arena, but confined Ilyanassa and Hydrobia to one-half of the dish. ... sediment thickness in an arena was found using the curvilinear regression equation:.
J. Exp. Mar. Biol. Ecol., 1985, Vol. 92, pp. 97-113 Elsevier
97
JEM 554
DISTURBANCE,
ZLYANASSA
EMIGRATION,
OBSOLETA
OF AN EPIFAUNAL
AND REFUGIA:
(Say), AFFECTS
THE
HOW
THE MUD
HABITAT
MZCRODEUTOPUS
AMPHIPOD,
SNAIL,
DISTRIBUTION
GRYLLOTALPA
(Costa)’
THEODORE Department
H. DEWITT
ofEcologyand Evolution, State
(Received
22 October
1984; revision
AND
JEFFREY
S. LEVINTON’
University of New York, Stony Brook, NY, 11794, V.S.A received
5 April 1985; accepted
21 June 1985)
Abstract: In the presence of the mud snail Ilyanassa obsoleta (Say), the tubicolous amphipod Microdeutopus gryllotalpa (Costa) emigrates to snail-free sediments, as demonstrated in laboratory and field experiments. Emigration occurs predominantly in the dark when the amphipod is most active. Unlike crevices, the thickness of sediments in which the amphipod is established offers no protection from snail disturbance. Emigration is shown to be caused by the disturbance generated by the snail’s plowing and burrowing across the sediment surface, and not a response to a reduction in the shared microfloral food supply. The crawling and burrowing of the smaller mud snail, Hydrobia totteni Morrison, does not disturb Microdeutopus, supporting the hypothesis that relative body sizes affects the ability of bioturbators/burrowers to disturb tube-dwellers. As the burrows ofMicrodeutopus extend only = 2 cm below the sediment surface, thick mud layers do not offer any refuge from Ilyanassa. However, very small solid surfaces (L l-2 mm in relief) to which the amphipods build tubes do provide some protection from Ilyanassa. In soft-sediment benthic communities, such small structures may provide significant refuge for small epifauna and shallow burrowing infauna escaping from small-scale, biogenic disturbance. Key words: disturbance; migration; refugia; habitat Microdeutopus gryllotalpa; amphipods; gastropods;
selection; benthic ecology; epifauna; Long Island Sound
Ilyanassa obsoleta;
INTRODUCTION
Disturbance has been implicated bottom communities. Disturbance
frequently as a major force structuring marine soft ranges in predictability and magnitude from
ephemeral catastrophes that wipe out hectares of biota (Boesch et al., 1976; Simon & Dauer, 1977; Sanders et al., 1980), to frequent, small-scale disruptions that defaunate patches less than a meter square in area (Van Blaricom, 1978; Woodin, 1978, 1981; Thistle, 1980). The burrowing activities of motile infauna and epifauna (Levinton, 1977; Wiltse, 1980; Brenchley, 1981; Wilson, 1981) and the deposition of reworked sediments by deposit-feeders disturb individual
(Rhoads & Young, 1970; Weinberg, 1979; Brenchley, 1981) can species on a smaller scale, leading to the local exclusion of affected,
’ Contribution No. 535 in Ecology * Author to whom correspondence 0022-0981/85/$03.30
0
1985 Elsevier
and Evolution at The State University should be addressed. Science
Publishers
B.V. (Biomedical
of New York at Stony Brook.
Division)
98
THEODORE
H. DEWITT AND JEFFREY S. LEVINTON
or target, species. The extent to which an individual is a member of a target species can depend on its motility (Brenchley, 198 1) and its size relative to the source of disturbance (Wilson, 198 1). Small tubicolous epifauna are particularly vulnerable to disturbance by bioturbators and burrowers (Grant, 1965; Mills, 1967; McCall, 1977; Weinberg, 1979; Wiltse, 1980; Brenchley, 198 1). Grant (1965) and Mills (1967) observed rapid changes in the densities of tubicolous amphipods and polychaetes, especially spionids, when herds of the mud snail, Zlyanussa obsoletu, moved across sandflats in Barnstable Harbor, Massachusetts. This mud snail is a dominant member of Western mid-Atlantic tidal flats where it may be found at densities of 200-1500 snails/m2 for most, if not all, of the year depending on the locality and the tidal height (Nichols & Robertson, 1979; Brenchley & Carlton, 1983; Levinton et al., 1985). In this paper, we report the results of experiments that demonstrate the rapid emigration of the epibenthic, tubicolous amphipod, Microdeutopus gryllotalpa away from the mud snail Zlyanassu obsoleta. Microdeutopus is a common, amphi-Atlantic, aorid amphipod found in the intertidal and subtidal in Europe but, to our knowledge, only subtidally in North America (Myers, 1969; Bousfield, 1973; Parker, 1975; pers. obs.). Furthermore, Microdeutopus is never collected on open mud flats, where ZZ@znassais ubiquitous, except in low salinity, estuarine creeks, where Zlyanassa is absent (Zajac & Whitlatch, 1982a, pers. comm.; Borowsky, pers. comm.; T. H. Dewitt, unpubl. data). In shallow, protected, subtidal habitats along the shores of Connecticut and Long Island, New York, Microdeutopus is readily collected in mussel clumps, on the blades of eel grass and branching macroalgae, and the fouling communities of piers and floating docks. In deep sediments, M, gryllota&a builds a U-shaped tube extending ~2 cm below the sediment surface. In sediments SO.5 cm thick, Microdeutopus constructs a straight tube, parallel to the underlying solid substratum, with the bottom of the tube attached to this solid surface. We argue here that the habitat distribution of Microdeutopus gryllotalpa may result from its emigration away from sources of disturbance such as that caused by the mud snail, Zlyanassa obsoleta, and the preference of Microdeutopus for, and its success in, crevice-rich refugia. There does not seem to be a refuge in depth for Microdeutopus as this amphipod’s tubes are constructed within the upper 1-2 cm of the sediment surface. However, we show that small solid substrata with vertical relief of l-2 mm do provide this amphipod with protection from Zlyanas.sa. We suggest that even very small, solid structures in benthic soft substrata may provide significant refugia for epifaunal and shallow-burrowing infauna from small-scale disturbances, such as those created by epifaunal snails. We also show that disturbance by smaller, epifaunal gastropods is insufficient to stimulate emigration in Microdeutopus, supporting Wilson’s (198 1) thesis that the relative sizes of interacting species affects the sensitivity of target species to disturbance from source species.
