barnacles, limpets and periwinkles: the effects of direct and indirect interactions on cyprid...

www.elsevier.com/locate/seares

Journal of Sea Research

Barnacles, limpets and periwinkles: the effects of direct and

indirect interactions on cyprid settlement and success

Sebastian P. Holmesa,*, Graham Walkerb, Jaap van der Meera

aRoyal Netherlands Institute for Sea Research (NIOZ), PO Box 59, 1790 AB, Den Burg, Texel, The NetherlandsbSchool of Ocean Sciences, University of Wales Bangor, Menai Bridge, Anglesey, LL59 5AB, UK

Received 14 August 2003; accepted 25 May 2004

Available online 6 November 2004

Abstract

Conventionally, direct interactions between species are considered to be the most important biological factors determining

community composition, structure and stability. However, it has been suggested that the indirect interactions occurring between

species may be as important. One area of ecology where the direct effects of one species on another have been well studied is in

the rocky intertidal. Examination of the effect of the presence of P. vulgata (limpets) and L. littorea (periwinkles) on the

settlement and development of S. balanoides (cyprids/barnacles), over a cyprid settlement season and some six months later, in

four different treatments (limpets only, limpets and periwinkles combined, periwinkles only and control (no animals)) revealed

the following: (1) that the presence of limpets increased cyprid settlement and recruitment success above treatments containing

no limpets; (2) that cyprid settlement and success were greatest on the limpets-only treatment, followed by the limpets-and-

periwinkles treatment, then by the control treatment and then by the periwinkles-only treatment; (3) that the initial effects

observed in the treatments were reflected in the long-term community structure; (4) that the effects of the treatments were

independent of variations in algal biomass between treatments, i.e. the effects were not indirectly mediated through a second

species (host); (5) that cyprid mortality was greatest on the periwinkles-only treatment; (6) that the source of the effect of

limpets on cyprid settlement appeared to originate indirectly through the action of their residual pedal mucus trails.

It is concluded that periwinkles can affect the settlement and success of barnacles directly through biological disturbance (i.e.

surface ablation). However, although limpets may have a direct negative effect on barnacle settlement and success, at low to

medium densities, limpets can positively indirectly influence the cyprid settlement and success. This effect operates at a factor

greater than that afforded by the direct negative effects of periwinkles in a mixed-species treatment. These results illustrate how the

indirect effects of one species on another can have a more important structuring effect than those derived from direct effects alone.

D 2004 Elsevier B.V. All rights reserved.

Keywords: Biological disturbance; Habitat facilitation; Indirect interactions; Littorina littorea; Patella vulgata; Path analysis; Pedal mucus;

Semibalanus balanoides

1385-1101/$ - s

doi:10.1016/j.se

* Correspon

+44 1624 83100

E-mail addr

53 (2005) 181–204

ee front matter D 2004 Elsevier B.V. All rights reserved.

ares.2004.05.004

ding author. Present address. Port Erin Marine Laboratory, Port Erin, Isle of Man, IM9 6JA, UK. Tel.: +44 1624 831000; fax:

1.

ess: [email protected] (S.P. Holmes).

S.P. Holmes et al. / Journal of Sea Research 53 (2005) 181–204182

1. Introduction

The importance of interactions between species in

forming and shaping the overall structure of a

community has been well documented since Clements

(1916) first proposed his monoclimax hypothesis. One

area that has received particular attention in the latter

part of the last century is the effect of predator and

prey interactions on community structure (see Con-

nell, 1961; Paine, 1966; Hall et al., 1970; Dayton,

1975; Menge, 1976; Lubchenco, 1978; Menge and

Lubchenco, 1981; Jernakoff, 1983; Menge et al.,

1986, 1994; Posey and Hines, 1991; Berlow and

Navarrete, 1997). These studies have suggested that

predators can act as a form of dbiological disturbanceT,dhaltingT, ddelayingT or ddeflectingT succession from

its normal path (Dayton, 1971). Much of the evidence

for this has been obtained from the study of grazer

(predator) and algae (prey) interactions in the rocky

intertidal (for reviews see Lubchenco and Gaines,

1981; Gaines and Lubchenco, 1982; Hawkins and

Hartnoll, 1983).

In contrast, the role of indirect interactions

between grazers and coexisting non-prey (CNP)

species has received relatively little attention, even

though these interactions may contribute greatly to

community structure and stability (Connell and

Slayter, 1977; Dugan, 1986; Van Tamelen, 1987;

Farrell, 1991; Geller, 1991; Menge, 1995). In the

few studies that have been carried out to examine

the effects of molluscan grazers on CNP species

there is little agreement. Some authors have sug-

gested that molluscan grazers can affect CNP species

directly through physical disturbance, i.e. through

abrasion/perturbation (Hatton, 1938; Lewis, 1954;

Connell, 1961; Dayton, 1971; Menge, 1976; Branch,

1981; Hartnoll and Hawkins, 1985; Miller and

Carefoot, 1989; Lohse, 1993; Berlow and Navarrete,

1997; Buschbaum, 2000, 2002). Other authors have

implied that the effects of molluscan grazers on CNP

species can be secondarily mediated through the

abundance of their prey (Jones, 1948; Lodge, 1948;

Burrows and Lodge, 1950; Southward, 1956; Lub-

chenco, 1982, 1984; Dethier and Duggins, 1984;

Van Tamelen, 1987; Farrell, 1991), whilst further

authors have suggested that molluscan grazers may

affect CNP species through their pedal mucus trails

(Raimondi, 1988; Johnson and Strathmann, 1989;

Proud, 1994; Holmes, 2002; Holmes et al., 2002). In

addition to this, there has been little congruity

between these studies as to both the effect of the

presence of molluscan grazers on CNP species (i.e.

inhibitory or stimulatory) and as to the nature of the

interactions (i.e. direct or indirect). Note that the

former interactions can be classified as allogenic

ecosystem engineering and the latter as autogeneic

ecosystem engineering (Jones et al., 1994).

One system that has proved itself useful in

examining community processes is the rocky inter-

tidal zone of sea shores (Connell, 1961; Paine, 1969;

Dayton, 1975; Menge, 1976, 1991; Lubchenco,

1978; Jernakoff, 1983, 1985; Hartnoll and Hawkins,

1985; Dean and Connell, 1987; Van Tamelen, 1987;

Berlow and Navarrete, 1997). The faunal composi-

tion of the mid-littoral zone of North Eastern

Atlantic Boreo-Arctic rocky shores can be dominated

by Patella vulgata L., Littorina littorea (L.) and

Semibalanus balanoides (L.) (Southward, 1958;

Moyse and Nelson-Smith, 1963). Both P. vulgata

and L. littorea are mobile grazing prosobranch

molluscs, and their role in determining the algal

composition of intertidal communities has been well

documented in the literature (for P. vulgata see

Jones, 1948; Lodge, 1948; Burrows and Lodge,

1950; Southward, 1956; Lewis, 1954; Hawkins,

1981; Hawkins and Hartnoll, 1983 and for L. littorea

see Lubchenco, 1978, 1982, 1984; Hunter and

Russell-Hunter, 1983; Wahl and Sonnichsen, 1992;

Wilhelmsen and Reise, 1994). In contrast, the

barnacle S. balanoides is a sessile filter feeder, the

adult (fixed) distribution of which is determined by

the settlement of its planktonic larval stage, the

cypris larva (Yule and Walker, 1987). Cyprid

settlement is affected by a variety of biological and

physical cues (for reviews see Crisp, 1974; Lewis,

1978; Yule and Walker, 1987). In particular, several

studies have shown that the algal composition of a

settlement surface can affect cyprid settlement. For

example, Maki et al. (1988, 1990, 1992) have found

that certain floral communities can negatively affect

barnacle cyprid settlement, whereas Wieczorek et al.

(1995) have found the reverse to be true (see also

Jones, 1948; Southward, 1956; Strathmann and

Branscomb, 1979; Strathmann et al., 1981; Hudon

et al., 1983; Wieczorek et al., 1995). Community

level interactions between barnacles, algae and

S.P. Holmes et al. / Journal of Sea Research 53 (2005) 181–204 183

grazers have been studied in many intertidal com-

munities (see Jones, 1948; Lodge, 1948; Burrows

and Lodge, 1950; Connell, 1961; Dayton, 1971;

Hawkins, 1981; Jernakoff, 1983; Hartnoll and

Hawkins, 1985; Van Tamelen, 1987; Farrell, 1991;

Berlow and Navarrete, 1997; Buschbaum, 2000,

2002). However, little quantitative analysis has been

carried out on the interaction between barnacles and

molluscan grazers in community formation (Van

Tamelen, 1987).

The present paper examines the effect of the

presence of two molluscan grazer species, P. vulgata

and L. littorea, on the settlement and recruitment

success of S. balanoides cyprids. Success is defined

as the ability of settled cyprids to successfully

metamorphose into juvenile barnacles and hence

recruit, whilst settlement is defined as the fixation of

an organism from the plankton on to a substratum

irrespective of whether or not a metamorphic change

has occurred. A direct interaction is defined as the

interaction occurring between two or more species

(components) which involves the immediate (direct)

transfer of energy/matter between the interacting

species. For example, the interaction between predator

and prey is considered to be a direct interaction.

Conversely, an indirect interaction is defined as the

interaction between species not involving the imme-

diate (direct) transfer of energy/matter and/or influ-

ence of the effector species on the affected species.

