The mismeasure of behavior: a natural history revision of prey preference in the banded tulip snail.
Snails (Natural history)
Snails (Physiological aspects)
Durham, Stephen R.
Dietl, Gregory P.
Visaggi, Christy C.
|Publication:||Name: Journal of Shellfish Research Publisher: National Shellfisheries Association, Inc. Audience: Academic Format: Magazine/Journal Subject: Biological sciences; Zoology and wildlife conservation Copyright: COPYRIGHT 2012 National Shellfisheries Association, Inc. ISSN: 0730-8000|
|Issue:||Date: April, 2012 Source Volume: 31 Source Issue: 1|
|Topic:||Event Code: 310 Science & research|
|Geographic:||Geographic Scope: United States Geographic Code: 1USA United States|
ABSTRACT The banded tulip snail, Fasciolaria (Cinctura) lilium
hunteria (Perry), is a predatory gastropod that is capable of subduing a
wide range of prey items using multiple attack behaviors. However, the
literature contains conflicting accounts of this predator's prey
preferences, which vary between the extremes of strong preferences for
snails over bivalves to opportunistic behavior in which prey are
incorporated into the diet based on their relative abundance in the
environment. Here we reexamine the extent to which prey items in the
diet ofF. hunteria are distributed in preference hierarchies to update
the natural history data on this molluscan predator. We tested F.
hunteria's preference between oysters, Crassostrea virginica
(Gmelin), and snails, Urosalpinx cinerea (Say), two ecologically
co-occurring prey items that require different attack behaviors to
subdue. Based on cost--benefit analyses, U. cinerea is more
energetically profitable than C. virginica, so we predicted that it
should be favored by tulip snails. We offered both prey simultaneously
to F. hunteria in a Y maze to test this hypothesis. Despite the vast
differences between the prey items in terms of potential biomass reward,
handling times, and risk to the predator, F. hunteria did not prefer
either C. virginica or U. cinerea, live or crushed. Our results suggest
F. hunteria has no strong preferences among prey items in its diet, and
is an opportunistic predator. This study is an example of the necessity
of revising natural history information at a time when accumulation of
such data is declining. In light of our results, we discuss the
importance of examining the sources of natural history information, and
of considering the time period and theoretical framework in which
natural history data were gathered and interpreted to prevent cascading
error effects resulting from the use of flawed natural history
KEY WORDS: prey preference, Faseiolaria, banded tulip snail, Y maze, predation, natural history
The Rebirth of Natural History
During the past century, biological research has shifted to emphasize the quantitative testing of hypotheses and theory, to the point of reducing basic descriptive and observational research on how organisms live and relate to one another (Futuyma 1998). However, recently, there has been a renewed interest in the importance of gathering natural history information (Bartholomew 1986, Greene & Losos 1988, Noss 1996a, Noss 1996b, Futuyma 1998, Grant 2000, Dayton & Sala 2001, Dayton 2003, Greene 2005, Hampton & Wheeler 2011).
Refocusing efforts on natural history research is vital because it produces the basic data that support our hypothesis testing and theoretical developments. It is also necessary to recognize that simply adding to our natural history information is not enough; it must be updated in light of new data and theory. This process includes reconciling disagreements about the behavior of species, because conflicting reports in the literature can have a negative impact on further research. Natural history data are also essential to our understanding of how ecosystems and communities function, and how we can best go about conserving ecosystems and the species that inhabit them.
The banded tulip snail, Fasciolaria (Cinctura) lilium hunteria (Perry), is an important gastropod predator in many near-shore marine ecosystems of the western Atlantic and Gulf of Mexico, ranging from North Carolina in the United States to the Yucatan Peninsula in Mexico. Banded tulip snails have been documented feeding on a variety of molluscs and onuphid worms, requiring multiple methods of attack (Wells 1958, Paine 1963). For instance, when preying on bivalves (e.g., Chione elevata (Say), Crassostrea virginica (Gmelin)), F. hunteria utilizes a wedging behavior that involves positioning on one valve of its prey and waiting until it opens to respire and feed, at which point F. hunteria wedges its apertural lip between the gaping valves, propping the bivalve open so that it can be eaten (Wells 1958, Dietl et al. 2010). However, to subdue a small gastropod prey (e.g., Nassarius vibex (Say), Urosalpinx cinerea (Say)), F. hunteria envelops its prey's shell and forces the operculum open to insert its proboscis (Wells 1958, Dietl et al. 2010).
The conclusions of two often-cited studies of F. hunteria feeding ecology disagree on the presence of prey preference in these predators. Wells (1958) examined the diet of F. hunteria experimentally and found a preference for the small predatory snail U. cinerea over the Eastern oyster (C. virginica). In contrast, Paine (1963) was not able to discern any prey preference by F. hunteria as part of his nearly year-long field observation of predatory gastropods on Baymouth Bar in Alligator Harbor, Florida. That these two studies present such drastically different results about this elementary aspect of tulip snail foraging behavior is a striking illustration of the need for more accurate natural history information.
Both studies assessed diet selection of the banded tulip snail before the development of optimal foraging theory during the late 1960s and early 1970s, which revolutionized ecologists' understanding of animal foraging behavior (Ydenberg et al. 2007). Here we reevaluate the presence of a prey preference in the banded tulip snail in the light of modern foraging theory.
Hodgson and Kitchell (1987) discussed the opposing foraging strategies of opportunism and optimal foraging as two hypotheses: the functional response hypothesis and the optimal foraging hypothesis. Functional response models predict that predator diets will "simply reflect encounter frequencies and the modifiers of probability of successful capture until satiation occurs" (i.e., opportunism, sensu Cody (1974); Wiens & Rotenberry (1979)), in contrast to optimal foraging models, which predict that a predator should always ignore low-ranked prey as long as high-ranked prey are available (Hodgson & Kitchell 1987, p. 324).
These two views have important implications for examining the prey preferences of predators. Although most predators have a preference on some level (Cody 1974, Kaspari & Joern 1993), if the optimal foraging hypothesis is accepted, then different prey types should be ranked by the predator based on energy profitability; however, if the functional response hypothesis is correct, "preference" should be correlated with the environmental abundance of the prey types in the predator's diet.
