Content uploaded by D.M. Parsons
Author content
All content in this area was uploaded by D.M. Parsons on Jul 03, 2014
Content may be subject to copyright.
MARINE ECOLOGY PROGRESS SERIES
Mar Ecol Prog Ser
Vol. 303: 235–244, 2005
Published November 21
INTRODUCTION
When an animal encounters a predation threat it
must make a choice between maintaining ‘normal’
activity patterns or changing activity to reduce the
probability of predation. Both decisions have poten-
tially detrimental consequences (Abrams 1993). Decid-
ing to maintain ‘normal’ activity could be a fatal choice
with direct implications, whereas a more conservative
decision would reduce predation risk but could indi-
rectly and negatively affect a host of functions includ-
ing activity patterns, vigilance, sheltering behavior,
social interactions, time spent feeding, growth and
reproductive potential (see review by Lima 1998).
These behavioral decisions are not without potentially
long-term consequences for the individual, population
or community.
Like natural predation events, marine fisheries can
expose target species to direct effects through the
removal of biomass, and indirect disturbance whereby
surviving animals modify some aspect of their biology.
The direct impact of biomass extraction is generally
the focus of fisheries management; however, its indi-
rect impacts and disturbances are largely ignored.
Some indirect impacts of fishing that have been
addressed include the capture of non-target species
(Davis 2002), habitat destruction caused by contact
fishing gears (Thrush et al. 2001), and trophic cascades
resulting from reduced target species abundance
(Babcock et al. 1999). Few studies, however, have
addressed the potential for fishing practices to indi-
rectly modify and disturb animal behavior. We do
know that (1) fishes subjected to capture and release
exhibit reduced growth and elevated mortality (Warner
1978, Loftus et al. 1988, Diggles & Ernst 1997), (2) fish
that escape through trawl nets have altered behaviors
that may increase their risk of predation (Ryer et al.
2004), and (3) dredge fishery discards can lead to
© Inter-Research 2005 · www.int-res.com*Email: darren.parsons@clear.net.nz
Indirect effects of recreational fishing on behavior of
the spiny lobster Panulirus argus
Darren M. Parsons*, David B. Eggleston
North Carolina State University, Department of Marine, Earth and Atmospheric Sciences, Raleigh, North Carolina 27695-8208, USA
ABSTRACT: Sublethal disturbance may lead to behavioral modifications that have detrimental con-
sequences for the individual. Sublethal disturbance is an indirect effect of fishing that has seldom
been examined. In summer 2003, we conducted surveys on the Caribbean spiny lobster Panulirus
argus in the Florida Keys, USA, before and after a 2 d mini-season exclusively for recreational sport-
divers to assess the frequency of injured lobsters. We also conducted (1) a tethering experiment to
assess the influence of disturbance and injury on predation-induced mortality, and (2) Y-maze labo-
ratory experiments to assess the effect of injury and disturbance on the shelter choice of conspecifics.
Injury surveys revealed an immediate increase (to 27.16%) in the percentage of injured lobsters from
pre- to post mini-season within patch reef habitats, but only a small increase in injured lobsters
amongst patch head habitats. Tethering trials revealed that injured lobsters were more likely to suc-
cumb to predation than uninjured controls, while the shelter choice experiments demonstrated that
the usual ability to attract conspecifics was altered by injury. Considering the ability of this fishery to
unintentionally injure >25% of the lobster population in patch reef habitats in just 2 d, the potential
population consequences to spiny lobsters must be examined.
KEY WORDS: Indirect behavioral disturbance · Sublethal injury · Unobserved mortality · Predation ·
Limb damage · Recreational fishing · Panulirus argus · Caribbean spiny lobster · Florida Keys
Resale or republication not permitted without written consent of the publisher
Mar Ecol Prog Ser 303: 235–244, 2005
behavioral modification of scavengers (Veale et al.
2000). What is lacking and needed, is quantification of
human disturbance events, the resulting behavioral
modifications by target species and, if altered be-
haviors lead to increased mortality, the mechanisms
underlying elevated mortality and population conse-
quences of such delayed mortality.
Recreational spiny lobster fisheries provide a unique
opportunity to quantify the effects of human distur-
bance on animal behavior and mortality because of
the temporally intense, pulsed nature of this human
disturbance (Eggleston et al. 2003), and because sport-
divers visually locate and attack lobsters in their dens.
Human disturbance of lobsters can result from either
(1) an unsuccessful attack, or (2) a successful capture
followed by release of a sub-legal lobster.
The Caribbean spiny lobster Panulirus argus sup-
ports important recreational and commercial fisheries
throughout its range (Lipcius & Eggleston 2000). P.
argus is highly gregarious, aggregating in crevices and
beneath sponges during the day (Herrnkind et al.