For example, the action of keystone species in freeing

up habitats for CNP species is considered an indirect

interaction. These definitions are in line with those of

Fath and Pattern (1998, 1999). The aims of the present

paper are as follows:

(1) to ascertain the effect (i.e. inhibitory, stimulatory

or none) of the presence of P. vulgata and L.

littorea on the settlement and success of S.

balanoides cyprids (a coexisting non-prey

(CNP) species);

(2) to evaluate whether or not the initial effects of

the presence of P. vulgata and/or L. littorea on

the settlement of the cyprids of S. balanoides,

have any bearing on the long-term structure of a

community;

(3) to determine the nature of the interactions (i.e.

direct or indirect) occurring between P. vulgata,

L. littorea and the cyprids of S. balanoides.

2. Materials and methods

2.1. Study sites

All investigations were carried out on the north-

eastern flank of the Isle of Cumbrae, Ayrshire,

Scotland. Two replicate experiments (blocks) were

run on two different shores (Fig. 1). The two sites

chosen were White Bay (National Grid Reference

(NGR) NS175, 592), a sheltered shore (Site 1) and

Skate Point (NGR NS163, 584), an exposed shore

(Site 2). Shores at both sites are composed of

bedding plains (4–110 m2 in size) of Old Red

(Devonian) sandstone, dipping north-easterly, with

ravines between the plains running parallel to the

shoreline.

Two similar bedding plains, one at each site, were

chosen. Both plains were approximately 90 m2 (~7�~13 m, width�length, respectively) with an angle of

dip of 38. The bedding plain at Site 1 was 2.45 m above

chart datum and the bedding plain at Site 2 was 2.23 m

above chart datum. The surface rugosity (Rz) of each

bedding plain was measured by taking 50 randomly

selected rock chips from each plain, and measuring

their Rz in the laboratory with a mechanical stylus,

using a method identical to that of Holmes et al.

(1997). Site 1 had meanFstandard error (SE) Rz of

120F50 Am and Site 2 a meanFSE Rz of 134F60 Am.

Both sites had an existing mid-littoral fauna of Semi-

balanus balanoides (L.), Patella vulgata L. (limpets)

and Littorina littorea (L.) (periwinkles). All experi-

ments were carried out for an initial period of 30 days

(during the cyprid settlement season) from 6 April

1995 to 6 May 1995 followed by a second observa-

tional visit some six months later, on 11 November

1995.

2.2. Site preparation

In order to minimise any edge effects (i.e.

variations in cyprid supply due to localised edge

hydrodynamics (see De Wolf, 1973) a 5�11 m

(length�width) uniform experimental area was des-

ignated in the centre of each bedding plain. The

experimental areas were subdivided into four, 2�5

m, experimental plots with each plot having a gap of

1 m from its nearest neighbour. These gaps acted as

buffer zones between experimental treatments.

Fig. 1. Location of study sites.


Before any manipulative work was carried out, the

average densities of P. vulgata and L. littorea were

determined at each site, by throwing one hundred

0.5�0.5 m quadrates over the experimental areas and

adjacent bedding plains. In addition to this, 30 rock

chips were taken at random from within each of the

experimental areas and analysed for their chloro-

phyll-a content using the HMSO (1986) hot meth-

anol extraction procedure. These rock chips were

used for both comparisons in the initial and final

algal composition between sites/treatments, and for

evaluating the efficiency of the cleaning method

applied to the shore surface.

Removal of all of the fauna and algae, with the

exception of limpets, from the experimental areas

was achieved as follows. The surface was first

scraped with a paint scraper followed by scrubbing

with a wire brush. This was visibly effective in

removing all of the large matter from the surface.

The effect of any remaining surface-bound arthropo-

din, i.e. to remove the effect of adult conspecifics on

cyprid settlement (see Knight-Jones, 1953) and/or

sub-surface biota was then minimised by scrubbing

the surface with 2 M NaOCl solution (Knight-Jones,

1953; Larman and Gabbott, 1975) have shown that

NaOCl, effectively neutralises arthropodin), then

rinsing it with copious amounts of seawater. The

surface was then finally heat sterilised using a butane

fuelled brazing torch (similar cleaning procedures

have been used by Dayton, 1971; Van Tamelen,

1987). Care was taken not to harm any of the

resident limpets within the experimental plot. Once

the surface was cleaned, 30 randomly selected rock

chips from each site were taken for chlorophyll-a

analysis and a further thirty rock chips from each site

were taken for examination by Scanning Electron

Microscopy (SEM). These rock chips were used in

conjunction with the first set of rock chips, i.e. those

taken before the surface was cleaned, to evaluate the

efficiency of the cleaning procedure. Sample prepa-

ration of the rock chips, for SEM analysis, was made

using the air drying procedure of Hill and Hawkins

(1990).

Each of the plots within the cleaned experimental

areas were then randomly allocated to one of the

four following experimental treatments: A: limpets

only, B: limpets and periwinkles C: periwinkles

only, and D: no animals whatsoever (Fig. 2). The

limpets for treatments A and B were the existing

inhabitants (for densities see Results). For treatments

C and D, the existing limpets were removed, and the

areas inhabited by them cleaned as outlined. The

number and shell lengths (anterior to posterior) of

the limpets in each of the treatments was also

Fig. 2. Schematic diagram of the experimental design at both sites.


recorded. Periwinkles were collected from the

adjacent bedding plains and applied at their natural

density (~400 ind m�2, see Results). In addition

2000 periwinkles were randomly collected from each

site and their total dry weight (shell+flesh; shells

cracked and dried at 40 8C until a constant dry weight

was obtained) recorded.

To prevent the intrusion of any unwanted fauna

within a treatment without introducing any potential

dcage effectsT (see Dayton, 1971; Cubit, 1984), the

treatments were isolated from each other and the

shore by enclosing each in a thin boundary, 25�5

mm (width�height, respectively), of Tree Tanglefoot

Pest Barrier (The Tanglefoot Company, Michigan,

USA). This has been shown to be a non-toxic agent

effective at preventing the intrusion of mobile

molluscan grazers (Geller, 1991). Fixed points, for

the monitoring of cyprid settlement, were made by

drilling seven holes within the interior 4 m2 of each

treatment. The holes were positioned on a random

basis with the proviso that no hole was less than 0.6

m from any other hole. Brass screws were secured

into the holes, after plugging with Rawlplugs, to

ensure that they could be relocated when necessary

(Fig. 2).

On daily visits to the sites, the Tanglefoot barriers

were renewed when necessary. The numbers of

periwinkles within the appropriate treatments were

maintained at natural levels by adding or removing

animals. Any other ingressed fauna were removed. L.

littorea were found to be the sole invasive organism,

particularly after stormy weather, requiring daily


removal to maintain treatment levels. Under normal

conditions removal rates ranged between 0–50 ind

m�2 d�1 for the appropriate treatments.

2.3. Observation of cyprid settlement and recruitment

Cyprid settlement and recruitment was monitored

daily by photographing each of the fixed monitoring

points in each treatment at both sites. The photo-

graphs were taken such that each image included a

minimum area of 400 cm2 defined by a radius of 12

cm from the base of each brass screw. In the

laboratory, the number of metamorphosed cyprids/

barnacles was counted on each photograph. Note that

these data are cumulative in that they include both

newly settled cyprids and cyprids which have settled

and successfully recruited to the shore during the

experimental period.

2.4. Calculation of the indices of algal biomass

Seven randomly selected rock chips from each

treatment, at each site, were taken for chlorophyll-a

analysis at days 1, 3, 7, 14 and 28. The chips were

stored at �30 8C in the dark prior to chlorophyll-a

extraction and extraction was made using the HMSO

(1986) hot methanol extraction procedure.

2.5. Observation of cyprid mortality

Cyprid mortality was observed in the following

way. On day 1 of the experimental period a

randomly selected area (0.25 m2) was marked within

each treatment, at both sites, using a non-toxic paint

marker (Edding 751 paint marker, Edding, Ger-

many), and photographed. In the laboratory, the

positions of the freshly settled cyprids within the

experimental area were marked (i.e. from the photo-

graph) onto acetate and the total number present

counted. Four days later the same areas were

photographed. The acetate prepared from the pre-

vious photograph was then laid over the new

photograph, and the number of successfully meta-

morphosed cyprids (i.e. juvenile barnacles), which

had developed from the original cyprids, noted. This

procedure was then repeated for a new batch of

freshly settled cyprids every three days, a further six

times.

2.6. Calculation of the amount of pedal mucus

produced by P. vulgata, the amount of pedal mucus

produced by L. littorea and rates of surface ablation

by both P. vulgata and L. littorea

2.6.1. L. littorea

Following the suggestion of Davies et al. (1992)

that periwinkles move ~2 m d�1, 1 m during

immersion and 1 m when exposed, the following

equations, from Davies et al. (1992), can be used to

calculate the quantity of pedal mucus produced by L.

littorea in an experimental plot:

y ¼ 8:71T10�7Tx 0:733ð Þ ð1Þ

Where y=the dry weight of the pedal mucus produced

(g) per mm travelled by an animal in seawater and

x=the whole animal dry weight (g).

y ¼ 1:184T10�6Tx 0:843ð Þ ð2Þ

Where y=the dry weight of the pedal mucus produced

(g) per mm travelled by an animal in 100% relative

humidity (RH) and x=the whole animal dry weight

(g).

That is the total amount of pedal mucus produced

by L. littorea (g 24 h�1 m�2)=the amount of pedal

mucus produced by a periwinkle of mean dry weight

for that site in 24 h (i.e. Eq. (1)+Eq. (2)) �400.

Similarly, the approximate area browsed in an

experimental plot by the periwinkles, in a 24 h period,

can be calculated from the following equations of

Davies et al. (1992):

y ¼ 6:412Tx 0:351ð Þ ð3Þ

Where y=the width of pedal mucus trail produced

(mm) by an animal in 100% RH and x=the whole

animal dry weight (g).

y ¼ 6:902Tx 0:281ð Þ ð4Þ

Where y=the width of the pedal mucus trail produced

(mm) by an animal in seawater and x=the whole

animal dry weight (g).

That is the total area browsed (cm2 24 h�1) by the

periwinkles m�2 in an experimental plot=(the width

(cm) of the pedal mucus trails produced by a

periwinkle of mean dry weight (cm), for that plot, in

100% RH (i.e. Eq. (3)) * 400 * the distance moved by

a periwinkle (i.e. 100 cm))+(the width (cm) of the


pedal mucus trails produced by a periwinkle of mean

dry weight (cm), for that plot, seawater (i.e. Eq. (4)) *

400 * the distance moved by a periwinkle (i.e.