MUD SNAILDISTURBANCEAND EPIFAUNALEMIGRATION
99
MATERIALS AND METHODS LABORATORYEXPERIMENTS Microdeutopus gryllotalpa (Costa) was kept in continuous culture in a recirculating sea-water system maintained in our laboratory. The seed population and periodic additions to the culture were collected from floating docks in Port Jefferson Harbor and in Flax Pond, a local salt marsh, both located on the north shore of Long Island, New York. Adult and juvenile amphipods were used for all of the following experiments as initial trials showed uniform responses by all size/age classes. The body sizes of Microdeutopus ranged from l-4 mm (length) and 0.3-1.0 mm (width). The mud snails Ilyanassa obsoleta (Say) and Hydrobia totteni Morrison were collected from mud flats in Flax Pond and maintained in our laboratory recirculating sea-water system. Moderately large mud snails (2-3 cm long Ilyanassa and 2-4 mm long Hydrobia) were employed in these experiments. Experiments to study amphipod emigration in the presence of It’yanassaor Hydrobia were conducted in battery dishes modified into experimental arenas by erecting a plastic screen (6.4 mm mesh for Zlyanassa, 1.8 mm mesh for Hydrobia) across the 18-cm diameter. This semi-permeable barrier allowed Microdeutopus to travel throughout the arena, but confined Ilyanassa and Hydrobia to one-half of the dish. Each dish was surrounded with a 20 cm high, 250-pm mesh screen and submerged z 17 cm into the recirculating sea-water system. This permitted constant exchange of water while confining all the contestants to the arenas. Sediment was sieved to < 250 pm and frozen for at least 24 h prior to being added to each dish. To achieve a sediment layer of equal and even thickness in each arena, the mud was thawed, mixed to a homogeneous consistency with a magnetic stirrer, and equal volume aliquots were pipetted into each dish. The mud was allowed to settle overnight before experimental animals were added to the arenas. Low densities of amphipods (0.086/cm2) were used to minimize intraspecific competition for space; this amounted to 22 amphipods per arena, 11 added to both sides of the central screen. Microdeutopus were introduced 24 h before the mud snails to allow the amphipods ample time to construct their tubes. Two It’yanassaor 200 Hydrobia were introduced to one side of each arena (except control arenas which had no snails). This resulted in snail densities of 0.016 IZyanassa/cm2 and 1.57 Hydrobia/cm2, which is well within the range of densities of both species found at Flax Pond (Levinton, 1979; Bianchi & Levinton, 1981) and elsewhere (Nichols & Roberston, 1979; Brenchley & Carlton, 1983). In each experiment, the snail-side of adjacent arenas was alternated to reduce the chance of phototaxic orientation by Microdeutopus. Unless otherwise noted, the experiments were run in our recirculating sea-water system at 17 “C under a 14 h light : 10 h dark lighting regime. The salinity was maintained at 24-28x,. At the end of each experimental period, arenas were removed from the sea-water system, and the sediment and amphipods from each side of the semipermeable barrier
100
THEODORE
H. DEWITT AND JEFFREY S. LEVINTON
were pipetted from the arena, washed through a 250- pm screen, and counted. A solid plastic sheet placed against the central barrier prevented amphipods from switching sides after the experiment was terminated. As there was no evidence of sediment from one side of an arena being transported to the other side, the sediments from both sides of each arena were combined, washed onto a pre-weighed filter, dried, and weighed. The sediment thickness in an arena was found using the curvilinear regression equation: sediment depth (mm) = 0.194 + 20.52X - 6.87X2, where X = dry wt sediment/area (in g/cm’) for sediments < 250 pm. This equation was determined using Flax Pond sediments collected from the same site as those used in this experiment. This relationship is highly significant (P 4 0.001) and accounts for 86 % of the variance (T. Dewitt, unpubl. data). Five experiments utilized these arenas. The impact of Ilyanassa and Hydrobia on emigration
Four arenas each were set up with 2 Zlyanassa, 200 Hydrobia, or no snails (control). After 3 days, the arenas were removed from the sea-water system and the amphipods counted. The effect of light on the rate of emigration due to disturbance
The sea-water system pool that held the arenas was bathed in constant fluorescent light, wrapped in constant darkness under heavy black plastic sheets, or exposed to a (14 h : 10 h) light/dark cycle. Each lighting regime was run for 48-60 h. Twelve experimental arenas were placed in the pool at the beginning of the experiment. Only ZZyanassa were used. At 6-, 12-, 24-, and 48-h intervals (and at 60 h in one case) three randomly selected dishes were removed from the pool, and the abundances of amphipods on each side of the central screen were assessed. For control experiments, three no-snail dishes under each lighting regime were removed after 24 h. The all-light and all-dark experiments were run twice, and the light/dark experiment was run once. The effect of sediment depth on emigration
Into each of six arenas was pipetted 1.1 mm (thin) and 9.0 mm (deep) of a homogeneous suspension of ~250 pm sediment. Zlyanassa were introduced into one side of each of four dishes for each treatment; the remaining two dishes of each treatment served as no-snail controls. After 24 h in a 14 h light: 10 h dark cycle, all dishes were removed and the amphipods counted. The sediments from two deep treatment dishes and two thin treatment dishes were dried and weighed for sediment thickness determination.