1975, Eggleston & Lipcius 1992, Eggleston & Dahl-
gren 2001), and foraging on gastropods, chitons and
bivalves in nearby seagrass beds and hard bottom
habitats at night (Cox et al. 1997). Ontogenetic habitat
shifts occur when juveniles move from inshore macro-
algal settlement habitats to back-reef crevice habitats,
and finally to offshore reefs as adults (Davis & Dodrill
1989). When spiny lobsters return to their dens from
nocturnal foraging, or exhibit larger scale ontogenetic
habitat shifts, they are guided by the olfactory cue of
conspecifics, resulting in the aggregation of indi-
viduals in these shelters (Ratchford & Eggleston 1998,
2000, Childress & Herrnkind 2001). Commercial fish-
ermen of the Florida Keys, exploit the gregarious
nature of spiny lobsters by baiting traps with live
juveniles (Hunt et al. 1986), whereas recreational
sport-divers exploit the gregarious nature of lobsters
by targeting dens with high densities of lobsters, and
coercing them into hand-nets with ‘tickle sticks’
(Eggleston et al. 2003). Divers disturb lobsters within a
den as they attempt to coerce the target individual out.
In addition, sub-legal lobsters that are captured, mea-
sured, and released will also be disturbed and poten-
tially injured. Therefore, the recreational fishery not
only extracts 22% of the total lobster harvest in the
Florida Keys each year (Hunt 2000), but probably
exerts major indirect effects through disturbance and
sublethal injury.
Approximately 50% of sub-legal and legal-sized
spiny lobsters in Biscayne Bay, Florida, possessed
injuries after the 1977 regular fishing season (com-
pared to 31% injury immediately before that season
opened in 1976; Davis 1981). These injuries reduced
growth rates, and increased the probability of death of
sub-legal lobsters before they recruited to the fishery
(Davis 1981). Since Davis’ (1981) study in the late
1970s, the intensity of the recreational fishery in the
Florida Keys has increased greatly. Most notably,
during a 2 d mini-season exclusively for recreational
sport-divers, which opens 1 wk prior to the opening of
the regular lobster fishing season for commercial and
recreational fishers during August to March (Eggle-
ston & Dahlgren 2001, Eggleston et al. 2003). For
example, the abundance of post mini-season lobsters
decreased by 55% in 1987 (Blonder et al. 1992), but
decreased by 80 to 90% after the 2000 and 2002 mini-
seasons (Eggleston & Dahlgren 2001, Eggleston et al.
2003). It seems reasonable to assume that the inci-
dence of lobster injury and disturbance has increased
accordingly, but this has remained untested.
In this study, we document the frequency of injured
lobsters immediately before and after the 2 d mini-
season for recreational sport-divers in the Florida
Keys. We also investigated 2 potential consequences of
disturbance and injury: (1) delayed predation-induced
mortality, and (2) effects on gregarious sheltering
behavior.
MATERIALS AND METHODS
Study sites. Lobster surveys and experiments were
conducted within the Great White Heron National
Wildlife Refuge on the NW side of the lower Florida
Keys, Florida (Fig. 1). Within this region we surveyed 2
types of coral reef habitat: (1) continuous coral reefs
and ledges, herein referred to as ‘patch reefs’, and (2)
discrete aggregations of boulder coral located within
shallow seagrass beds, herein referred to as ‘patch
heads’ (see Eggleston et al. 2003 for description of
coral patch reef and patch head habitats in this region).
236
0 10 Kilometers
N
Key West
Sugarloaf
Key
Cudjoe
Channel
Waltz Key
Basin
Tur k ey
Basin
Florida
Florida Keys
Content Keys
Snipe Point
Little Crane
Key
Fig. 1. Florida, USA (inset) and Florida Keys, showing patch
reef (q) and patch head (Q) study sites
Parsons & Eggleston: Indirect effects of recreational fishing on spiny lobster
Patch reefs were located along a nearly continuous line
running NE to SW and parallel to the lower Florida
Keys (Fig. 1). Patch heads were located within discrete
basins, such as Waltz Key Basin and Turkey Basin,
along the edge of Cudjoe Channel, and in the area
surrounding Little Crane Key (Fig. 1).
Sublethal lobster injury. SCUBA divers quantified
the proportion and distribution of injured lobsters 4 to
6 d before and after the 2 d mini-season for sport-
divers during July and August 2003. Injuries were
defined as those lobsters missing >50% of an antenna,
missing legs, or with damage (i.e. crushed exoskele-
ton) to the abdomen, cephalothorax or supraorbital
horns (Davis 1981). The size of injured lobsters was
estimated visually (without capture) using a 50 cm long
PVC pipe marked with 1 cm increments. All divers
conducting these surveys had previously practiced
estimating the size of plastic lobster models while
diving, until estimates were accurate to within 1 cm
carapace length (CL). When a lobster was partially
obscured within a crevice, a tickle stick was used to
gently coerce the lobster into view so that size could
be estimated.