Based on these definitions, F. hunteria is an optimal forager according to Wells (1958), but a functional response-type predator according to Paine (1963). This disagreement led us to use a cost benefit analysis (Kitchell et al. 1981, Kelley 1988) to predict which prey would be chosen by F. hunteria based on energy profitability. We tested this prediction using a Y maze (Michelson 1960, Chadwick & Thorpe 1981, Bonsdorff & Vahl 1982, O'Sullivan et al. 1987, Simon & Barnes 1996).
MATERIALS AND METHODS
We began by producing cost-benefit curves for each prey type used by Wells (1958): the Eastern oyster, C. virginica, and the oyster drill, U. cinerea. All C. virginica used in our study were collected from an intertidal flat near Masonboro Island, NC (34[degrees]10'40.13" N, 77[degrees]50'32.57" W), and U. cinerea were collected from an oyster reef beneath the main bridge onto Wrightsville Beach, NC (34[degrees]13'8.21" N, 77[degrees]48'38.98" W).
The cost-benefit curves depict prey size versus prey handling time that are standardized by a measure of energy content. We used grams of ash-free dry mass (AFDM) as our energy proxy (e.g., Kingsley-Smith et al. 2009). To estimate AFDM for C. virginica, we used 22 oysters ranging in size from 26.9-95.6 mm in shell height. Shell height refers to the distance between the umbo and the growth margin. The oysters were processed at the Cornell University Stable Isotope Laboratory in Ithaca, NY, in November 2010 and January 2011. The soft bodies were dried and weighed before and after burning in a muffle furnace at 500[degrees]C for 4 h. AFDM was calculated as the postburned mass (measured in grams), subtracted from the preburned mass (measured in grams) for each specimen. The natural logarithms of these values were regressed against the logarithms of oyster height to produce a regression equation. The AFDM for U. cinerea was determined using the same method, with the exception that they were not removed from their shells and the analysis was carried out at Dartmouth College in Hanover, NH, during May 2009 and June 2009. Shell size for U. cinerea refers to the distance from the apex of the shell to the tip of the siphonal canal.
The second type of information needed to produce cost-benefit curves was handling time data for both prey types. To gather these data we confined a single F. hunteria in a 19-L aquarium with 1 prey individual, and used a Wingscapes Plantcam time-lapse camera to capture an image of the tank once every minute until the prey was consumed. Time-lapse trials were conducted from February 2010 to October 2010 at the Paleontological Research Institution in Ithaca, NY; and November 2010 to May 2011 at the Center for Marine Science at the University of North Carolina Wilmington. During these trials, salinity ranged from 25-32 [per thousand], and water temperatures ranged from 19-26[degrees]C. A predation event was defined as the time between the F. hunteria climbing onto its prey and abandoning the consumed prey. The 15 F. hunteria used ranged in size from 55.3-92.8 mm, and the prey ranged in size from 19.5-29.8 mm for U. cinerea, and 29.8-65.0 mm for C. virginica. After each test, we analyzed the photos to determine the approximate handling time for each predation event. The aquarium bottom was covered in a shallow layer (~1 cm) of well-sorted beach sand to approximate natural conditions more closely. The prey U. cinerea were allowed to roam the tank freely, whereas C. virginica were glued to rocks using a 2-part epoxy to simulate more accurately prey cemented to a reef (Harper 1991).
With these data, we produced cost-benefit regressions of the ratio of handling time and biomass against prey size (e.g., Kelley 1988). We also regressed predator size (measured in millimeters) versus handling time (measured in minutes) divided by prey size (measured in millimeters) to examine changes in handling time for each prey with different sized F. hunteria, as a test of the robustness of our prediction. We expected larger, stronger F. hunteria to open oysters more easily than smaller individuals, but F. hunteria of all sizes were expected to subdue U. cinerea prey easily.
Our test of F. hunteria's preference for either C. virginica or U. cinerea prey was conducted using a Y maze. The acute chemosensory abilities of marine gastropods are well known, and play central roles in foraging, mating, and predator avoidance (Kohn 1961, Croll 1983, Rahman et al. 2000, Weissburg et al. 2002, Ferner & Weissburg 2005). The Y-maze design took advantage of the sensory capabilities of F. hunteria to allow determination of prey choice before any actual predation event took place.
The Y-maze apparatus was constructed using ~7.6-cm diameter PVC piping. Each branch was 60 cm long, for a total flow distance of 120 cm. The apparatus was calibrated to a flow speed of approximately 4.4 cm/sec into the downstream branch of the Y maze, corresponding to a flow speed of 2.2 cm/sec from each upstream branch. These flow velocities are within the natural range typically observed on intertidal mud flats, which can vary between 0 cm/sec and 45 cm/sec, depending on tidal conditions (Finelli et al. 2000). Furthermore, Ferner and Weissburg (2005) found that the whelk Busycon carica (Gmelin) tracked odor plumes equally well at flow velocities of 1.5-5 cm/sec. Because F. hunteria and B. carica are ecologically similar predators, we considered our flow speeds to be appropriate for F. hunteria. Evenness of flow between the two upstream branches and the absence of eddying from one branch into the other were monitored by periodically dripping different food colorings into each upstream branch and observing the subsequent flows. Ferner and Weissburg (2005) suggested that flow turbulence is another important factor in determining tracking success by some marine predators, but they did not find a statistically significant difference in tracking success by B. carica in turbulent flows versus laminar flows, so turbulence was not considered a likely inhibition to chemoreception by F. hunteria in this study. To give the banded tulip snails traction and to simulate more accurately their natural environment, the bottom of the Y maze was covered with ~1 cm of beach sand. Tests were also conducted away from direct light to avoid branch choice bias associated with shadows along the walls of the Y maze.
We conducted all preference tests in July 2009 and October 2009 at the Center for Marine Science at the University of North Carolina Wilmington using running, flow-through seawater from Masonboro Sound. During July 2009, salinity ranged from 26- 34 [per thousand], and water temperature ranged from 24-32[degrees]C. Salinity and water temperature ranged from 30-37 [per thousand] and 20-24[degrees]C, respectively, during October 2009. We used 22 F. hunteria that ranged in size from 64-100 mm. Individual F. hunteria were not starved prior to preference testing, and were fed C. virginica and U. cinerea periodically between trials.