MUD SNAIL DISTURBANCE
AND EPIFAUNAL
EMIGRATION
101
The effect of crevices on emigration
On the snail side of four arenas, two layers of 6-mm square mesh plastic screen were glued to the bottom of the dish to create crevices. The final vertical relief of the screen was % 3 mm. Four other arenas had no crevices and four more arenas were used as no-snail controls. Into each bowl was pipetted 50 ml of mud. Again, only Zlyanassawas employed. After 24 h, all arenas were removed from the sea-water system and the amphipods counted. To determine whether competition for food, rather than disturbance, might be the basis for emigration, the abundance of microalgae was measured on the snail and no-snail sides of the no-crevice control arenas in the fourth experiment. Sediment microalgae are highly preferred as food by Microdeutopus (Dewitt, 1985). If microalgae are significantly more abundant on the no-snail side of the dish, then emigration due to low food cannot be excluded as the proximate force driving emigration. Microalgal densities were measured by quantitative chlorophyll a extraction and fluorometry using the procedures of Vollenweider (1969) and Strickland & Parsons (1969). The impact of Ilyanassa on colonization by Microdeutopus
Zero, one, two, or three mud snails were placed in snail-inclusion arenas (14 cm diameter Petri dish base, covered with 5 mm of < 250 pm sediment, and surrounded with 2-mm square mesh screen 20 cm high) which had been set up in the Microdeutopus culture tank. The screen surrounding each arena allowed free passage of Mcrodeutopus into and out of each. Four replicas of each snail-abundance treatment were created, and all 16 arenas were arranged in a randomized block design. The experiment ran for 72 h, after which the number of amphipods having colonized the sediments inside the arenas was determined. FIELD EXPERIMENT
To test whether Microdeutopus moves away from IZyanassa in the field, a caging experiment was run in the shallow subtidal (lo-20 cm below MLW) along a tidal creek bank at Flax Pond in August 1983. Eight 0.25 m square by 15 cm high inclusion/exclusion cages were constructed of two layers of 1 cm mesh aluminum screen. Microdeutopus could easily pass through the walls of the cage, but Ilyunussa could not. The top 2.5 cm of the outer screen of each cage was curled sharply outwards, and the inner layer curled inwards, to create a barrier that IZyanassacould not climb over from inside or outside the cages. The cages were arrayed in two rows of four cages each, spaced = 25 cm apart. A 60 cm high, 12 mm square mesh fence surrounded the experimental plot to reduce fouling by drift algae. All algae, animals, and sediment particles >, 3 mm were removed from inside each cage to a depth of 10 cm, and the cages were left fallow for 1 wk to allow the sediment to consolidate and reestablish its microflora. Four cages (two in each row, staggered to reduce position effects) were designated
102
THEODORE H. DEWITT AND JEFFREY S. LEVINTON
as “ + snail” cages, the other four as “no-snail” controls. An open-ended, 12.5 cm diameter coffee can was placed in the center of each cage to confine the 200 amphipods that were introduced into the retainer. The amphipods were collected from Flax Pond and our laboratory culture. Ten 1Iyana.s.~ collected in the immediate vicinity were placed around this retainer in the “ + snail” cages. An hour after adding the amphipods, the coffee-can retainers were removed and the experiment began. Six days later, seven 3.8 cm diameter by 8 cm deep cores were taken from each cage. The cores were frozen upon transfer to the laboratory, and later thawed, sieved through a 500~pm screen, and sorted for amphipods.
RESULTS LABORATORY EXPERIMENTS
Effect of Ilyanassa and Hydrobia on emigration
After 3 days of exposure to either Zlyanassaobsoleta or Hydrobia totteni,the amphipods moved away from Zlyunassa only (Fig. 1.). While more amphipods accumulated on the “no-snail” side of the arena, there was no significant difference in numbers of Microdeutopus gryllotalpa between the “ + snail” and “no-snail” treatments for Hydrobia (G-test for goodness-of-fit to a 1: 1 ratio; P > 0.05). Significantly more amphipods were found on the “no-snail” sides of the arenas in the ZZyanussaexperiments (G-test; a 5 100.0
0
Snails (Or Left)
z 900
a
No Snails (or Right)
k
800
g 700 cr, I 60.0 :: w 500 5
40.0
$ 5
30.0 20.0
2
I0.C,-
!
0.0 Ilyanassa
Hydrobia
S Fig. 1. Emigration of Microdeutopus gryllotalpa in the presence of the mudsnails Ilyanassa obsoleta and Hydrobia totteni: values are means ( f 1 SD) of the proportion of the amphipod population found on each side of each arena; amphipods show significant (P I 0.01) negative association with Ilyanassa, but no such pattern holds when associated with Hydrobia; in no-snail control arenas, the amphipods show no significant preference for either the left or right sides of the dishes.
MUD
SNAIL
DISTURBANCE
AND
EPIFAUNAL
103
EMIGRATION
P I 0.01). In the control experiment (no snails present) there was no preference for the right or left halves of the arenas (G-test; P > 0.05). Effect of illumination on emigration
Amphipods move away from Ilyanassa at a significantly greater rate in the dark than in the light (Fig. 2). Amphipods experiencing a light/dark cycle show the same pattern of emigration as individuals in all light and all dark conditions: they move relatively little
u 80.0 F
2 70.0 0
i Q
60.0
2 0 2
50.0 40.0
I
I
I
40
20 TIME
I
I
60
(hours)
Fig. 2. The effect of illumination on Microdeutopus gryllotulpa emigration in the presence of mud snails: regression lines and values are the percent of amphipod populations found on the “no-snail” half of each experimental arena at time intervals following the addition of snails to each arena under three lighting regimes:24hoflight(~and......),14hlight:lOhdatk(~and----),and24hofdark(+ and--); all three slopes are significantly different from one another (P (: 0.05); at time 0, all treatments began at “50% population allopatric”.
in the light but quite noticeably in the dark. This is particularly evident from the samples taken within the first 24 h. A test of the equality of the slopes of the regression lines for these three illumination treatments reveals them to be significantly different from each other (P I 0.05). Those amphipods remaining on the “ + snail” sides of the arenas often built their tubes against the base of the 6.4-mm mesh screen dividing the dish. Such a position probably provides protection from disturbance in a manner analogous to more pronounced crevices, as is described later. Effect of sediment depth on emigration
The thickness of the mud layer overlying a solid substratum does not affect emigration of Microdeutopus away from Zlyanassa (Fig. 3). Nearly as many amphipods in deep (9.0 mm) mud (68.39% k 12.30 SD of the population) move away from the mudsnails as do amphipods in thin (1.1 mm) mud (74.8% f 11.68 SD of the population). While
104
THEODORE H. DEWITT AND JEFFREY S. LEVINTON
there is a significant movement away from the mudsnails in both deep and thin sediments (G-test for goodness-of-fit to a 1: 1 ratio; P ~2O.Ol), there is no significant difference in the magnitude ofthe emigration in these two sediment thickness treatments (t-test on arcsin transformed percentages; P > 0.05). In control arenas lacking snails, the amphipods distributed themselves approximately equally on both sides of the screen partition, regardless of the sediment thickness (t-test; P > 0.05).
