Phenotypic selection in an intertidal snail: effects of a catastrophic storm
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
MARINE ECOLOGY PROGRESS SERIES
Published May 22
Mar Ecol Prog Ser
Phenotypic selection in an intertidal snail:
effects of a catastrophic storm
Geoffrey C. Trussell*
Department of ZoologyICenter for Marine Biology, University of New Hampshire, Durham, New Hampshire 03824, USA
ABSTRACT: Littorina obtusata exhibits clear morphological variation (e.g. shell height, shell length,
and aperture area) among shores differentially exposed to wave energies. Selection imposed by the
hydrodynamic environment is often invoked to explain the correlation between morphology and wave
exposure in intertidal organisms, but rarely is this hypothesis tested. I examined the effects of a cata-
strophic storm on the shell length and relative shell height and aperture area of L. obtusata populations
on 2 protected and 1 wave-exposed shore in New England (USA) to test this hypothesis. Snails sampled
after the storm had relatively squatter shells than those sampled before the storm, which is consistent
with the pattern in shell height found in natural populations. The rapid shift supports the hypothesis
that shell morphology on wave-exposed shores reflects, in part, selection imposed by hydrodynamic
stress. Compared to morphological differences found between natural populations, shifts in relative
shell height after the storm were small. Hence, despite its magnitude, the impact of this storm may be
limited by available variation in relative shell height, especially if selection pressures operating during
'typical' conditions have depleted variation in this trait. There were 2 surprising results. First, relative
aperture area decreased after the storm. This result is counter-intuitive because one would expect a
shift to larger apertures, which can accomodate a larger adhesive foot. Second, a decrease in shell
length after the storm was found only at the wave-exposed site, suggesting that the effects of the storm
were more pronounced on this population. Because a shift in relative shell height and aperture area
occurred at all 3 sites, these traits appear to be more sensitive to large storms than is shell length. The
shifts in both relative shell height and aperture area may reflect differences in the hydrodynamic prop-
erties of shells collected before and after the storm, but I suggest that the shifts were mediated by the
ability of snails to avoid free-stream flows by hiding in sheltered crevices. Both reduced shell height
and aperture area may be advantageous when trying to fit into sheltered microhabitats during periods
of increased hydrodynamic stress. In particular, the shift in aperture area may reflect the premium on
successful crevice use over conventional adhesion during periods of extreme wave energies. Finally,
while recent work on intertidal snails has emphasized the importance of phenotypic plasticity to mor-
phological differentiation among populations, my results suggest that selection can assume a more
prominent role during unusual events like this storm.
KEY WORDS: Catastrophic storm - Hydrodynamic forces . Littorina obtusata . Morphological varia-
tion . Phenotypic and natural selection . Wave energies
INTRODUCTION cies (Kitching et al. 1966, Palumbi 1984, 1986, Denny et
al. 1985, Etter 1988a, Trussell et al. 1993). One conse-
Wave energy on intertidal shores varies considerably quence of increased wave energy is the increased risk
in both space and time and is thought to exert a strong of dislodgement for organisms living in the intertidal
influence on morphological variation i n intertidal spe- zone. The risk of dislodgement may interfere with var-
ious aspects of a wave-swept existence (Menge 1974,
Menge 1978a, b, Denny et al. 1985, Brown & Quinn
'Present address: School of Marine ScienceNIMS, College of
William and Mary, Gloucester Point, Virginia 23062, USA. 1988, Etter 1989), and may be lethal if individuals are
E-mail: gct@vuns.edu. Correspondence: 222 Upland Road, removed from their typical environment (Etter 1988a)
Cambridge, Massachusetts 02140, USA or if wave forces exceed the individual's structural
Q Inter-Research 1997
Resale of full article not permitted74 Mar Ecol Prog Ser 151: 73-79, 1997
strength (Denny 1988). The magnitude of hydrody- heights reaching -11 m (Davis & Dolan 1992); wave
namic forces (drag, lift, and acceleration reaction) is heights of 10 m were observed at the Canoe Beach
expected to be greatest on wave-exposed shores Cove study site (G. Trussell & J. Witman pers. obs.).
