppen et al. (1996), nor Holyoak et al. (1999), note polar body
formation in L. digitalis. Similarly, polar body formation is not described in P. vulgata
(Smith, 1935; Dodd, 1957) or A. testudinalis (Kessel, 1964). Although the evolutionary
significance of this structure might be negligible, the omission of such a conspicuous
developmental feature is vexing. At the very least, polar body formation was practically
useful, in the absence of egg membrane lifting, as early evidence of fertilization. I was
unable to see lifting of the egg membrane in L. digitalis, as described by Koppen et al.
Presence of the telotroch (anal tuft) is inconsistently reported in the
patellogastropod literature (but see Smith, 1935; Proctor, 1968), yet this structure might
play an important sensory role in trochophores. The "sprint" behavior described above
(also noted by Kessel, 1964) must be facilitated by the ability of trochophores to receive
physical and/or optical stimuli. Smith (1935) proposed such a role for refractive bodies
in P. vulgata, due to their ciliation and optically refractive properties. The anal tuft and
refractive bodies bear several features in common: a) they appear on the extreme (but
opposite) tips of trochophores, b) they are protrusions that bear non-motile cilia, and c)
they disappear during formation of the veliger stage. Structural similarities of refractive
bodies and the telotroch suggest a common function, and it is possible that both are
sensory bodies in patellogastropod trochophores. This idea is bolstered by the fact that
haliotid trochophores, from the closely-related pleurotomarioidea, possess neither
22
refractive bodies nor telotrochs and do not display the same sprint behavior (Crofts, 1937;
Kay, pers. obsv.). In contrast, the reflexive tendency ofveligers to retract into their
protoconchs is a behavior typical of this larval form (Garstang, cited in Hardy, 1958).
Although no feeding experiments were conducted with veligers of 1. digitalis and
1. asmi, they appear lecithotrophic. This conclusion is based upon the lack of a post-oral
ciliary band and a well-defined mouth on the velum, both of which are present and
conspicuous in planktotrophic gastropods (Hickman, 1992; Pederson & Page, 2000).
Further evidence of their lecithotrophy is derived from the velar aspect ratio (VAR)
scheme of Moran (1999). Both species had VAR's (VAR=width of velum divided by
total length of larva) conservatively estimated between 0.66 and 0.85, which are well
within the range of VAR values for lecithotrophic taxa surveyed by Moran. Additionally,
the larvae of L. digitalis and L. asmi raised during this study were not provided food, but
experienced negligible mortality during latter stages of culture.
The series of lines on larval shells from competent 1. digitalis and L. asmi does
not extend over the entire surface of the protoconch. This sculpture pattern contrasts with
western Pacific lottiids described by Amio (1963), whose larval shells are completely
sculptured from the posterior to the aperture. It is possible that these sutures were present
in early stage veligers, but they were lost during subsequent shell building and/or
maintenance processes. Most likely, these sculpture lines are a product of the larval shell
growth pattern, but lack the defensive value of protoconch features present in gastropod
larvae with longer planktonic developments (Hickman, 1999). These patterns might be
23
of use to students of zooplankton, since protoconch morphology is reliably employed in
family-level identification of eastern Pacific gastropod veligers (Goddard, in press).
In addition to any diagnostic utility, the phylogenetic value of protoconch
morphology has been considered for some haliotids (Hayashi, 1983), trochaceans
(Hickman, 1992), and opisthobranchs (Thompson, 1961). Although prosobranch
phylogeny at the ordinal and familiar level is based primarily upon adult features
(Lindberg, 1988; Bieler, 1992), protoconch morphology has potential for resolution of
lower-order relationships (Hickman, 1992). Such endeavor is dependant upon a vast
library of published images for any given taxon, but unfortunately no such images exist
(to my knowledge) for the patellogastropoda.
In postlarval L. asmi and L. digitalis, the protoconch appears to be displaced
clockwise relative to the adult shell's long axis (Fig. 5). This offset has been described in
other patellogastropods (Morse, 1910; Thompson, 1912; Smith 1935) and in a fissurellid
keyhole limpet (Pemet, 1997). Because this alignment is not established during initial
shell growth (Figs. 4C&F), and because the protoconch is soon lost to abrasion (pers.
obsv; also noted by Morse, 1910; Pernet, 1997), any functional value of this alignment
should be dismissed. The most logical interpretation of this phenomenon, in the absence
of function, is as a developmental vestige of coiled ancestry. Indeed, Lindberg (1988)
has cited this feature in his assertion that the patellogastropod ancestor bore a coiled
teloconch. Alternatively, Haszprunar (1988) argues compellingly that the ancestor
resembled extant limpets and had a bilaterally symmetrical adult shell.