100 cm)).

2.6.2. P. vulgata

If it is assumed that the daily period of tidal

immersion experienced by the experimental areas is

12 h, and that during this time all of the limpets are

fully mobile, then the amount of pedal mucus

produced by P. vulgata in an experimental plot can

be calculated from the following equation of Davies et

al. (1990a):

y ¼ 8:51T10�7Tx 1:903ð Þ

Where y=the dry weight of pedal mucus (g) produced

per hour in seawater when the animal is mobile and

x=the anterior posterior shell length (mm) of the

animal.

That is the total amount of pedal mucus produced

by P. vulgata (g 24 h�1)=the sum of the individual

amounts produced by each animal over 24 h, for that

experimental plot.

A rough estimate of the area browsed over by P.

vulgata, in an experimental plot in a 24 h period can

be made by dividing the wet weight of pedal mucus

produced in 24 h, by the average thickness of the

pedal mucus produced by a limpet, providing the

following assumptions are made:

(1) that the density of P. vulgata pedal mucus is

approximately equal to the density of water/

seawater (i.e. 1 g=~1 cm3);

(2) that the dry weight of the pedal mucus produced

by P. vulgata pedal mucus is ~8% of its wet

weight (Davies et al., 1990b; see also Grenon

and Walker, 1980). That is:

A ¼ DWP=8ð ÞT100ð Þ=TT

Where A=area browsed (cm2 24 h�1), DWP=(dry

weight of pedal mucus produced (g 24 h�1) and

TT=trail thickness (0.001 cm, see Holmes et al., 2002).

2.7. Statistical analysis of the experimental results

Because of the inherent pseudoreplication pre-

sent in the experimental design, the mean data

obtained over the whole experimental/survey period

were analysed using a randomised block design

(i.e. each site represents a block). Hence, in the

analysis of variances (ANOVAs) performed, three

factors were designated: site (1, 2), the presence (0,

1) of limpets in a treatment and the presence (0, 1)

of periwinkles in a treatment. It should be noted

that site is considered to be a random factor and

hence it has no interaction, and the latter two

factors fixed. In this way comparison of the effects

of one treatment to another can be made in the

absence of post-hoc analyses. Schematics are

presented to aid the reader to interpret the results

of the ANOVAs.

Comparisons between the effects of the treatments

and the effects of the algal biomass on cyprid

settlement were made as above, with the exception

that site was defined as a fixed factor (i.e. to test for

possible interaction effects). This choice has no

effect on the tests for the effects of the treatments.

In order to explore whether cyprid densities were

affected by the absence/presence of the two inverte-

brate species, or, alternatively, indirectly through

changes in algal biomass, a path analysis was

performed. Path analysis allows the formulation of

alternative structural models, which can easily be

summarised in the form of a path diagram. A path

diagram shows the causal relationships, where the

strength of each relationship (or path) is indicated by

a path coefficient. Path coefficients are basically

standard partial regression coefficients (Sokal and

Rohlf, 1981). Calculation of the relative effects of

the treatments on both cyprid settlement and algal

biomass was made by dividing the pooled mean data

for the treatments by that obtained for the control

treatment.

3. Results

3.1. Faunal densities and efficiency of the cleaning

procedure

3.1.1. P. vulgata

Comparison of the densities of P. vulgata between

sites using a Students’ t-test revealed no differences

(t=1.24, P=0.15). The meanFstandard error (SE)

density of P. vulgata at site 1 was 26.8F0.8, and


25.8F0.7 ind m�2 at Site 2. The individual (counted)

density of limpets in treatments A and B at Site 1

was 29 and 30 ind m�2, respectively, and 32 and 31

Fig. 3. MeanFSE daily number o

ind m�2 for treatments A and B at Site 2,

respectively. Analysis of the data for the shell

lengths of P. vulgata between treatments and sites

f cyprids (barnacles) settled.


by ANOVA revealed that there were no differences

in the shell lengths of P. vulgata between sites

(F=2.77, P=0.10), and/or treatments (F=1.73

P=0.163). The meanFSE overall shell length of P.

vulgata at either site was 3.67F0.08 cm.

3.1.2. L. littorea

Comparisons of the densities of L. littorea between

sites using a Students’ t-test revealed no differences

(t=0.78, P=0.84). The meanFSE density of L. littorea

at Site 1 was 416F18, and 393F8 ind m�2 at Site 2.

Analysis of the mean dry weights (shell+flesh)

recorded for L. littorea between sites, using a t-test,

revealed that Site 2 periwinkles were heavier than Site

1 periwinkles (t=23.69, P=0.001). The meanFSE dry

weight of periwinkles at Site 2 was 1.39F0.02 and

0.78F0.01 g at Site 1.

3.2. Evaluation of the surface cleaning procedure

The meanFSE initial chlorophyll-a concentrations

of the rock chips at Sites 1 and 2 were 8.13F3.40 and

12.87F5.20 Ag chlorophyll-a cm�2, respectively.

Once the surfaces had been cleaned, the chlorophyll-

a concentration FSE measured from a second set of

rock chips dropped to 1.67F0.52 for Site 1 and

1.80F0.48 Ag chlorophyll-a cm�2 for Site 2. Analysis

Fig. 4. Polynomial regression (cubic) F95% confidence intervals for the ef

of these data using a Kruskal–Wallis test showed that

there were no differences in the chlorophyll-a

concentrations between sites (H=0.04, P=0.832), and

that the cleaning procedure was effective in reducing

the chlorophyll-a concentration of both sites

(H=51.75, P=0.001). Examination of the rock chips

taken after the application of the cleansing procedure,

by Scanning Electron Microscopy, failed to reveal the

presence of any epilithic organisms. The cleansing

procedure appears effective in removing epilithic

organisms but ineffective in removing endolithic

organisms, as indicated by the residual chlorophyll-a

concentration of the rock chips, taken after the

application of the cleaning procedure.

3.3. Observation of cyprid settlement and success

Unfortunately some of the films used during the

experimental period were accidentally damaged and

failed to process (develop) properly. This resulted in

only 19 sampling days out of a possible 31 for Site 1

and 18 for Site 2 (see Fig. 3). Observation of the

settlement profile of the cyprids over the settlement

season (Fig. 3) reveals the following:

(1) that initially cyprid settlement and survival was

uniform across all treatments;

fects of the treatments (pooled data) on cyprid (barnacle) settlement.


(2) that after ~7 days had elapsed more cyprids

appeared to be settling and surviving on treat-

ments containing limpets (i.e. A and B) than

treatments without;

(3) that by day 15 of the experimental period,

cyprid settlement and success was always

greatest on treatment A followed by B, then

by D and then by treatment C. Treatments A

and B were thereafter always different to

treatments C and D and to each other, with

treatments C and D generally remaining differ-

ent to each other. That is, the presence of

periwinkles in a treatment in the absence of

limpets depressed cyprid settlement below that

Table 1

Results of the ANOVA for pooled mean cyprid/barnacle settlement/succe

Cyprid success 6 April–6 May

Source d.f. S.S.

Site (R) 1 0.511

Limpets present (F) 1 0.484

Littorina present (F) 1 0.074

Limpet�Littorina 1 0.001

Residual 3 0.052

Total 7 1.122

F or R in parenthesis indicates a fixed or random factor, respectively.

significant result at P=0.01 or less.

Schematic of treatment differences as derived from the ANOVA

Treatments A (limpets only) B (limpets and periwink

Means (FSE) 29.56F1.40 23.68F1.10

The number of cyprids settled cm�2 d�1 decreases from left to right

treatments at P=0.05 or less.

Barnacle success some 6 months later (1 November)

Source d.f. S.S.

Site (R) 1 0.659




Residual 3 0.003

Total 7 1.353

F or R in parenthesis indicates a fixed or random factor, respectively. ***


Treatments A (limpets only) B (limpets and periwink

Means (F SE) 28.74F1.87 22.41F1.39

The number of barnacles present cm�2 decreases from left to right.

treatments at P=0.05 or less.

of the control treatment (D). Further illustra-

tion of the period at which differences were

generated between the treatments, and their

persistence over time can be made by poly-

nomial (cubic) regression of the data (Fig. 4).

From the plot it can be seen that after ~8 days

have elapsed, treatments A and B were

consistently different to each other and all

other treatments, treatment A having greater

cyprid settlement and survival than treatment

B. In comparison, treatments C and D

remained approximately equal until day 18 of

the experimental period. Thereafter all treat-

ments were different from each other, treat-

ss data (natural-log-transformed)

M.S. F P

0.511

0.484 279.516 0.001***

0.074 42.831 0.007**

0.001 0.565 0.507

0.002

*** denotes a significant result at P=0.001 or less, ** denotes a

les) D (no animals) C (periwinkles only)

17.74F1.11 14.70F0.97

. A break in the line indicates a significant difference between

M.S. F P

0.659

0.592 675.583 0.001***

0.098 111.783 0.001***

0.001 1.558 0.301

0.001

denotes a significant result at P=0.001 or less.

les) D (no animals) C (periwinkles only)

16.43F1.09 13.44F0.79

A break in the line indicates a significant difference between


ment D having greater cyprid settlement than

treatment C.

Analysis of the pooled mean natural-log-trans-

formed cyprid count data (April–May) revealed that

cyprid settlement and success was different between

all treatments, with greatest cyprid settlement

occurring on treatment A (limpets only) followed

by B (limpets and periwinkles) then D (no animals)

and then C (periwinkles only) (Table 1). In effect

the presence of limpets on a treatment increased

cyprid settlement, whereas the presence of periwin-

kles decreased it, the positive effect of the presence

of limpets outweighing the negative effect of the

presence of periwinkles. Correspondingly, analysis

of the pooled mean natural-log-transformed cyprid

survival data (i.e. the number of remaining bar-

Fig. 5. MeanFSE chlorophyll-a con

nacles) some six months after the experimental

period, using ANOVA, revealed exactly the same

pattern (Table 1). That is, barnacle density was

greatest on treatment A, followed by B, then D, and

then C. This suggests that the initial effects of the

treatments on cyprid settlement and success are

reflected in the long-term structure of a community.