Patch reef survey sites were chosen by superimpos-
ing a 1 km × 1 km grid over the continuous patch reef
between Snipe Point and Content Keys (Fig. 1). From
this grid, 6 cells were randomly selected. At each patch
reef cell location, the research boat was anchored as
close as possible to the middle of the grid cell using a
differential GPS, whereupon 2 divers would survey
areas in opposite directions from the boat. Each diver
swam in the same direction before and after the 2 d
mini-season. Divers used a 50 cm length of PVC pipe to
estimate the width of the area searched, swimming
along a nearly straight line away from the boat for
10 min, then surfacing and visually estimating the dis-
tance traveled from the boat. Distances estimated in
this manner are not significantly different from dis-
tances measured with a boat based differential GPS
(Eggleston et al. 2003). In the current study, the aver-
age area searched for a given 10 min survey before the
mini-season was greater (380 m
2
) than the average sur-
vey distance after the mini-season (285 m
2
), but these
areas were not significantly different (paired t-test;
df = 11, t = 1.96, p-value >0.07). Therefore, lobster
abundance was standardized to density (no. m
–2
), and
the proportion of injured lobsters was compared be-
tween pre- and post mini-season.
Patch heads surveyed in this study were chosen from
those measured in 2001 and 2002 for overall lobster
loss from before to after the mini-season (Eggleston et
al. 2003). Divers surveyed patch heads as described
above, but searched the entire patch rather than using
a transect approach (Eggleston et al. 2003). A total of
16 patch head clusters were surveyed from 4 separate
locations: (1) Waltz Key Basin, (2) Turkey Basin,
(3) Cudjoe Channel and (4) Little Crane Key (Fig. 1).
The area of each patch head was calculated by mea-
suring the radius of the coral patch head with a ruler
and multiplying by πr
2
. The response variable mea-
sured for coral patch heads was the proportion of
injured lobsters within each survey site.
Effects of sublethal injury and disturbance on preda-
tion-induced mortality. Tethering experiments assessed
the relative effects of disturbance and sublethal injury
on predation-induced mortality of lobsters in the field.
We hypothesized that injured lobsters would suffer
higher predation-induced mortality than control lob-
sters because leaking body fluids would attract preda-
tors and potentially repulse conspecifics (Ratchford
1999), thereby reducing the benefits of group defense
(Herrnkind et al. 2001). Each tether consisted of a
60 cm long monofilament line anchored to the sub-
stratum by a metal stake adjacent to a natural coral
crevice. The other end of the tether was attached to a
lobster by a cable-tie harness, which was secured
around the cephalothorax between the second and
third walking legs and fastened to the carapace using
cyanoacrylate glue (Eggleston et al. 1990, Lipcius et al.
1998). The tether was long enough for the lobster to
retreat into its shelter, but not so long that the lobster
would became entangled. The minimum legal size of
spiny lobsters that can be harvested in Florida is
76 mm CL. Thus, we chose spiny lobsters with an aver-
age CL of 74 mm to mimic sub-legal lobsters that
are often captured, measured and released by sport-
divers. At each of the 4 patch head locations, 3 lobsters
were tethered to their own individual patch heads.
Each lobster at a given location was randomly assigned
to 1 of 3 treatments: (1) disturbed, whereby a lobster
was prodded with a tickle stick for 1 min each day;
(2) injured, whereby a lobster was initially injured by
removing 1 antenna and the back 2 walking legs on
one side of the cephalothorax; (3) control, with no
disturbance or injury. The methods used in this study
met established university, national and international
guidelines for the ethical treatment of invertebrate
animals.
If daily observations revealed that a given lobster
was alive, the experiment was continued. If the lobster
was absent and the tethering harness intact, the re-
sponse was recorded as an escape. If part of the cara-
pace remained attached to the tether or if the mono-
filament had been snapped, then the response was
recorded as a predation event. Daily observations con-
tinued until the lobster had escaped, died, or until 5 d
had elapsed. After 5 d, it was assumed that starvation
might confound results. When an experiment at a
patch head was discontinued for one of the above rea-
sons, another lobster would be tethered to the same
237
Mar Ecol Prog Ser 303: 235–244, 2005
patch head the next day and a treatment would be
applied at random. In this way, 4 replicates were
conducted within Cudjoe Channel, 3 at Turkey Basin,
2 around Little Crane Key and 1 at Waltz Key Basin,
making a total of 10 replicates for each treatment over
a 17 d period. To avoid human interference, these
experiments were conducted when the fishery was
closed. The response variable used was the probability
of an individual within a treatment group surviving
the full duration of the experiment (5 d). This was
estimated with a survival function derived from the
Kaplan-Meier procedure (Kaplan & Meier 1958). Stan-
dard errors for this survival probability were calculated
as described by Pollock et al. (1989). These survival
probabilities were then compared between treatments
using a 1-tailed normal distribution Z-test.