because mean and maximum water velocities and I evaluated the effects of the storm on Littorina
accelerations are greater (Denny 1985, 1988, Denny et obtusata shell morphology by collecting pre- and post-
al. 1985, Denny & Gaines 1990). On wave-exposed storm samples from a wave-exposed and 2 protected
shores natural selection is expected to favor traits that shores. In considering shell shape. I focused on shell
reduce the risk of dislodgement. height relative to shell length because previous work
The risk of dislodgement for intertidal snails is (Trussell et al. 1993, Trussell 1997) on several popula-
dependent on a number of factors, including flow tions found obvious differences in relative shell height
velocity and acceleration, shell shape and size, and the that were associated with wave-exposure. I also made
adhesive strength of the foot (Etter 1988a, Trussell et measurements of aperture length and width to calcu-
al. 1993, Trussell 1997). For snails to resist dislodge- late aperture area, which can serve as a predictor of
ment, their adhesive ability, which is a function of foot snail adhesive ability (Palmer 1992). Finally, I evalu-
size (Trussell et al. 1993. Trussell 1997), must exceed ated the storm's effect on snail size by using shell
the magnitude of the hydrodynamic forces acting on length as my size criterion.
the snail. Wave-imparted hydrodynamic forces act
directly on the snail's shell so shell shape and size can
reduce the risk of dislodgement in 2 ways. Changes in MATERIALS AND METHODS
shell shape and size may (1) directly alter the magni-
tude of hydrodynamic forces acting on the shell, and I made pre-storm measurements of shell height,
(2) influence a snail's ability to hide from free-stream shell length, aperture length and width to character-
flows in sheltered cracks or crevices. ize shell traits and tenacity of a wave-exposed
Morphological variation in intertidal snails is often [Pemaquid Point, Maine (43"50' N, 69" 39' 12" W ) ] and
correlated with differences in wave energies (Kitching 2 protected [South Harpswell, Maine (43' 43' 54" N,
et al. 1966, Kitching 1976, Kitching & Lockwood 1974, 70" 02' W); Canoe Beach Cove, Nahant, Massachu-
Raffaelli 1978, Crothers 1983, Johannesson 1986, Etter setts (42" 25' 42" N, 70" 55' 48" W)] populations of Lit-
1988a, Trussell et al. 1993, Trussell 1996, 1997). In gen- torina obtusata in New England. Detailed descriptions
eral, predation by crab predators on protected shores of these sites are provided elsewhere (Trussell et al.
and dislodgement by wave forces on wave-exposed 1993, Trussell 1997). Shell height was measured as
shores are believed to be the major forces driving habi- the maximum distance perpendicular to the plane of
tat-specific differences in shell morphology (Kitching the aperture and the highest point of the shell and
et al. 1966, Reimchen 1982, Palmer 1985, Trussell shell length as the maximum dimension of the shell
1997). Typically, the shells of snails on wave-exposed parallel to the plane of the aperture (see Fig 1 in
shores are thinner, squatter relative to shell length, Trussell et al. 1993). Aperture width (AW) was mea-
smaller-sized, and larger-apertured than those of pro- sured as the maximum dimension of the aperture run-
tected conspecifics. In addition, wave-exposed snails ning perpendicular to the axis of shell length. Aper-
typically have a larger adhesive foot than similarly- ture length (AL) was measured as the maximum
sized protected conspecifics and are therefore able to dimension of the aperture along the same axis as the
resist greater dislodgement forces. Hence, the above shell length measurement. Aperture area was calcu-
traits are expected to reduce the risk of d.islodgement lated from these measurements assuming an elliptical
for snails on wave-exposed shores. shape using the formula 0.25n (AW X AL) (Miller
It is clear that wave energies, particularly on wave- 1974, Lowell 1986). Measurements were made with
exposed shores or during large storms, are an impor- digital calipers to the nearest 0.01 mm. Snails were
tant agent of disturbance in intertidal communities sampled (Table 1) from 0.25 m2 quadrats blindly
(Suchanek 1978, 1981, Sousa 1979, 1984, Paine & tossed on each shore at the same tidal height 1-1.5 to
Levin 1981).Disturbances are discrete even.ts that can 2 m mean low water (MLW)].