Prior to experimentation, several different F. hunteria individuals were given a choice of prey or raw seawater and all chose the branch containing prey, verifying our setup. For the experimental trials, an F. hunteria chose between C. virginica or U. cinerea prey. Both crushed prey and live prey were used to account for the possibility that live oysters may not be open and producing effluent throughout the test. When using crushed prey, 1 small C. virginica (24-60 mm) and 3-6 U. cinerea (14-26 mm), depending on oyster size, were used. During the live tests, 2 small C. virginiea (24-40 mm) and 6 U. cinerea (12-25 mm) were used to increase effluent volume. During every trial, the prey groups were each confined to a mesh bag and submerged in 1 of the 2 upstream branches. Each prey type was kept in the same mesh bag throughout the trial period. The C. virginica branch was determined by a coin toss, and we switched the prey branches after a successful test. Crushed prey were not used for more than 3 tests, no matter the outcome, to keep effluent concentrations from becoming too low. Live prey were used until a result was achieved, but were not used for more than 2 successful tests in a row.
Tests were allowed to run for up to 10 min, unless the F. hunteria was in motion at 10 min. A branch choice was made by a banded tulip snail as soon as the apex of the snail's shell passed the branch divide. After a test was completed, the F. hunteria was removed and the prey branches were changed if necessary. Between each test, the substrate in the Y maze was mixed to clear odors from the sand bed (Ferner & Weissburg 2005), and then the apparatus was allowed to run for at least 2 min to clear any effluent from the preceding test from the maze. This time was deemed sufficient based on how long it took the apparatus to clear the food coloring during the flow monitoring described earlier.
The snail U. cinerea had an average AFDM of 0.086 g, and C. virginica's average AFDM was 0.199 g. The biomass regressions produced from these data are given in Figure 1. During the handling time analysis, 8 predation events were observed with U. cinerea, and 10 with C. virginica (Table 1). On average, it took F. hunteria about 131 min to subdue and consume 1 U. cinerea, with a minimum handling time of 65 min and a maximum of 225 min. In contrast, F. hunteria handled C. virginica individuals for an average time of approximately 701 min, with a minimum and maximum handling time of 96 min and 2,790 min, respectively.
Our cost benefit curves demonstrated that U. cinerea is a more energetically favorable prey item than C. virginica (Fig. 2). Aside from 1 large oyster that was killed and consumed quickly (120 min), the U. cinerea prey items typically had a lower ratio of handling time (cost) to biomass (benefit) in comparison with C. virginica prey.
When predator length (measured in millimeters) was regressed against the ratio of handling time (measured in minutes) and prey size (measured in millimeters; Fig. 3), we found that C. virginica prey become more favorable in terms of handling time as F. hunteria grow larger, but U. cinerea prey take approximately the same amount of time to consume for a wide size range of F. hunteria.
A total of 140 trials were run with the Y maze--89 with live prey and 51 using crushed prey. Of these trials, F. hunteria chose a prey item 43 times. Of the successful trials, 22 occurred while using live prey and 21 with crushed prey. During trials in which F. hunteria made no prey choice, they most often exhibited 1 of 3 behaviors: they did not move at all; they moved downstream, away from the prey items; or they attempted to bury into the sediment. These snails were considered "unmotivated foragers" (sensu Ferner and Weissburg 2005). During the 22 choice events using live prey, F. hunteria chose each prey 50% of the time: C. virginiea 11 times and U. cinerea 11 times, resulting in a choice ratio of 1.00 (Table 2). Of 21 results with crushed prey, F. hunteria chose C. virginica in 43% of tests and U. einerea in 57% of tests, or 9 C. virginica and 12 U. cinerea, producing a choice ratio of 0.75 (Table 2).
Before testing our prediction, based on the cost benefit analysis, that F. hunteria prefers U. cinerea prey, we checked for treatment biases in our data set. Differences between prey choice results from the trial periods in July 2009 and October 2009 were not significant (Crushed prey: Fisher's exact test, P = 0.670; Live prey: Fisher's exact test, P = 1.000; Table 2), so we pooled the data from the 2 months for analysis. Similarly, a chi-square test suggested no significant difference between prey choice data from trials using crushed versus live prey (chi-square = 0.22, df = 1, P = 0.639; Table 2), so we pooled the live and crushed data from the 2 trial periods for our analysis of prey choice. After confirming no bias in our treatments, a chi-square test demonstrated that our results show a significant deviation from our expectation of a preference for U. cinerea (chi-square = 26.06, df= 1, P < 0.0001; Table 2).
[FIGURE 1 OMITTED]
In addition, of the 15 F. hunteria with which we had more than 1 successful trial, 10 individuals (67%) chose each branch and each prey in at least 1 of the tests, and only 2 individuals (13 %) chose the same prey in each test. The branch choice ratio, right branch to left branch, was I. 15, further suggesting a lack of choice bias by F. hunteria (53% and 47%, respectively; chi-square = 0.12, df= 1, P = 0.733; Table 2). These results are consistent with a lack of preference by F. hunteria for either U. cinerea or C. virginica prey.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
Wells (1958) or Paine (1963)?
Our results lend support to Paine's (1963) observations of opportunism in F. hunteria and do not agree with those of Wells (1958). Despite the fact that our cost benefit analysis showed that U. cinerea is the more energetically favorable prey item, F. hunteria chose prey randomly between U. cinerea and C. virginica, resulting in a prey choice ratio of approximately 1.00. Paine (1963) also observed F. hunteria feeding on almost a 50:50 ratio of gastropods and bivalves (9 F. hunteria preying on bivalves and 10 F. hunteria feeding on gastropods), as well as onuphid worms, barnacles, and carrion. Our study thus resolves the conflicting reports of prey preference by Paine (1963) and Wells (1958).