5
100.0
iT
a
90.0
c
Cl Snails lZ$ No Snails
ueep
I bin
Deep
1hm
TREATMENTS Fig. 3. The effect of sediment depth on Mmdeutopus emigration in the presence of the mudsnail, ZtJIanassa obsole?c~values are means ( f 1 SE) of the proportion of the amphipod population found on both sides of each arena for each sediment depth treatment; N = 4 for each of the experiments with snails (those on the left of the figure) and N = 3 for each of the two “no-snail” control experiments (on the right); significantly fewer amphipods are present on the snail side of the arena in both Deep and Thin sediment treatments; however, there is no significant difference in the proportion of the amphipods that emigrate in the Deep of Thin treatments (t-test on arcsin transformed data; significance is evaluated at P i 0.05); there is no significant preference for either side of the control (no snails) arenas (goodness-of-fit G-test; P > 0.05); the experiments ran for ~24 h under a 14 h light: 10 h dark regime,
The sediments in the deep treatment were appro~mately nine times thicker than those in the thin treatment (9.04 mm [ + 0.297 SD] vs. 1.07 mm [ + 0.099 SD]). ~~y~~~~~~ can easily burrow deeper than 5 cm under the sediment surface (Brenchley & Carlton, 1983) and some individuals burrowed to the bottom of the deep sediment arenas (9 mm). While most IIyanussa crawl on the surface of the sediment, plowing a furrow z 1-2 mm deep, we have observed others cruising, slowly, l-2 cm below the surface of very soft sediments. The U-shaped tubes of Mcrodeutopus gryllota~a rarely extend > l-2 cm below the sediment surface.
MUD
SNAIL
DISTURBANCE
AND EPIFAUNAL
EMIGRATION
105
Effect of crevices on emigration
In the presence of crevices, relatively few Microdeutopus moved away from Zlyanassa, as they did when crevices were absent (Fig. 4). When crevices were present but mud snails absent, most of the amphipods settled on the crevice side of the arena. In these
2 100.0 u
lx
Q: 90.0
k
80.0
:
70.0
J, I
60.0
2w
50.0
5
400
$ 30.0 $
20.0
2
IO.0
0 a
0.0
8
+ Crevice +Snail
- Crevice +Snail
+ Crevice -Snail
-Crevice -Snail
TREATMENTS Fig. 4. The effect of crevices on the emigration of Microdeutopus in the presence of mud snails: values are the means ( + 1 SE) of the proportion of the amphipod populations found on both sides of each arena for each experimental treatment; N = 4 for all four experiments; from left to right, the four experimental treatments are: crevices and snails present; crevices absent, snails present; crevices present, snails absent; crevices and snails absent; for each experiment, the left bar represents the treatment having crevices and/or snails, and the right bar represents the vacant half of each arena, containing only a 2-mm layer of sediment; the “no crevices, no snails” control on the far right has only the 2-mm mud layer; significantly more amphipods remain on the snail side of the arenas when crevices are present than when crevices are absent (t-test on arcsin transformed data; P < 0.05); when snails are absent, amphipods show a highly significant preference for the crevice half of the arenas (goodness-of-fit G-test to a 1: 1 ratio; P < 0.01); when snails and crevices are absent there is no significant preference for either half of the arenas; the experiment ran for -24 h under a 14 h light : 10 h dark regime.
three treatments, significantly more amphipods were found on one side of each arena (t-test on arcsin transformed percentages; P < 0.05). As in the sediment thickness experiment, when no snails and no crevices are present, the amphipods are distributed approximately equally on both sides of the arena (t-test on arcsin transformed percentages; P > 0.05). Effect of mudsnails on sediment microflora
There was no significant difference in the chlorophyll a concentrations on the “ + snail” and “no-snail” sides of the four arenas used in the “snails, no crevices” treatment in the previous experiment (Fig. 4) (analysis of variance; P > 0.05). The
THEODORE
106
H. DEWITT
AND JEFFREY
S. LEVINTON
chlorophyll u concentration on the snail side of the arenas was 0.0261 pg Chl a/mg sediment ( + 0.0061 SD) and from the “no-snail” sides 0.0260 pg Chl a/mg sediment ( + 0.0059 SD). While Ilyanussa can significantly deplete benthic microalgae (Bianchi & Levinton, 198 1; Connor et al., 1982), the effect of 24 h of grazing on the algal standing stock was minimal. Despite the lack of evidence of food depletion, the amphipods still moved away from the mud snails (Fig. 4). This strongly suggests that Microdeutopus emigration in the presence of Ilyunussu is not caused by a decrease in the food concentration. With longer periods of coexistence, Zlyanuxsu would probably significantly deplete the microalgal standing stock, which would probably further increase the emigration rate of Microdeutopus gryllotu& (Dewitt, 1985). Effect of Ilyanassa on colonization The number of amphipods in each snail-inclusion arena is inversely proportional to the density of mud snails occupying that arena (Table I). A mean of 135.25 (k 13.89) amphipods had successfully colonized empty arenas, while only 67.75 (& 15.51 SD) amphipods colonized arenas containing three ZZyunumu. This inverse relationship is significant (linear regression, b = - 25.95; P I 0.01). Many of the tubes were attached to the sides of the Petri dish base of the arenas; however, the proportion of the inhabitants doing so was not recorded. TABLE I The effect
of mudsnail,
Ilyanassa obsoleta, density on colonization gryllotalpa.
of the amphipod,
Microdeutopus
Amphipod abundance (colonists) Number of snails 0
Regression
equation:
1 2 3 No. of amphipods
SE
Mean
135.25 13.89 118.0 4.43 61.0 13.40 67.15 15.51 = 134.425 - 25.95 (No. of mud Analysis
Source
Replicates
d.f.
