Testing the role of spines as predatory defense.

ABSTRACT Spines are frequently considered to be an important physical defense against predators. This experiment shows that spines do not always function successfully as physical protection. Using natural predators and prey (the drilling muricid gastropod Nucella lamellosa (Gmelin 1791) and the mussel Mytilus trossulus (Gould 1850)) this experiment used artificial spines in different configurations and densities to try to assess the role of spines as predatory defense. The presence of spines did not inhibit the predator from choosing ornamented prey. Greater spine density did not improve the probability of surviving a predator encounter. Although mean handling time increased to some extent with higher spine density, the outcome of the encounter was the same. Presence or density of spines did not deter predators from attacking prey, and experimental prey with greater spine-density experienced greater mortality than did specimens with fewer spines. Drillhole dimensions differed slightly between the treatments, probably as a consequence of difficulties for the gastropod to maneuver its accessory boring organ (ABO) in between the spines.

KEY WORDS: predator-prey, spines, defense, Nucella lamellosa, Mytilus trossulus, San Juan Island 

INTRODUCTION

Predation is arguably a major driving force in evolution. Predatory behavior probably originated many times and fossil evidence of predation is documented in the oldest skeletal organisms (Bengtson & Zhao 1992), though predation itself is likely to be much older (Bengtson 2002). During recent years much attention has been devoted not only to document predation or failed predation in the fossil record, but also to understand the tempo and modus of predator-prey interactions through time. This effort is reflected in a series of thematic studies and symposia (e.g., Kowalewski & Kelley 2002, Kelley et al. 2003). One of the earliest and most influential papers exploring coevolution between predators and their prey described the so-called Mesozoic Marine Revolution (MMR) (Vermeij 1977). Vermeij (1977, 1978, 1982, 1987)proposed that predation increased through geologic time, a result of an escalation in predatory intensity and a resulting response from the prey. The theory of escalation is that an organism's enemies, as a group, become more dangerous through time, and that prey must respond to this. In support of his theory, Vermeij found that gastropod shells changed from simple unornamented forms in the Palaeozoic to robust, thick and ornamented in the Mesozoic, and that, at the same time, new and innovative groups of molluscivore predators emerged. Hence, Vermeij inferred that the increase in gastropod ornament was an evolutionary response to increasing predation pressure from the new predators.

Changes in morphology in predators and their prey as a result of coevolution and/or escalation is the immediate point to address in analytical studies of these types of biotic interactions. Particularly for the prey species, it is important to explain whether morphological features have an antipredatory function or serve some other purpose related to, for instance, environment or sexual behavior. Ornamentation in the form of spines is one of several ways of protecting an organism against predators. Although many animals and plants have spines, the nature of the spines and their function varies greatly among and within taxa. In shelly organisms spines are regularly considered to be of importance in protecting the organism against predators, either acting as direct defense (e.g., Palmer 1979, Vermeij 1987, Stone 1998), as a way of increasing handling time (e.g., Miller & LaBarbera 1995), or as camouflage (e.g., Vance 1978, Feifarek 1987). Spines are also reported as serving several other functions such as anchorage (e.g., Stanley 1970, Stenzel 1971) or a way of decreasing the mass/area ratio, which prevents organisms living on soft substrates to sink into the substrate (e.g., Grant 1966 but see Leighton 2000 for an alternative view), and for increasing surface/volume ratio for respiration. Savazzi (1999) contains many interesting chapters on this type of functional morphology of invertebrates. Inbar and Lev-Yadun (2005) noted that spines in the animal and plant kingdoms often are highly conspicuous and therefore should be considered as a warning sign similar to coloration, odor, and so on. Studies of highly spinose bivalve shells have shown disruption of attacks by boring predators (Beatty & Rollins 2002).