The second response variable observed in this ex-
periment was the number of conspecifics co-occupying
the patch head den with each tethered lobster. This
number was recorded daily and adjusted to represent
a percentage of the number of lobsters within that den
when the experiment was initiated. We hypothesized
that the average number of lobsters co-occupying a
den with disturbed and injured lobsters would be sig-
nificantly lower than the number of conspecifics occu-
pying a den with control lobsters because the injured
and disturbed lobsters would release body fluids or
possibly stress hormones that could negate the effect
of their natural attraction odor (Ratchford & Eggleston
1998, Ratchford 1999). The daily percent initial occu-
pancy by conspecifics was averaged for the duration
of each tethering trial and analyzed with a 1-way
ANOVA, with experimental treatment (injured, dis-
turbed and control) as the class variable. The assump-
tions of ANOVA were met for these data (normality:
Kolmogorov-Smirnov test, p > 0.15; homogeneity of
variance: Levene’s test, p > 0.07).
Effects of disturbance and injury on sheltering
behavior. To assess how disturbance and injury modi-
fied normal sheltering behavior of Panulirus argus,
a Y-maze, laboratory tank system was constructed
(Fig. 2). We built 3 identical 235 l Y-maze, flow-
through seawater tanks (1.9 m × 0.61 m × 0.28 m) to
allow 3 trials to be conducted simultaneously. Seawa-
ter was pumped into a 190 l header tank, which fed six
38 l header tanks via adjustable valves. Seawater
drained from pairs of these 38 l header tanks into both
sides of each Y-maze tank, after free falling ~10 cm to
ensure that no audible cues were transmitted directly
from the header tank to the Y-maze (Ratchford &
Eggleston 1998). The only shelters available within the
Y-maze tank were 2 concrete blocks placed on either
side of the corners nearest the header tanks (Fig. 2).
Seawater flowed through the Y-maze tanks, through a
drainpipe, and then exited into a nearby canal. Water
flow into each arm of the Y-maze was set at ~1.5 l min
–1
(Ratchford & Eggleston 1998). To check the direction-
ality of flow in the Y-maze, fluorescein dye was added
to 1 header tank and the flow observed. In all cases,
the dye would flow into the start area, and not the
other arm of the Y-maze, before flowing out through
the drainpipe (Fig. 2).
Experimental trials were initiated in the evening
(~20:00 h) from July 29 to August 14, 2003. A trial
began by randomly choosing one of the paired 38 l
header tanks to contain a lobster (hereafter referred to
as the ‘treatment lobster’). Another lobster (the ‘exper-
imental lobster’) was then placed in the start area of
each Y-maze tank (Fig. 2). Lobsters with an average
CL of 75 mm were used in this experiment because
they represented a sub-legal size that was expected to
be most frequently handled and released by sport-
divers, and were a close match to the average lobster
size used in the field-tethering experiment. The
Y-maze was then left undisturbed, with water con-
stantly flowing until 04:30 h, after which one of each of
the 3 treatments (disturbed, injured and control) was
238
Concrete
b
lock
shelter
Adjustable
valve
Treatment
lobster
Experimental
Lobster in
‘start area’
Divider
b
oard
Header
tank
Inflow
hosing
Outflow
hosin
g
Seawater
supply
Fig. 2. Experimental Y-maze arena used to examine shelter-
ing behaviour of Panulirus argus in response to injury and
disturbance. Treatment (control, disturbance, or injury) was
applied to lobster in header tank and shelter occupancy of
experimental lobster was recorded in concrete block shelters
Parsons & Eggleston: Indirect effects of recreational fishing on spiny lobster
randomly assigned to 1 lobster in each of the header
tanks. The disturbed treatment was conducted by
prodding the lobster with a tickle stick for ~1 min; the
injured treatment was conducted by breaking 1 an-
tenna and the last 2 walking legs on one side of the
lobster, while the control lobster was left undisturbed.
This procedure was conducted under red light to
ensure that all other lobsters remained undisturbed
(Cummins et al. 1984).
The next morning before 08:00 h, the shelter chosen
by the experimental lobster was recorded. For each
trial, the response was defined as either positive or
negative. If the experimental lobster was sheltering in
the concrete block downstream of the occupied header
tank, then the experimental lobster’s shelter choice
was positively associated with a conspecific in the
header tank. If the experimental lobster was sheltering
downstream of the unoccupied header tank, the exper-
imental lobster’s shelter choice was negatively associ-
ated with a conspecific in the header tank. If the exper-
imental lobster was not located within either concrete
block shelter, then that individual trial was omitted
from the results. A total of 11 disturbed, 13 injured and
10 control trials were successfully conducted. Data
were analyzed by comparing the observed proportion
of positive shelter choices for each treatment to the
proportion expected by random choice (0.5). Differ-
ences between observed and expected shelter choices
were assessed using 1-tailed binomial tests (Zar 1984).