displace or kill individuals or populations, and are Soon after this sampling a catastrophic storm
often defined by their intensity of impact and fre- occurred from October 28 to 31, 1991. In addition to
quency of occurrence (Sousa 1984). Here, I examine direct observation of wave heights near 10 m at the
the impact of a catastrophic storm on the shell mor- Canoe Beach Cove site, I obtained wind speed data
phology of the intertidal snail Littonna obtusata. This recorded during the storm and the latter half of 1991
storm was classified as a '50-year storm' and is one of (Fig. 1) Wind speeds, in addition to fetch and the
the 1.argest on record for the Gulf of Maine ( D a v ~ s& duration that the wind blows, can be used to estimate
Dolan 1992) The storm lasted for 144 h and had wave wave heights and wave forces experienced by inter-Trussell. Phenotyplc selection in an ~ntertidalsna11 75
collected 1 to 3 wk after the storm from each popula-
tion at the same tidal height and location as pre-storm
-
5 14 - samples (Table 1).
E - S t a t i s t i c a l a n a l y s e s . I conducted analyses of covari-
zg 12-
l
ance (ANCOVA) to determine (1) pre-storm differ-
V) ences in relative shell height and aperture area be-
.- tween the 3 study sites, and (2) differences in relative
shell height and aperture area between pre- a n d post-
storm samples for each population. In all ANCOVAs I
W 8- used shell length as the covariate to adjust for size
M
111
L - effects on shell height and aperture area. For those
>
6 ANCOVAs Involving aperture area, both shell length
6 -
and aperture area were loglo transformed. Otherwise,
-
analyses were performed on untransformed data using
4 . - ,-- .--- - , - , t . +
-
,
-. Tvwe I11 sums of sauares (Wllkinson 19891 I also used
'L
1 5 9 13 17 21
JUh' JUL AUG SEP OCT NOV ANOVA on shell length data to examine size differ-
Time (Weeks) ences between populations and before and after the
storm
Fig 1 Average weekly w ~ n dspeeds (&SE)for the latter half of
1991 Maximum wind speeds of 80 mph (129 km h ' ) a n d
wave helghts of -10 m were recorded at Nahdnt, I\/lassachu-
setts (where the Canoe Beach Cove s ~ t 1s e located) Data were RESULTS
recorded at the Portland, Maine International Jetport and
obtained from the N a t ~ o n a Occ'anlc
l and Atmospheric Adnun- Pre-storm comparisons
istration, Ashevillc, North Cdrolina
The ANCOVA on pre-storm samples from the 3 sites
tidal organisms (Denny 1988). Due to the magnitude indicated homogeneity of slopes (Fi2,139, = 1.95; p >
of this storm, storm surges elevated water velocities 0.05) and that snails from both protected sites (Canoe
and accelerations far above their typlcal levels on Beach Cove and South Harpswell) had significantly
each shore Because wave-imparted forces are taller shells in terms of relative shell height than wave-
directly proportional to water velocity and accelera- -
exposed (Pemaquid Point) conspeclfics (F,,,,,,, -
tion (Denny et al. 1985), snails must have experienced 150.27; p < 0.00001; Fig. 2, Table 2). ANOVA on shell
greater hydrodynamic forces during this storm than length found that protected snails fom both sites were
during conditions of normal wave energy. Hence, the longer than those from the wave-exposed site (F,,, = ,,,,
probability of snail dislodgement was increased on 18.21;p < 0.00001).
each shore. Pre-storm comparison of aperture area relative to
To test the effect of increased wave energies on shell shell length revealed honlogeneous slopes (F12,139, =
morphology, I repeated measurements of shell height, 1.83;p > 0 05) and that snails from both protected sites
shell length and aperture length and width using the had smaller apertures than conspecifics from the
methods described above on post-storm samples of wave-exposed site (F,,,,,,, = 37.06; p < 0.00001; Fig 3,
snails from each population. Post-storm samples were Table 2).