Reconciling Wells (1958)
Weak Preferences and Prey Switching
Clearly our data and those of Paine (1963) suggest that F. hunteria is an opportunist, but how, then, can the results of Wells (1958) be explained? Wells' results do suggest his F. hunteria had a weak preference for U. cinerea over C. virginica prey. Wells (1958) offered groups of F. hunteria 4 different ratios of U. cinerea to C. virginica prey: 1:2, 1:1, 2:1, and 4:1. More U. cinerea than C. virginica were consumed during the interval when both prey were available to F. hunteria in all 4 of Wells' treatments; and in the 2:1 and 4:1 treatments, all U. cinerea were consumed at least a day before any C. virginica were killed. Such a result is puzzling because as an opportunistic predator we would expect F. hunteria to switch prey as the relative abundance reversed, considering the attack rate should approximate the predators' encounter rates with prey. However, Wells' (1958) results and ours are not mutually exclusive.
First, we must distinguish between weak preferences and strong preferences (sensu Murdoch 1969). According to Murdoch (1969), weak preferences can be changed easily throughout the lifetime of an individual. In other words, weak preferences are not innate; they have no genetic basis. The snail U. cinerea, itself a predatory gastropod, is one example of a forager known to develop weak preferences (Wood 1968). Wood (1968) pointed out that repeated ingestion of prey tissue is required to condition U. cinerea when feeding on barnacles, mussels, and oysters, and that although U. cinerea "can and apparently does develop transient predilections for specific prey, it is not held rigidly to such preferences by genetic limitation" (p. 315). In contrast, a strong preference is one that is difficult or impossible to change via conditioning and, we add, is at least partially genetically determined. A good example of this type of preference is that of the predatory drilling snail Nucella canaliculata (Duclos) for the mussel Mytilus trossulus (Gould) along the coast of Washington and Oregon (Sanford et al. 2003). Even given the year-long duration of the study by Sanford et al. (2003), several N. canaliculata proved incapable of learning to prey on the very similar mussel Mytilus californianus (Conrad). As opportunistic predators, it seems likely that F. hunteria may be susceptible to development of a weak prey preference by conditioning, as described by Murdoch (1969), and we suggest that it was likely this type of preference that Wells (1958) observed.
Weak preferences may arise in opportunistic foragers because experience and learning occupy the role in determining prey choice that is filled genetically for individuals with strong preferences. Although the mechanisms for learning in gastropods are not completely understood, when repeatedly exposed to a single prey type, many predatory gastropods improve their feeding efficiency on prey with experience (Wood 1968, Hughes & Croy 1993). Furthermore, the number of predation events needed to improve feeding efficiency on novel prey is remarkably consistent between very disparate predator types: 5-10 events for fishes, crabs, and snails (Hughes et al. 1992), suggesting that learning, to some extent, is a common capability in marine foragers.
Memory is another factor, related to learning, that is likely to be important in the formation of weak preferences. The rule of thumb mechanism (Bergelson 1985) is one example of how memory may play a role. Bergelson's (1985) "rule of thumb" states that a predator should choose the prey it most recently successfully captured for its next meal. This mechanism could influence weak preferences as long as snails have sufficient memory to recall at least their previous meal. Indeed, Belisle and Cresswell (1997) model results demonstrate that even a very limited memory capacity (5-20 feeding events) is sufficient to facilitate discriminatory foraging behavior.
Fasciolaria hunteria in Wells' (1958) treatments had ample opportunity to consume 5 10 U. cinerea, and thus, potential to develop a weak preference for them. If Wells' banded tulip snails were offered the choice between U. cinerea and C. virginica prey in our Y maze immediately after his study, we would expect them to have preferred U. cinerea. However, by design, the Y maze precludes ingestive conditioning of the predators--a process that both Wood (1968) and Hughes and Croy (1993) suggest is important for forming preferences in gastropods--so it is not surprising that F. hunteria in our study demonstrated no preference for either prey item.
Effects of Prey Abundance
Another important consideration when examining Wells' (1958) results is the relative ease with which F. hunteria can find U. cinerea prey in enclosures. The snail U. cinerea is possibly far more difficult to locate in F. hunteria's natural habitat. Certainly, the natural ratio of U. cinerea to C. virginica in Masonboro Sound, NC, likely is far less than 1:2. Although densities can be quite high (106-947/[m.sup.2]), most reported densities of U. cinerea on oyster reefs are relatively low (6-29/[m.sup.2]), suggesting U. cinerea/ C. virginica ratios may fluctuate in nature, but that high densities of U. einerea prey are patchy and relatively uncommon (see Carriker (1955, p. 17) and White and Wilson (1996, p. 560), for discussions of U. cinerea densities on oyster reefs).
In contrast to U. cinerea densities, oyster reefs can easily support densities of hundreds to more than 1,000 oysters/[m.sup.2] (Schulte et al. 2009). Although densities vary dramatically depending on numerous environmental factors, in addition to harvesting intensity by humans (Schulte et al. 2009), the generally higher range of C. virginica densities relative to the range of U. einerea densities suggests that the treatments used by Wells (1958) are probably unlikely to occur in a natural oyster reef environment except in localized, temporally restricted situations.
When U. cinerea and C. virginica prey are enclosed with F. hunteria, search times become unrealistically short, but it is still important to consider handling times. Our handling time data suggest that, on average, C. virginica takes almost 7 times longer to consume than U. cinerea. Thus, F. hunteria in Wells' (1958) treatments that chose to prey on oysters almost certainly took longer to eat them, and individuals that successfully opened an oyster may have been more thoroughly satiated compared with their U. cinerea-attacking conspecifics, depending on the sizes of C. virginica offered by Wells (1958). In addition to handling time and relative biomass differences, another factor that may have influenced the lack of switching by F. hunteria in Wells' (1958) treatments is the dangerous nature of C. virginiea prey relative to U. cinerea, because during wedging, F. hunteria risk breaking their own shell (C. virginica as dangerous prey are discussed in the next section).
