Download - Quantification of radular marks as a method for estimating grazing of intertidal gastropods on rocky shores

Journal of Experimental Marine Biology and EcologyŽ .258 2001 155–171

www.elsevier.nlrlocaterjembe

Quantification of radular marks as a method forestimating grazing of intertidal gastropods on

rocky shores

R.E. Forrest, M.G. Chapman), A.J. UnderwoodCentre for Research on Ecological Impacts of Coastal Cities, Marine Ecology Laboratories, A11,

UniÕersity of Sydney, Sydney, NSW 2006, Australia

Received 25 May 2000; received in revised form 8 December 2000; accepted 5 January 2001

Abstract

Wax discs have been used previously on intertidal rocky shores to record the grazing activityof gastropods. This study has evaluated this methodology for recording grazing of four commonintertidal microalgal grazers on intertidal shores in New South Wales, Australia. In the laboratory,

Ž .the four species examined—the patellid limpet, Cellana tramoserica Sowerby , the trochid,Ž .Austrocochlea porcata A. Adams , the neritid, Nerita atramentosa Reeve and the littorinid,

Ž .Bembicium nanum Lamarck —made distinctive marks in the wax. These allowed identificationof each species or combinations of species grazing over the different discs. Field experimentsshowed that the intensity of grazing, as indicated by the mean number of scratches per disc, waspositively related to the number of gastropods in the surrounding area during low tide for C.tramoserica. The number of scratches per disc in any area was correlated with the percentage of

Ždiscs scratched. The relationship for C. tramoserica was found at two scales—in sites approxi-. Ž .mately 3=3 m and also in plots 50=50 cm within sites. Therefore, densities that were

measured when these limpets were inactive during low tide provided good estimates of grazingactivity during high tide. This is largely because these limpets do not move far between wherethey rest and where they feed. The amount of microalgal food in the vicinity was not correlatedwith density, nor with grazing intensity. No relationship between density and grazing intensity was

Žfound for N. atramentosa, although experiments were only done in the field at one spatial scale in.sites, 3=3 m . Results obtained in the laboratory and in the field show that wax discs are useful

to distinguish grazing by different species of gastropods on Australian rocky shores and allowtests of hypotheses about grazing activity at different spatial scales. q 2001 Elsevier Science B.V.All rights reserved.

Keywords: Gastropods; Grazing; Microalgae; Quantification; Rocky shores

) Corresponding author. Tel.: q61-2-9351-4778; fax: q61-2-9351-6713.Ž .E-mail address: [email protected] M.G. Chapman .

0022-0981r01r$ - see front matter q2001 Elsevier Science B.V. All rights reserved.Ž .PII: S0022-0981 01 00212-X

( )R.E. Forrest et al.rJ. Exp. Mar. Biol. Ecol. 258 2001 155–171156

1. Introduction

Grazing by large herbivorous gastropods on intertidal shores has important effects onŽthe distribution and abundance of algae Castenholz, 1961; Nicotri, 1977; Lubchenco,

1980, 1983; Underwood, 1980, 1984a; Lubchenco and Gaines, 1981; Underwood and.Jernakoff, 1981, 1984; Hawkins and Hartnoll, 1983; Jernakoff, 1985a,b . The cover of

algae, in turn, influences many other components of intertidal assemblages, for example,Žby providing shelter from desiccation during low tide Hruby and Norton, 1979;

. ŽBrawley and Johnson, 1991 , overgrowing and killing other sessile organisms Connell,.1961; Dayton, 1971; Denley and Underwood, 1979; Jernakoff, 1985c , modifying

Ž . Žecological processes, such as competition Kastendiek, 1982 or predation Underwood,. Ž .1999 , or pre-empting space and thereby preventing recruitment Hawkins, 1983 .

Because of such strong direct and indirect effects of grazing, there is much interest inspatial and temporal patterns of distribution and activity of intertidal gastropods. Grazingbehaviour is not, however, easy to measure directly with many levels of spatial andtemporal replication. Most procedures that have been used to measure grazing directly

Ž .are time-consuming e.g. Boyden and Zeldis, 1979; Kitting, 1979; Petraitis, 1992 andonly allow relatively few animals to be measured during any single period of foraging.Therefore, the intensity of grazing in an area has usually been estimated indirectly, bymeasuring the movement of animals, assuming that there are straightforward relation-

Ž .ships between movement and grazing. It is usually assumed e.g. Branch, 1971 thatwhen snails and limpets move, they are feeding.

A common, but even more indirect measure of grazing is the local density of grazersŽMackay and Underwood, 1977; Underwood, 1977; Chelazzi et al., 1983; Levings and

.Garrity, 1983; Chapman, 1995 . For logistic reasons, it is usual to record densitiesduring low tide when the animals are relatively inactive and the habitat is easily

Žaccessible e.g. Underwood, 1976a, 1978, 1984b; Creese, 1980; Creese and Underwood,.1982; Marshall and Keough, 1994 , even though many animals feed at other states of the

Ž .tide reviewed by Underwood, 1979; Hawkins and Hartnoll, 1983 . The density ofgrazers is not always a reliable measure of grazing in an area, especially for species that

Žfeed far from their positions during low tide Cook and Cook, 1981; Levings and.Garrity, 1983; Chelazzi et al., 1990; reviewed by Chapman and Underwood, 1992 .

ŽIntertidal gastropods are often patchily distributed at very small spatial scales cm–m;.Chapman, 1994; Underwood and Chapman, 1996, 1998 and this may result in

concentrated patterns of grazing, especially if the animals do not move far whenŽ .foraging Chapman, 1995 . Limited extent of movement while foraging may concentrate

Žgrazing within very small patches of habitat Mackay and Underwood, 1977; Under-.wood, 1977; Chapman, 2000a , either because of the distribution of particular microhab-

Ž .itats Creese, 1982; Chapman, 1995, 2000b or because of patchiness in the microalgalŽ .food supply Nicotri, 1977; Underwood, 1984a,b; MacLulich, 1987 . Therefore, if

density of animals is to be used to infer intensity of grazing, care must be taken tomeasure densities at appropriate spatial scales. Such scales are, however, not alwayseasy to identify.