The present paper explores the theme further through a modified experimental setup based on the work by Stone (1998). Stone (1998) showed that varying spine lengths in epifaunal bivalves (the mytilids Perna viridis and Septifer virgatus and the chamid Chama reflexa) resulted in predator avoidance to relative degrees. This deterrence was apparent in the subtropical gastropods used in that experiment (the muricids Thais luteostoma, T. clavigera, and Morula musiva) but appeared to have little or no effect on extraoral feeding by asteroids (Coscinasterias acutispina). Stone (1998) set up four different experiments where bivalve prey were offered to the predators in a choice between unornamented individuals and individuals with varying degrees of shell ornamentation. The ornamentation, in the form of artificial polyethylene spines, was attached to the bivalve shells in varying arrangements. Experiment 4 in Stone's study was designed to show whether increasingly spinose prey had an increasingly inhibitory effect on shell boring muricids. Stone cemented 20 individual prey to the bottom of an aquarium (randomly right or left valve). He then attached polyethylene spines to the free valve using epoxy resin, and as a control unornamented prey were dotted with the resin. The density of the spines was kept constant but the lengths of the spines varied (1 mm, 2 mm, and 4 mm). The results suggested that prey with less expressed ornamentation were preferred ahead of highly spinose prey, and Stone therefore concluded that the primary effect of more pronounced spinosity was that of preventing direct access to the shell of the bivalve. The present study further explores experimentally the effects of adding spines to spineless taxa to test how effective spines are in protecting an organism from predators. By placing artificial spines on the mussel Mytilus trossulus I wanted to record the effects spines had on predation by the muricid gastropod Nucella lamellosa. Three questions are of primary interest here: would the gastropod be able to handle the mussel as normal and consume it without problems? Would it take longer for the gastropod to kill prey with a higher spine density? Would the spines have such an effect that successful predation is impossible? The null hypothesis is that the presence of spines will not deter predators.

MATERIALS AND METHODS

The study was carried out at Friday Harbor Laboratories, located in Friday Harbor on San Juan Island off the coast of northwestern United States. Two organisms were used in the experiment, the predatory gastropod Nucella lamellosa (Gmelin 1791), and the mussel My tilus trossulus (Gould 1850) acting as prey.

Specimens of Nucella were collected from the rocky shoreline along Cattle Point located in the southernmost part of San Juan Island. Here, clusters of up to 30 individuals of different sizes could be found within cracks and crevices in the rocks in the intertidal zone. Usually, gastropods from Cattle Point prey mainly on barnacles; however, observations in the area suggest that they are prone to prey on Mytilus when/where barnacles are less common (pers. obs.).

The mussels were collected from Argyle Creek, a narrow tidal stream that runs between two saltwater habitats, the Argyle Lagoon and North Bay, on the eastern part of San Juan Island. Argyle Creek carries a unique combination of sub to intertidal species within a limited area. The creek never runs completely dry during the tidal cycles, suggesting that the predators can hunt prey constantly (Kowalewski 2004). It can also be noted that shell growth in Mytilus can take place continuously because they rarely become exposed. As a consequence, Mytilus collected from Argyle Creek may be morphologically different from other habitats (Kowalewski 2004).

The fact that predators and prey are collected from different sites (and that they, as mentioned, may be morphologically different between the sites) was taken into consideration in project design. However, the same species co-occur at Argyle Creek and Cattle Point and collecting differently from the two sites was done to minimize the human effects on the ecosystems at the sites.

Experimental Design

To study the effects of artificial spines on mussels on gastropod predation, three experiments using similar size sea-tables were set up. Each sea-table had a steady and continuous flow of seawater at a temperature around 13[degrees]C to 14[degrees]C, thus reflecting natural conditions. In each of the three sea-tables 45 specimens of the predator Nucella lamellosa were placed. Each gastropod was measured in terms of width, length and aperture dimensions to the nearest 0.1 mm using a caliper. The gastropods were distributed so that each sea-table held 45 individuals, statistically similar in size. Each individual was marked with a number so that behavior could be monitored. All gastropods were starved for eight days prior to the start of the experiment.