Furthermore, results of Murdoch (1969), who examined switching between 2 weakly preferred prey items by the predatory gastropod Acanthinucella spirata (Blainville), showed that switching does not always result when the relative abundance of 2 prey items in a predator's environment become reversed as a result of consumption of the more abundant prey. Murdoch (1969) hypothesized that for switching to occur, there must be an opportunity for the predator to become trained on the abundant prey type, such as may occur in patchy environments. Thus, Murdoch (1969) suggested, a switching pattern will not develop if 2 prey are distributed in such a way that encounters with both are relatively frequent, because the altemating training period does not have an effect large enough to make the proportion of the most abundant prey item larger in the predator's diet than in the total prey population. Based on Murdoch's (1969) findings, Wells' (1958) treatment conditions were not conducive to the development of prey switching because of their short duration (U. cinerea prey were never present for more than 5 days, and all trials were stopped after 8 days), and both prey items were present in relatively large numbers. Thus, the preference observed by Wells (1958) was likely weaker than his data seem to suggest.
Effects of Dangerous Prey
It is important to highlight that C. virginica and U. cinerea are not equal prey. For our analysis we focused on differences in biomass and handling time, but another dimension of inequality between these prey items is the risk they pose to F. hunteria. Dietl et al. (2010) examined wedging-induced repair scars on F. hunteria caused by predation on C. virginica and found that shell damage to the predator occurred in 22% of successful attacks, and also that 49% of cases of shell damage from wedging occurred during unsuccessful attacks. In contrast, F. hunteria predation on gastropods, including U. cinerea, does not pose a risk of damage to F. hunteria's shell and, once caught, U. cinerea appears to be a certain meal for F. hunteria (Dietl et al. 2010, pers. obs.).
These observations suggest that C. virginica is a more dangerous prey than U. cinerea, and support for this can be found in our data. For instance, the scattered nature of the data points in the C. virginica curve in Figure 2 reflect the varied lengths of time it took for C. virginica to gape and allow the waiting F. hunteria to begin wedging. This variation illustrates that predation by wedging is, to some extent, controlled by the prey, and the documentation of failed predation attempts by Dietl et al. (2010) and our handling time data demonstrate that wedging behavior offers an uncertain meal for lengthy handling time commitment and potential shell damage.
We also observed a trend toward decreasing C. virginica handling time with increasing F. hunteria shell length (Fig. 3). This trend suggests that larger, more powerful F. hunteria are more capable of wedging oysters than smaller individuals and juveniles. In contrast, the relationship between handling time and predator shell length for U. cinema prey indicates that U. cinerea were almost equally vulnerable to all F. hunteria sizes tested.
Criticisms of Binary Choice
Although some researchers have proposed problems with binary choice experiments (Schuck-Paim & Kacelnik 2007, Pavlic & Passino 2010), we believe that their criticisms do not apply to our system. Pavlic and Passino (2010) discuss how binary, mutually exclusive choice experiments may not test realistic scenarios because simultaneous encounters with prey are rare in nature, and even when they occur they may not be mutually exclusive. Pavlic and Passino (2010) also suggest that in binary choice experiments, a predator will select the prey with the lower handling time first.
Neither criticism applies to our system because simultaneous encounters with the effluent trails of C. virginica and U. cinerea must occur relatively frequently for F. hunteria inhabiting oyster reef environments where these 2 prey items coexist (e.g., as is the case for bees foraging on flowers (Waddington & Holden 1979)). Furthermore, for our system, it is likely that a simultaneous encounter with C. virginica and U. cinerea will be mutually exclusive because U. cinerea is a mobile prey item, capable of fleeing should F. hunteria choose to prey on the immobile C. virginica first. Handling times for U. cinerea were also less than those for C. virginica on average, such that we would expect F. hunteria always to choose U. cinema first ifPavlic and Passino (2010) are correct, but this was not the case (Table 2).
We did consider testing our prediction using serial tank experiments offering each prey to F. hunteria individually, and in a third combined prey treatment, as suggested by Underwood and Clarke (2005), but analysis of results from this method is complicated ifa mix of mobile and immobile prey is used, as was the case for our system. We also considered offering prey simultaneously to F. hunteria and analyzing the data based on order of prey consumed (sensu Taplin (2007)), but the methods of Taplin (2007) required monitoring the order of consumption of prey items in the group, which in our view potentially mixed effects such as differences in handling time between prey items and encounter rates, depending on both the relative mobility of the prey types being tested and changes in relative abundance as prey were depleted. For these reasons, we decided to use the Y maze.
Another criticism by Pavlic and Passino (2010) and SchuckPaim and Kacelnik (2007) is that the common practice of preceding preference experiments by starving the study subjects may broaden artificially the subjects' diets, and bias them toward the food option with the lowest handling time. However, in some groups, diet breadth actually increases as foragers near satiation, indicating preexperimental feeding may not necessarily guarantee the average natural response to a prey choice (Richards 1983). In the case of our study, although F. hunteria were fed infrequently in the laboratory, they were not deliberately starved for our preference tests. Furthermore, the almost 50:50 ratio of C. virginiea to U. cinerea chosen in the Y maze would not be expected if the F. hunteria were biased toward U. cinerea because of its lower handling time.
Experimental Design Assumptions
Gastropods are known to follow the effluent trails of their prey and to react defensively to those of their predators (Croll 1983, Kohn 1961, Rahman et al. 2000, Weissburg et al. 2002, Ferner & Weissburg 2005), yet we know little about how gastropods process these important signals. We assumed F. hunteria was capable of sensing multiple effluent trails at once and did not process them serially (i.e., F. hunteria are capable of choosing among multiple effluent trails).
We also assumed that predation risk would not affect the prey preference we observed. Although more work is necessary to support this assumption further, we believe it is justified because preliminary data suggest predator effluent depresses feeding rates by F. hunteria, but does not appear to induce prey preference changes (unpubl. data). More work is needed to understand how snails respond to effluent plumes from both prey and predators.
Last, we assumed for this study that a third prey option would not affect the preference for C. virginica or U. cinerea. The diet of F. hunteria is broad (Wells 1958, Paine 1963), and more prey choices should be tested, such as onuphid worms, to examine the generality of our conclusions.