..-
50 L. digitalisC'0 •0 0 L. asmi ••...-- 40 •x
•--
"'0 30Q)
en •• •co 20 • •Q) •Q)
"- 00en 10
0) 0 0 •0) 68 •Q) 0 •'+-
0
~ -_._--~. -------,----- .~~--_.- ~,--'---~--'-------------'------~~-'---r---~
8 10 12 14 16 18 20 22 24 26
Size of female (mm)
FIGURE 1. The total number of eggs released as a function of shell
length for individual females induced to spawn in the laboratory.
24
G
e
c
---pb
25
FIGURE 2. Light micrographs of developmental stages in [offia asmi, reared at
13oC. (A) fertilized ovum displaying polar body (pb) (0.5 hr.). (B) first cleavage (1 hr.).
(C) Gastrula surmounted by persistent polar body (9hrs.). (D) Completely formed
trochophore (20 hrs.). ac, apical acilia; pc. prototrochal cilia; pg, prototrochal girdle; rb,
refractive bodies; tel, telotroch. (E) Early pretorsional veliger (35 hrs.) with incomplete
larval shell (Is) and telotroch still visible beneath the foot rudiment (fr). (F) Post-torsional
veliger (62 hrs.) with larval retractor muscles (Irm) and operculum (0) completely
formed. Larvae withdraw into protoconch at this stage. mp, metapodium; pr, propodium;
pre, propodial cilia; v. velum; vc, velar cilia. (G) Competent veliger (5 days). Velar cilia
are present. but not visible in this photograph. c1. cephalic tentacle; e. eye; L foot. (H)
Postlarval individual. arrow marks the break between larval and adult shells. a, anus; as,
adult shell: fg. foregut; r. radula. All scale bars are 100 Ilm.
A B
fet
mic
~-mac
26
FIGURE 3. Light micrographs of developmental stages in LOffia digitalis, reared
at 13°C. (A) Fertilized ovum displaying polar body (pb) (0.5 hr.). m, egg membrane. (B)
Eight cell stage (2.5 hrs.). mic, micromere; mac, macromere. (C) Newly formed
trochophore with stomodaeum (sto) below prototrochal girdle (pg). tel, telotroch. (24.5
hrs. at 8.50 C). (D) Late stage trochophore (35 hrs.) with larval shell formation
underway. ac, apical cilia; 1'1', foot rudiment pg. prototrochal girdle; tel, telotroch. (E)
Pretorsional veliger vV'ith telotroch visible posterior to the foot rudiment Cfr) (45 hrs.). Is,
larval shell. (F) Post-torsional veliger (62 hI's.) with larval retractor muscles (lrm) and
operculum (0) completely formed. Larvae withdraw into protoconch at this stage. mp,
metapodium; pr, propodium: v, velum: vc. \'CIaI' cilia. (G) Competent veliger (5.5 days).
e, eye; ct, cephalic tentacle: 1". foot; fet, tirst epipodial tentacle: vm, visceral mass. (H)
Postlarval individual (15 days) with \vell clcwloped and vascularized radular muscles
(rm). fg, foregut: hg, hindgut: r. radula: s. stomach. Arro\v marks the break between
larval and adult shells. All scale bars arc 100 ~lm.
27
FlCiURE.:+. Scanning electron micrographs of: (A) LOllia digifalis larval shell in
lateral view. rotated slightly to expose the shell aperature. (B) Bottom view of L. digitalis
lanai shell. (C) L digitalis postlarva. atTO\\ marks break between adult and larval shells.
(0) Lollia asm! larval shell in lateral \·iew. rotated slightly to expose the shell aperature.
(E) Bottom vie\\ of L asm! larval shdl. (F) L (/.\lI1i postlana. alTO\\ marks break between
adult ancllarval shells. Scale bars are 50 ~lm.
28
FIGURE 5. Scanning electron micrographs of postlarval (A) Latria digitalis and
(B) Latria asmi. Note the slight clockwise offset of larval shell, relative to the long axis
of the adult shell. Scale bars are 200 flm.
TABLE 1. Laboratory spawning summary for
Lotria digitalis and L asmi.