Calculation of the effects of the treatments on

cyprid settlement and success in comparison to the

control treatment (D) revealed that treatment A

(limpets only) could increase cyprid settlement by a

multiple of 1.67 whereas treatment C (periwinkles

only) decreased it by a multiple of 0.83. If it is

assumed that the presence of limpets only has a

positive effect, and the presence of periwinkles a

negative effect, then the positive effect of limpets

on cyprid settlement outweighed the negative effect

centration for each treatment.


of periwinkles in treatment B (mixed species),

resulting in an increase in cyprid settlement by a

multiple of 1.33.

3.4. Indices of algal biomass

The film containing the photographs of the rock

chips (i.e. for measurement of the surface area of

the rock chips) for day 14 (i.e. 21 April 1995) was

lost, so no data are presented for that sampling

date. From a meanFSE initial algal biomass of

1.67F0.52 Ag chlorophyll-a cm�2 at Site 1 and

1.80F0.48 Ag chlorophyll-a cm�2 at Site 2 the

algal biomass of all treatments, at both sites, had

increased by the end of the experimental period

(Fig. 5). Comparison of the algal biomass of the

treatments at the end of the experimental period and

their initial chlorophyll-a concentrations (i.e. that

before any experimental work was carried out),

using Kruskal–Wallis tests, revealed differences in

the dbeforeT and dafterT chlorophyll-a concentrations

between treatments (H=22.71, P=0.001 for Site 1

and H=15.71, P=0.001 for Site 2). Analysis using

ANOVA of the pooled mean natural-log-trans-

formed chlorophyll-a data of the treatments revealed

that treatments A (limpets only) and D (no animals)

were different from treatments B (limpets and

periwinkles) and C (periwinkles only) but not from

each other (Table 2, Fig. 5). It would appear that

the presence of periwinkles in a treatment consis-

tently reduced the algal biomass of that treatment

Table 2

Results of the ANOVA for the pooled mean algal biomass data (natural-l

Source d.f. S.S.

Site (R) 1 0.158




Residual 3 0.035

Total 7 1.248

F or R in parenthesis indicates a fixed or random factor, respectively. **


Treatments C (periwinkles only) B (limpets and periw

Means (FSE) 2.51F0.18 2.28F0.18

Chlorophyll-a concentration increases from left to right. A break in the l

or less.

over time, with reference to treatments without

periwinkles. The presence of limpets in a treatment

appears to have no dmajorT effect on algal biomass.

Calculation of the effects of the statistically differ-

ent treatments on algal biomass, relative to the

control treatment (D), determined that treatments C

(periwinkles only) and B (mixed species) decreased

algal biomass by multiples of 0.48 and 0.44,

respectively.

3.5. Average mortality of cyprids

Analysis of the pooled mean natural-log-trans-

formed mortality data by ANOVA revealed that

mortality on treatment C (periwinkles only) was

greater than that on all other treatments, which were

not different to each other (Table 3).

3.6. The amount of pedal mucus produced by P.

vulgata, the amount of pedal mucus produced by L.

littorea and the area ablated by both P. vulgata and by

L. littorea

3.6.1. L. littorea

Using the equations of Davies et al. (1992), as

outlined, the amountFSE of pedal mucus produced

(dry weight) by 400 ind d�1 was calculated as

0.668F0.008 g pedal mucus d�1 for Site 1, and

1.071F0.015 g pedal mucus d�1 for Site 2. Therefore,

Site 2 periwinkles produce more pedal mucus than Site

1 periwinkles. Using the second set of equations, as

og-transformed)

M.S. F P

0.158

0.014 1.203 0.353

1.037 87.773 0.003**

0.004 0.033 0.868

0.012

denotes a significant result at P=0.01 or less.

inkles) A (limpets only) D (no animals)

4.84F1.05 5.17F0.99

ine indicates a significant difference between treatments at P=0.05

Table 3

Results of the ANOVA for the pooled mean mortality data (natural-log-transformed)

Source d.f. S.S. M.S. F P

Site (R) 1 0.003 0.003

Limpets present (F) 1 0.039 0.039 31.515 0.011*

Littorina present (F) 1 0.102 0.102 83.235 0.003**

Limpet�Littorina 1 0.057 0.057 46.484 0.006**

Residual 3 0.004 0.001

Total 7 0.205

F or R in parenthesis indicates a fixed or random factor, respectively.* denotes a significant result at P=0.05 or less, ** denotes a significant

result at P=0.01 or less.


Treatments D (no animals) A (limpets only) B (limpets and periwinkles) C (periwinkles only)

Mean (FSE) % mortality. 33F5 34F5 36F4 49F8

Cyprid mortality increases from left to right. A break in the line indicates a significant difference between the treatments at P=0.05 or less.

Table 4

Results of the ANOVA for the combined effect of treatments on

chlorophyll-a concentration and cyprid settlement (pooled mean

natural-log-transformed data)


Cyprids

Site (F) 1 0.511 0.511 294.995 0.001***

Limpets present (F) 1 0.484 0.484 279.516 0.001***


Limpet�Littorina 1 0.001 0.001 0.565 0.507

Residual 3 0.005 0.002

Total 7 1.076


Algae

Site (F) 1 0.158 0.158 13.375 0.035*

Limpets present (F) 1 0.014 0.014 1.203 0.353


Limpet�Littorina 1 0.001 0.001 0.033 0.868

Residual 3 0.035 0.012

Total 7 1.245

F in parenthesis indicates a fixed factor.

* Denotes a significant result at 0.05 or less.

** Denotes a significant result at 0.01 or less.

*** Denotes a significant result at P=0.001 or less.


outlined in Materials and Methods, the minimalFSE

area browsed in a 24 h period by 400 ind was

calculated as 5.26F0.07 m�2 for Site 1, and

5.25F0.07 m�2 for Site 2. Note that these calculated

figures represent a complete browsal of the exper-

imental surface area ~ every 5 h for both sites.

3.6.2. P. vulgata

As mentioned in Section 3.1.a, there were no

differences in the density of limpets between treat-

ments, and/or sites (meanFSE density=26.2F0.8

limpets m�2). The overall meanFSE shell length

was 3.67F0.08 cm for both sites. Using the equation

of Davies et al. (1990a,b), as outlined, the amountFSE of pedal mucus (dry weight) produced was

calculated as 0.255F0.005 g pedal mucus d�1 for

all appropriate treatments, at both sites. Using the

second equation, as outlined, the total areaFSE

browsed by the limpets in a 24 h period was

calculated as 0.3188F0.0069 m2. In terms of the

time taken by limpets to browse the experimental

area this equates to a complete browsal of the

experimental surface area by the limpets, ~ every 3

days assuming that they do not re-browse previously

browsed areas.

3.7. Analysis of the combined effects of algal biomass

and the experimental treatments on cyprid settlement

From the initial analyses the presence of limpets

had a significant and large positive effect on cyprid

densities, but no effect on algal density (Table 4

and Fig. 6). In contrast, periwinkles showed a

small negative effect on cyprid densities and had a

huge negative effect on algal density (Table 4 and

Fig. 6). Both treatment effects were significant but

no significant interactions between limpet and

Fig. 7. Schematic of two different path analysis models.


periwinkles were observable. For the path analysis

two alternative models were considered. The first

model resembled the ANOVA model, but ignores

the interaction (which in the ANOVA was small

and non-significant). The second model assumes

that the effect of periwinkles on cyprid larvae

works indirectly through algae. Both models

performed almost equally well in terms of coef-

ficient of determination (R2): 0.997 for the first

model and 0.987 for the second. The path diagrams

(Fig. 7) reveal that the negative effect of periwin-

kles on cyprid larvae (a path coefficient of �0.263

in the first model) could work almost entirely

through the algae, if algal density was responsible

for positively affecting cyprid settlement. However,

given that treatments A and D have approximately

equal, high algal densities but very different results

in terms of cyprid settlement, whereas treatments B

and C have low algal densities with different

results in terms of cyprid settlement. It is apparent

therefore, as in Fig. 6 and Model 1, that the

presence of limpets is responsible for increasing

cyprid settlement irrespective of algal concentration.

The derived effect in the 2nd path analysis model

for the possible effect of algae mediated by the

action of periwinkles in this case is anomalous.

Fig. 6. Plot of the relationship between algal biomass and cyprid

settlement as determined by the treatments at each site (data log

transformed).

The indirect effects on the settlement of cyprid

larvae through the algae mediated by site and

limpets are ineffectual.

4. Discussion

One of the ultimate goals in ecology is to discover

general truths about how communities and ecosystems

are regulated (Menge, 1991). To this end, the results

of the experiments presented here have shown that the

presence of P. vulgata can increase cyprid settlement

and success whereas the presence of L. littorea can

decrease it, the positive effect of the presence of P.

vulgata outweighing the negative effect of the

presence of L. littorea.

In contrast to these results, Lewis (1954), Connell

(1961), Dayton (1971), Menge (1976), Choat (1977),

Denley and Underwood (1979), Hawkins (1983),

Paine (1981), Farrell (1988), Miller and Carefoot

(1989), Safriel et al. (1994) and Kim (1997) have all

shown that limpets can have a negative effect on

newly settled barnacle species by either dbulldozingT


recruits off the rock surface or inadvertently eating

recruits whilst grazing algae. Although limpets do

undoubtedly consume (see Jones, 1948; Miller and

Carefoot, 1989) and dislodge (Lewis, 1954; Connell,

1961; Dayton, 1971; Menge, 1976; Choat, 1977;

Denley and Underwood, 1979; Hawkins, 1983; Paine,

1981; Farrell, 1988; Kim, 1997) newly settled cyprids,

the importance of physical (biological) disturbance by

limpets will fundamentally depend upon four factors.