In addition, 1-tailed binomial unconditional p-values
were calculated using 2 individual Fisher’s exact tests
to allow for comparison with the control treatment
(Berger 1996). Where multiple comparisons were made,
Type I error was controlled for by Bonferroni correction.
RESULTS
Sublethal lobster injury
During July 2003, both coral patch reef and patch
head habitats contained an extremely low percentage
of injured lobsters before the 2 d recreational mini-
season (0 and 0.15%, respectively). After the mini-
season the percentage of injured lobsters had increased
to 27.16% on patch reefs, and 3.77% on patch heads.
Effects of sublethal injury and disturbance on
predation-induced mortality
The mean percent survival (Kaplan-Meier statistic)
of tethered lobsters in coral patch heads prior to the
mini-season declined with increasing intensity of the
experimental treatment (Fig. 3). The only statistically
significant difference in percent survival, however,
was between injured and control lobsters (1-tailed Z
test
(α = 0.05)
; Z
obs
= 1.80, Z
crit
= 1.65). The percent survival
of disturbed lobsters was not significantly different
than survival of injured (1-tailed Z test
(α = 0.05)
; Z
obs
=
0.97, Z
crit
= 1.65) or control (1-tailed Z test
(α = 0.05)
; Z
obs
=
0.35, Z
crit
= 1.65) lobsters.
While applying the daily disturbance and injury
treatments to the tethered lobsters, the number of co-
resident Panulirus argus was recorded to determine
any potential avoidance behavior or decreased attrac-
tion by untethered conspecifics residing within coral
patch heads. There was a decrease in the daily propor-
tion of lobsters co-resident with control and injured
lobsters, but not with disturbed lobsters (Fig. 4). Sig-
nificantly more co-resident lobsters remained with dis-
turbed than injured lobsters (1-way ANOVA; df = 2,
F = 3.87, p < 0.04). However, neither the injured nor
disturbed treatments were significantly different from
the control (Tukey’s multiple comparison test).
Effects of disturbance and injury on sheltering
behavior
The strength of attraction by Panulirus argus to con-
specific odors decreased as the intensity of the treat-
ment increased (Fig. 5). For example, lobsters chose
shelter receiving water from an undisturbed conspe-
cific 80% of the time, which was significantly greater
than random (binomial test; n = 10, p = 0.04). The per-
cent of lobsters attracted to conspecifics was not differ-
ent than random in the disturbed (55%: binomial test;
239
0
50
100
Control Disturbed Injured
Experimental treatment
% survival
A
AB
B
Fig. 3. Panulirus argus. Percent survival of lobsters tethered to
coral patch heads as a function of disturbance (prodding with
a tickle stick daily) or injury (loss of 1 antenna and 2 walking
legs). Means ± SE are shown; n = 10 for each treatment. Treat-
ments with matching letters denote no significant differ-
ence upon pairwise comparison. See ‘Results’ for outcome
of statistical tests
Mar Ecol Prog Ser 303: 235–244, 2005
n = 11, p = 0.23) and injured (38%: binomial test; n =
13, p = 0.16) treatments. The percent of lobsters
attracted to control lobsters was significantly higher
than the percent of lobsters attracted to injured lob-
sters at the 0.05 level of significance and provided
some evidence of a significant difference at the
0.025 level, after accounting for multiple comparisons
(Fisher’s exact test; unconditional p-value = 0.03). The
control versus disturbed (p-value = 0.13) comparison
revealed no significant difference in the percentage
of lobsters attracted (Fig. 5).
DISCUSSION
This study illustrates that sublethal disturbances by
recreational sport-divers can increase the frequency of
injured lobsters, alter shelter choice behavior, and in-
crease predation-induced mortality of injured lobsters.
Examples of behavioral disturbance in the marine
environment are rare, although tourism activities like
dolphin-watching are known to alter marine mammal
behavior (Constantine et al. 2004). The current study is
one of the first (as far as we are aware) to demonstrate
that recreational fishing disturbance may increase
mortality of the target species through a modification
of prey behavior. In commercial fisheries, behavioral
impairment caused by escape through trawl cod-ends
may elevate mortality of juvenile walleye pollock
Theragra chalcogramma exposed to the lingcod pre-
dator Ophiodon elongatus (Ryer 2002). In terrestrial
systems, redistribution and increased nocturnal activity
of white-tailed deer Odocoileus virginianus exposed to
hunters has been hypothesized to increase predation
by the Florida panther Felis concolor coryi (Kilgo et al.