4 4 4 4 snails)
of variance
MS
Among
snail densities 3 13716.750 Error 12 1416.458 Posthoc comparison of means by the Tukey-Kramer method 0 Snails 1 Snail 2 Snails 3 Snails
(Sokal
F
Significance
9.6838
P < 0.005
& Rohlf,
1981)
MUD SNAIL DISTURBANCE
AND EPIFAUNAL
EMIGRATION
107
Field experiment
More amphipods remained in the snail exclusion cages than in the inclusion cages (Fig. 5) after 6 days. Over this period, there was a 46.9% net decline in the amphipod 0.35
r
-z 0 0.30 5: \ 2 0.25 : 5 z
0.20
%
0.15
8 a 0.10 I r$ 0.05 0.00 Initial
Snails
No Snails
TREATMENTS Fig. 5. Field experiment: displacement ofMicro&utopus in Zlyanussa inclusion (snails present) cages relative to snail exclusion cages; values are the mean ( f 1 SE) densities of amphipods found in the inclusion and exclusion cages; N = 4 cages for both treatments; the amphipod densities for each cage are based on the pooled numbers of Microdeutopus collected in seven cores taken in each cage; the initial density of Microdeutopus is shown on the left; significantly more amphipods remained in the snail exclosure cages (r-test on untransformed data; P < 0.05) after 6 days.
density in the snail exclusion cages (starting density 0.32/cm2; final density 0.17 [ & 0.054 sD]/cm2). In the snail inclusion cages, there was a net decline of 84.4% (final density 0.05 [ + 0.042 sD]/cm2). These results corroborate the laboratory experiments which demonstrated the displacement and emigration of Microdeutopus in the presence of ZZyunu.ssa.This field experiment also supports the finding that deeper sediments do not provide a refuge from disturbance, as the muds in the experimental area are over 0.5 m deep (personal observation). Hydrobiu totteni collected in the cores were present at low density (0.253 [ + 0.5 14 sol/cm’); no significant difference in Hydrobia densities existed between the exclusion or inclusion cages.
DISCUSSION
Laboratory and field experiments demonstrate the differences in emigration of the tubicolous amphipod, Microdeutopus gryllotalpa, in response to disturbance caused by the mudsnails Zlyunassa obsoletu and Hydrobiu totteni. In the presence of the larger mud snail, Zlyanassa, Microdeutopus emigrates to sediments lacking this snail, while in the
108
THEODORE
H. DEWITT AND JEFFREY S. LEVINTON
presence of the smaller snail, Hydrobia, the amphipod shows no significant emigration (Fig. 1). Disturbance from packs of roving Zlyanassa obsoleta has been implicated in other studies as causing the emigration or death of several epibenthic invertebrates, thus creating allopatric distributions between this mudsnail and other species. Sanders et al. (1962), Grant (1965), and Mills (1967) attribute major changes in the species composition of sandflats at Barnstable Harbor, Massachusetts, to disturbance created by herds of Zlyanassa. Epifaunal species (especially tubicolous polychaetes and amphipods) were particularly susceptible, while deeper infaunal species were relatively unaffected. Hunt (198 1) reports a similar decline in epifaunal and shallow burrowing species in the presence of I. obsoletu on a North Carolina mudflat, possibly due to larval mortality (via ingestion or burial) or larval avoidance of the mudsnail. Levinton et al. (1985) credit disturbance by Zfyanassa with creating a bimodal zonation of the smaller mud snail, Hydrobia totteni, on Long Island, New York, mudflats: Hydrobia occupies areas shallower or deeper than Zlyanassa. Emigration of Hydrobiu in the presence of Zlyanassa was readily observed in laboratory and field experiments utilizing arenas and cages similar to those employed in the present study (Levinton et al., 1985). Ironically, ZZyanassa appears to get its come-uppance when it overlaps with the introduced periwinkle, Littorina littorea (Brenchley & Carlton, 1983). They report that the mud snails emigrate from intertidal hard substrates to mud and sand flats to escape disturbance caused by Littorina grazing on the shells of Zlyanassa. The emigration of Microdeutopus is a response to disturbance from Zlyanassa and not due to a decrease in food level. Both Microdeutopus and Zbanassa are deposit-feeders and utilize sediment microalgae as a major source of their nutrition (Curtis & Hurd, 1979; Pace etal., 1979; Connor & Edgar, 1982; Dewitt, 1985), and Zlyanassa can significantly reduce the microalgal standing stock of the sediment (Bianchi & Levinton, 1981; Connor et al., 1982). Although Microdeutopus will emigrate from low-food sediments (Dewitt, 1985), the emigration witnessed in these experiments occurred before the mud snails had time to significantly lower the microalgal concentration in the sediments. Thus, emigration by Microdeutopus in the presence of Zlyanassa is proximally due to disturbance and not the result of exploitation competition. Ultimately, however, competition for food and space could be invoked in combination with disturbance to explain the allopatric distribution of Microdeutopus and ZZyanassaas both species share similar foods and habitats. ZZyanassaalso significantly hinders colonization of sediments by adult and juvenile amphipods (Fig. 5). Although most Microdeutopus moved away from ZZyanussain the short-term laboratory experiments, several amphipods successfully colonized the snailinclusion arenas containing mud snails. It is not impossible for these two epifaunal species to coexist, but populations of Microdeutopus would be greatly reduced in the presence of even moderate densities of Zlyanassa. Evidence gained from these experiments suggests that the primary means by which ZIyanassa causes a decrease in the population size of colonists is disturbance. As Microdeutopus gryllotaIpa has marked abilities to discriminate among habitat types (Dewitt, 1985) it is also very possible that
MUD SNAIL DISTURBANCE
AND EPIFAUNAL
EMlGRATION
109
this amphipod actively avoids settling near Zlyatlassa. Hunt (198 1) also suggested that settling larvae may avoid Z~~ff~u~~a when colonizing soft sediments, although further experiments are needed to subst~tiate this hypothesis. Thus, not only can this mud snail stimulate epifaunal emigration, it may also reduce the rate of larval and adult colonization of the benthos through disturbance or differential avoidance by colonists. Some aspects of the physical structure of the sediment habitat (e.g., sediment depth and topographic relief) have a demonstrable effect on the rate of emigration of amphipods while in the presence of mud snails. Both sediment depth and topo~aphic relief strongly influence habitat preference, colonization rate, search time during colonization, and emigration rate (independent of disturbance) (Dewitt, 1985). While both deep mud and crevices can provide Microdeutopus with refuge from predation (Dewitt, 1985), they do not equally offer protection from Zlyanassa. Crevices do provide a refuge from mud snail dist~b~ce (Fig. 41, but sediment thickness does not affect the rate of emi~ation due to disturbance (Fig. 3). The latter result is corroborated by the field experiment where sediments are many centimeters (if not meters) deep, and significantly more amphipods emigrate from cages containing mud snails than from cages lacking snails (Fig. 5). Established, deep-burrowing infauna do not seem to be affected by the presence of Z~~~~~~u,although recruitment from the water column may be di~nished by the snail. The tubes of Microdeutopus rarely penetrate > 2 cm below the sediment surface, while Zf~~~nussareadily burrows deeper than this (Brenchley & Carlton, 1983, pers. obs.). When not burrowing, ZZyunassucrawls across the sediment surface, plowing a furrow a millimeter or two deep. During this crawling or burrowing, mud snails can easily break apart the tube of~~c~~eu~opus or bury its openings. When a tube is built inside a crevice or hole smaller than the shortest linear dimension of the mud snail (e.g., the shell width of Zlyunassa; FZ1 cm in these experiments), the tube is protected from the bulldozing of the mud snail. These results suggest that solid substrata serving as refugia for epifaunal invertebrates do not need to be very large structures relative to the size of animal at risk. In fact, a barrier only = 2 mm high, or a crevice 5 5 mm wide is suf%ient to shelter this tubicolous amphipod from ZZyc~~~ssa. Topographic relief on this scale is generally available on most beaches and mudflats in the form of pebbles, shell fragments, decaying vegetation, or jetsam. The extent to which such structures are utilized as refuges is presently unknown, but they are potentially as important ecologically as larger emergent structures (e.g., Woodin, 1978, 1981). If an amphipod’s tube is disturbed, the animal will retreat to the center of the tube for several seconds before leaving the tube or resuming its primary position just inside either tube opening. If sufficiently disturbed, the amphipod will emerge rapidly from the tube and scurry across the sediment, usually towards a nearby emergent structure, if present. Relatively few amphipods moved during illuminated experimental periods, yet they markedly moved away from Zlya~assain the dark and under 14 h : 10 h li~ht/dark regimes (Figs. l-4). Microdeutopus gryllofalpa is generally more active nocturnally than diurnally, with more individuals out of their tubes, on the sediment surface and in the
110
THEODORE
H. DEWITT AND JEFFREY S. LEVINTON
water column at night than during the day (Dewitt, 1985). Mcrodeutopus and many other invertebrates emerge from the benthos at night when the risk of being eaten by visual predators is minimal (Levinton, 1971; Hobson & Chess, 1976, 1977; Alldredge & King, 1977, 1980; Robertson & Howard, 1978). As a response to disturbance, there are at least two non-exclusive explanations for why Microdeutopus has greater dispersal at night than during the day: (1) these amphipods do not leave their tubes during the day because of the risk of being eaten by diurnal predators (especially fish); and (2) only those ~p~pods which emerge from their tubes, for reasons independent of being disturbed, move away from mud snails. These hypotheses have yet to be tested. Just how much disturbance is necessary to stimulate emigration in Microdeutopus is unresolved. Clearly, the intensity lies between that wrought by Ilyunussa and that by Hydrobiu. The crawling and burrowing behavior of Hydrobia is very similar to that of ~~anassa, except that Hydrob~a do not burrow as deeply, and they are considerably smaller than Zlyanassa (2-4 mm long vs. 2-2.5 cm long). These results are important to the debate concerning the role of biotubators and burrowers in excluding tubebuilders from a benthic community. This debate emerged from the “trophic group amensalism” studies of Rhoads & Young (1970) and carried forward in several later studies, including Woodin (1974, 1978), McCall (1977), Orth (1977), Weinberg (1979), Wiltse (1980), Brenchley (1981), and Wilson (1981, 1984). As argued by Wilson (1981, 1984), the difference in size between bioturbators/burrowers and other epifaunal species in benthic communities determines the susceptibility of the target species to disturbance by the bioturbators/burrowers as much as the living position (Woodin, 1974) or motility (Brenchley, 1981) of the target species. The difference in response by ~icrudeutopus g~~lota~a to I~anassa and Hydrobia supports this hypothesis. I~anassa is only one of several biological disturbers in Long Island Sound coastal mudflats. Herds of the fiddler crabs Ucapugnax and U. pugihtor make regular forays onto lower intertidal and shallow subtidal flats with the ebbing tides; large numbers of the errant polychaete Nereis succinea slither across the sediments shared by Ilyanassa, Hydrobia, and Microdeutopus. The role of these other species as biot~bators has yet to be examine (although Commito, 1982, found that nereid polychaetes caused significant reductions in local population sizes of the tubicolous amphipod Corophium volutatorthrough predation or disturbance). Microdeutopus gryllotalpa is clearly able to live in the shallow water habitats occupied by Ilyanassa obsoieta. This ~phipod is found on shallow mudflats in low salinity estuaries of Long Island Sound (Zajac & Whitlatch, 1982a,b; T. H. Dewitt, pers. obs.) and intertidally in Europe (Myers, 1969) where Z!y~nussa does not occur, and will readily establish breeding colonies in laboratory mudflat microcosms. However, along the coastline of Long Island, New York, we have rarely encountered Microdeutopus on any mudflats, yet we collect this amphipod regularly from crevice-rich substrata, such as mussel clumps, the blades of seagrass and branching algae, and from the undersides of floating docks. Ilyanassa is ubiquitous on the intertidal flats of Long Island. Given Micrudeutopus’s reaction (emigration and possibly avoidance) to Zlyanussa and its
MUD SNAIL DISTURBANCE AND EPIFAUNAL
EMIGRATION
111
preference for crevice-rich microhabitats (which serve as refugia from disturbance and predation), we believe that ~~ana~~a excludes ~~crode~topus ~~liota~a from shallow water mudflats in New England and Long Island where the two species are potentially sympatric. However, given that small solid substrata provide refugia from disturbance in laboratory microcosms, one would expect Microdeutopus, and other tubicolous epifauna to be more abundant on mudflats inhabited by Ilyalaassa. Perhaps these micro-refugia (e.g., pebbles, shells, and other debris) are undetected by colonists or provide less adequate protection than those used in our laboratory experiments. While their potential for providing refuge exists, their realized importance in this regard is presently unknown. Nonetheless, as other tubicolous amphipods (Mills, 1967; Hunt, 1981), tubicolous polychaetes (Sanders et al., 1962; Grant, 1965; Hunt, 1981) and smaller mudsnails (Levinton et al., 1985) show similar allopatric distributions and emi~ation responses to I. o~~oleta, the disturb~ce caused by this mud snail as it bulldozes across North American mudflats must be recognized as a major force shaping the composition of soft sediment, epifaunal communities.