These are:

(1) the density of the grazers (see Hatton, 1938;

Connell, 1961; Dayton, 1971; Lubchenco, 1978,

1982; Miller and Carefoot, 1989; Berlow and

Navarrete, 1997; Jarrett, 1997);

(2) the mobility of the grazers (for this study P.

vulgata browsed over an area of 0.32 m2 d�1);

(3) the levels of settlement and recruitment experi-

enced by a site, both temporally and spatially

(see Pyefinch, 1948; De Wolf, 1973; Wethey,

1983, 1984; Berlow and Navarrete, 1997);

(4) the ability of the cyprids (i.e. a CNP species) to

occupy spatial, temporal and morphological

refuges (see Connell, 1961; Lubchenco and

Menge, 1978; Miller and Carefoot, 1989; Geller,

1991; Bertness et al., 1992; Berlow and Nav-

arrete, 1997).

For example, Connell (1961), whilst working on

shores on the Isle of Cumbrae similar to those used

here, determined that two limpets, each of 50 mm

shell length, caged in a cleared area of shore (0.0345

m2) could reduce the success of S. balanoides cyprids

by a factor of ~0.5 during a cyprid settlement season.

For an individual limpet of mean shell length (i.e. 36.7

mm) (i.e. as in these experiments) this equates to a

decrease in cyprid settlement by a multiple of 0.8 per

unit area browsed by that limpet. That is, the negative

effect of each limpet on cyprid settlement for this

paper is equal to 1–(2 / 0.5 / (total area browsed by the

limpets used by Connell (1961), calculated from the

equations in Materials and Methods (436 cm2)) /

experimental area (345 cm2)).

If it is assumed that:

(1) the number of cyprids settled in any area

browsed by limpets is reduced by a factor of

0.8;

(2) the limpets will not re-browse a previously

browsed area within each 24 h period unless

forced to by spatial considerations;

(3) the area browsed by the limpets is adequately

described by the equations in the materials and

methods;

(4) that following the browsing of an area, that area

can increase cyprid settlement by a factor of 1.67

(i.e. the mean factorial increase in cyprid

settlement recorded for treatment A),

then a model can be formulated which illustrates

the approximate density-dependent effects of P.

vulgata on cyprid settlement. A plot of the various

effects of different limpet densities on the settlement

of cyprids is given in Fig. 8. In this illustration, a

constant daily additive recruitment of 100 cyprids

m�2 was provided as the input into the system and

on day 1 following initial settlement, only negative

effects were considered. Thereafter both positive

and negative effects were allowed to interact to

predict the subsequent levels of cyprid recruitment.

From the plot, it can be seen that at low limpet

densities the settlement of cyprids is increased in

comparison to the input level of recruitment (i.e. the

no animals (control) treatment). However, as limpet

density increases (i.e. such that the total area

browsed by the limpets increases), then cyprid

settlement is negatively affected by the presence

of the limpets.

Pyefinch (1948) and Knight-Jones (1953) have

shown that cyprid mortality through limpet disturb-

ance is restricted to the first 24 h following

settlement, i.e. until metamorphosis has occurred.

Similarly, Barnes and Powell (1950) and Crisp

(1961) are of the opinion that once growth of the

juvenile barnacle base has begun, cyprids are

immune to the effects of disturbance by limpets.

Personal observations of the treatments, during the

experimental period, suggest that for this study

cyprid mortality through disturbance by the limpets

was restricted to a radius of 1–2 cm around the home

scar of individual limpets. If it is assumed that both

the calculations of limpet motility are accurate and

that the presence of limpets within a treatment does

in some way enhance cyprid settlement and success,

any cyprids occupying the 0.68 m2 area not abraded

by the limpets, in a day, will have successfully

Fig. 8. Illustration of the density dependent effect of P. vulgata on cyprid settlement.


metamorphosed and will be immune to future

disturbance by the limpets. This proposal is indi-

rectly evidenced by the results obtained for the levels

of cyprid mortality recorded within each treatment.

That is, cyprid mortality was highest on treatment C

(L. littorea only) whilst treatments A, B and D had

similar levels. In addition to this, any surface spaces

that are opened up through biological disturbance by

the limpets will attract more recruits than other

treatments, provided that space is not re-browsed

within 24 h in the following ways:

(1) through the stimulatory effects of arthropodin

and cyprid dfootprintsT left behind by the

abraded cyprids (see Knight-Jones, 1953; Lar-

man and Gabbott, 1975; Moyse and Hui, 1981;

Yule and Walker, 1984, 1987; Dineen and Hines,

1992; Clare et al., 1994), i.e. the previous

presence of the cyprids will initiate a settlement

cascade (Wethey, 1986);

(2) by providing free space for new recruits to settle

on, in the face of intraspecific competition for

space (see Crisp, 1961; Moyse and Hui, 1981;

Bertness et al., 1992) and the propensity of

cyprids to settle according to their age (see

Lucas et al., 1979; Jarrett, 1997).

Bertness et al. (1983), Proud (1994) and Buschbaum

(2000, 2002) have examined the effect of the presence

of L. littorea on cyprid settlement. These authors, in

agreement with the results recorded here, found that

cyprid settlement was reduced in the presence of L.

littorea. Bertness et al. (1983) and Buschbaum (2000,

2002) ascribed such results to physical disturbance of

the newly settled cyprids by the periwinkles. However,

Proud (1994) and Holmes (2002) have found that

surfaces that have been pre-conditioned with the pedal

mucus produced by L. littorea can increase the

settlement of S. balanoides cyprids (see also Raimondi,

1988; Johnson and Strathmann, 1989; Holmes et al.,

2002). This suggests, in a similar manner as for P.

vulgata, that there may be some density-dependent

relationship between the positive and negative effects

of the presence of L. littorea on cyprid settlement.

Examination of the effect of the presence of another

non-predatory snail species (Gibbula cineraria) on

cyprid settlement has revealed no effect (Turner and

Todd, 1991; Gosselin and Qian, 1996). Nevertheless, it

should be noted that the role of predatory gastropod

species (e.g. Nucella lapillus) in determining barnacle

population structure has been well documented in the

literature (for examples see Dayton, 1971; Miller and

Carefoot, 1989; Proud, 1994).

The results presented here determined that cyprid

mortality was greatest on treatment C (i.e. L.

littorea only) with no differences occurring between

the other treatments. For this study, it appears that

the levels of cyprid mortality through limpet

presence are negligible and that instead, it is the

presence of L. littorea in isolation that has a

dramatic effect upon cyprid survival (see Busch-

baum, 2000, 2002). Although limpets can graze and


consume cyprids (Jones, 1948; Denley and Under-

wood, 1979; Miller and Carefoot, 1989) L. littorea

with its taenioglossan radula is physically unable to

graze cyprids from a shore surface (Steneck and

Watling, 1982). Bertness et al. (1983) and Busch-

baum (2000, 2002) have recorded similar results to

the ones found here for the effect of the presence of

periwinkles on cyprid mortality. They concluded

that the increase in mortality was attributable to the

biological disturbance of the cyprids by the peri-

winkles. Alternatively, the pedal mucus or some

metabolite produced by L. littorea may be effecting

cyprid settlement. However, although Johnson and

Strathmann (1989) and Proud (1994) have both

shown that the pedal mucus of some prosobranch

species can have a negative effect upon cyprid

settlement, both Proud (1994) and Holmes (2002)

have found that a surface conditioned with the pedal

mucus produced by L. littorea can increase S.

balanoides cyprid settlement, in addition to which

Holmes (2002) recorded no effect for various pedal

extracts of L. littorea on settlement (see also

Raimondi, 1988; Johnson and Strathmann, 1989).

Similarly, Wahl and Sonnichsen (1992) have found

that the pedal mucus produced by L. littorea, and or

the presence of P. vulgata in the absence of grazing,

has no observable biocidal effect upon algal

propagule settlement and development. Therefore,

given that as shown here L. littorea can have a

negative effect on cyprid settlement, the observed

affect of their presence on cyprid settlement cannot

originate from the pedal mucus produced by L.

littorea, and hence must either originate from

biological disturbance, the most likely explanation,

or from some metabolite produced by them.

The studies of Jones (1948), Southward (1956)

and Lubchenco (1978, 1982, 1984) have all shown

that grazing by either P. vulgata or L. littorea can

lead to the formation of distinctly different algal

assemblages. In line with these findings, analysis of

the index of algal biomass for each treatment

revealed differences in the chlorophyll-a concen-

tration between treatments. That is, treatments B

(limpets and periwinkles) and C (periwinkles only)

had a consistently lower algal biomass, for the

whole experimental period, than treatments A

(limpets only) or D (no animals), but not from

each other. However, in contrast to the effects of L.

littorea, either in isolation or in combination with P.

vulgata, P. vulgata appeared to have little effect on

the algal biomass of a treatment. The effective role

of L. littorea in reducing algal biomass has been

well documented in the literature (Lubchenco, 1978,

1982, 1984; Jernakoff, 1983, 1985; Geller, 1991;

see Norton et al., 1990 for review). However, in

contrast to the results recorded here for P. vulgata,

many authors have ascribed a similar role to limpets

(see Jones, 1948; Lodge, 1948; Southward, 1956;

Connell, 1961; see Branch, 1981; Hawkins and

Hartnoll, 1983 for reviews). Although limpets are

extremely important in shaping algal successional

and colonisation processes, their role in determining

algal biomass will fundamentally depend upon the

four factors previously ascribed for the potential of

limpets to act as a biological disturbance on CNP

species. Note that Dayton (1975), Lubchenco and

Menge (1978), Jernakoff (1983, 1985) and Geller

(1991) have recorded no effect of limpets on algal

biomass.