Recently, a new method for recording grazing by gastropods in the field wasdeveloped, which can address directly the links between local densities, extent of

( )R.E. Forrest et al.rJ. Exp. Mar. Biol. Ecol. 258 2001 155–171 157

Ž .movement and intensity of grazing Thompson et al., 1997 . This method used waxdiscs, inserted into the rocky intertidal substratum, to record the radular scratches ofgrazing gastropods, thus providing a quantitative measure of grazing intensity. Becauseradular morphology varies among species, different species of gastropod can leaveidentifiable scratches in the surface of the wax. Discs can be placed in many sitessimultaneously and left for varying periods of time, thereby providing direct quantitativemeasures of grazing of a variety of species in different places and at different times, to

Ž .test directly various hypotheses about intensity of grazing Thompson et al., 1997 .This study examines the applicability of this technique for measuring grazing of

gastropods on Australian rocky shores. First, laboratory experiments were done to testŽ .the hypothesis that the patellid limpet, Cellana tramoserica Sowerby , the trochid,

Ž .Austrocochlea porcata A. Adams , the neritid, Nerita atramentosa Reeve and theŽ .littorinid, Bembicium nanum Lamarck , which have different types of radulae, would

leave distinctive and identifiable scratches when feeding. These species are among thelargest and most common intertidal herbivorous gastropods in New South Wales andhave been the subject of many previous investigations into their feeding behaviour,

Ž .movements and distribution e.g. Underwood, 1975, 1976a,b, 1977, 1978 . They feed inŽthe same areas on the same microalgal food, for which they often compete Underwood,

.1984b .Wax discs were also used in field experiments to test hypotheses about the relation-

ship between density of gastropods and intensity of grazing. The first field experimentexamined the model that densities of N. atramentosa and C. tramoserica, measuredduring low tide in sites approximately 3=3 m, would be reliable indirect estimates ofthe relative amount of grazing activity in these sites during high tides.

Ž . Ž .Microalgal food Underwood, 1984a,b , feeding Mackay and Underwood, 1977 andŽ .density of limpets Underwood and Chapman, 1996, 1998 are all patchy at small spatial

Ž .scales. The model was therefore proposed that density in very small plots 50=50 cmŽ .within various sites 3=3 m sites, as used before would correlate better with intensity

of grazing than was the case at the larger scale of sites. A second field experiment wasŽ .therefore done using C. tramoserica, to examine the models that: 1 local density

during low tide is a better estimate of grazing activity at small scales than at largerŽ .scales, 2 densities of inactive animals during low tide are a reliable measure of

Ž .densities of active feeding animals in the same area during high tide and 3 grazingintensity is related to the amount of microalgae in the immediate vicinity.

2. Materials and methods

2.1. Preparation of the wax discs

Ž .Plastic holders 14- mm diameter were placed upside-down on glass and filled withŽ . Ž . Žhot 958C dental wax Boral Investo via the hole in the base of the holder after

.Thompson et al., 1997 . To avoid uneven cooling of the wax, the glass was kept hot by


resting it over a bath of hot water. After the holders had been filled, the glass andholders were refrigerated for 1 h. They were then allowed to return to room temperature,after which, holders were removed from the glass, resulting in wax discs with a smooth,shiny surface. The wax-filled holders could then be inserted into holes drilled into therock-surface. The discs fitted snugly into the holes, with the surface of the wax flushwith the rock surface.

2.2. Laboratory experiments

The first laboratory experiment tested the hypothesis that different species of grazerswould leave distinctive and identifiable marks in the wax discs. Twelve wax discs were

Ž .embedded into the top surfaces of each of four sandstone plates 25=25 cm . One platewas placed into each of four aquaria, each containing 12 individuals of either C.tramoserica, N. atramentosa, B. nanum or A. porcata. Discs were removed from theaquaria after 7 days, and the experiment was repeated. The discs were examined for

Ž .scratches under a light microscope and also using scanning electron microscopy SEM .To test the hypothesis that marks from different species could be identified when discswere scratched by more than one species, a similar experiment was done usingcombinations of two of the four species in each aquarium. All six possible combinationsof two species were tested. Again, discs were removed after 7 days and examined undera light microscope. Finally, to test the hypothesis that the marks could be used reliablyto identify the species making the marks, blind tests were conducted with six volunteers.Volunteers were given a set of SEM images of the marks made by different species and20 discs that had been scraped by one or two species. They were asked to identify whichspecies had scraped each of the discs.

2.3. Field experiments

The experiments were done on intertidal shores in the Cape Banks Scientific MarineŽResearch Area, New South Wales, Australia described in detail in Underwood et al.,

.1983 . The first experiment tested the hypothesis that the number of scratches on discswould be significantly larger in places with large densities of gastropods, becausedensity was considered an indirect measure of the intensity of grazing in a given area.

ŽFor each of the two species, C. tramoserica and N. atramentosa, two sites each.approximately 3=3 m with large densities of animals and two sites with small

densities were chosen. Density was estimated from counts in 10 AlargeB randomlyŽ .placed quadrats 50=50 cm in each site. Densities of N. atramentosa were measured

on Days 1, 3 and 8 of the experiment; densities of C. tramoserica were measured onDays 1 and 6. Sites were at similar heights on the shore and did not noticeably differ interms of topographical complexity or cover of macroalgae. Sites varied in their distancesfrom one another, but the average distance between sites was approximately 10 m.