Table 1. Dates of sample collection for 3 populations of Llttonna obtilsata before a n d after the storm event For p o p u l a t ~ o n shav-
lng multiple sample dates for both before and after the storm (i.e Canoe Beach Cove and South Harpswell) samples for each
period were pooled
Before storm After storm Time after storm ( d ) Overall time d ~ f f e r e n c e( d )
Pemaquid Point, Maine (wave-exposed)
Oct 24, 1991 ( N = 50) Nov 22, 1991 [N = 46)
C a n o e Beach Cove. Massachusetts (protected)
Sep 20, 1991 ( N = 20) Nov 8 , 1991 ( N = 45)
Oct 10, 1991 (N = 25) Nov 9 , 1991 ( N = 45)
South Harpswell, Maine (protected)
Oct 16, 1991 (N = 25) Nov 11, 1991 (N = 83)
Oct 24. 1991 ( N = 25)76 Mar Ecol Prog Ser 151. 73-79, 1997
Table 2. Ljttorina obtilsata. Summary of least-squares regres- Before Storm
sion statistics and range of shell lengths of snails in samples
collected before (BS) and after (AS) the storm. Linear mea-
surements are in m111and area measurements in mm2
-E
g 1.45
d
Regression R2 N Range of shell
E
lengths in sample
2 1.40
Shell height (Y) vs shell length (X): 2
a
C
Canoe Beach Cove
1.35
BS Y = O.51X + 1.42 0.94 45
.4S Y = 0 . 5 1 X + 1.31 0.94 90 F-
V)
South Harpswell $ 1.30
BS Y = O.55X+ 1.07 0.97 50 6.88-14.19 4
AS Y = 0.56X + 0.86 0.98 88 4.72-14.23 3
Pemaquid Point 1.25
South Harpswell Canoe Beach Cove Pemaquid Point
BS Y = O.52X + 0.70 0.95 50 6.72-10.78 Protected Protected Exposed
AS Y = 0.52X + 0.57 0.93 46 6.40-10.66
Fig. 3. Littorina obtusata. Adjusted mean log aperture area
Aperture area (Y) vs shell length (X):
(+SE) from ANCOVA for snalls sampled before the storm
Canoe Beach Cove
from a wave-exposed and 2 protected shores in New Eng-
BS Y = 4 . 7 8 X - 21.91 0.95 45 Same as above
land, USA. See Table 1 for sampling dates and Table 2 for
AS Y = 4 . 5 2 X - 19.94 0.94 90 Same as above
regression statistics and sample slzes
South Harpswell
BS Y = 4 . 7 5 X - 21.12 0.94 50 Same as above
AS" Y = 4.89X - 24.28 0.94 45 6.61-14.01 Before Storm
Pemaquid Point After Storm
BS Y = 5 . 0 3 X - 21.12 0.95 50 Same as above
AS Y = 4 . 4 5 X - 18.15 0.95 46 Same as above
dAperture area measurements were not made for all snails
collected after the storm at South Harpsivell
Before Storm
South Harpswell Canoe Beach Cove
I
Pemaquid Point
Protected Protected Exposed
Fig. 4. Littonna obtusata. ridlusted mean shell height (*SE)
from ANCOVA for snalls sampled before and after the storm
Protected Protected ~xposed at each study site. See Table 1 for sampling dates and Table 2
for regression statist~csand sample sizes
Fig. 2. Littorina obtusata. Adjusted mean shell height (?SE)
from ANCOVA for s n a ~ l ssampled before the storm from a
wave-exposed and 2 protected shores in Neiv England, USA. p Z 0.55). Snails measured before the storm from
See Table 1 for sampling dates and Table 2 for regression sta-
1 9.42; p < 0.005; Fig. 41,
Canoe Beach Cove ( F 1 1 , t 3 2 =
tistics and sample sizes
South Harpswell = 12.33; p 0.001; Fig. 41, and
Pemaquid Point = 13.06; p c 0 001; Fig 4 ) all had
Pre-storm versus post-storm comparisons taller shells relative to shell length than conspecifics
measured from each site after the storm.
ANCOVAs on pre- and post-storm measu.rements of ANCOVAs on aperture area as a function of shell
shell height as a function of shell length for each site length found that the apertures of post-storm samples
found that slopes in all cases were homogeneous (all were smaller than those of pre-storm samples collectedTrussell: Phenotypic select1
- '*p < 0.01
m
-'
Before Storm
After Storm
(Fig. 4). Hence, the natural experiment supports the
hypothesis that among population variation in relative
shell height is partly d u e to differences in wave energy
among the 3 study sites. Compared to the morphologi-
cal differences found between natural populations,
shifts in relative shell height after the storm were
""p < 0.0001 small. Despite the magnitude of the storm recorded in
this study, its impact on natural populations may have
been limited by the available variation in relatlve shell
height, especially if selection pressures operating dur-
ing 'typical' conditions have depleted variation in this
tralt.