If it is assumed that the method for measuring the

algal biomass of the treatments was reasonably

accurate, then the differences in algal biomass

recorded between the treatments can be explained as

follows. The results have shown that both the biomass

(i.e. 400 ind m�2 as opposed to ~26 ind m�2) and the

area browsed (5.25 m2 d�1 as opposed to 0.32 m2

d�1) in a treatment by L. littorea is much greater than

the corresponding biomass and area browsed by P.

vulgata. Given that the metabolic rate of L. littorea is

greater than that of P. vulgata (for L. littorea see

Newell and Pye, 1971; Shumway et al., 1993; and for

P. vulgata see Gompel, 1937; Davies, 1966; Jones,

1968), it is likely that the periwinkles are grazing the

majority of algae both recruited to, and growing on

the surface of the treatments containing them. P.

vulgata in contrast, although having a slight effect on

the algal biomass of the treatments is consuming,

relative to the periwinkles, little of the algae both

recruited to and growing on the surface of a treatment.

This is illustrated by the inability of P. vulgata to

reduce the algal biomass of treatments A and B below

those of treatments D and C, respectively.

As outlined, grazers can affect CNP species

settlement and success directly through biological

disturbance (Hatton, 1938; Connell, 1961; Dayton,

1971; Menge, 1976; Miller and Carefoot, 1989;


Lohse, 1993; Berlow and Navarrete, 1997) and

indirectly from the following sources:

(1) by reducing the algal biomass of a substratum

(Jones, 1948; Lodge, 1948; Burrows and Lodge,

1950; Southward, 1956; Dethier and Duggins,

1984; Van Tamelen, 1987);

(2) by generating microalgal assemblages, on the

substratum, which are either inhibitory or

stimulatory to cyprid settlement (Strathmann

and Branscomb, 1979; Strathmann et al., 1981;

Hudon et al., 1983; Maki et al., 1988, 1990,

1992; Wieczorek et al., 1995);

(3) through their pedal mucus trails (Raimondi,

1988; Johnson and Strathmann, 1989; Proud,

1994; Holmes, 2002; Holmes et al., 2002).

A diagram of both the nature/source of these

interactions and their direction (i.e. positive or

negative) is given in Fig. 9.

An additional source of an effect so far not

considered in this paper is that of the potential effects

of grazer-produced metabolites on cyprid settlement.

That is, the studies of De Silva (1962), Williams

Fig. 9. Schematic diagram for all of the possible direct and indirect intera

(1964), Gee (1965), Hadfield and Scheuer (1985),

Morse (1991) and Mokady et al. (1992) have shown

that the presence of one species can have an indirect

effect on the settlement of another species, through

some chemotaxic effect of the latter species’ metab-

olites. Although discussion will be made of the

potential effects of such metabolites there are no

observations within the literature pertaining to any

effect of such metabolites produced by P. vulgata or

L. littorea.

From the results (i.e. the analysis of the combined

effects of algal biomass and experimental treatments

on cyprid settlement) it was ascertained that the algal

biomass of the treatments had no effect on cyprid

settlement. Correspondingly, from the discussion on

cyprid mortality it was determined that the negative

effect of the presence of L. littorea on cyprid

settlement may originate from biological disturbance

(see Fig. 10). However, in consideration of the

possible biocidal/chemotaxic affects of metabolites

produced by L. littorea, if the metabolites were

responsible for the observed effects, then cyprid

settlement on Site 2 for treatment C should be about

half of the corresponding treatment on Site 1. N.B

ctions for the effect of the presence of grazers on cyprid settlement.

Fig. 10. Schematic diagram outlining the possible individual effects, both direct and indirect, of each grazer species on cyprid settlement.


assuming that the biomass and quantity of pedal

mucus produced by Site 2 periwinkles are about twice

that produced by the periwinkles from Site 1, and

hence that L. littorea from Site 2 produce about twice

the amount of metabolites. As this is evidently not the

case, it is reasonable to assume that biological

disturbance alone, in line with the observations of

Bertness et al. (1983) and Buschbaum (2000, 2002),

can explain the observed effects for the presence of L.

littorea (see also Proud, 1994).

In contrast to L. littorea, the positive effect of the

presence of P. vulgata on cyprid settlement must

originate from either their pedal mucus trails and/or

some positive chemotactic effect of their metabolites

on cyprid settlement (see Fig. 8). That is, the algal

biomass of the treatments had no effect on cyprid

settlement and in addition, biological disturbance by

the limpets would produce a negative rather than

positive effect. Raimondi (1988), Johnson and

Strathmann (1989), Proud (1994) and Holmes


(2002) have all shown that the pedal mucus

produced by some molluscs can have a positive

effect on cyprid settlement (see also Holmes et al.,

2002). In particular, Holmes (2002) has shown that

the pedal mucus produced by P. vulgata can increase

cyprid settlement by a factor of ~ 6 in the laboratory

and by a factor of ~ 4 in the field, which would

suggest that the pedal mucus trails left behind by P.

vulgata are acting as the major source of stimulus

(see also Holmes et al., 2002).

Johnson and Strathmann (1989) and Proud (1994)

have suggested that the source of effect of pedal

mucus on settlement is chemotactic in origin. Alter-

natively, Zobell (1938, 1939a), Holmes (2002) and

Holmes et al. (2002) have suggested that the mucoid

matrices produced by organisms may increase the

settlement of organisms by physically entrapping

them in the mucus matrix (see also Angst, 1923;

Zobell and Allen, 1933, 1935; Zobell, 1939b).

Ignoring any potential chemotaxic effects of either

the pedal mucus and/or metabolites, if the pedal

mucus produced by P. vulgata is acting in the

manner described by Holmes (2002) and Holmes et

al. (2002), then the presence of P. vulgata in a

treatment would increase the likelihood of cyprid

settlement by increasing the contact time for the

cyprids with the substratum (see Crisp, 1974;

Rittschof et al., 1984; Yule and Walker, 1984,

1987). Similarly, for the mixed treatment plots, the

adhesive effect of P. vulgata pedal mucus on cyprids

would afford the cyprids some protection from

dislodgement by browsing L. littorea. If the increase

in cyprid settlement for treatments containing P.

vulgata does originate from either some chemotactic

property of pedal mucus and/or the metabolites

produced by P. vulgata, the strength of this cue

acting on cyprid settlement is greater than the

amount of biological disturbance generated by L.

littorea. That is, the autogenic ecosystem engineering

effects of P. vulgata far outweigh the allogenic

ecosystem engineering effects of L. littorea. How-

ever, for the experiments in this paper it is

impossible to distinguish between the potential effect

of pedal mucus and those of metabolites produced

by P. vulgata. Whatever the mechanism is of

increased cyprid settlement for the presence of P.

vulgata, any source of increased cyprid contact with

a surface will result in a positive feedback settlement

cascade reflected in successively greater numbers of

cyprids settling on the surface.

In summary, P. vulgata acts indirectly to increase

S. balanoides cyprid settlement and success. In

contrast, L. littorea acts directly to decrease S.

balanoides cyprid settlement and success through

biological disturbance. The initial effects of these

animals on cyprid settlement are reflected in the long-

term community structure. Such results are important,

not only for rocky shore ecology, but for ecology in

general as they illustrate the influence that previously

overlooked indirect effects can have on both deter-

mining the structure of community and ameliorating

direct effects, which have previously been considered

the dominant structuring factor.

Acknowledgements

We would like to thank Professor J. Davenport

for the generous use of the facilities at Millport

Marine Station and Ms. C. Sturgess for her

assistance with the field work and counting the

barnacles. In addition we would like to note that this

research was funded by the Ecology Centre, Sunder-

land, UK.

References

Angst, E.C., 1923. The Fouling of Ship Bottoms by Bacteria.

United States Navy Department, Washington.

Barnes, H., Powell, H.T., 1950. The development, general

morphology and subsequent elimination of barnacle popula-

tions, Balanus crenatus and Balanus balanoides, after a heavy

initial settlement. J. Anim. Ecol. 19, 175–179.

Berlow, E.L., Navarrete, S.A., 1997. Spatial and temporal variation

in intertidal community organisation: Lessons from repeating

field experiments. J. Exp. Mar. Biol. Ecol. 214, 195–229.

Bertness, M.D., Yund, P.O., Brown, A.F., 1983. Snail grazing and

the abundance of algal crusts on a sheltered New-England rocky

beach. J. Exp. Mar. Biol. Ecol. 71, 147–164.

Bertness, M.D., Gaines, S.D., Stephens, E.G., Yund, P.O., 1992.

Components of recruitment in populations of the acorn barnacle

Semibalanus balanoides (Linnaeus). J. Exp. Mar. Biol. Ecol.

156, 199–215.

Branch, G.M., 1981. The biology of limpets: physical factors,

energy flow and ecological interactions. Ann. Rev. Ocean. Mar.

Biol. 19, 235–380.

Buschbaum, C., 2000. Direct and indirect effects of Littorina

littorea (L.) on barnacles growing on mussel beds in the

Wadden Sea. Hydrobiologia 440, 119–128.


Buschbaum, C., 2002. Predation on barnacles of intertidal and

subtidal mussel beds in the Wadden Sea. Helgol. Mar. Res. 56,

37–43.

Burrows, E.M., Lodge, S.M., 1950. Notes on the inter-relationships

of Patella, Balanus, and Fucus on a semi-exposed coast. Rep.

Mar. Biol. St. Port Erin 62, 30–34.

Choat, J.H., 1977. The influence of sessile organisms on the

population biology of three species of acmaeid limpets. J. Exp.

Mar. Biol. Ecol. 26, 1–26.

Clare, A.S., Freet, R.K., McClary, M., 1994. On the antennular

secretion of the cyprid of Balanus amphitrite amphitrite, and

its role as a settlement pheromone. J. Mar. Biol. Ass. UK 74,

243–250.

Clements, F.E., 1916. Plant succession, Publication vol. 242.

Carnegie Institute of Washington, Washington, District of

Columbia.

Connell, J.H., 1961. Effects of competition, predation by Thais

lapillus, and other factors on natural populations of the barnacle

Balanus balanoides. Ecol. Monogr. 31, 61–104.

Connell, J.H., Slayter, R.O., 1977. Mechanisms of succession in

natural communities and their role in community stability and

organisation. Am. Nat. 111, 1119–1144.

Crisp, D.J., 1961. Territorial behaviour in barnacle settlement. J.