In each site, four arrays of nine regularly spaced discs were inserted into the rock.Each array was a grid with approximately 5-cm spacing between adjacent discs. Arrays


were approximately 1–3-m apart and were haphazardly placed on relatively flat patchesof substratum, which were free of foliose algae. In sites with N. atramentosa, discs wereplaced in each site on the same day and were left for 8 days. In sites with C.tramoserica, due to poor weather, discs could only be placed in one site with largedensity and one site with small density on the first day. These discs were collected after6 days, when the experiment was repeated in two new sites. This period of time waschosen because previous experiments with other species indicated that a period of 7–14

Ždays was optimal i.e. discs should be scratched but not yet saturated with scratches;.Thompson et al., 1997 . All discs were returned to the laboratory, examined under the

light microscope and the scratches made on each disc by the species of interest werecounted.

The second experiment, designed to test the hypothesis that there would be a closerpositive relationship between number of scratches per disc and density of gastropods atsmaller spatial scales within sites than at the scale of sites, was done using only C.

Ž .tramoserica. Six sites approximately 3=3 m were chosen, three with large and threewith small densities of limpets, measured as described previously. Density measured

Ž .using AlargeB quadrats 50=50 cm over the whole site was referred to as density at thelarger scale. Density was measured on four occasions for over 5 weeks to ensure that thedensities were maintained prior to and during the experiment. Density was also

Ž .measured in three plots 50=50 cm within each site, which were haphazardly chosenand approximately 0.5–1m apart. In each plot, limpets were counted in four replicate

Ž .AsmallB quadrats 13=13 cm . Density measured using AsmallB quadrats, in plots, wasreferred to as density at the smaller scale. The experiment started after ensuring that thedensities in sites had remained consistent for 3 weeks. Two replicate arrays of nine wax

Ž .discs each array approximately 13=13 cm were placed into each plot and left in thefield for 1 week. They were then returned to the laboratory, where the scratches made byC. tramoserica on each disc were counted as previously. The mean number of scratchesper disc and mean density at the larger and at the smaller spatial scale were recorded. Inboth cases, the mean density was the average of the last two times sampled, i.e. whilethe discs were in the field.

To test the hypothesis that density of limpets during low tide would be correlatedwith density during high tide, limpets within a 1 m diameter circle around each of two

Ž .plots with the circle centred on the plot in each site were counted during low tide andduring high tide on each of 2 days. On the first occasion, not all plots were reached bythe tide and data were only obtained from seven plots.

To test the hypothesis that there is a positive relationship between amount of grazingand amount of food on the substratum, microalgal samples were taken from four

Ž . Žreplicate patches approximately 5=5 cm in each of two plots in each site as. Ždescribed by Underwood, 1984a . Because encrusting macroalgae notably Hildenbran-

Ž . .dia rubra Sommerfelt Meneghini , which were abundant in these study sites, can causeŽ .large inaccuracies in estimates of microalgae Underwood, 1984a , all of the patches

were scrubbed with wire brushes and metal scrapers to remove macroalgae prior to theŽexperiment. Patches were left to recover their microflora for 2 weeks generally

.sufficient time; MacLulich, 1986 before the samples were taken. At this time, there wasno evidence of recolonisation by H. rubra. The surface of the rock was then scraped


down to bare rock in each of the four patches in each plot and the area of each scrapedpatch was traced onto clear plastic. The scrapings were placed in 90% acetone, kept inthe dark on ice and taken to the laboratory. Each sample was then finely ground,centrifuged and weighed before the chlorophyll was measured using a spectrophotom-eter. Three different wavelengths were used to determine concentration of chlorophyll

Ž Ž . .from mixed microalgal assemblages see Underwood 1984a for details , which wereconverted to mg chlorophyllrg rockrcm2.

3. Results

3.1. Laboratory experiments

Each species left distinctive scratches in the wax. Although the marks were visible toŽ .the naked eye, magnification =10 was necessary to identify the distinguishing features

Ž .of each species. C. tramoserica made long ;1 mm , straight scratches, often in groupsŽ .of four, in a AW-likeB formation Fig. 1a . These marks were most likely caused by four

of the long, sharp lateral teeth of the docoglossan radula contacting the substratum witheach rasp. B. nanum made smaller, shallower marks in discrete groups, each represent-

Ž .ing one rasp of the substratum by the taenioglossan radula Fig. 1b . A. porcata and N.atramentosa made marks similar to each other, that were characterised by sets of fineparallel scratches in a fingerprint-like formation, probably made by the sweeping actionof the many fine marginal teeth of the rhipidoglossan radula. In addition, marks of A.

Ž .porcata showed pairs of ApinpricksB between the AfingerprintsB Fig. 1c . The marks ofN. atramentosa were distinguished by deep, lens-shaped gouges that overlaid the

Ž .AfingerprintsB Fig. 1d . For each of these two species, it seems likely that one rasp ofthe substratum is represented by two AfingerprintsB, made by the right and left sets ofmarginal teeth, with the more central lateral teeth as the source of the pinpricks or

Ž Ž .gouges see Fretter and Graham 1962 for a general description of the different.morphotypes of radulae . The scratches in the wax could be distinguished when the discs

were scraped by any two species, unless the disc was extensively scratched by C.tramoserica, which could destroy the marks of other species.

All volunteers correctly identified at least 65% of the marks; four of them correctlyidentified at least 80% of the marks. Most errors were due to failure to notice that a dischad been scratched by more than one species, rather than wrongly identifying the marksthat were seen.

3.2. Field experiments

For C. tramoserica, sites with large densities had significantly more limpets thanŽsites with small densities Fs28.62, 1 and 2 df , P-0.05, after eliminating interaction

.terms; Table 1a . Although the two levels of density did not significantly differ for N.