ACKNOWLEDGEMENTS
We wish to thank G. Lopez, G. Hechtel, B. Borowsky, and two anonymous reviewers for their useful criticism of the manuscript. We are very grateful for the assistance of M. Campbell, M. Jones, R. Cort and M. Weissburg. This project comprises a part of the dissertation research of T. Dewitt and was supported by a National Science Foundation grant to J. S. Levinton.
REFERENCES ALLDREDGE,A. L. & J.M. KING, 1977. Distribution, abundance, and substrate preferences of demersal reef zooplankton at Lizard Island Lagoon, Great Barrier Reef. A&r. Eiol., Vol. 41, pp. 317-333. ALLDREDGE,A. L. & J. M. KING, 1980. Effects of moonlight on the vertical migration patterns of demersal zooplankton. J. Exp. Mar. Biol. Ecol., Vol. 44, pp. 133-156. BIANCHI,T. & J. S. LEVINTON,1981. Nutrition and food limitation ofdeposit-feeders. II. Differential effects of Hydrobia totteniand Ilyanassa obsoleta on the microbial community. J. Mar. Res., Vol. 39, pp. 547-556. BOESCH, D. F., R. J. DIAZ & R. W. VIRNSTEIN, 1976. Effects of tropical storm Agnes on soft-bottom macrobenthic communities of the James and York estuaries and the Lower Chesapeake Bay. Chesapeake Sci., Vol. 17, pp. 246-259. BOUSFIELD,E. L., 1973. Shallow-eater ga~~a~dea~ ~mph~oda of New England. Cornell University Press, Ithaca, N.Y., 312 pp. BRENCHLEY,G.A., 1981. Disturbance and community structure: an experimental study of bioturbation in marine soft-bottom environments. J. Mar. Res., Vol. 39, pp. 767-790. BRENCHLEY,G. A. & J. T. CARLTON,1983. Competitive displacement of native mud snails by introduced periwinkles in the New England intertidal zone. Bioi. Bufl. (Woods Hole, Mass.), Vol. 165, pp. 543-558. COMMITO,J. A., 1982. Importance of predation by infaunal polychaetes in controlling the structure of a soft-bottom community in Maine, U.S.A. Mar. Bioi., Vol. 68, pp. 77-81. CONNOR,M. S. & R. K. EDGAR, 1982. Selective grazing by the mud snail Ilyanassa obsoleta. Oecologia (Berlin), Vol. 53, pp. 271-275. CONNOR,M. S., J. M. TEAL,& I. VALIELA,1982. The effect of feeding of mud snails, Ilyanassa obsoleta (Say),
112
THEODORE
H. DEWITT
AND JEFFREY
S. LEVINTON
on the structure and metabolism of a laboratory benthic algal community. J. Exp. Mar. Biol. Ecol., Vol. 65, pp. 29-45. CURTIS, L. A. & L.E. HURD, 1979. On the broad nutritional requirements of the mud snail, Ilyunassa (Nassarius) obsoleta (Say), and its polytrophic role in the food web. J. Exp. Mar. Biol. Ecol., Vol. 41, pp. 289-297. DEWITT, T. H., 1985. The behavior and ecology of migration and colonization of the tubicolous benthic amphipod Microdeutopus gryllotalpa. Ph.D. dissertation, State University of New York, Stony Brook, N.Y., 266 pp. GRANT, D.C., 1965. Specific diversity in the infauna of an intertidal sand community. Ph.D. dissertation, Yale University, New Haven, Conn., 114 pp. HOBSON,E. S. & J. R. CHESS, 1976.Trophic interactions among fishes and zooplankters near shore at Santa Catalina Island, California. Fish. Bull. NOAA, Vol. 74, pp. 567-598. HOBSON, E.S. & J.R. CHESS, 1977. Trophic relationships among fishes and plankton in the lagoon at Enewetok Atoll, Marshall Islands. Fish. Bull. NOAA, Vol. 77, pp. 275-280. HUNT, J. H., 1981. The importance of adult-larval interactions in determining abundance patterns of soft-sediment infauna. Masters thesis, University of North Carolina, Chapel Hill, N.C., 59 pp. LEVINTON,J. S., 1971. Control of tellinacean (Mollusca: Bivalvia) feeding behavior by predation. Limnol. Oceanogr., Vol. 6, pp. 660-662. LEVINTON,J. S., 1977. Ecology of shallow water deposit-feeding communities. In, Ecology of marine benthos, edited by B.C. Coull, University of South Carolina Press, Columbia, S.C., pp 191-227. LEVINTON,J.S., 1979. The effect of density upon deposit-feeding populations: movement, feeding and floating of Hydrobia ventrosa Montagu (Gastropoda, Prosobranchia). 0ecolq:a (Berlin), Vol. 43, pp. 27-39. LEVINTON,J. S., S. STEWART& T.H. DEWITT, 1985. Complex interactions of a deposit feeder with its resources. 