Canoe Beach Cove Pemaqulcl Point
Protected Exposed
Selection against relatively taller shells by increased
hydrodynamic forces may reflect 2 factors. Flrst, a
Fig. 5. Littorina obtusata. Adjusted mean log aperture area decrease in relative shell height may reduce the shell's
(*SE) from ANCOVA for snails sampled before and after the coefficients of lift and/or drag; theoretically a reduction
storm at each study site. See Table 1 for sampling dates and in either of these coefficients could reduce the magni-
Table 2 for regression statistics and sample sizes
tude of hydrodynamic forces acting on the snail's shell
a n d thus the risk of dislodgement. For example, Dud-
ley (1985) found that limpets from highly wave-
from Canoe Beach Cove (FI,,,321 = 8.85; p < 0.01; Fig. 5), exposed environments had shells that experienced
South Harpswell (F,,,,,, = 8.47; p < 0.01; Fig. 5), and lower relative drag than those of limpets from shel-
Pemaquid Point (E, :., = 71.66; p < 0.00001; Fig. 5). In tered environments (see also Branch & Marsh 1978).
all cases slopes were homogeneous (all p at least Denny (1989) also reported a limpet shell shape that
10.20). reduced drag but concluded that the conditions favor-
ANOVA on shell length for samples collected before ing streamlining, such as predictable flow direction
a n d after the storm detected a shift to smaller-sized and the absence of upstream objects (Vogel 1981), are
shells at the wave-exposed site (Pemaquid Point: unlikely to occur in the field. In addition, since both
F,,,,,,= 6.92; p < 0.01). However, no differences in pre- studies were conducted in laboratory flumes with ori-
versus post-storm shell length was detected at both entation of shells into flow controlled, their relevance
protected sites (Canoe Beach Cove: F[,,,,,,= 0.04; p > to field conditions is difficult to determine.
0.05; South Harpswell: F11,;161 = 3.01; p > 0.05). A second and perhaps more likely scenario is that
relatively squatter shells will increase the availability
of sheltered microhabitats for snails during periods of
DISCUSSION intense wave energy. Exclusion from sheltered cracks
and crevices may force taller-shelled snails to bear the
Wave energy is believed to be an important selective full brunt of storm surge. Crevice use in intertidal
force on many intertidal snail attributes. Although snails is common and the size structure of some snail
snail growth rates (Janson 1982, Brown & Quinn 1988), populations is coupled with microhabitat size a n d
reproductive output (Etter 1988b), and physiological availability. For example, Emson & Faller-Fritsch
stress (Etter 1989) are known to vary with gradients in (1976) increased the size structure of Littor-ina rudis
wave-exposure, substantial emphasis has considered populations by increasing crevice size, and Rafaelli &
the influence of wave energy on snail shell form (Kitch- Hughes (1978) found a tight rel.ationship between the
ing et al. 1966, Kitching & Lockwood 1974, Palmer sizes of L. rudis a n d L. neritoides and the size of avail-
1985, Johannesson 1986, Trussell et al. 1993, Trussell able crevices. Interestingly, the relationship docu-
1996, 1997). mented by Rafaelli & Hughes (1978) held for wave-
I found that the shells of protected snails were taller exposed shores but not protected shores suggesting
and smaller apertured relative to shell length than that crevice size a n d availability a r e more important on
wave-exposed conspecifics (Figs. 2 & 3). For relative wave-exposed shores in reducing the effects of wave-
shell height, the dramatic increase in water velocity imparted forces. L. obtusata on wave-exposed shores,
and acceleration durlng the storm reinforced the pat- especially after periods of increased wave action, are
tern of relatively squatter shells as wave exposure typically found in or near cracks and crevices, while
increased. Snails measured from each population after protected-shore snails are typically found on the inter-
the storm had significantly squatter shells relative to tidal alga Ascophyllum nodosum (author's pers. obs.).