1998). The current study has shown that sublethal
disturbances can have detrimental consequences for
individuals, and we recommend quantification of
similar disturbance events in other fisheries.
Sublethal lobster injury
The percentage of injured lobsters increased from
0 to 27% in coral patch reefs during the sport-diver
mini-season. While previous studies have indicated a
high percentage of injury resulting from the cumula-
tive and often long-term effects of recreational and
commercial fishing (an increase from 31 to 50% over a
7 mo fishing season; Davis 1981), the current study has
shown that a substantial increase in the frequency of
injury can occur over only 2 d. Since injuries to fishery
target species can reduce survival (Ryer et al. 2004),
the frequency at which they occur and the subsequent
increase in mortality rate should be incorporated into
population and fishery stock assessment models. Alter-
natively, on coral patch heads we did not observe a
large increase in the percentage of injured lobsters
from pre- to post mini-season. This was opposite to the
pattern of injury frequency we expected given that
sport-diver fishing effort can be 10-fold greater on
coral patch heads than on coral patch reefs (Eggleston
et al. 2003).
It is unclear why the percentage of injured lobsters
was low in coral patch head habitats compared to coral
patch reefs. Physical differences between patch heads
and patch reefs may result in easier capture of lobster
at patch heads, thereby reducing the chance of injur-
240
0
0.5
1
1.5
Control Disturbed Injured
Treatment
Daily proportion of co-residents
remaining
AB BA
Fig. 4. Panulirus argus. Mean (±SE) daily proportion of original
number of co-resident lobsters remaining in coral patch heads
during tethering experiments. This was determined by counting
initial number of co-resident lobsters within the same coral
patch head as the tethered lobster and daily recounting them.
Dashed line: null hypothesis of no change in number of co-
resident lobsters over time. Treatments with matching letters
denote no significant difference upon pairwise comparison
0
50
100
Control Disturbed Injured
Treatment
% attracted
n = 10
p = 0.04 *
n = 11
p = 0.23
n = 13
p = 0.16
AABB
Fig. 5. Panulirus argus. Results of Y-maze laboratory experi-
ment testing effects of disturbance and injury on attractive-
ness of experimental lobsters to conspecifics. p-values were
based on a 1-tailed binomial test (Zar 1984), where null prob-
ability of choosing a shelter with conspecific odor = 0.5; *: sig-
nificant at α = 0.05. Treatments without matching letters
denote borderline significant difference at α = 0.025 upon
pairwise comparison using multiple Fisher’s exact tests
Parsons & Eggleston: Indirect effects of recreational fishing on spiny lobster
ing lobsters. For example, patch heads are generally
surrounded by sand and seagrass. Lobsters that escape
an initial capture attempt by divers will typically tail-
flip into another den in the same patch head, or onto
the surrounding sand and seagrass. Once a lobster is in
the seagrass it is easily captured by placing a hand-net
behind it and coercing it into the net with a tickle stick
(D. M. Parsons pers. obs.). Thus, the attack success by
divers on patch heads may be higher than on patch
reefs, resulting in fewer remaining injured lobsters.
Another explanation might involve the dispersion of
injured lobsters away from disturbed dens and toward
alternative, potentially suboptimal shelters. These
hypotheses remain to be tested.
Effects of sublethal injury and disturbance on
predation-induced mortality
The increased mortality of injured lobsters in the
tethering study, and the elevated frequency of injury
after the mini-season, suggest that lobster injury in-
creases overall mortality rates on lobsters. While esti-
mates of mortality may be inflated due to experimental
artefacts common to all tethering experiments (Peter-
son & Black 1994), it is the relative differences be-
tween treatments that are of interest here. We did not
observe any differences in artefacts between treat-
ments, such as a higher probability of lobsters tangling
in the tether for some treatments.
The higher predation rate on injured versus control
lobsters in the field could be due to 3 non-mutually
exclusive mechanisms. (1) When an injury occurs flesh
is exposed, releasing a mixture of organic compounds.
These compounds include amino acids, which are
known chemical attractants of foraging predators
(Zimmer et al. 1999). Therefore, predators with a good
olfactory sense, such as sharks (Hamlett 1999), could
follow this odor trail and more easily locate their prey.
(2) The defensive capacity of a lobster may be reduced
by injury. If legs are removed it may restrict a lobster’s
mobility and therefore ability to escape. Removal of
the antenna, which are used to whip potential preda-
tors (Herrnkind et al. 2001), may limit a lobster’s ability
to successfully fend off attackers. (3) Injured lobsters
may lose the benefits of group defense (Herrnkind et
al. 2001) when co-resident lobsters abandon the den.