Geographical Variation and Coevolution
Although our data and those of Paine (1963) both suggest F. hunteria is an opportunistic predator without a strong preference hierarchy, its catholic diet across the species range does not necessarily mean that our conclusions apply to all populations within the range. Although we are unaware of a species that switches from opportunist to optimal forager between populations, there is evidence for geographical shifts in prey preference by predatory molluscs (Sanford et al. 2003), and diet variation between populations of a single species inhabiting different environments (Kohn 1968).
Despite their narrow geographical scope, our results pose interesting questions about how F. hunteria is interacting with its prey from a coevolutionary perspective. The hypothesis of coevolutionary alternation (Thompson 2005) postulates that, over evolutionary time, as natural selection favors increasingly effective defenses in a predator's current preferred prey, the predator's preference hierarchies will shift to emphasize in its diet another prey item with lower defenses. Because there is a fitness cost to maintaining most defenses, Thompson (2005) hypothesizes that a formerly preferred prey item's high levels of defense will be selected against as predation pressure decreases, until it again becomes a favorable prey item for the predator, resulting in an alternating pattern of prey preference.
One of the assumptions of the coevolutionary alternation hypothesis is that preference hierarchies are genetically determined, but our study demonstrates that not all foragers' diets are arranged in preference hierarchies. At first glance, F. hunteria seems to be a promising subject for studying coevolutionary alternation; with a broad diet, there are plenty of options among which to switch. However, our results currently do not support a preference hierarchy in the diets ofF. hunteria from Masonboro Inlet, NC, suggesting that the coevolutionary alternation hypothesis may not be applied to this predator-prey system. In addition, as discussed by Murdoch (1969), prey switching is not necessarily evidence of a preference hierarchy, nor is a lack of prey switching necessarily indicative of a strong preference for that prey. Further research on F. hunteria populations across a wider geographical range is required to establish the generality of our conclusions and to examine the coevolutionary implications of predator--prey interactions in the absence of genetically based preference hierarchies.
Amending Natural History
Our analysis serves as a case study of the importance of keeping old natural history data current. Updating natural history information and resolving conflicting accounts in the literature form an essential foundation on which future hypotheses and theory will be based. Ecological theory has vastly increased our understanding of animal behavior in recent decades, and although not all advances have required thorough knowledge of many species, "comparative studies that rely on knowing the ... idiosyncrasies of organisms are among our most important tests of general hypotheses" (Futuyma 1998, p. 3). Thus, continued progress in foraging theory depends on the accumulation of accurate natural history data.
Just as taxonomy is updated continuously to reflect new advances in species identification techniques and theories, we must treat natural history data the same way--updating and reinterpreting it in light of current theory. In the case of our system, 2 studies published only half a decade apart disagreed on F. hunteria prey preference, and both predated optimal foraging theory. This discrepancy was a clear case in which the data interpretation required revision, and use of the wrong study could potentially have misled researchers. Bortolus (2008) cautioned against "error cascades" resulting from the use of flawed taxonomy; seemingly minor mistakes in species identification can propagate into an error in experimental ecology and on into errors in large-scale ecological studies or environmental management that can have serious direct effects on scientific knowledge and ecosystems. Inaccuracies in natural history, like those in taxonomy, are likely to have cascading error effects because natural history data underlie so much of our theory and ecological understanding. Indeed, Hampton and Wheeler (2011) describe "the observation and description of nature as the foundational step in science" (p. 1).
It is easy to envision how an incomplete understanding of predator-prey interactions could confound management of the natural systems. For instance, F. hunteria are important predators on oyster reefs--ecosystems that are functionally extinct in many locations worldwide and are currently the subject of intense restoration efforts (Beck et al. 2011). In this context, categorizing F. hunteria as an opportunist versus a specialist predator could influence management of these habitats because the role of such a conspicuous predator in ecosystem function would be interpreted differently depending on whether F. hunteria is assumed to feed intensely on one prey type or more diffusely on many prey.
Other cases of contradictory natural history reports undoubtedly exist in the literature. It is important for biologists to compare the sources of their natural history information, and to consider when, and in what theoretical framework, the data were gathered and interpreted. This vigilance is necessary to prevent the propagation of misinformation that can easily mislead other researchers, and may ultimately directly affect environmental management and conservation efforts.
In conclusion, our results support a lack of a strong prey preference by F. hunteria for either C. virginica or U. cinerea, suggesting these gastropod predators do not exhibit a prey preference hierarchy. The banded tulip snail behaves opportunistically, and our results support the field observations of Paine (1963) in Florida. Although more prey species need to be tested and further study must be conducted using individuals from across the geographic range of F. hunteria, our methods may hold promise for examining opportunism more broadly, as well as the evolutionary (and coevolutionary) implications of opportunism, using an important and conspicuous nearshore marine predatorprey system. Last, our study demonstrates the importance of continuing to gather natural history data, of updating existing natural history observations in light of new theories and discoveries, and of resolving conflicting reports of the behaviors of animals in the literature to facilitate future research.
We thank Brad Taylor and Ryan Calsbeek of Dartmouth College for valuable comments and discussions during the early stages of this project, and Christopher Finelli of the University of North Carolina Wilmington for advice on calculating flow velocities. We are also grateful to Ron Moore, Rob Deanes, and Troy Alphin for help with Y-maze setup, to Brad Parnell and Dana Friend for help with monitoring handling time trials, and to the Benthic Lab at the University of North Carolina Wilmington for providing salinity and water temperature data (www.ncoystermonitoring.org). This study was supported in part by a grant from the National Science Foundation (NSF EAR 055109).
Bartholomew, G. A. 1986. The role of natural history in contemporary biology. Bioscience 36:324-329.
Beck, M. W., R. D. Brumbaugh, L. Airoldi, A. Carranza, L. D. Coen, C. Crawford, O. Defeo, G. J. Edgar, B. Hancock, M. C. Kay, H. S. Lenihan, M. W. Luckenbach, C. L. Toropova, G. Zhang & X. Guo. 2011. Oyster reefs at risk and recommendations for conservation, restoration, and management. Bioscience 61:107-116.
Belisle, C. & J. Cresswell. 1997. The effects of a limited memory capacity on foraging behavior. Theor. Popul. Biol. 52:78-90.