shell length than those measured before the storm If size-related exclusion from sheltered microhabitats78 Mar Ecol Prog Ser 151. 73-79, 1997
is the mechanism accounting for my results, then stationary until conditions improve, especially since
crevice size and availability may be important on both Miller (197-1)showed that the ability of snails to rcmain
wave-exposed a n d protected shores during unusually attached to the substratum (tenacity) is reduced when
intense storms. crawling. Hence, the risk of dislodgement for migrat-
The crevice use hypothesis also explains the peculiar ing snails during intense storms would be greater than
shift to smaller apertures after the storm in each popu- that for snails remaining in crevices. Fletcher (1990)
lation (Fig. 5). Intuitively, since aperture area is a reli- detected no seasonal effects on the vertical zonation of
able predictor of tenacity, one would expect a shift to Littorina obtusata suggesting that pronounced sea-
larger apertures in response to increased hydrody- sonal migrations do not occur. Despite my arguments,
namic stress. This logic has been used to account for these and other snail behaviors may be involved and
the typically larger apertures of wave-exposed snails, more research is needed to determine t h e ~ rpotential
which a r e presumably capable of accomodating a influence on the morphological changes I observed.
larger adhesive foot. However, during unusually large In summary, the precise mechanism responsible for
storms the ability to exploit sheltered microhabitats is the morphological shifts I observed remains unclear.
probably paramount in reducing the risk of d~slodge- However, I suggest that both relatlve shell height and
ment because it is unlikely that snails are able to with- aperture area, or factors related to them (Endler 1986),
stand such intense free-stream flows via adhesion. The have a n important role in reducing the wave-imparted
shift to both relatively squatter shells a n d smaller aper- hydrodynamic forces acting on snails during large
tures probably reflects the premium on successful storms. Moreover, the shift in relative shell height is
exploitation of sheltered crevices during larger storms. consistent with differences observed among natural
Despite this discussion, much remains to be learned populations experiencing different wave energies.
about how a n d why the shift in aperture area occurred. This result suggests that increased water velocities and
As expected, a shift to shorter shells after the storm accelerations may, directly or indirectly, serve as
occurred at the wave-exposed study site. Shorter shells potent selective forces on shell morphology that con-
could directly reduce the magnitude of hydrodynamic tribute to variation in gastropod shell form on shores
forces (Denny et al. 1985) a n d , like relatively squatter differentially exposed to wave action. Although phe-
shells a n d smaller apertures, may improve the ability notypic plasticity in snall traits such as foot size (Etter
to hide in sheltered microhabitats. However, this shift 1988a, Trussell 1997) a n d shell thickness (Appleton &
did not occur at the protected sites. These results imply Palmer 1988, Palmer 1990, Trussell 1996) have demon-
that the effects of the storm were only pronounced strated that selection is not the sole factor responsible
enough at the wave-exposed site to elicit a shift to for morphological differences among populations, my
shorter shells. If so, then it appears that relative shell results suggest that selection can assume a more
height a n d aperture area are more sensitive to the prominent role during unusual events like this storm
effects of large storms than shell length because the (Bumpus 1899, Boag & Grant 1981, Endler 1986).
shift in both traits occurred at all 3 study sites.
Although the correlation between the change In rel- Acknowledgements. T h ~ sresearch was supported by grants
from Sigma Xi, the Scientific Research Society, the Hawaiian
ative shell height a n d aperture area a n d the storm sug-
Malacological Society. and the Lerner-Gray Fund of the
gests that selection occurred, the lack of direct evi- American Museum of Natural History. Support during writing
dence of storm-induced mortality requires that my was provided by a School of Marine Sclence Fellowship (Col-
results be viewed with caution. However, if snails were Iege of Willlam & Mary). I thank S. Genovese, T Kocher, R.
able to survive being swept to the subtidal, I think the Olson, W. Sobczak, and G. Vermeij for helpful comments on
earlier drafts. I also thank J. Ayers and J. Witman for provid-
amount of time that samples were taken after the storm ing me with laboratory space during my course of study at
(Table 1) was probably sufficient for survivors to return Northeastern University's Marine Science Center. Special
to the intertidal zone. thanks to R. Etter and R. Palmer for t h e ~ insight
r and extensive
While both pre- and post-storm samples were taken contribution to this work.