This third mechanism was probably not important in
determining patterns of predator-induced mortality in
this study, since co-resident lobsters abandoned dens
containing tethered lobsters that were exposed to both
control and injury treatments. At time-scales greater
than the duration of this study, an additional mecha-
nism for higher mortality of injured lobsters would be
increased energetic costs associated with injury (Juanes
& Smith 1995). Increased energetic demand may cause
an animal to take greater risks to obtain food and
reduce its ability to escape predators (Sinclair & Arcese
1995), further elevating predation.
Effects of disturbance and injury on sheltering
behavior
The results from our Y-maze experiments indicated
that lobsters are attracted to dens containing con-
specifics, as has been previously observed (Ratchford &
Eggleston 1998). This strong tendency for lobsters to
seek shelter emitting the odor of conspecifics appeared
to be reduced when the treatment lobster was injured.
Moreover, reduced attractiveness by conspecifics toward
injured lobsters could alter the ecological relationship
between gregariousness, predation risk and shelter
choice observed in Panulirus argus (Eggleston & Lipcius
1992), and reduced social interaction could lessen the
benefits of group defense (Herrnkind et al. 2001).
Lobsters find suitable shelter faster when there are
conspecifics within a shelter than when they are absent
(Childress & Herrnkind 2001). Therefore, when high
numbers of injured lobsters exist in a den, the cues lob-
sters use to find dens may be reduced. This reduction in
the ‘guide effect’ (Childress & Herrnkind 2001) may
alter local patterns of habitat use, even for lobsters that
are not disturbed or injured. For example, undisturbed
lobsters may be stranded in soft sediment habitats in
the absence of correct cues from uninjured lobsters
(Childress & Herrnkind 2001). Moreover, lowered
success in finding suitable daytime shelter may expose
lobsters to a relatively high risk of predation mortality.
There are 2 potential mechanisms explaining the re-
duced attraction behavior observed in our Y-maze
choice experiments. Either production of the normal
attractive odor (Ratchford & Eggleston 1998) may be
reduced or cease when lobsters are disturbed or in-
jured, or disturbance and injury may initiate the release
of a disturbance odor. Results of a similar experiment
also indicated that disturbance may promote the re-
lease of an odor that alters the sheltering behavior of
conspecifics (Ratchford 1999). If such a disturbance
odor exists, it could serve as a pheromone that is unin-
tentionally released under stressful conditions. It would
provide no benefit to the message sender, but broad-
cast information about predation risk (Chivers & Smith
1998). Since the disturbed and injured treatments in the
current study did not produce significant avoidance, it
is likely that if a disturbance odor exists, it does not op-
erate to the exclusion of the attraction odor. This would
indicate that either the production of the attraction odor
was reduced or terminated, or that the attraction and
disturbance odors were competing with one another.
241
Mar Ecol Prog Ser 303: 235–244, 2005
FINAL REMARKS
Reduced growth rates of sub-legal lobsters caused
by injury may prolong the time they spend in smaller
size classes, resulting in an additional 22% of sub-legal
lobsters succumbing to natural predation before they
enter the fishery (Davis 1981). Our study suggests that
injured lobsters will also be exposed to a higher daily
rate of predation, which could further decrease the
proportion of individuals eventually recruiting to the
fishery. Spiny lobsters also appear to alter their shelter-
ing behavior in the presence of injured conspecifics,
which may affect survival. The population-level and
fishery implications of these indirect disturbances
remain unknown, but some level of unobserved mor-
tality is likely to exist. Detrimental effects of human
disturbance have been demonstrated at the population
level amongst various terrestrial species (e.g. mule
deer, Yarmoloy et al. 1988; caribou, Harrington &
Veitch 1992; pink footed geese, Madsen 1994), but
less frequently in marine fisheries where the indirect
effects of human disturbance and the direct effects of
fishing itself may be difficult to separate. Matthews
(2001), however, estimated that exposure and confine-
ment of juvenile Panulirus argus used as attractants in
traps reduces recruitment to the fishery by 0.646 mil-
lion individuals annually. If the population-level conse-
quences of the human disturbance events described in
the current study can also be quantified, management
attempts to reduce the frequency of avoidable injuries
to sub-legal lobsters and/or account for this unob-
served mortality may prove beneficial.
With respect to the potential impact of indirect
effects in other recreational lobster fisheries, recre-
ational catch data are rare and there have been no
investigations, that we are aware of, addressing the
indirect effects of recreational fishermen. In South
Africa, Western Australia and New Zealand, annual
recreational catches of spiny lobsters are ~379, 630 and
300 t, respectively (Cockcroft & Mackenzie 1997,
Melville-Smith et al. 2001, K. J. Sullivan unpubl. data).