()

R.E

.Forrestet

al.rJ.E

xp.Mar.B

iol.Ecol.258

2001155

–171

161

Ž .Fig. 1. Scanning electron micrograph showing radular scratches in wax discs made by a C. tramoserica, with arrow indicating four scratches in the AW-likeBŽ . Ž . Ž .formation, b B. nanum, with arrow indicating marks made by one radular rasp, c A. porcata, with arrow indicating a set of ApinpricksB and d N. atramentosa,

with arrow indicating lens-shaped gouges. Arrows in all images indicate probable direction of movement of feeding animal. Scale bars1 mm.


Table 1Ž . Ž .a Mean S.E. densities per square metre averaged over all times; note that the two species were measured in

Ž . Ž .different sites at each level of density; b mean S.E. number of scratches per disc

Large density Small density

Site 1 Site 2 Site 3 Site 4

( )aŽ . Ž . Ž . Ž .C. tramoserica 47.0 3.1 58.8 5.0 21.0 4.9 21.6 2.8Ž . Ž . Ž . Ž .N. atramentosa 167.2 11.9 115.7 12.5 51.6 6.1 44.3 7.3

( )bŽ . Ž . Ž . Ž .C. tramoserica 21.7 5.9 60.6 10.7 14.7 4.4 19.7 5.3Ž . Ž . Ž . Ž .N. atramentosa 16.4 5.3 1.9 0.8 13.3 4.6 8.9 2.4

Žatramentosa Fs12.95, 1 and 2 df , P)0.05, after eliminating interaction terms; Table.1a , the sites with large densities of snails did contain a consistently larger average

Ž .number of snails than sites with small densities Table 1a . Densities in sites did notŽsignificantly differ among times of sampling for either species Fs5.03, 1 and 2 df ,

P)0.05; Fs4.35, 2 and 4 df , P)0.05, for C. tramoserica and N. atramentosa,.respectively .

In the first experiment, some discs were lost. Therefore, discs were removed atrandom from relevant arrays to give ns7 for N. atramentosa and ns6 for C.tramoserica. Scratches were counted in the following way: one scratch by N. atramen-tosa was a single AfingerprintB and its associated gouges. Although this probably onlyrepresents half of one rasp of the radula, these marks were discrete and easy to count.One scratch by C. tramoserica was four gouges in the AWB pattern, which probablyrepresents one radular rasp. This pattern was also fairly discrete, except where grazingwas intense. When this was the case, the number of scratches was estimated by countingevery fourth mark. New marks made by C. tramoserica could be distinguished fromunderlying older marks and also from the scratches made by other species.

There were no significant differences in the mean number of scratches per discŽbetween the two levels of density for either species Fs2.66, 1 and 2 df , P)0.05;

Fs0.21, 1 and 2 df , P)0.05, for C. tramoserica and N. atramentosa, respectively;.data transformed to natural logarithms . Despite the lack of significant differences, sites

with large density of C. tramoserica did have a greater mean number of scratches perŽ .disc than sites with small density Table 1b . This was not, however, the case for N.

Ž .atramentosa Table 1b . Components of variation, at the scale of discs within arrays andthe scale of arrays in each plot, were calculated from the mean square estimatesŽ .Underwood, 1997 . Most of the variability was identified at the scale of discs within

Žarrays 74% and 59% of the variation for N. atramentosa and C. tramoserica,. Žrespectively and among arrays in each site 22% and 21% for N. atramentosa and C.

.tramoserica, respectively . Over all sites, there was a significant positive correlationbetween the mean number of scratches per disc and the percentage of discs scratched in

Žeach array Spearman’s rs0.80, Spearman’s rs0.90 for N. atramentosa and C..tramoserica, respectively, 14 df , P-0.01 .


In the second field experiment, there was a significant difference between the twolevels of density measured at the larger scale, despite Site 1 being significantly smaller

Ž .than the other two large density sites Table 2a; Fig. 2a . As in the previous experiment,Ž .densities in sites did not significantly differ among times of sampling Table 2a .

Because density measured at the smaller scale was measured in fixed quadrats, thesedata could not be analysed for temporal variability. Separate analyses were thereforedone for each time, separately examining variability among plots, sites and the twolevels of density. In these analyses, sites with large densities were significantly differentfrom sites with small densities at each time and there were no significant differences

Ž .among sites at each level of density Table 2b . There were some significant differencesŽ .among plots within sites at each time SNK tests, P-0.05 , due to one plot differing

from the others in one or two sites, but these differences were rare. Therefore, incontrast to what was predicted from the apparent patchy distribution of limpets within

Ž .sites, density was generally similar among the replicate plots in each site Fig. 3a .As in the first experiment, some discs were heavily scraped by C. tramoserica,

whereas others were sparsely or not scraped. Again, some discs were lost and it wasnecessary to reduce n to 7. In contrast to the previous experiment, wax discs in sites

Table 2Ž . Ž .C. tramoserica: a analysis of mean densities per square metre, measured in sites approximately 3=3 m

Ž .using AlargeB 50=50 cm quadrats, during the second field experiment; ns10; data untransformed.) )) Ž .nssP)0.05, P-0.05, P-0.01. b Summary of analyses of mean density per square metres,

Ž . Ž .measured in plots 50=50 cm using AsmallB 13=13 cm quadrats, at four times during the secondexperiment; ns4; data untransformed

Ž .a

Source df MS F P))aDensitysD 1 80960.3 52.95))Ž . Ž .Site Density sS D 4 1528.9 6.77

TimessT 3 630.8 2.80 nsD=T 3 115.3 0.51 ns

Ž .T=S D 12 225.7 0.65 nsResidual 21 345.4Cochran’s test Cs0.11, ns

Ž .a

Source df Time 1 Time 2 Time 3 Time 4

F P F P F P F P)) ) ) ))DensitysD 1 57.76 16.00 8.16 21.43

Ž . Ž .Site Density sS D 4 0.20 ns 0.55 ns 1.59 ns 0.43 ns) )Ž Ž ..Plot S D 12 2.23 1.30 ns 2.25 1.55 ns

Residual 54Cochran’s test Cs0.14, ns Cs0.17, ns Cs0.22, ns Cs0.14, ns

a Ž . Ž .Density tested against Site Density after elimination of D=T and T=S D , P)0.25.