Each sea-table also contained 30 individuals of Mytilus trossulus, which were also statistically similar in size. Barnacles, limpets, and other epibionts were removed from the mussels by careful scraping with a blunt knife. Any mussels damaged during this procedure were removed and not used in the experiment. Heavily encrusted mussels were not used in the experiment. Thus, the likelihood of the mussels being affected in a way that would make them more susceptible to predation was small. The length and width of each individual mussel was measured to the nearest 0.1 mm. Thirty specimens (Treatment 1) were dotted on both valves with a two component epoxy (Ace Marine Epoxy 1065804) to rule out any biases in predatory behavior caused by the adhesive used. On the following 30 specimens (Treatment 2), five spines were glued to each valve of the mussel in a square-like configuration with the fifth spine located in the center (Fig. 1). In a similar approach, eight spines were placed on each valve of the remaining 30 specimens (Treatment 3) (Fig. 1). The spine configuration used in this experiment was adopted after a pilot study. Using a much higher spine density (~15-20 spines depending on the size of the mussel, but completely covering the shell surface) resulted in a complete physical deterrence of predators. The aim of this study, however, was to observe differences in behavior owing to differences in spine density, not to document what density completely prevented the gastropods from consuming the bivalves. The spines used for every "spinose" mussel were cut from a plastic mesh and were 6-8 mm in length. Each spine had a rounded base 2 mm in diameter and a constricted, oval tip 1-3 mm in diameter. In contrast to Stone (1998), I kept spine lengths constant and varied the densities. Handling time outside the water for each mussel was the same and similar amount of adhesive was used on all specimens. A pilot study was carried out before the experiment started and it showed that the gastropods seemed to be little affected by the experimental conditions and started feeding immediately. As such, they are excellent organisms to study under laboratory conditions.

[FIGURE 1 OMITTED]

All three experiments started simultaneously by randomly placing predators and prey on the bottom of the sea-tables. The gastropods were allowed to hunt freely and choose their prey at will. The mussels were separated from each other and were not allowed to clump together as clumping behavior has been shown to inhibit predators that need to orient their prey during the attack (e.g., Bertness & Grosholz 1985, Cote & Jelnikar 1999). Thus, this approach tested only the effect of the spines. Whenever a Nucella latched onto a mussel, both predator and prey were separated from the others by a plastic cage. This was to monitor and study the behavior and individual response among the gastropods to the mussels. Time from latching through consumption and release was noted to record any differences in handling time between the different spine densities.

When a Mytilus had been killed and eaten and the gastropod had abandoned it, the mussel was removed from the tank and the Nucella was allowed to continue hunting. Every mussel killed was photographed and drillhole diameter was measured using Scion Image software. The experiment was stopped after 12 days.

RESULTS

Mussels without ornamentation were attacked by the gastropods without special attention given to the adhesive, consequently, the adhesive used did not affect the ability of gastropods to detect, attack, and consume the mussels. Most drillholes were located between the epoxy patches, but in some cases the gastropods had drilled straight through the epoxy. A similar amount of adhesive was used on every mussel so any toxic effect the adhesive might have had on the prey can be ruled out.

Treatment 1

The number of kills of normal mussels (without spines) was 43% or 13 individuals. This result was assumed to represent close to natural conditions and used as a control for the rest of the experiments. Mean handling time per kill, from latching to release, was 82 h. Three gastropods completed two kills each, which resulted in a total number of 10 individual drilling gastropods.

Treatment 2

In this experiment, with five spines on each valve, the number of successful kills was 50% that is, 15 individual mussels were killed and consumed. All drilled mussels were attacked by different gastropods and the mean handling time was 108 h. One mussel was found consumed unobserved and without a drillhole present.

Treatment 3

The highest spine density, eight spines on each valve, had a kill rate of 77% or 23 individuals. Twenty-one successful attacks by different gastropods with one gastropod killing two mussels were recorded. The mean handling time for this spine configuration was 90 h. One mussel was consumed unnoticed.