2. Field and laboratory experiments on interference between Hydrobia totfeni and Ilyanassa obsoleta (Gastropoda) and its possible relationship to seasonal shiRs in zonation on mudflats. Mar. Ecol. Prog. Ser., Vol. 22, pp. 53-58. MCCALL, P. L., 1977. Community patterns and adaptive strategies of the infaunal benthos of Long Island Sound. J. Mar. Res., Vol. 35, pp. 221-265. MILLS, E.L., 1967. The biology of an ampeliscid amphipod crustacean sibling species pair. J. Fish. Res. Board Can., Vol. 24, pp. 305-355. MYERS,A.A., 1969. A revision of the amphipod genus Microdeutopus Costa (Gammaridea : Aoridae). Bull. Br. Mus. (Nat. Hi%) Zool., Vol. 17, pp. 93-148. NICHOLS,J.A. & J.R. ROBERTSON,1979. Field evidence that the eastern mud snail, Ilyanassa obsoleta, influences nematode community structure. Nautilus, Vol. 93, pp. 44-46. ORTH, R.J., 1977. The importance of sediment stability in seagrass communities. In, Ecology of marine benfhos, edited by B.C. Coull, University of South Carolina Press, Columbia, S.C., pp. 28 l-300. PACE, M. L., S. SHIMMEL,& W. M. DARLEY,1979. The effect of grazing by a gastropod, Nassarius obsolerus, on the benthic microbial community of a salt marsh. Estuarine Coastal Mar. Sci., Vol. 9, pp. 121-134. PARKER,R. H., 1975. The study of benthic communities : a model and a review. Elsevier, Amsterdam, 279 pp. RHOADS, D. C. & D. K. YOUNG, 1970. The influence of deposit feeding organisms on sediment structure and community trophic structure. J. Mar. Res., Vol. 28, pp. 150-178. ROBERTSON,A.I. & R.K. HOWARD, 1978. Die1 trophic interactions between vertically migrating zooplankton and their fish predators in an eelgrass community. Mar. Biol., Vol. 48, pp. 207-213. SANDERS,H. L., E. M. GOUDSMIT,E. L. MILLS & G. R. HAMPSON, 1962. A study of the intertidal fauna of Barnstable Harbor, Massachusetts. Limnol. Oceanogr., Vol. 7, pp. 63-79. SANDERS,H.L., J.F. GRASSLE, G.R. HAMPSON,L. S. MORSE, S. GARNER-PRICE& C.C. JONES, 1980. Anatomy of an oil spill: long-term effects from the grounding of the barge Florida off West Falmouth, Massachusetts. J. Mar. Res., Vol. 38, pp. 265-380. SIMON,J. L & D. M. DAUER, 1977. Reestablishment of a benthic community following natural defaunation. In, Ecology of marine benthos, edited by B.C. Coull, University of South Carolina Press, Columbia, S.C., pp. 139-154. SOKAL, R.R. & F.J. ROHLF, 1981. Biometry. W.H. Freeman & Co, San Francisco, Calif., 2nd edition, 859 pp. STRICKLAND, J. D. H. & T. R. PARSONS,1969. A practical handbook of seawater analysis. Bull. Fish. Res. Board Can., No. 167, 3 11 pp.
MUD SNAIL DISTURBANCE
AND EPIFAUNAL
EMIGRATION
113
THISTLE, D., 1980. Natural physical disturbances and communities of marine soft bottoms. Mar. Ecol. Prog. Ser., Vol. 6, pp. 223-l 18. VAN BLARICOM, G.R., 1978. Disturbance, predation, and resource allocation in a high energy sublittoral sand-bottom ecosystem: experimental analyses of critical structuring processes for the infaunal community. Ph.D. dissertation, University of California, San Diego. VOLLENWEIDER, R. A., 1969. A manual on the methods for measuring primary productivity in aquatic environments. I.B.P. Handbook No. 12. Blackwell Scientific Publications, Oxford, 213 pp. WEINBERG, J. R., 1979. Ecological determinants of spionid distributions within dense patches of depositfeeding polychaete Axiothella rubrocincta. Mar. Ecol. Prog. Ser., Vol. 1, pp. 301-314. WILSON, JR., W.H., 1981. Sediment-mediated interactions in a densely populated infaunal community: the effects of the polychaete Abarenicola pacifica. J. Mar. Res., Vol. 39, pp. 735-748. WILSON, JR., W. H., 1984. An experimental analysis of spatial competition in a dense infaunal community: the importance of relative effects. Estunrine Coastal Shelf Sci., Vol. 18, pp. 673-684. WILTSE, W. I., 1980. Effects ofPolinices duplicatus (Gastropoda : Naticidae) on infaunal community structure at Barnstable Harbor, Massachusetts. Mar. Biol., Vol. 56, pp. 301-310. WOODIN, S.A., 1974. Polychaete abundance patterns in a marine soft-sediment environment: the importance of biological interactions. Ecol. Monogr., Vol. 44, pp. 171-187. WOODIN, S.A., 1978. Refuges, disturbance, and community structure: a marine soft-bottom example. Ecology, Vol. 59, pp. 274-284. WOODIN, S.A., 1981. Disturbance and community structure in a shallow water sand flat. Ecology, Vol. 62, pp. 1052-1066. ZAIAC, R.N. & R.B. WHITLATCH, 1982a. Responses of estuarine infauna to disturbance. I. Spatial and temporal variation of initial recolonization. Mar. Ecol. Prog. Ser., Vol. 10, pp. l-14. ZAJAC, R.N. & R. B. WHITLATCH, 1982b. Responses of estuarine infauna to disturbance. II. Spatial and temporal variation of succession. Mar. Ecol. Prog. Ser., Vol. 10, pp. 15-27.