Ž .Fig. 2. MeanqS.E.: a density of C. tramoserica per square metre, measured in sites, using AlargeB quadratsŽ . Ž . Ž .50=50 cm and b number of scratches per disc per site in three sites with large densities solid bars and

Ž .three sites with small densities hashed bars of limpets.

with large densities of limpets had significantly more scratches per disc than discs insites with small densities, although there were also significant differences among sitesŽ .Table 3; Fig. 2b . The number of scratches was variable among discs in the same array,

Ž .but not between arrays in each plot, nor among plots in each site Table 3; Fig. 3b .There was a strong correlation between mean density of limpets measured at the larger

Ž .scale and number of scratches per disc per site Spearman’s rs0.89, 4 df , P-0.05and between mean density measured at the smaller scale and number of scratches per

Ž .disc per plot Spearman’s rs0.70, 16 df , P-0.01 . As in the previous experiment,there was a strong positive correlation between number of scratches per disc and

Ž .percentage of discs scratched per array Spearman’s rs0.88, 34 df , P-0.01 . Inaddition, the number of limpets in each plot during low and during high tide at each time

Žwas strongly positively correlated Spearman’s rs0.99, 5 df , P-0.01; Spearman’s.rs0.96, 10 df , P-0.01, for Times 1 and 2, respectively .


Ž .Fig. 3. MeanqS.E.: a density of C. tramoserica per square metre, measured in plots, using AsmallB quadratsŽ . Ž .13=13 cm , in each of three plots in three sites with large densities solid bars and three plots with small

Ž . Ž .densities hashed bars of limpets, b number of scratches per disc per plot, in each of three plots in three sitesŽ . Ž . Ž .with large densities solid bars and three plots with small densities hashed bars of limpets and c

meanqS.E. mg chlorophyllrg rockrcm2 of rock surface in each of two plots in three sites with largeŽ . Ž .densities solid bars and three plots with small densities hashed bars of limpets.

The model that intensity of feeding is related to the amount of food available was notŽ .supported. Although the amount of chlorophyll varied among plots Fig. 3c , there were

no correlations between amount of chlorophyll and mean number of scratches per plot


Table 3Ž .C. tramoserica: analysis of mean number of scratches per disc; ns7; transformations 6xq1 . nssP)0.05,

)P-0.05

Source df MS F P)DensitysD 1 511.7 10.96)Ž Ž ..Site DensitysS D 4 46.7 3.98

Ž Ž ..Plot D=SsP D=S 12 11.7 1.96 nsŽ .Array D=S=P 18 6.0 1.14 ns

Res 216 5.2Ž .Cochran’s test Cs0.08, ns after transformation

Ž .Spearman’s rs0.06; 10 df ; P)0.05 , the amount of chlorophyll and number ofŽ .scratches per limpet per plot Spearman’s rs0.30; 10 df ; P)0.05 , or the amount of

Ž .chlorophyll and density of limpets per plot Spearman’s rsy0.04; 10 df ; P)0.05 .

4. Discussion

This study has demonstrated that wax discs, placed on intertidal rocky shores in NewSouth Wales, were useful for measuring grazing in areas where many species of largegrazing gastropods feed in the same areas on the shore. More specifically, for the limpet,C. tramoserica, wax discs proved to be useful for testing hypotheses about therelationship between density of grazers in an area during low tide and intensity ofgrazing during high tide.

ŽIn the laboratory, the four species investigated C. tramoserica, B. nanum, A..porcata and N. atramentosa made distinctive scratches in the wax that were easy to see

at relatively low magnification under a light microscope. The marks enabled identifica-Žtion of the different species, even those with relatively similar radulae see also

.Thompson et al., 1997 and when the discs had been scratched by more than onespecies. From the results of the laboratory experiments, where many discs were heavilyscratched and animals were observed readily grazing over discs, it was concluded thatwax discs would be able to provide a reliable estimate of grazing activity in the field. Itshould be noted that implicit in this conclusion was the assumption that animals indifferent sites and plots would respond to encounters with discs in a similar way, thusleaving comparable records of their grazing on the wax discs.

In the field, strong correlations between the percentage of discs scratched and thenumber of scratches per disc per array showed that either of these two measures could

Ž .be used to estimate grazing activity. Thompson et al. 1997 showed similar relation-ships between the area of disc scratched and the proportion of discs scratched, although

Ž .they did not record the number of scratches per disc. Thompson et al. 1997 found areaof discs scratched to be a good measure of grazing but acknowledged that it couldunderestimate grazing where parts of a disc had been scratched multiple times. Thepresent study addresses this problem by counting the number of scratches per discbecause new scratches could clearly be seen overlaying previous marks. Thompson et al.


Ž .1997 showed large variation in the amount of scraping among individual discs withinan array and this was also found here. Whether this was due to limpets avoidingparticular discs, or grazing in different intensities at very small spatial scales and

Žtherefore AmissingB certain patches in an array even though these may only be.centimetres apart , is not known. Unquantified observations in the laboratory suggested

that individuals of each species appeared to change their path when encountering few ofthe wax discs, while readily moving over most of them. It is not known what features ofthe wax discs caused these differences in behaviour, but it is important to ensure that theedge of the disc is flush with the rock surface and does not form a small AlipB. Thissmall-scale variability in grazing intensity, at scales of centimetres and metres, refocusesattention on the need to consider patterns and processes at the scales at which theindividual animals operate, when attempting to understand ecological processes at larger

Ž .spatial scales Wiens, 1976; Chapman, 1995; Underwood and Chapman, 1996, 1998 .Nevertheless, despite such small-scale variability, the amount of grazing, as measured

using wax discs, was positively correlated with the density of the limpet, C. tramoserica,in sites as well as in plots. Although the first experiment showed most variation ingrazing at the scale of arrays, there were more scratches per disc in the sites with largerdensities of limpets than in the sites with fewer limpets, although these differences werenot significant. In this experiment, the two sites within each level of density werenecessarily sampled during different periods of time, due to poor weather conditions andthe logistics of setting up the experiments. Therefore, the differences between sites ineach level of density were confounded by different experimental times. Foraging in C.