The gastropods involved in the killings did not differ in size between the tanks (Table 1). In Treatment 1 the mean size (length) of gastropods that killed and consumed mussels was 45.6 mm. Treatment 2 had a mean size of 46.2 mm and in Treatment 3 the gastropods were 45.5 mm in size. However, there was a significant difference in prey size of the mussels consumed. Nucella killed and consumed smaller prey in the tank with normal mussels than in the tanks with spinose mussels (Table 2). There was no difference in size of prey taken between mussels in Treatments 2 and 3.

Individuals with highest spine density (Treatment 3) were killed more frequently than those with lower density (Treatment 2). Spinose mussels were killed to a greater degree than non-spinose mussels (Fishers exact, P = 0.02). Drillhole diameter is smaller in the prey with higher spine density (Fig. 2).

[FIGURE 2 OMITTED]

Handling times between the experiments also differed. Treatment l, normal mussels, had the expected lowest handling time. Treatment 2 had the longest handling time and Treatment 3 intermediate handling times (Table 2; Fig. 3). The success rate was close to 100% in all treatments. With few exceptions, all the mussels that were attacked were also killed and consumed.

[FIGURE 3 OMITTED]

DISCUSSION

Prey resists predators through a range of physiological, morphological, and behavioral adaptations (Vermeij 1982, 1987). In nature, bivalves show a great variety in spine density and length. The "thorny" or "spiny" oysters (Family Spondylidae), for example, have spines varying in numbers from very few to hundreds. Consequently, spines may also serve other functions, but in many cases they effectively deter predators (Kohn 1999). Stone (1998) showed that varying spine lengths in epifaunal bivalves resulted in predator avoidance to various degrees. By grinding down spines and knobs on molluscs Palmer (1979) demonstrated that the vulnerability to crushing by the predatory puffer fish increased. This is also seen in the fossil record, demonstrated for example in Devonian brachiopods where spinose species have a lower percentage of complete drillholes than morphologically similar nonspinose species (Leighton 2001). Muricids are, along with naticids, probably the best-understood predatory gastropods and their habits are well known (Carriker & Yochelson 1968). The muricid used in this experiment is a carnivorous gastropod that subdues the prey and drills a hole through the shell eating the soft tissues inside (Kitchell et al. 1981). The snails use a chemomechanical process to drill through bivalve shells. The holes are made by a combination of secretion of acidic solution from the accessory boring organ (ABO), to soften the shell, and scraping the radula sectionally in quadrants of a circle passing from left to right to remove the loosened material (Ziegelmeier 1954, Carriker & Van Zandt 1967). The ABO of the muricids is located in the anterior region of the sole of the foot (Clelland & Saleuddin 2000). In the present experiment the spines were blunt and could therefore not repel predators directly but the presence of spines on a prey should obstruct the foot of the muricid to reach the shell and hinder killing prey by drilling. The spines were placed so that the gastropods could not fit between the spines. To be able to drill, the gastropods were required to twist and turn their foot into a favorable position. Hence, the spines would make the bivalve less susceptible for the predators. Unlike naticids, muricids do not use their foot to grapple and hold their prey.

The expected outcome from this study was that higher spine density would result in longer handling time and therefore also lower kill rates. However, what can be concluded is that Nucella is capable of detecting, drilling, and killing highly ornamented prey. The presence of spines did not inhibit choice of prey. In fact, the number of kills was highest in the mussels with the highest spine density. When comparing normal mussels to those with highest spine density, we can conclude that the spinose mussels for some reason were preferred by the predators. An obvious reason for this may be that the gastropods were too inexperienced with ornamented prey to realize that they were a poor prey choice. Over time therefore, the gastropods would perhaps learn to avoid this type of prey (if, indeed, they were a poor prey choice). One can also speculate that the mussels were inhibited by the spines and could not easily maneuver themselves, or escape the gastropods. However, monitoring the sea-tables gave no obvious indication for this. As discussed in Vermeij et al. (1989), something that might be of importance is the fact that under laboratory conditions the gastropods can give more attention to the prey. Given that the predators need not worry about predators of their own or physical disturbance (wave-action etc.), laboratory conditions can actually increase the frequency of successful attacks.