Date #spawned/ Lunar
(2000) Stimulus # attempted Phase
Lot/ia digitalis:
2/12 HP 2/30 3/4
B 7/30
3/2 HP 3/24 New
B 8/24
3/6 B+TS 10/20 New
3/21 B+TS 4/45 Full
3/27 B+HP+TS 0/50 1/4
4/6 B+HP+TS 3/100 New
4/14 B+TS 11/50 3/4-Full
4/25 B+TS 7/50 1/4
LOffia asmi:
4/14 B+TS 2/20 3/4-Full
4/20 VOL 10/20 Full
4/25 B+HP+TS 11/40 1/4
5/23 HP 0/19 1/4
5/28 HP+TS 1/20 1/4
6/6 B+HP+TS 2/50 3/4
6/16 HP 3/50 Full
B 0/50 Full
Laboratory spawning of animals collected from
the field. Spawning stimuli were hydrogen
peroxide (HP), vigorous bubbling (B), and
thermal shocking (TS) (see text for details).
VOL = volunteer spawning event.
29
30
TAS LE 2. Summary of development in Lotiia asm i at 13.0° C, and L. digitalis at 8.5 & 13.0° C
Developmental event or stage
Time since fertilization
L. asmi (13°) L. digitalis (13°) L. digitalis (8.5°)
Polar body formation
Ist cleavage
2nd cleavage
3'd cleavage
4th cleavage
Ciliated gastrulae
Apical cilia appear, polar body lost
Swimming trochophores
Telotroch (anal tuft) formed and ciliated
Protoconch initiated / protoconch completed
Larval retractor muscles
Torsion started / torsion completed
Operculum fully formed
First epipodial tentacle
Eyespots, ciliation of metapodium
Propodium formation / cilia on propodium
Cephalic tentacles begin to elongate
Cephalic tentacles bearing 3+ tubercles
Larvae crawl (behavioral competence)
Vela lost
Food visible in gut
Adult shell growth
Animals Imm long
0.25-0.5 hr
I hr
1.5 hr
2 hr
2.5 hr
9hr
10.5 hr
15 hr
28 hr/40-43 hr
36 hr P
41 hr/43-48 hr
52 hr
57-59 hr
61-62 hr
64-70 hr/70 hr
3 d
4.25-4.5 d
5.25 d
6.5 d
6.5 d
7-8 d
-40-50 d
0.25-0.5 hr
1.25 hr
1.75 hr
2.5 hr
3 hr
9.5 hr
II hr
14-16 hr
16.5 hr
29 hr/44-48 hr
48 hr/50-52 hr
57 hr
58-60 hr
63.5-64 hr
64-70 hr/70 hr
3.5 d
4.5 d
5.5 d
6.5 d
6.5 d
7-8 d
-35-40 d
0.25-0.5 hr
1.5hr
2.5-3 hr
16 hr
17-20 hr
24.5 hr
38 hr/88 hr
64 hr P
88 hr/ 94 hr
4.6 d
5.6 d
5.8 d
5.9 d
8.6 d
9.1 d
10.5 d
II d
11.5-12d
Times for 130 C observations were averaged from two rearing events for both L. asmi and L. digitalis.
P = visible under polarized light, but not white light.
TABLE 3. Egg sizes of free spawning northeastern Pacific Lottiids.
31
Species Egg size
(/lm)
Source
Lottia asmi
Lottia digitalis
Lottia giganteat
Lottia pelta
Notoacmea insesssa
Tectura persona
Tectura scutum
Acmaea testudinalis*
133
147
156
197
130
128-144
185
150
200
150
185
140
(This study)
(Holyoak et aI., 1999)
(This study)
(Koppen et aI., 1996)
(Daly, 1975)
(Hadfield and Strathmann, 1996)
(Nicotri, 1974)
(Proctor, 1968)
(Strathmann, 1987)
(Karp, 1970)
(Nicotri, 1974)
(Kessel, 1964)
t Egg sizes measured from dissected gonads, not from spawned
gametes.
*spawning described in a western Atlantic specimen, but this
species occurs in the northeastern Pacific (Lindberg, 1979).
...
32
BRIDGE I
As demonstrated in Chapter II, adult Lotfia digitalis from South Cove, Cape
Arago, free spawn gametes that require ~5-6 days to develop and reach metamorphic
competence. During this planktonic period, it is likely that tidal mixing and nearshore
currents disperse larvae away from the parental habitat. Larvae must ultimately rejoin
adult conspecifics, however, if they are to form part of a reproductive population and
contribute their genetic material to subsequent generations.