at the same tidal height a n d location on each shore,
other factors such as storm or seasonally-induced LITERATURE CITED
migration downshore out of my sampling areas could
also explain the morphological shifts I detected. For Appleton RD.Palmer AR (1988) Water-borne stimuli released
example, harsh weather conditions a r e known to by crabs and damaged prey induce more predator-resis-
tant shells in a marine gastropod. Proc Natl Acad Sci USA
restrict foraging behavior in Nucella lapillus (Burrows 85:4387-4391
& Hughes 1989).I think it unlikely, however, that snails Boay PT, Grant PR (1981) Intense natural selection in a popu-
would migrate large distances downshore during lation of Darwin's finches (Geospizinae)in the Galapogos.
unusually intense storms. Instead. I would expect Science 214:82-85
Branch GM, Marsh A C (19781 Tenacity and shell shape in six
snails to seek the nearest availa.ble crevice and remainTrussell: Phenotypic selection in an intertidal snail 79
Patella species: adaptive features. J Exp Mar Biol Ecol 34: forms in Nucella. J Anim Ecol 35:113-126
111-130 Lowell RB (1986) C r a b predation on limpets: predator behav-
Brown KM, Quinn J F (1988) T h e effect of wave disturbance ior a n d defensive features of the shell morphology of t h e
on growth In three specles of intel-t~dalgastropods prey. Biol Bull 17 1.577-596
Oecologia 75.420-425 Menge BA (1974) Effect of wave action and competition on
Bumpus HC (1899) T h e elimination of the unflt a s illustrated brooding a n d reproductive effort in the seastar Leptaste-
by the introduced sparrow, Passer domesticus. Mass. Bio- rias hexactis. Ecology 55:84-93
logical Lectures. Marine Biological Laboratory, Woods Menge BA (1978a) Predation intensity in a rocky intertidal
Hole. Ginn a n d Co., Boston, p 209-226 community. Relation b e t w e r n predator foraging activity
Burrows MT, Hughes RN (1989) Natural foraging of the dog- and environmental harshness Oecologia 34.1-16
whelk, Nucella lapillus (Linndeus); the weather and Menge BA (1978b) Predation lntens~tyin a rocky intertidal
whether to feed. J Mollusc Stud 55:285-295 community. Effect of a n algal canopy, wave action a n d
Crothers J H (1983) Some observations on shell shape varia- desiccation o n predator feeding rates. O e c o l o g ~ a34: 17-35
tion in North American populat~onsof Nrlcella lapillus (L.) Miller SL (1974) Adaptive design of locomotion and foot form
B101 J Lin Soc 19:237-274 in Prosobranch gastropods. J Exp Mar Biol Ecol 1499-156
Davls RE, Dolan RJ (1992) The 'All Hallo\vs' Eve' coastal Paine RT, Levin SA (1981) l n t e r t ~ d a llandscapes. d ~ s t u r b a n c e
storm-October, 1991. J Coastal Res 8:978-983 a n d the dynamics of pattern. Ecol Monogr 51:145-178
Denny hlW (1985) Waves forces on intertidal organisms: a Palmer AR (1985) Adaptive value of shell variation in Thais
case study. Limnol Oceanogr 30.1 171-1 187 lamellosa: effect of thick shells on vulnerability to a n d
Denny MW (1988) Biology and the mechanics of the wave- thelr preference by crabs. Veliger 27:349-356
swept environment. Princeton University Press, Princeton Palmer AR (1992) lntraspecific vanatlon in three species of
Denny MW (1989) A limpet shell shape that reduces drag: rocky shore gastropods from Hong Kong. correlations
laboratory demonstration of a hydrodynamic mechanism among habitats a n d a companson with temperate species.
a n d an exploration of its effectiveness in nature. C a n J In: Morton B ( e d ) T h e marine flora a n d fauna of Hong
Zoo1 67:2098-2106 Kong and southern China 111. Hong Kong University Press.