The paper by Stone (1998) was the setup for the present work and the results here contradict in some ways those presented by Stone. Using similar experimental setups, but slightly different approaches, the present work shows that spines need not be an effective means of predator deterrence. Reasons for this may, naturally, lie within the different approaches. Stone (1998) showed that longer spine lengths resulted in predator avoidance; the longest spines Stone used were 4 mm long and in the present experiment the spines were 6-8 mm long (long enough to possibly repel predators according to that study) but arranged in different configurations and densities. Despite this, the only support for spines as predatory defense is the slightly increased handling times. Increased handling times is of great importance because prey that are more costly for the predator to consume are more likely to be avoided by a predator (Vermeij 1987). Mean handling time increased slightly in the spinose mussels and this might be seen as evidence for predator deterrence. However, this result may occur for a multitude of reasons (e.g., shell thickness, size of predator, size of prey, location of drillhole, competition from other predators, and similar). The fact that the gastropods consumed larger prey in Treatment 1 and 2 might as well reflect longer handling times if we assume, for example, that larger mussels have thicker shells.

The fact that drillhole diameter is smaller in prey with the highest spine density may be because the gastropods cannot extend their foot completely between the spines and therefore are forced to drill a smaller hole, perhaps using only the ABO. The greater drilling time for Treatment 2 is probably caused by the fact that the presence of spines inhibits the gastropods from maneuvering their foot. The handling time decreases again in Treatment 3 and this is probably connected to a smaller drillhole size. A smaller drillhole is most likely drilled faster and compensates for the problems the presence of spines may cause. Typically, there is a direct relationship between the size of the predator and the size of the drillhole they produce (Harper 2005), but no such connection was found here.

None of the normal (Treatment 1) mussels were drilled in the area between the valves, or on the umbo, but instead all drillholes were located on the surface of the valve. On the other hand the mussels with spines did show some preferences to have been drilled in the area between the valves and on the umbo most likely because of the presence of spines on other parts of the shell. However, location of the drillholes apparently did not affect the end result of the attack; the same numbers of nonspinose and spinose mussels were killed. Nonetheless, the gastropods are skilled enough to twist and turn their foot and ABO into favorable positions, making them highly adaptable predators, unaffected by this type of spines. In support of the conclusions made in this study, Wieters and Navarrete (1998) showed that feeding preferences among Nucella spp. is affected by local prey population, suggesting that natural differences in relative abundance of prey are important and that the gastropods are flexible predators that can consume highly ornamented prey if they are given as the only choice.

CONCLUSION

The outcome of this study shows that spines do not always provide protection from predators. Although handling times increased slightly the kill rate increased as well. It is clear that spines may serve other functions apart from predator deterrence, and one can speculate that the reason why spines have not evolved on Mytilus is simply because these bivalves would not benefit from having them. Individual traits of an organism cannot always be studied in isolation; the overall effectiveness of a prey item at deterring predation may be the result of complex interactions of several different attributes. The study shows, however, that Nucella are highly adaptable predators and that spines have little effect on their ability to consume prey, at least in the configurations evaluated here. Presence or density of spines did not deter predators from attacking prey and prey with greater spine-density experienced greater mortality than did specimens with fewer or no spines. Their varied natural diet supports the fact that they are likely to prey-switch if necessary. Spine geometry, size, and density are aspects that need to be taken into consideration when evaluating the functional significance of spines. Therefore, future research should be directed towards handling times and to evaluate a variety of spine configurations and spine lengths and their effects on different predators further.