It can be hypothesized that two general phenomena are operative in the
recruitment of L. digitalis into adult habitat. First, larvae might settle directly into adult
habitat. Second, post-settlement processes, such as migration or differential mortality,
might alter patterns of random settlement, or differential settlement into habitats outside
the adult habitat, and thereby determine the adult distribution pattern that is observed.
Chapter III addresses the importance of settlement vs. post-settlement processes in the
recruitment of L. digitalis into high intertidal habitat.
33
CHAPTER III
LABORATORY EVIDENCE THAT NATURAL RECRUITMENT IN THE LIMPET
Lottia digitalis (PATELLOGASTROPODA: LOTTIIDAE) IS DRIVEN BY
SETTLEMENT RATHER THAN POST-SETTLEMENT PROCESSES
Introduction
Many patellogastropod limpet species release gametes that ultimately develop
into planktonic ve1iger larvae (Lebour 1937; Fretter and Graham 1962; Strathmann
1987). After a brief pelagic life-stage, these larvae settle onto appropriate surfaces where
they metamorphose and commit to a benthic existence. During the benthic life stage,
many limpet species form strict habitat associations with particular algal species (Yonge
1962; Proctor 1968; Choat and Black 1979; Munoz and Santelices 1989), other
gastropods (Eickenberry & Wickizer 1964; Scheibling et al. 1990), and tidal zones
(Haven 1972; Branch 1976; Underwood 1979). An interesting dynamic of the early life
history in these limpets, in light of their habitat fidelity and planktonic larval dispersal, is
the recruitment ofjuveniles into highly specific adult habitats. With regard to this
dynamic, two general patterns emerge: 1) some limpet species are reported to settle
directly int%nto the adult habitat (Graham and Fretter 1947; Proctor 1968; Scheibling et
al. 1990), 2) others recruit indirectly by settling outside, and then migrating into, the adult
habitat (Corpuz 1981; McGrath 1992; and Delany et al. 1998). A third possible
34
recruitment pattern is that adjacent habitats experience equal settlement, but the habitats
impose differential postlarval mortality and thereby determine juvenile/adult
distributions. Although this pattern has been observed for some taxa (Raimondi 1988;
Rowley 1989), it has not been reported for recruitment of intertidal limpets.
Studies of limpet recruitment are typically based on length-frequency
observations, such that size-class structures are compared for adjacent habitat types (e.g.
Frank 1965; Breen 1972; Quinn 1988; McGrath 1992; Delany et al. 1998). These studies
are usually dependent upon observations made in the field, and they interpret the
presence and/or absence of juveniles in different habitats as evidence of recruitment
patterns. As such, their accuracy is limited because many months may pass between
initial settlement and the time at which juveniles are large enough to be visible and
identifiable in the field (Kay, Chapter IV of this thesis; Frank 1965; Parry 1982; Quinn
1988). During the interval between settlement and census, juvenile distributions across
(or within) adjacent habitats can change due to differential mortality or movement
(Graham and Fretter 1947; Keough and Downes 1982; Connell 1985; Rowley 1989;
Rodriguez et al. 1993). Consequently, and with particular regard to mobile organisms,
juvenile distributions at the time of census may not correspond to settlement locations.
Recruitment studies based on length frequency observations alone, although valuable and
informative, are not always unequivocal in their identification of settlement sites and
early juvenile habitats.
For many benthic marine invertebrates with planktonic larvae, settlement tends to
be a nonrandom process triggered by environmental cues (e.g. Burke 1983; Butman
\-
35
1987; Rodriguez et al. 1993). Environmental cues are typically associated with juvenile
or adult habitat, and are assumed to assist in successful recruitment. In the laboratory,
where competent larvae can be manipulated under controlled conditions, settlement cues
for a given taxon can be explored and identified. Indeed, laboratory studies have
identified the preferred settlement substrata of numerous gastropods (Hadfield 1978;
Morse and Morse 1984; Lambert and Todd 1994; Duame et al. 1999; Krug and Manzi
1999). In one such study, Proctor (1968) was able to settle larvae of the patellogastropod
limpet Acmaea incessa onto fragments of its algal host Egregia menziessii, but was
unable to settle larvae onto other non-host algae or bacterial films.