Exp. Biol. 38, 429–446.

Crisp, D.J., 1974. Factors influencing the settlement of marine

invertebrate larvae. In: Grant, P.T., Mackie, A.M. (Eds.),

Chemoreception in Marine Organisms. Academic Press, Lon-

don, pp. 177–265.

Cubit, J.D., 1984. Herbivory and seasonal abundance of algae on a

high rocky intertidal shore. Ecology 65, 1904–1917.

Davies, P.S., 1966. Physiological ecology of Patella. I. the effect of

body size and temperature on metabolic rate. J. Mar. Biol. Ass.

UK 46, 647–658.

Davies, M.S., Hawkins, S.J., Jones, H.D., 1990a. Mucus production

and physiological energetics in Patella vulgata L. J. Moll. Stud.

56, 499–503.

Davies, M.S., Jones, H.D., Hawkins, S.J., 1990b. Seasonal variation

in the composition of pedal mucus from Patella vulgata L.

J. Exp. Mar. Biol. Ecol. 144, 101–112.

Davies, M.S., Hawkins, S.J., Jones, H.D., 1992. Pedal mucus and its

influence on the microbial food-supply of 2 intertidal gastro-

pods, Patella vulgata L and Littorina littorea (L). J. Exp. Mar.

Biol. Ecol. 161, 57–77.

Dayton, P.K., 1971. Competition, disturbance, and community

organisation: the provision and subsequent utilisation of

space in a rocky intertidal community. Ecol. Monogr. 41,

351–388.

Dayton, P.K., 1975. Experimental evaluation of ecological domi-

nance in a rocky intertidal algal community. Ecol. Monogr. 45,

137–159.

De Wolf, P., 1973. Ecological observations on the mechanisms of

dispersal of barnacle larvae during planktonic life and settling.

Neth. J. Sea. Res. 6, 1–129.

Dean, R.L., Connell, J.H., 1987. Marine-invertebrates in an algal

succession. 2. Tests of hypotheses to explain changes in

diversity with succession. J. Exp. Mar. Biol. Ecol. 109,

217–247.

Denley, E.J., Underwood, A.J., 1979. Experiments on factors

influencing settlement, survival, and growth of two species of

barnacles in New South Wales. J. Exp. Mar. Biol. Ecol. 36,

269–293.

De Silva, P.H.D.H., 1962. Experiments on choice of substrata by

Spirorbis larvae (Serpulidae). J. Exp. Mar. Biol. Ecol. 39,

483–490.

Dethier, M.N., Duggins, D.O., 1984. An indirect commensalism

between marine herbivores and the importance of competitive

hierarchies. Am. Nat. 124, 205–219.

Dineen, J.F., Hines, A.H., 1992. Interactive effects of salinity and

adult extract upon settlement of the estuarine barnacle

Balanus improvisus (Darwin, 1854). J. Exp. Mar. Biol. Ecol.

156, 239–252.

Dugan, M.L., 1986. Three-way interactions: barnacles, limpets, and

algae in a Sonoran Desert rocky intertidal community. Am. Nat.

127, 293–316.

Farrell, T.M., 1988. Community stability: effects of limpet removal

and reintroduction in a rocky intertidal community. Oceanologia

75, 190–197.

Farrell, T.M., 1991. Models and mechanisms of succession, An

example from a rocky intertidal community. Ecol. Monogr. 61,

95–113.

Fath, B.D., Pattern, B.C., 1998. Network synergism: emergence of

positive relations in ecological systems. Ecol. Model. 107,

127–143.

Fath, B.D., Pattern, B.C., 1999. Review of the foundations of

network environ analysis. Ecosystems 2, 167–199.

Gaines, S.D., Lubchenco, J., 1982. A unified approach to marine

plant-herbivore interactions. 2. Biogeography. Ann. Rev. Eco.

Syst. 13, 111–138.

Gee, J.M., 1965. Chemical stimulation of settlement in

larvae of Spirorbis rupestris (Serpulidae). Anim. Behav. 13,

181–186.

Geller, J.B., 1991. Gastropod grazers and algal colonisation on a

rocky shore in Northern California, the importance of the body

size of grazers. J. Exp. Mar. Biol. Ecol. 150, 1–17.

Gompel, M., 1937. Recherches sur la consommation d’oxygene de

quelques animaux aquatiques littoraux. C.R. Academia of

Science, Paris 205, 816–819.

Gosselin, L.A., Qian, P.Y., 1996. Early postsettlement mortality of

an intertidal barnacle, a critical period for survival. Mar. Ecol.

Prog. Ser. 135, 69–75.

Grenon, J.F., Walker, G., 1980. Biochemical and rheological

properties of the pedal mucus of the limpet Patella vulgata L.

Comp. Biochem. Phys. 66, 451–458.

Hadfield, M.G., Scheuer, D., 1985. Evidence for a soluble

metamorphic inducer in Phestilla: ecological, chemical and

biological data. Bull. Mar. Sci. 37, 556–566.

Hall, D., Cooper, W., Werner, E.E., 1970. An experimental

approach to the production dynamics and structure of fresh-

water animal communities. Limnol. Oceangr. 15, 839–928.

Hartnoll, R.G., Hawkins, S.J., 1985. Patchiness and fluctuations on

moderately exposed rocky shores. Ophelia 24, 53–63.

Hatton, H., 1938. Essais de bionomie explicative sur quelques

especes intercotidales d’algues et d’animaux. Ann. Inst. Ocean.,

Monaco 17, 241–348.


Hawkins, S.J., 1981. The influence of Patella grazing on the fucoid/

barnacle mosaic on moderately exposed rocky shores. Kiel.

Meer. Sond. 5, 537–543.

Hawkins, S.J., 1983. Interactions of Patella and macroalgae with

settling Semibalanus balanoides (L). J. Exp. Mar. Biol. Ecol.

71, 55–72.

Hawkins, S.J., Hartnoll, R.G., 1983. Grazing of intertidal algae

by marine-invertebrates. Ann. Rev. Ocean. Mar. Biol. 21,

195–282.

Hill, A.S., Hawkins, S.J., 1990. An investigation of methods for

sampling microbial films on rocky shores. J. Mar. Biol. Ass. UK

70, 77–88.

HMSO, 1986. The determination of chlorophyll a in aquatic

environments. Methods for the Examination of Water and

Associated Materials vol. 4.. HMSO, London, pp. 51–109.

Holmes, S.P., 2002. The effect of pedal mucus on barnacle cyprid

settlement: a source for indirect interactions in the rocky

intertidal? J. Mar. Biol. Ass. UK 82, 1–13.

Holmes, S.P., Sturgess, C.J., Davies, M.S., 1997. The effect of rock-

type on the settlement of Balanus balanoides (L.) cyprids.

Biofouling 11, 137–147.

Holmes, S.P., Cherrill, A., Davies, M.S., 2002. The surface

characteristics of pedal mucus: a potential aid to the settlement

of marine organisms? J. Mar. Biol. Ass. UK 82, 131–139.

Hudon, C., Bourget, E., Legendre, P., 1983. An integrated study of

the factors influencing the choice of the settling site of

Balanus crenatus cyprid larvae. Can. J. Fish. Aquat. Sci. 40,

1186–1194.

Hunter, R.D., Russell-Hunter, W.D., 1983. Bioenergetic and

community changes in intertidal aufwuchs grazed by Littorina

littorea. Ecology 64, 761–769.

Jarrett, J.N., 1997. Temporal variation in substratum specificity of

Semibalanus balanoides (Linnaeus) cyprids. J. Exp. Mar. Biol.

Ecol. 211, 103–114.

Jernakoff, P., 1983. Factors affecting the recruitment of algae in a

midshore region dominated by barnacles. J. Exp. Mar. Biol.

Ecol. 67, 17–31.

Jernakoff, P., 1985. An experimental evaluation of the influence of

barnacles, crevices and seasonal patterns of grazing on algal

diversity and cover in an intertidal barnacle zone. J. Exp. Mar.

Biol. Ecol. 88, 287–302.

Johnson, L.E., Strathmann, R.R., 1989. Settling barnacle larvae

avoid substrata previously occupied by a mobile predator.


Jones, N.S., 1948. Observations on the biology of Patella vulgata

at Port St. Mary, Isle of Man. Proc. Liverpool Biol. Soc. 56,

60–77.

Jones, H.D., 1968. Some aspects of heart function in Patella

vulgata L. Nature 217, 1170–1172.

Jones, C.G., Lawton, J.H., Shachakm, M., 1994. Organisms as

ecosystem engineers. Okios 69, 373–386.

Kim, J.H., 1997. The role of herbivory, and direct and indirect

interactions, in algal succession. J. Exp. Mar. Biol. Ecol. 217,

119–135.

Knight-Jones, EW., 1953. Laboratory experiments on gregarious-

ness during setting in Balanus balanoides and other barnacles.

J. Exp. Biol. 30, 584–599.

Larman, V.N., Gabbott, P.A., 1975. Settlement of cyprid larvae of

Balanus balanoides and Elminius modestus induced by extracts

of adult barnacles and other marine animals. J. Mar. Biol. Ass.

UK 55, 183–190.

Lewis, J.R., 1954. Observations on a high-level population of

limpets. J. Anim. Ecol. 23, 85–100.

Lewis, C.A., 1978. A review of substratum selection in free-living

and symbiotic cirripeds. In: Chia, F.-S., Rice, M.E. (Eds.),

Settlement and Metamorphosis of Marine Invertebrate Larvae.

Elsevier, New York, pp. 219–234.

Lodge, S.M., 1948. Algal growth in the absence of Patella on an

experimental ship in the foreshore, Port St. Mary, Isle of Man.

Proc. Liverpool Biol. Soc. 56, 78–85.

Lohse, D.P., 1993. The effects of substratum type on the population

dynamics of three common intertidal animals. J. Exp. Mar. Biol.

Ecol. 173, 133–154.

Lubchenco, J., 1978. Plant species diversity in a marine intertidal

community: importance of herbivore food preference and algal

competitive abilities. Am. Nat. 112, 23–39.