ACKNOWLEDGMENTS

This study was carried out in the summer of 2004 as a part of the course Predator-prey interactions: Experimental and field approaches in Biology and Paleontology at Friday Harbor Laboratories. The author appreciates the help and encouragement from the organizers of the course; Lindsey Leighton and Michat Kowalewski. They have been invaluable in their support. Kowalewski is also thanked for help with the statistics. Rich Krause provided help whenever it was needed, big or small issues alike. Jan Ove Ebbestad, Michael Streng and Graham Budd (Uppsala University) read and commented earlier drafts of the manuscript, and a review by Gregory Dietl (Yale University) erased some of the remaining question marks. Gratefully accepted grants from Friday Harbor Laboratories, Ingegerd och Viking Olof Bjorks hedersledamotsstipendium (Vastmanlands-Dala nation) and Otterborgs donationsfond, Geologiska sektionen, Uppsala University made the trip to USA possible. Last, but not least the author thanks Carlos, Emily, Jen, John, Karen, Kate, Sabrina and Una for help and support.

LITERATURE CITED

Beatty, W. L. & H. B. Rollins. 2002. The role of spinose ornament in predator deterrence: the bivalve Arcinella, Pinecrest (Pliocene) of Florida. Abstracts with Programs--Geological Society of America 34:356-357.

Bengtson, S. 2002. Origins and early evolution of predation. Paleontological Society Papers 8:289-317.

Bengtson, S. & Y. Zhao. 1992. Predatorial borings in Late Precambrian mineralized exoskeletons. Science 257:367-369.

Bertness, M. D. & E. Grosholz. 1985. Population dynamics of the ribbed mussel, Geukensia demissa: the costs and benefits of a clumped distribution. Oecologia 67:192-204.

Carriker, M. R. & E. L. Yochelson. 1968. Recent gastropod boreholes and Ordovician cylindrical borings. United States Geological Survey Professional Paper 593-B:1-26.

Carriker, M. R. & D. Van Zandt. 1967. Gastropod Urosalpinx: pH of accessory boring organ while boring. Science 158:920-922.

Cote, I. M. & E. Jelnikar. 1999. Predator-induced clumping behaviour in mussels (Mytilus edulis Linnaeus). J. Exp. Mar. Biol. Ecol. 235:201-211.

Clelland, E. S. & A. S. M. Saleuddin. 2000. Vacuolar-type ATPase in the accessory boring organ of Nucella lamellosa (Gmelin) (Mollusca: Gastropoda): role in shell penetration. Biol. Bull. 198: 272-283.

Feifarek, B. P. 1987. Spines and epibionts as antipredator defenses in the thorny oyster Spondylus americanus Hermann. J. Exp. Mar. Biol. Ecol. 105:39-56.

Grant, R. E. 1966. Spine arrangement and life habits of the productoid brachiopod Waagenoconcha. J. Paleontol. 40:1063-1069.

Harper, E. M. 2005. Recognising predator-prey interactions in the fossil record. Geol. Today 21:191-196.

Inbar, M. & S. Lev-Yadun. 2005. Conspicuous and aposematic spines in the animal kingdom. Naturwissenschaften 92:170 172.

Kelley, P. H., M. Kowalewski, & T. A. Hansen, editors. 2003. Predator-prey interactions in the fossil record. Topics in Geobiology 20. New York: Kluwer Academic Publishers Group. 484 pp.

Kitchell, J. A., C. H. Boggs, J. F. Kitchell & J. A. Rice. 1981. Prey selection by naticid gastropods: experimental tests and application to the fossil record. Paleobiology 7:533-552.

Kohn, A. J. 1999. Anti-predator defences of shelled gastropods. In: E. Savazzi, editor. Functional morphology of the invertebrate skeleton. Chichester: John Wiley and Sons. pp. 169-181.

Kowalewski, M. 2004. Drill holes produced by the predatory gastropod Nucella lamellosa (muricidae): Palaeobiological and ecological implications. J. Molluscan Stud. 70:359-370.

Kowalewski, M., & P. H. Kelley, editors. 2002. The fossil record of predation. The Paleontological Society Papers 8. New Haven: The Paleontological Society, Yale University Reprographics & Imaging Services. 398 pp.

Leighton, L. R. 2000. Environmental distribution of spinose brachiopods from the Devonian of New York: test of the substrate hypothesis. Palaios 15:184-193.

Leighton, L. R. 2001. New example of Devonian predatory boreholes and the influence of brachiopod spines on predator success. Palaeoecology 165:53-69.