Denny MW, D a n ~ e TL,l Koehl MAR (1985) Mechanical h m ~ t s Hong Kong, p 649-691
to slze in wave-swept organisms. Ecol Monogr 55:69-102 P a l u m b ~ SR (1984) Tactics of accl~mation:morphological
Denny MW, Gaines SD (1990) O n the prediction of maximal changes of sponges in a n unpredictable environment.
intertidal wave forces. Limnol Oceanogr 35:l-15 Science 225:1478-1480
Dudley R (1985) Fluid-dynamic drag of limpet shells. Veliger Palumbi SR (1986) HOW body plans limit acclimation:
28.6-13 responses of a demosponge to wave force Ecology 67:
Emson RH, Faller-Fritsch RJ (1976) An e x p e r m e n t a l Invest]- 208-214
gation Into the effect of crevlce availability on abundance Raffaelli DG (1978) The relat~onshipbetwcrn shell ~njuries,
and size structure In a population of Liltorina rudis shell thickness a n d habitat characteristics of the intertidal
(Maton):Gastropoda: Prosobranchia. J Exp Mar B ~ o lEcol snail Littorina rudis Maton. J Mollusc Stud 44:166-170
23:285-297 Raffaelli DG, Hughes RN (1978)T h e effect of crevice size and
Endler JA (1986) Natural selection In the wild Pr~ncetonUni- a\,cl~labilityon populations of L~ttorinarudjs a n d Littorina
verslty Press, Princeton ner~toides.J A n ~ mEcol 74:71-83
Etter RJ (1988a) Asymmetrical developmental plasticity in a n Reimchen TR (1982) Shell size divergence in Littorina mariae
intertidal snail. Evolution 42:322-334 a n d L. obtusata a n d predation. C a n J Zoo1 60:687-695
Etter RJ (1988b) Physiological stress a n d color polymorphism Sousa WP (1979) Experimental investigations of disturbance
in the intert~dal snail Nucella lapillus Evolution 42 and ecological succession in a rocky intertidal algal com-
660-680 munity Ecol Monogr 49:227-254
Etter RJ (1989) Life history variation in the intertidal snail Sousa WP (1984) T h c role of disturbance in natural communi-
Nucella lapillus across a wave-exposure gradient. Ecology ties. Annu Rev Ecol Syst 15:353-391
70:1857-1876 Suchanek TH (1978) T h e ecology of Mytllus edulis L. in
Fletcher CR (1992) Distribution of the flat periwinkles Ljtto- exposed rocky intertidal communities. J Exp Mar Biol Ecol
rlna obtusata (L.) dnd L. lnanae Sacchi & Rastelli (Gas- 31:105-120
tropoda: Prosobranchia): relat~onshipto the slope of the Suchanek TH (1981) T h e role of d ~ s t u r b a n c ein the evolution
shore. Proc Int Symp Littorinid Biol 3:119-126 of life history strategies in the intertidal mussels Mytilus
Janson K (1982) Genetic and environmental effects on the rdulis and Mylilus californianus. Oecologia 50: 143-152
growth ratcm of Littorina saxatilis Mar Biol 69-73-78 Trussell GC (1996) Phenotypic plasticity in a n intertidal snail:
Johanesson B I 1986) Shell morphology of L ~ t t o r ~ nsaxatilis
a the I-oleof a common crab predator. Evolution 50:448-454
Ollvi. the relative importance of physical factors a n d pre- Trussell GC (1997) Phenotypic plasticity in the foot size of a n
d a t ~ o nJ. Exp Mar Biol Ecol 102:183-195 intertidal snail. Ecology 78: 1033-1 048
Kitching JA (1976) Distribution and changes in shell form of Trussell GC, Johnson AS. Rudolph SG. Gilfillan ES (1993)
Thais spp. (Gastropoda) near Barnfield, B.C. J Exp Mar Resistance to dislodgment: habitat a n d size-specific differ-
B101Ecol 23:109-126 ences in morphology and tenacity in an intertidal snail.
Kitchlng JA, Lockwood J (1974) Observations on shell form Mar Ecol Prog Ser 100 135-144
and its ecological significance In thalsid gastropods of the W~lkinsonG (1989) SYSTAT. the system for statistics SYSTAT
genus Lepsiella in K e w Zealand. Mar Biol28:131-144 Inc. Evanston, 1L
Kitching JA, Muntz L, Ebling FJ (1966)T h e ecology of Lough Vogel S (1981) Llt(3 in movlng fluids. Princeton University
lne XV. The ecological significance of shell and body Press, Prlnceton
This article was presented by K L. Heck J r [Senior Editorial Manuscript first received. A4arch 26, 1996
Advisor), Dauphin Island, Alabama, USA Revised version accepted: March 13, 1.997You can also read