Bergelson, J. M. 1985. A mechanistic interpretation of prey selection by Anaxjunius larvae (Odenata: Aeschnidae). Ecology 66:1699-1705.
Bonsdorff, E. & O. Vahl. 1982. Food preference of the sea urchins Echinus acutus and E. esculentus. Mar. Behav. Physiol. 8:243-248.
Bortolus, A. 2008. Error cascades in the biological sciences: the unwanted consequences of using bad taxonomy in ecology. Ambio 37:114-118.
Carriker, M. R., 1955. Critical review of biology and control of oyster drills Urosalpinx and Eupleura. Spec. Sci. Rep. Fish. 148. 150 pp. Chadwick, S. R. & J. P. Thorpe. 1981. An investigation of some aspects of bryozoan predation by dorid nudibranchs (Mollusca: Opisthobranchia). In: G. P. Larwood & C. Nielsen, editors. Recent and fossil Bryozoa. Fredensborg: Olsen & Olsen. pp. 51-58.
Cody, M. L. 1974. Competition and the structure of bird communities. Princeton, N J: Princeton University Press. 326 pp. Croll, R. N. 1983. Gastropod chemoreception. Biol. Rev. Camb. Philos. Soc. 58:293-319.
Dayton, P. K. 2003. The importance of the natural sciences to conservation. Am. Nat. 162:1-13.
Dayton, P. K. & E. Sala. 2001. Natural history: the sense of wonder, creativity and progress in ecology. Sci. Mar. 65:199-206.
Dietl, G. P., S. R. Durham & P. H. Kelley. 2010. Shell repair as a reliable indicator of bivalve predation by shell-wedging gastropods in the fossil record. Palaeogeogr. Palaeoclimatol. Palaeoecol. 296:174-184.
Ferner, M. C. & M. J. Weissburg. 2005. Slow-moving predatory gastropods track prey odors in fast and turbulent flow. J. Exp. Biol. 208:809-819.
Finelli, C. M., N. D. Pentcheff, R. K. Zimmer & D. S. Wethey. 2000. Physical constraints on ecological processes: a field test of odormediated foraging. Ecology 81:784-797.
Futuyma, D. J. 1998. Wherefore and whither the naturalist? Am. Nat. 151:1-6.
Grant, P. R. 2000. What does it mean to be a naturalist at the end of the twentieth century? Am. Nat. 155:1-12.
Greene, H. W. 2005. Organisms in nature as a central focus for biology. Trends Ecol. Evol. 20:23-27.
Greene, H. W. & J. B. Losos. 1988. Systematics, natural history, and conservation. Bioscience 38:458-462.
Harper, E. M. 1991. The role of predation in the evolution of cementation in bivalves. Palaeontology 34:455-460.
Hampton, S. E. & T. A. Wheeler. 2011. Fostering the rebirth of natural history. Biol. Lett. DOI:10.1098/rsbl.2011.0777.
Hodgson, J. R. & J. F. Kitchell. 1987. Opportunistic foraging by largemouth bass (Micropterus salmoides). Am. Midl. Nat. 118:323-336.
Hughes, R. N. & M. I. Croy. 1993. An experimental analysis of frequency-dependent predation (switching) in the 15-spined stickleback, Spinachia spinachia. J. Anim. Ecol. 62:341-352.
Hughes, R. N., M. J. Kaiser, P. A. Mackney & K. Warburton. 1992. Optimizing foraging behavior through learning. J. Fish Biol. 41: 77-91.
Kaspari, M. & A. Joern. 1993. Prey choice by three insectivorous grassland birds: reevaluating opportunism. Oikos 68:414-430.
Kelley, P. H. 1988. Predation by Miocene gastropods of the Chesapeake group: stereotyped and predictable. Palaios 3:436-448.
Kingsley-Smith, P. R., H. D. Harwell, M. L. Kellogg, S. M. Allen, S. K. Allen, Jr., D. W. Meritt, K. T. Paynter, Jr. & M. W. Luckenbach. 2009. Survival and growth of triploid Crassostrea virginica (Gmelin,
1791) and C. ariakensis (Fujita, 1913) in bottom environments of Chesapeake Bay: implications for an introduction. J. Shellfish Res. 28:164-184.
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. 1961. Chemoreception in gastropod mollusks. Am. Zool. 1:291-308.
Kohn, A. J. 1968. Microhabitats, abundance and food of Conus on atoll reefs in the Maldive and Chagos Islands. Ecology 49:1046-1062.
Michelson, E. H. 1960. Chemoreception in the snail Australorbis glabratus. Am. J. Trop. Med. Hyg. 9:480-487.
Murdoch, W. W. 1969. Switching in general predators: experiments on predator specificity and stability of prey populations. Ecological Monographs 39:335-354.
Noss, R. F. 1996a. Conservation or convenience? Conserv. Biol. 10: 921-922.
Noss, R. F. 1996b. The naturalists are dying off. Conserv. Biol. 10:1-3.
O'Sullivan, J. B., R. R. McConnaughey & M. E. Huber. 1987. A bloodsucking snail: the Cooper's nutmeg, Cancellaria cooperi Gabb, parasitizes the California electric ray, Torpedo californica Ayres. Biol. Bull. 172:362-366.
Paine, R. T. 1963. Trophic relationships of 8 sympatric predatory gastropods. Ecology 44:63-73.
Pavlic, T. P. & K. M. Passino. 2010. When rate maximization is impulsive. Behav. Ecol. Sociobiol. 64:1255-1265.
Rahman, Y. J., R. B. Forward, Jr. & D. Rittschof. 2000. Responses of mud snails and periwinkles to environmental odors and disaccharide mimics offish odor. J. Chem. Ecol. 26:679-695.
Richards, L. J. 1983. Hunger and the optimal diet. Am. Nat. 122:326-334.
Sanford, E., M. S. Roth, G. C. Johns, J. P. Wares & G. N. Somero. 2003. Local selection and latitudinal variation in a marine predator-prey interaction. Science 300:1135-1137.
Schuck-Paim, C. & A. Kacelnik. 2007. Choice processes in multialternative decision making. Behav. Ecol. 18:541-550.