Miller, D. J. & M. LaBarbera. 1995. Effects of foliaceous varices on the mechanical properties of Chicoreus dilectus (Gastropoda: Muricidae). J. Zool. 236:151-160.

Palmer, A. R. 1979. Fish predation and the evolution of gastropod shell sculpture: experimental and geographic evidence. Evolution Int. J. Org. Evolution 33:697-713.

Savazzi, E., editor. 1999. Functional morphology of the invertebrate skeleton. Chichester: John Wiley and Sons. 706 pp.

Stanley, S. M. 1970. Relation of shellform to life habits of the Bivalvia (Mollusca). Geological Society of America Memoir 25:1-269.

Stenzel, H. B. 1971. Oysters. In: R. C. Moore, editor. Treatise on invertebrate paleontology, part N, Mollusca. Geological Society of America, Lawrence: University of Kansas Press. pp. 953-1224.

Stone, H. M. I. 1998. On predator deterrence by pronounced shell ornament in epifaunal bivalves. Palaeontology 41:1051 1068.

Vance, R. R. 1978. A mutualistic interaction between a sessile marine clam and its epibionts. Ecology 59:679-685.

Vermeij, G. J. 1977. The Mesozoic marine revolution: evidence from snails, predators and grazers. Palaeobiology 3:245-258.

Vermeij, G. J. 1978. Biogeography and adaptation patterns of marine life. Cambridge, Massachusetts: Harvard University Press. 332 pp.

Vermeij, G. J. 1982. Unsuccessful predation and evolution. Am. Nat. 120:701-720.

Vermeij, G. J. 1987. Evolution and escalation: an ecological history of life. New Jersey: Princeton University Press, Princeton. 527 pp.

Vermeij, G. J., E. C. Dudley & E. Zipser. 1989. Successful and unsuccessful drilling predation in recent pelecypods. Veliger 32:266-273.

Wieters, E. A. & S. A. Navarrete. 1998. Spatial variability in prey preferences if intertidal whelks Nucella canaliculata and Nucella emarginata. J. Exp. Mar. Biol. Ecol. 222:133-148.

Ziegelmeier, E. 1954. Beobachtungen fiber den Nahrungserwerb bei der Naticid Lunatia nitida Donavan (Gastropoda, Prosobranchia). Helgol. Wiss. Meeresunters 5:1-33.

TABLE 1.

Number of kills; sizes of the gastropods that killed; drillhole
diameter. Note that the gastropods that killed did not differ in size
between the tanks. n = 30 for Mytilus trossulus and n = 45 for
Nucella lamellosa.

                     Mean Size of   Mean Size of     Mean
                      Gastropods     Mussels in    Drillhole     Mean
            No. of   that Killed     Experiment    Diameter    Handling
Treatment   Kills        (mm)           (mm)         (mm)      Time (h)

    1         13         45.6           43.9          2.2         82
    2         15         46.2           44.6          2.2        108
    3         23         45.5           44.2          1.8         90

TABLE 2.

Number of kills (0 = no kill; 1 = kill); sizes of the mussels
involved; standard deviation and median of the mussels. Note that
Nucella lamellosa killed and consumed smaller prey in the tank
with normal mussels than in the tanks with spinose mussels. There
was no difference in size of prey taken between mussels in
Treatments 2 and 3.

                         No. of     Mean Size
             Kill/No    Kills/No    Of Mussels
Treatment     Kill       Kills         (mm)       STD     Median

    1           0          17          46.0       6.44     45.3
    1           1          13          41.1       4.28     39.5
    2           0          15          44.5       3.92     45.1
    2           1          15          44.7       4.02     44.6
    3           0           7          43.6       3.64     42.5
    3           1          23          44.4       3.14     44.5

SEBASTIAN WILLMAN

Uppsala University, Department of Earth Sciences, Palaeobiology, Villavagen 16, SE-752 36 Uppsala, Sweden

E-mail: [e-mail]