Žtramoserica and other intertidal gastropods is variable from time to time Mackay andUnderwood, 1977; Underwood, 1979; Hawkins and Hartnoll, 1983; Crowe, 1999;

.Chapman, 2000a and temporal variability potentially increased the apparent variabilitybetween sites. Nevertheless, in the second experiment, the number of scratches per disc

Ž .was strongly correlated with the mean density of limpets in sites tens of metres apartŽ .and also in plots separated by only a metre or so in each site . Additionally, in the

second experiment, discs in sites with large densities of limpets had a significantlygreater mean number of scratches than discs in sites with small densities. There were nosignificant differences among plots within sites, but neither were there significantdifferences in density among plots. In contrast to what had been predicted, the generally

Žgreat variability in densities of intertidal gastropods at small spatial scales cm–m;.Underwood and Chapman, 1996 was not found here.

Densities of C. tramoserica in plots during low tide strongly reflected local densitiesduring high tide, indicating that small-scale densities of limpets during low tide are validmeasures of relative densities during high tide and probably a reliable estimate of thenumber of animals feeding in an area. Similar results at this scale have been shown for

Ž .Littorina unifasciata Gray Chapman, 1995 , although spatial patterns of densities ofintertidal animals during low and high tide have not been extensively examined, exceptfor species that show distinctive patterns of behaviour associated with either periods of

Žactivity or inactivity. These include such behaviours as clustering during low tide e.g.. ŽMoulton, 1962; Moran, 1985; Chapman and Underwood, 1996 , homing Mackay and

.Underwood, 1977; Cook and Cook, 1978; Chelazzi et al., 1990 or zonal feedingŽ .migrations Levings and Garrity, 1983 . Our results indicate that densities of C.


tramoserica measured during low tide, are good indicators of intensity of grazing duringhigh tide and this is so at scales of metres to tens of metres. Of course, there may beseasonal or other temporal differences in the particular relationship between amount ofgrazing and density during low tide. To ensure the reliability of this as a generalrelationship requires analyses under different weather conditions, which was not the aimof this study.

Ž .Mackay and Underwood 1977 showed that C. tramoserica showed greater rates ofemigration from and smaller rates of immigration into sites with reduced amounts ofmicroalgae, showing a positive correlation between amount of food and density oflimpets. In the experiments described here, in contrast to what was predicted, there was

Ž .no correlation between the amount of microalgae measured as amount of chlorophylland either density of limpets, or the amount of grazing. Grazing and other feedingbehaviour can be very variable in intertidal habitats because such factors as the vagaries

Ž .of weather Moran, 1985; Burrows and Hughes, 1989; Chapman, 2000a , avoidance ofŽ .predators Levings and Garrity, 1983 or the distribution and availability of microhabi-

Ž .tats Underwood, 1977; Fairweather, 1988; Chapman, 2000a . Such factors may be moreimportant than the quantity of food in determining foraging behaviour of microalgalfeeding gastropods, at least over relatively short periods of time. In addition, other

Ž .grazers snails and starfish were present in the experimental areas and may havecontributed to obscuring any relationship between standing stock of microalgal food andgrazing by C. tramoserica. Nevertheless, the combined results of these experimentsclearly demonstrate that C. tramoserica graze in proportion to the densities at whichthey are found during low tide on this shore.

Results for N. atramentosa were less convincing, showing that for this species,density was not a good estimate of grazing intensity, at least among sites, tens of metresapart. N. atramentosa is very restricted in its use of particular microhabitats during low

Ž .tide Underwood, 1976b and therefore, it is likely that the large spatial scale investi-gated in this study was not appropriate for this species. In addition, this species is

Ž .sometimes inactive from one low tide to the next unpublished data . The small meannumber of scratches per disc found in one of the sites with large density of snails mayreflect low rates of foraging in this site during the experiment. It appears that the waxdisc methodology could be very useful in identifying the spatial and temporal scale atwhich this species grazes outside of the microhabitats in which it is generally foundduring low tide.

Ž .This study extends the findings of Thompson et al. 1997 that arrays of wax discs onintertidal shores are useful tools for testing specific hypotheses about the factorsaffecting the foraging of intertidal gastropods. They measure grazing activity directlyand can be used to distinguish among different species feeding in the same areas.Importantly, they are relatively cheap to deploy and monitor, thus allowing adequatereplication in time and space to quantify accurately patterns of variability in foragingand the spatial scales over which animals feed. In addition, as discussed by Thompson et

Ž .al. 1997 , they can be used to measure grazing under conditions when it may beŽ .difficult to access the study sites e.g. at night or when the seas are rough . This

methodology therefore has great potential to contribute to new models about the role ofgrazing in rocky intertidal communities.


Acknowledgements

The study was supported by funds from the Australian Research Council and theCentre for Research on Ecological Impacts of Coastal Cities. Scanning electron microg-raphy was done at the Australian Key Centre for Microscopy and Microanalysis,University of Sydney. Thanks to the staff and students at the Centre who assisted withdifferent phases of this project, especially those who gave their time to assist with thelaboratory experiment. V. Mathews assisted with the images. P. Archambault, B.Kelaher, R. Thompson and an anonymous referee offered constructive comment on an

[ ]earlier draft of this manuscript. RW

References

Boyden, C.R., Zeldis, J.R., 1979. Preliminary observations using an attached microphone sensor to studyfeeding behaviour of an intertidal limpet. Est. Coast. Shelf. Sci. 9, 759–769.