Lottia digitalis (Rathke 1833) is a patellogastropod limpet (family Lottiidae - see
Lindberg 1986) that is found high in the intertidal. L. digitalis occur as (at least) two
different ecotypes, which are readily distinguishable based on shell morphology, that
occupy rocky substrata or the stalked barnacle Pollicipes polymerus (See Byers 1989;
Lindberg and Pearse 1990). Whereas the recruitment of L. digitalis onto P. polymerus
has not been investigated, distributional studies based on length-frequency data suggest
that L. digitalis settles and recruits directly into rocky, high-intertidal adult habitat (Frank
1965; Breen 1972). Two factors, however, suggest that L. digitalis larvae might settle
lower in the rocky intertidal, and then undertake an upward postlarval migration. First,
the youngest individuals observed by Frank (1965) and Breen (1972) were consistently
seen at lower elevations within the adult habitat range, but were many months old by the
time they were visible. Second, a similar indirect recruitment pattern has been
hypothesized for other species of limpets that live high in the rocky intertidal (Corpuz
36
1981; Delany et al. 1998). Previous descriptions of L. digitalis recruitment do not
address the possibility of indirect recruitment in this species.
This paper reports on laboratory settlement experiments conducted with larvae of
LoUia digitalis, which were exposed to natural substrata collected from various tidal
heights near and within an adult habitat. The purpose of these experiments was to
identify likely settlement surfaces, and the height at which they occur on the shore, for
larvae of L. digitalis. Information regarding settlement preferences presented in this
report, when considered in tandem with size distributions previously observed by Frank
(1965) and Breen (1972), addresses the relative importance of initial settlement versus
post-settlement processes during the recruitment of L. digitalis into rocky, high-intertidal
habitat.
Additionally, I exposed larvae of L. digitalis to live specimens of P. polymerus to
determine if the former settles (and possibly recruits) directly onto this host. Given the
results of this experiment, and the fact that the larvae I used were reared from adults
taken from rocky habitat, I was able to address the possibility of a genetic basis for L.
digitalis habitat "preference" (P. polymerus vs. rocky substrata) at the time of larval
settlement.
Materials and Methods
Adult LoUia digitalis were collected from South Cove, Cape Arago, OR. (430 18'
N; 1240 24' W) during February and March 2000. In order to identify likely sites for
natural settlement, larvae spawned from these adults were reared to metamorphic
37
competence as described by Kay (Chapter II of this thesis), and in the laboratory they
were exposed to a variety of field-collected substrata. Three of these substrata were rock
fragments removed from, and representative of: dominant rocky surfaces available for
intertidal settlement within and immediately below the site of adult collection.
Fragments of the three rock types, termed High, Mid, and CCA, were removed with
hammer and chisel or by hand, and were returned to the Oregon Institute of Marine
Biology (OIMB) in Charleston, OR., where settlement experiments took place.
The High rock was collected from an adult L. digitalis boulder-field habitat, at a
tidal height between I and 2 meters above mean lower low water (MLLW). Rock chips
were taken from intermediate levels within the adult habitat, at about 1.5 m above
MLLW. High rock was course sandstone, with a texture similar to 100-grit sandpaper,
and was heavily colonized by diatoms and a filamentous green alga that was ubiquitous
in winter throughout the adult habitat. This alga was also present on the shells of many
adult L. digitalis at my site. The adult habitat was dominated exclusively by boulders of
this rock type at the site of collection, which extended ~75 meters along the coast. Other
lottiid species within this habitat zone included Lottia paradigitalis (moderately abundant
~1.0 - 1.4 m above MLLW), as well as Lottia pelta and Tectura persona (both were
common but present in very low abundance). An individual Macclintockia scabra was
also seen in this zone.
The Mid rock was collected from a tidal level 0.5 - 1.0 m above MLLW, a height
at which the acmaeid limpet Tectura scutum was the dominant limpet species. L.
paradigitalis and L. pelta also occurred in this zone but were less abundant than T
38
scutum. The CCA substratum was collected from 0.0-0.5 m above MLLWas pebbles
that were encrusted by a crustose corraline alga (CCA) within the family Corallinaceae
(see Steneck and Paine 1986). Although I could not identify this alga, it was distinct and
common at the South Cove site. At this tidal height the lottiids Lottia asmi (on their hosts
Tegulafunebralis) and T. scutum, as well as the acmaeid Acmaea mitra, were common
but present in very low adundance.