Schulte, D. M., R. P. Burke & R. N. Lipcius. 2009. Unprecedented restoration of a native oyster metapopulation. Science 325:1124-1128.
Simon, T. W. & K. Barnes. 1996. Olfaction and prey search in the carnivorous leech Haemopis marmorata. J. Exp. Biol. 199:2041-2051.
Taplin, R. H. 2007. Experimental design and analysis to investigate predator preferences for prey. J. Exp. Mar. Biol. Ecol. 344:116-122.
Thompson, J. N. 2005. The geographic mosaic of coevolution. Chicago, IL: The University of Chicago Press. 443 pp.
Underwood, A. J. & K. R. Clarke. 2005. Solving some statistical problems in analyses of experiments on choices of food and on associations with habitat. J. Exp. Mar. Biol. Ecol. 318:227-237.
Waddington, K. D. & L. R. Holden. 1979. Optimal foraging: on flower selection by bees. Am. Nat. 114:179-196.
Weissburg, M. J., M. C. Ferner, D. P. Pisut & D. L. Smee. 2002. Ecological consequences of chemically mediated prey perception. J. Chem. Ecol. 28:1953-1970.
Wells, H. W. 1958. Predation of pelecypods and gastropods by Fasciolaria hunteria (Perry). Bull. Mar. Sci. 8:152-166.
White, M. E. & E. A. Wilson. 1996. Predators, pests, and competitors. In: V. S. Kennedy, R. I. E. Newell, & A. F. Eble, editors. The Eastern oyster: Crassostrea virginica. College Park, MD: Maryland Sea Grant College. pp. 559-579.
Wiens, J. A. & J. T. Rotenberry. 1979. Diet niche relationships among North American grassland and shrubsteppe birds. Oecologia 42: 253-292.
Wood, L. 1968. Physiological and ecological aspects of prey selection by the marine gastropod Urosalpinx cinerea (Prosobranchia: Muricidae). Malacologio 6:267-320.
Ydenberg, R. C., J. S. Brown & D. W. Stephens. 2007. Foraging: an overview. In: D. W. Stephens, J. S. Brown & R. C. Ydenberg, editors. Foraging: behavior and ecology. Chicago, IL: University of Chicago Press. pp. 1-28.
STEPHEN R. DURHAM, (1) * ([dagger]) GREGORY P. DIETL, (1) AND CHRISTY C. VISAGGI (2)
(1) Paleontological Research Institution, 1259 Trumansburg Road, Ithaca, NY, 14850; (2) Department of Biology and Marine Biology, University of North Carolina Wilmington, 601 S. College Road, Wilmington, NC, 28403
* Corresponding author. E-mail: email@example.com
([dagger]) Current address: Department of Earth and Atmospheric Sciences, 1142 Snee Hall, Cornell University, Ithaca, NY 14853
TABLE 1. Summary of handling time observations and biomass estimates from biomass regression equations. Prey Prey F. hunteria Handling Biomass Species Size (mm) Size (mm) Time (min) (g AFDM) U. cinerea 22.7 79.9 65 0.062 U. cinerea 21.0 92.8 66 0.047 U. cinerea 21.6 79.9 67 0.052 U. cinerea 24.9 91.9 140 0.086 U. cinerea 21.5 58.1 225 0.051 U. cinerea 22.7 55.3 136 0.061 U. cinerea 29.8 68.4 202 0.159 U. cinerea 22.7 63.9 144 0.061 C. virginica 62.7 78.5 810 0.146 C. virginica 29.8 63.9 845 0.019 C. virginica 44.3 55.3 1,230 0.056 C. virginica 55.6 58.1 2,790 0.105 C. virginica 35.8 78.3 244 0.031 C. virginica 36.3 83.2 488 0.033 C. virginica 30.0 74.6 96 0.019 C. virginica 65.0 81.8 120 0.162 C. virginica 34.3 68.3 244 0.028 C. virginica 37.8 56.9 145 0.036 TABLE 2. Summary of preference test results. Crushed Prey F. hunteria F. hunteria Oyster Prey Time Specimen Size (mm) Branch Choice (min:sec) 10 85.8 L U 6:46 1 77.5 R U 6:56 3 70.6 R U 5:00 4 75.6 L U 7:00 10 85.8 R C 9:25 7 79.2 L C 4:56 9 64.0 L C 12:44 4 75.6 R C 9:43 8 69.1 L C 12:05 3 70.6 R U 4:23 11 87.7 L U 3:47 13 100.4 R C 8:07 12 83.0 R U 5:48 11 87.7 L U 2:17 12 83.0 R C 4:34 18 89.7 R U 8:40 18 89.7 R U 11:30 12 83.0 L C 11:23 21 88.0 R U 14:22 20 89.4 L U 21:09 20 89.4 R C 10:44 Live Prey F. hunteria F. hunteria Oyster Prey Time Specimen Size (mm) Branch Choice (min:sec) 5 73.7 R U 4:42 8 69.1 L C 2:31 6 81.8 L U 6:46 7 79.2 R U 4:18 3 70.6 L C 1:50 8 69.l R U 2:15 5 73.7 L C 8:24 6 81.8 R C 3:51 1 77.5 R C 5:57 6 81.8 L U 4:59 17 92.8 R U 3:52 11 87.7 L C 4:49 11 87.7 L U 3:54 15 86.6 R C 10:40 18 89.7 R U 11:21 18 89.7 L C 5:22 12 83.0 L U 11:58 12 83.0 R U 10:39 13 100.4 L C 11:56 15 86.6 R C 6:25 21 88.0 R C 12:20 11 87.7 L U 5:52 Shaded results are from July 2009, unshaded results are from October 2009. Oyster Drill (U. cinerea) Branch data are simply the opposite of the Oyster (C. virginica) Branch category, so are not listed here. Results are listed in the order in which they were acquired during our trials. (C = Crassostrea virginica prey, U = Urosalpinx cinerea prey, R = right branch, L = left branch).
|Gale Copyright:||Copyright 2012 Gale, Cengage Learning. All rights reserved.|