Branch, G.M., 1971. The ecology of Patella Linnaeus from the Cape Peninsula, South Africa: 1. Zonation,movements and feeding. Zool. Afr. 6, 1–38.

Brawley, S.H., Johnson, L.E., 1991. Survival of fucoid embryos in the intertidal zone depends upondevelopmental stage and microhabitat. J. Phycol. 27, 179–186.

Ž .Burrows, M.T., Hughes, R.N., 1989. Natural foraging of the dogwhelk, Nucella lapillus Linnaeus : theweather and whether to feed. J. Moll. Stud. 55, 285–295.

Castenholz, R.W., 1961. The effect of grazing on littoral diatom populations. Ecology 42, 783–794.Chapman, M.G., 1994. Small- and broad-scale patterns of distribution of the upper-shore littorinid Nodilitto-

rina pyramidalis in New South Wales. Aust. J. Ecol. 19, 83–95.Chapman, M.G., 1995. Aggregation of the littorinid snail Littorina unifasciata in New South Wales, Australia.

Mar. Ecol. Prog. Ser. 126, 191–202.Chapman, M.G., 2000a. Variability of foraging in highshore habitats: dealing with unpredictability. Hydrobi-

ologia 426, 75–87.Chapman, M.G., 2000b. A comparative study of differences among species and patches of habitat on

movement of three species of intertidal gastropods. J. Exp. Mar. Biol. Ecol. 244, 181–201.Chapman, M.G., Underwood, A.J., 1992. Foraging behaviour of marine benthic grazers. In: John, D.M.,

Ž .Hawkins, S.J., Price, J.H. Eds , Plant–Animal Interactions in the Marine Benthos, Systematics Associa-tion Special Volume No. 46. Clarendon Press, Oxford, pp. 289–317.

Chapman, M.G., Underwood, A.J., 1996. Influences of tidal conditions, temperature and desiccation onpatterns of aggregation of the high-shore periwinkle, Littorina unifasciata, in New South Wales, Australia.J. Exp. Mar. Biol. Ecol. 196, 213–237.

Chelazzi, G., Focardi, S., Deneubourg, J.L., 1983. A comparative study on the movement patterns of twoŽ .sympatric tropical chitons Mollusca: Polyplacophora . Mar. Biol. 74, 115–125.

Chelazzi, G., Della Santini, P., Parpagnoli, D., 1990. The role of trail following in the homing of intertidalchitons: a comparison between three Acanthopleura spp. Mar. Biol. 105, 445–450.

Connell, J.H., 1961. The effect of competition, predation by Thais lapillus and other factors on the distributionof the barnacle Chthamalus stellatus. Ecology 42, 710–723.

Cook, S.B., Cook, C.B., 1978. Tidal amplitude and activity in the pulmonate limpets Siphonaria nemoralisŽ . Ž .Gould and S. alternata Say . J. Exp. Mar. Biol. Ecol. 35, 119–136.

Cook, S.B., Cook, C.B., 1981. Activity patterns in Siphonaria populations: heading choice and effects of sizeand grazing interval. J. Exp. Mar. Biol. Ecol. 49, 69–80.

Creese, R.G., 1980. An analysis of distribution and abundance of populations of the high-shore limpet,Ž . Ž .Notoacmea petterdi Tennison–Woods . Oecologia Berlin 335, 252–260.

Creese, R.G., 1982. Distribution and abundance of the acmaeid limpet Patelloida latistrigata, and itsŽ .interaction with barnacles. Oecologia Berlin 52, 85–96.


Creese, R.G., Underwood, A.J., 1982. Analysis of inter- and intraspecfic competition among intertidal limpetsŽ .with different methods of feeding. Oecologia Berlin 53, 337–346.

Crowe, T.P., 1999. Limits to generality: seasonal and temporal variation in dispersal of an intertidal gastropod.J. Exp. Mar. Biol. Ecol. 232, 177–196.

Dayton, P.K., 1971. Competition, disturbance, and community organization: the provision and subsequentutilisation of space in a rocky intertidal community. Ecol. Monogr. 41, 351–389.

Denley, E.J., Underwood, A.J., 1979. Experiments on factors influencing settlement, survival and growth oftwo species of barnacles in New South Wales. J. Exp. Mar. Biol. Ecol. 36, 269–293.

Ž .Fairweather, P.G., 1988. Movements of intertidal whelks Morula marginalba and Thais orbita in relation toavailability of prey and shelter. Mar. Biol. 100, 63–68.

Fretter, V., Graham, A., 1962. British Prosobranch Molluscs. The Ray Society, Dorking, 755.Ž .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 invertebrates. Annu. Rev. Oceanogr. Mar.

Biol. 21, 195–282.Hruby, T., Norton, T.A., 1979. Algal colonization on rocky shores in the Firth of Clyde. J. Ecol. 67, 65–77.Jernakoff, P., 1985a. The effect of overgrowth by algae on the survival of the intertidal barnacle Tesseropora

Ž .rosea Krauss . J. Exp. Mar. Biol. Ecol. 94, 89–98.Jernakoff, P., 1985b. An experimental evaluation of the influence of barnacles, crevices and the seasonal

patterns of grazing on the algal diversity and cover in an intertidal barnacle zone. J. Exp. Mar. Biol. Ecol.88, 287–302.

Jernakoff, P., 1985c. Interactions between the limpet Patelloida latistrigata and algae on an intertidal rockplatform. Mar. Ecol. Prog. Ser. 23, 71–78.

Kastendiek, J., 1982. Competitor-mediated coexistence: interactions among three species of benthic macroal-gae. J. Exp. Mar. Biol. Ecol. 62, 201–210.