Lubchenco, J., 1982. Effects of grazers and algal competitors on

fucoid colonisation in tide pools. J. Phycol. 18, 544–550.

Lubchenco, J., 1984. Relative importance of competition vs

predation during early seaweed succession in New England.

Am. Zoo. 24, 129.

Lubchenco, J., Gaines, S.D., 1981. A unified approach to marine

plant-herbivore interactions. 1. Populations and communities.

Ann. Rev. Eco. Syst. 12, 405–437.

Lubchenco, J., Menge, B.A., 1978. Community development and

persistence in a low rocky intertidal zone. Ecol. Monogr. 48,

67–94.

Lucas, M.I., Walker, G., Holland, D.L., Crisp, D.J., 1979. An

energy budget for the free-swimming and metamorphosing

larvae of Balanus balanoides (Crustacea: Cirripedia). Mar. Biol.

55, 221–229.

Maki, J.S., Rittschof, D., Costlow, J.D., Mitchell, R., 1988.

Inhibition of attachment of larval barnacles, Balanus amphitrite,

by bacterial surface-films. Mar. Biol. 97, 199–206.

Maki, J.S., Rittschof, D., Samuelsson, M.O., Szewzyk, U., Yule,

A.B., Kjelleberg, S., Costlow, J.D., Mitchell, R., 1990. Effect of

marine-bacteria and their exopolymers on the attachment of

barnacle cypris larvae. Bull. Mar. Sci. 46, 499–511.

Maki, J.S., Rittschof, D., Mitchell, R., 1992. Inhibition of larval

barnacle attachment to bacterial films: an investigation of

physical factors. Micro. Ecol. 23, 97–106.

Menge, B.A., 1976. Organisation of the New England Rocky

intertidal community: role of predation, competition and

environmental heterogeneity. Ecol. Monogr. 46, 355–393.

Menge, B.A., 1991. Relative importance of recruitment and other

causes of variation in rocky intertidal community structure.


Menge, B.A., 1995. Indirect effects in marine rocky intertidal

interaction webs: patterns and importance. Ecol. Monogr. 65,

21–74.

Menge, B.A., Lubchenco, J., 1981. Community organisation in

temperate and tropical rocky inter-tidal habitats—prey refuges

in relation to consumer pressure-gradients. Ecol. Monogr. 51,

429–450.


Menge, B.A., Lubchenco, J., Ashkenas, L.R., Ramsey, F., 1986.

Experimental separation of effects of consumers on sessile prey

in the low zone of a rocky shore in the bay of Panama - direct

and indirect consequences of food web complexity. J. Exp. Mar.

Biol. Ecol. 100, 225–269.

Menge, B.A., Berlow, E.L., Blanchette, C.A., Navarrete, S.A.,

Yamada, S.B., 1994. The keystone species concept - variation in

interaction strength in a rocky intertidal habitat. Ecol. Monogr.

64, 249–286.

Miller, K.M., Carefoot, T.H., 1989. The role of spatial and size

refuges in the interaction between juvenile barnacles and

grazing limpets. J. Exp. Mar. Biol. Ecol. 134, 157–174.

Mokady, O., Arazi, G., Bonar, B., Loya, Y., 1992. Settlement and

metamorphosis specificity of Lithophaga simplex Iredale

(Bivalvia: Mytilidae) on Red Sea corals. J. Exp. Mar. Biol.

Ecol. 162, 243–251.

Morse, A.N.C., 1991. How do planktonic larvae know where to

settle. Am. Sci. 79, 154–167.

Moyse, J., Hui, E., 1981. Avoidance by Balanus balanoides cyprids

of settlement on conspecific adults. J. Mar. Biol. Ass. UK 61,

449–460.

Moyse, J., Nelson-Smith, A., 1963. Zonation of animals and plants

on rocky shores around Dale, Pembrokeshire. Field Stud. 1,

1–31.

Newell, R.C., Pye, V.I., 1971. Variations in the relationship between

oxygen consumption, body size and summated tissue metabo-

lism in the winkle Littorina littorea. J. Mar. Biol. Ass. UK 51,

315–338.

Norton, T.A., Hawkins, S.J., Manley, N.L., Williams, G.A., Watson,

D.C., 1990. Scraping a living—a review of littorinid grazing.

Hydrobiologia 193, 117–138.

Paine, R.T., 1966. Food web complexity and species diversity. Am.

Nat. 100, 65–75.

Paine, R.T., 1969. The Pisaster-Tegula interaction: prey patches,

predator food preference, and intertidal community structure.

Ecology 50, 950–961.

Paine, R.T., 1981. Barnacle ecology—is competition important—

the forgotten roles of disturbance and predation. Paleobiology 7,

553–560.

Posey, M.H., Hines, A.H., 1991. Complex predator-prey interac-

tions within an estuarine benthic community. Ecology 72,

2155–2169.

Proud, S.V., 1994. Tributylin pollution and the bioindicator Nucella

lapillus: population recovery and community level responses,

Ph.D. Dissertation. University of Liverpool, Liverpool.

Pyefinch, K.A., 1948. Notes on the biology of cirripedes. J. Mar.

Biol. Ass. UK 27, 464–503.

Raimondi, P.T., 1988. Settlement cues and determination of

the vertical limit of an intertidal barnacle. Ecology 69,

400–407.

Rittschof, D., Branscomb, E.S., Costlow, J.D., 1984. Settlement

and behaviour in relation to flow and surface in larval

barnacles, Balanus amphitrite Darwin. J. Exp. Mar. Biol.

Ecol. 82, 131–146.

Safriel, U.N., Erez, N., Keasar, T., 1994. How do limpets maintain

barnacle-free submerged artificial surfaces. Bull. Mar. Sci. 54,

17–23.

Shumway, S.E., Lesser, M.P., Crisp, D.J., 1993. Specific dynamic

action demonstrated in the herbivorous marine periwinkles,

Littorina littorea L and Littorina obtusata L (Mollusca,

Gastropoda). Comp. Biochem. Phys. 106, 391–395.

Sokal, R.R., Rohlf, J.F., 1981. Biometry. W.H. Freeman & Co.,

New York.

Southward, A.J., 1956. The population balance between limpets and

seaweeds on wave-beaten rocky shores. Rep. Mar. Biol. St. Port

Erin 68, 20–29.

Southward, A.J., 1958. The zonation of plants and animals on rocky

sea shores. Biol. Rev. 33, 137–177.

Steneck, R.S., Watling, L., 1982. Feeding capabilities and limitation

of herbivorous molluscs: a functional group approach. Mar.

Biol. 68, 299–319.

Strathmann, R.R., Branscomb, E.S., 1979. Adequacy of cues

to favourable sites used by settling larvae of two

intertidal barnacles. In: Stancyzk, S.E. (Ed.), Reproductive

Ecology of Marine Invertebrates. University of South

Carolina, pp. 77–89.

Strathmann, R.R., Branscomb, E.S., Vedder, K., 1981. Fatal errors

in set as a cost of dispersal and the influence of intertidal flora

on set of barnacles. Oceanologia 48, 13–18.

Turner, S.J., Todd, C., 1991. The effects of Gibbula cineraria (L.)

and Asterias rubens L. on developing epifaunal assemblages.


Van Tamelen, P.G., 1987. Early successional mechanisms in the

rocky intertidal. The role of direct and indirect interactions.


Wahl, M., Sonnichsen, H., 1992. Marine epibiosis IV. The

periwinkle Littorina littorea lacks typical antifouling defences

why are some populations so little fouled? Mar. Ecol. Prog. Ser.

88, 225–235.

Wethey, D.S., 1983. Spatial pattern in barnacle settlement—

daily changes during the settlement season. Am. Zoo. 23,

1021.

Wethey, D.S., 1984. Spatial pattern in barnacle settlement: day to

day changes during the settlement season. J. Mar. Biol. Ass. UK

64, 687–698.

Wethey, D.S., 1986. Ranking of settlement cues by barnacle

larvae—influence of surface contour. Bull. Mar. Sci. 39,

393–400.

Wieczorek, S.K., Clare, A.S., Todd, C.D., 1995. Inhibitory and

facilitatory effects of microbial films on settlement of

Balanus amphitrite amphitrite larvae. Mar. Ecol. Prog. Ser.

119, 221–228.

Wilhelmsen, U., Reise, K., 1994. Grazing on green-algae by the

periwinkle Littorina littorea in the Wadden Sea. Helgol7nderMeeresunters 48, 233–242.

Williams, G.B., 1964. The effect of extracts of Fucus serratus in

promoting the settlement of larvae of Spirorbis borealis

(Polychaeta). J. Mar. Biol. Ass. UK 44, 397–414.

Yule, A.B., Walker, G., 1984. The temporary adhesion of barnacle

cyprids: effects of some differing surface characteristics. J. Mar.

Biol. Ass. UK 64, 429–439.

Yule, A.B., Walker, G., 1987. Adhesion in barnacles. Southward,

A.J. Crustacean Issues Barnacle Biology vol. 5. A.A. Balkema,

Rotterdam, pp. 389–404.


Zobell, C.E., 1938. The sequence of events in the fouling of

submerged surfaces. Fed. Paint, Varnish Prod. Clubs 178,

379–385.

Zobell, C.E., 1939a. The role of bacteria in the fouling of

submerged surfaces. Biol. Bull. 77, 302.

Zobell, C.E., 1939b. Fouling of submerged surfaces and possible

preventive procedures. The biological approach to the

preparation of antifouling paints. Paint, Oil, Chem. Rev.

101, 74–77.

Zobell, C.E., Allen, E.C., 1933. Attachment of bacteria to

submerged slides. Proc. Soc. Experim. Biol. Med. 30,

1409–1411.

Zobell, C.E., Allen, E.C., 1935. The significance of marine bacteria

in the fouling of submerged surfaces. J. Bact. 29, 239–251.

barnacles, limpets and periwinkles: the effects of direct and indirect interactions on cyprid...

Documents