Fragments of the three rock types were placed in 175ml (5 cm tall, 6.5 cm base)
Pyrex custard dishes, both in the presence and absence of adult L. digitalis mucus, such
that six settlement treatments were created (High, High/Adults, Mid, Mid/Adults, CCA,
CCA/Adults). To prepare the three Adults treatment groups, adult L. digitalis were
allowed to crawl on the rocks for 24 hrs. and were removed immediately prior to the
experiment in order to avoid "bulldozing" of settled larvae. Each of the six treatment
groups, in addition to a sterile control dish, was replicated six times (total number of
dishes = 7 treatments x 6 replicates = 42 dishes). Replicate dishes were then filled with
0.45Ilm-filtered seawater (FSW), and 50 L. digitalis larvae were individually pipetted
into each settlement dish. All dishes were then haphazardly arranged in a 13.0oC
incubator under a 12: 12 Light:Dark photoperiod.
At three and six days after larvae were introduced to settlement vessels, each dish
was examined under a dissecting microscope for 25 minutes. Recovered individuals
were scored as: dead, unsettled, settled, vela shed, and adult shell growth (the latter two
conditions are sequential events during metamorphosis). Inevitably, some individuals
remained unaccounted for after the 25-minute search period. Results are presented as the
39
percentage of larvae that fall into each of the five categories (dead, unsettled, settled, vela
shed, adult shell growth). In order to calculate these percentages, the number of larvae in
each category was divided by 50 (the total number of larvae introduced to each dish),
rather than the total number of larvae recovered (i.e. some number < 50). In most dishes
recovery success was ~70%, because larvae were often difficult to find due to their small
size. Larvae that were unaccounted for probably had not metamorphosed, because those
individuals with adult shell growth were relatively large and easy to see. This experiment
was run twice using offspring from two pairs of parents, spawned separately on February
12 and on March 2. Larvae from the February 12 and March 2 spawning events were
introduced to settlement surfaces on February 18 and March 8, respectively. Growth
rates of newly settled juveniles that were spawned on March 2, and settled onto
High/Adults and Mid/Adults substrata, were followed for 70 days after fertilization.
Growth rates were compared for 3 dishes in both treatments that each contained 5
juveniles (number ofjuveniles per treatment = 3 dishes x 5 individuals/dish = 15). These
juveniles were reared at 13.00 C and were immersed in 32%0 seawater throughout the 70-
day period.
In addition to the High, Mid and CCA rocky substrata, 1. digitalis larvae were
exposed to the macroalgae Ulva sp., Enteromorpha contorta, Alaria marginata, and
Polysiphonia sp., which were common in protected habitats near the site of adult 1.
digitalis collection. The design of this experiment was identical to that detailed above,
with the exception that a positive control (High rock) accompanied the negative control
(0.45!lm FSW) (total dishes = 6 treatments x 6 replicates = 36 dishes). To limit algal
40
tissue damage and/or decomposition that might have artificially deterred settlement, algae
were placed in dishes as whole organisms or non-severed blades detached at a node. For
A. marginata this was not possible, due to its large size, and so settlement surfaces from
this species were pruned from the vegetative thallus. This experiment was run twice
using offspring from two different parental pairs, which were spawned on March 2 and
on March 6. Larvae from the March 2 and March 6 spawning events were introduced to
settlement surfaces on March 8 and March 12, respectively. The larvae spawned on
March 2 were siblings of those larvae exposed to the rocky substrata described above.
The results of the experiments described above (see Results, Figs. I and 2)
indicated that specific settlement cues on the High rock probably included one or more of
the following: 1) physical/chemical properties of the High rock, 2) benthic diatoms on
the High rock's surface, 3) the filamentous green alga (FGA) observed on the High
rock's surface, 4) the mucus or some other biochemical signature of adult L. digitalis. In
order to distinguish among these possibilities, I conducted a settlement experiment in
which larvae were exposed to each of these potential cues. Treatment groups for this
experiment, which were termed SterHigh (Sterilized High rock), Diatom, FGA, and
Mucus, were prepared in the following manner: To prepare the SterHigh treatment,
newly-collected High rock fragments were autoclaved in seawater for 15 minutes,
scrubbed with a toothbrush, and then autoclaved a second time. To prepare the Diatom
and FGA treatments, these algae were individually isolated from the High rock surface,
and cultured in 50 ml tubes containg f/2 medium with silica (Diatom) and without silica
(FGA). The diatom isolated was Navicula sp., and was 24.4 Jlm long by 12.2 Jlm wide (5