Kitting, C.L., 1979. The use of feeding noises to determine the algal food being consumed by individualŽ .intertidal molluscs. Oecologia Berlin 40, 1–17.

Levings, S.C., Garrity, S.D., 1983. Diel and tidal movement of two co-occurring neritid snails: differences ingrazing patterns on a tropical rocky shore. J. Exp. Mar. Biol. Ecol. 67, 261–278.

Lubchenco, J., 1980. Algal zonation in the New England rocky intertidal community: an experimentalanalysis. Ecology 61, 333–344.

Lubchenco, J., 1983. Littorina and Fucus: effects of herbivores, substratum heterogeneity, and plant escapesduring succession. Ecology 64, 1116–1123.

Lubchenco, J., Gaines, S.D., 1981. A unified approach to marine plant–herbivore interactions: I. Populationsand communities. Annu. Rev. Ecol. Syst. 12, 405–437.

Mackay, D.A., Underwood, A.J., 1977. Experimental studies on homing in the intertidal patellid limpetŽ . Ž .Cellana tramoserica Sowerby . Oecologia Berlin 30, 215–237.

MacLulich, J.H., 1986. Experimental evaluation of methods for sampling and assaying intertidal epilithicmicroalgae. Mar. Ecol. Prog. Ser. 34, 275–280.

MacLulich, J.H., 1987. Variations in the density and variety of intertidal epilithic microflora. Mar. Ecol. Prog.Ser. 40, 285–293.

Marshall, P.A., Keough, M.J., 1994. Asymmetry in intraspecific competition in the limpet Cellana tramoser-Ž .ica Sowerby . J. Exp. Mar. Biol. Ecol. 177, 121–138.

Moran, M.J., 1985. The timing and significance of sheltering and foraging behaviour of the predatoryŽ .gastropod Morula marginalba Blainville Muricidae . J. Exp. Mar. Biol. Ecol. 93, 103–114.

Moulton, J.S., 1962. Intertidal clustering of an intertidal gastropod. Biol. Bull. Mar. Biol. Lab., Woods Hole,Mass. 123, 170–178.

Nicotri, M.E., 1977. Grazing effects of four intertidal herbivores on the microflora. Ecology 58, 1020–1032.Petraitis, P.S., 1992. Effects of body size and water temperature on grazing rates of four intertidal gastropods.

Aust. J. Ecol. 17, 409–414.Thompson, R.C., Johnson, L.E., Hawkins, S.J., 1997. A method for spatial and temporal assessment of

gastropod grazing intensity in the field: the use of radula scrapes on wax surfaces. J. Exp. Mar. Biol. Ecol.218, 63–76.


Underwood, A.J., 1975. Intertidal zonation of prosobranch gastropods: analysis of densities of four coexistingspecies. J. Exp. Mar. Biol. Ecol. 19, 197–216.

Underwood, A.J., 1976a. Analysis of patterns of dispersion of intertidal prosobranch gastropods in relation toŽ .macroalgae and rock pools. Oecologia Berlin 25, 145–154.

Underwood, A.J., 1976b. Comparative studies on the biology of Nerita atramentosa Reeve, BembiciumŽ . Ž . Ž .nanum Lamarck and Cellana tramoserica Sowerby Gastropoda: Prosobranchia in S.E. Australia. J.

Exp. Mar. Biol. Ecol. 18, 153–172.Underwood, A.J., 1977. Movements of intertidal gastropods. J. Exp. Mar. Biol. Ecol. 26, 191–201.Underwood, A.J., 1978. An experimental evaluation of competition between three species of intertidal

Ž .prosobranch gastropods. Oecologia Berlin 33, 185–208.Underwood, A.J., 1979. The ecology of intertidal gastropods. Adv. Mar. Biol. 16, 111–210.Underwood, A.J., 1980. The effects of grazing by gastropods and physical factors on the upper limits of

Ž .distribution of intertidal macroalgae. Oecologia Berlin 46, 201–213.Underwood, A.J., 1984a. The vertical distribution and seasonal abundance of intertidal microalgae on a rocky

shore in New South Wales. J. Exp. Mar. Biol. Ecol. 78, 199–220.Underwood, A.J., 1984b. Vertical and seasonal patterns in competition for microalgae between intertidal

Ž .gastropods. Oecologia Berlin 64, 211–222.Underwood, A.J., 1997. Experiments in Ecology: Their Logical Design and Interpretation Using Analysis of

Variance. Cambridge Univ. Press, Cambridge, 504.Underwood, A.J., 1999. Physical disturbances and their direct effect on an indirect effect: responses of an

intertidal assemblage to a severe storm. J. Exp. Mar. Biol. Ecol. 232, 125–140.Underwood, A.J., Chapman, M.G., 1996. Scales of spatial patterns of distribution of intertidal invertebrates.

Ž .Oecologia Berlin 107, 212–224.Underwood, A.J., Chapman, M.G., 1998. Spatial analyses of intertidal assemblages on sheltered rocky shores.

Aust. J. Ecol. 23, 138–157.Underwood, A.J., Jernakoff, P., 1981. Effects of interactions between algae and grazing gastropods on the

Ž .structure of a lowshore algal community. Oecologia Berlin 48, 221–233.Underwood, A.J., Jernakoff, P., 1984. The effects of tidal height, wave-exposure, seasonality and rock pools

on grazing and the distribution of intertidal macroalgae in New South Wales. J. Exp. Mar. Biol. Ecol. 75,71–96.

Underwood, A.J., Denley, E.J., Moran, M.J., 1983. Experimental analyses of the structure and dynamics of theŽ .midshore intertidal communities in New South Wales. Oecologia Berlin 56, 202–219.

Wiens, J.A., 1976. Population responses to patchy environments. Annu. Rev. Ecol. Syst. 7, 81–120.

Download - Quantification of radular marks as a method for estimating grazing of intertidal gastropods on rocky shores

Top Related