This represents 25, 4.8 and ~10% of the total annual
landings in the respective countries. In contrast, the
Floridian recreational fishery lands 225 t in the first 2 d,
and accounts for ~40% of the total annual landings
(Sharp et al. 2005). In all but one of these fisheries,
there have been recent increases in the percentage of
the total catch taken by recreational fishermen: an
increase of ~10% between 1993 and 2001 in Florida
(Sharp et al. 2005), ~18% between 1991 and 1995 in
South Africa (Cockcroft & Mackenzie 1997), and ~3%
between 1986 and 1998 in Western Australia (Melville-
Smith et al. 2001). Whether the increasing proportion
of landings by the recreational fishery will correspond-
ingly increase the frequency of injured lobsters is
unclear. Regardless of the percentage of catch ob-
tained by recreational fishermen, spiny lobster fish-
eries throughout the world may have large unobserved
mortalities associated with commercial diving and
trapping. Commercial fishers in the Florida Keys also
have a high potential to contribute indirect effects to
the lobster population, since sub-legal lobsters are
used as attractants and traps are set for long periods
(up to a month) without escape gaps (Matthews 2001).
In other countries, traps usually have escape gaps and
soak for ~1 d (Phillips et al. 1994, D. M. Parsons pers.
obs.). The differences in fishing methods suggest that
indirect effects are probably most prevalent in the
Floridian fishery, but may still be of concern to other
spiny lobster fisheries. The frequency of indirect
impacts on a given fishery will depend on the relative
proportion of different capture methods within a fish-
ery, and the potential for sublethal injuries posed by
each method. While indirect impacts should be
assessed in all fisheries, of particular interest is how
the magnitude of indirect impacts will change as some
fisheries become dominated by recreational effort
(Coleman et al. 2004).
Acknowledgements. Thanks to G. Bell, M. Darcy and G. Plaia
for assistance with field work; M. Childress, J. Hightower,
S. Thrush, T. Wollcot and 2 anonymous reviewers for com-
ments on the manuscript; and K. Pollock, and N. Tolimieri for
statistical assistance. We thank J. Sobel of The Oceans
Conservancy for facilitating the research funding and his
enthusiastic administration of this project; and A. Gude of the
US Fish & Wildlife Service for logistical support. Funding for
this project was provided by a Challenge Cost-Share Agree-
ment between The Oceans Conservancy and the US Fish &
Wildlife Service for Contracts 1448-40181-99-6 and 1448-
40181-00-6143, as well as the Fund for Sustainable Fisheries,
North Carolina State University. This research was conducted
with a research and education permit granted by the Florida
Keys National Marine Sanctuary Program (FKNMS-2002-061).
LITERATURE CITED
Abrams PA (1993) Why predation rate should not be propor-
tional to predator density. Ecology 74:726–733
Babcock RC, Kelly S, Shears NT, Walker JW, Willis TJ (1999)
Changes in community structure in temperate marine
reserves. Mar Ecol Prog Ser 189:125–134
Berger RL (1996) More powerful tests from confidence
interval p values. Am Stat 50:314–318
Blonder BI, Hunt JH, Forcucci D, Lyons WG (1992) Effects of
recreational and commercial fishing on spiny lobster
abundance at Looe Key National Marine Sanctuary. In:
Waugh GT, Godwin MH, Kau SM (eds) Proc 41st Gulf
Carib Fish Inst 41:487–491
Childress MJ, Herrnkind WF (2001) The guide effect influ-
ence on the gregariousness of juvenile Caribbean spiny
lobsters. Anim Behav 62:465–472
Chivers DP, Smith RJF (1998) Chemical alarm signalling in
aquatic predator–prey systems: a review and prospectus.
Ecoscience 5:338–352
Cockcroft AC, Mackenzie AJ (1997) The recreational fishery
242
Parsons & Eggleston: Indirect effects of recreational fishing on spiny lobster
for west coast rock lobster Jasus lalandii in South Africa.
S Afr J Mar Sci 18:75–84
Coleman FC, Figueira WF, Ueland JS, Crowder LB (2004) The
impact of United States recreational fisheries on marine
fish populations. Science 305:1958–1960
Constantine R, Brunton DH, Dennis T (2004) Dolphin-watching
tour boats change bottlenose dolphin (Tursiops truncatus)
behaviour. Biol Conserv 117:299–307
Cox C, Hunt JH, Lyons WG, Davis GE (1997) Nocturnal
foraging of the Caribbean spiny lobster (Panulirus argus)
on offshore reefs of Florida, USA. Mar Freshw Res 48:
671–679
Cummins DR, Chen DM, Goldsmith TH (1984) Spectral sen-
sitivity of the spiny lobster, Panulirus argus. Biol Bull
(Woods Hole) 166:269–276
Davis GE (1981) Effects of injuries on spiny lobster, Panulirus
argus, and implications for fishery management. Fish Bull
(Seattle) 78:979–984
Davis GE, Dodrill JW (1989) Recreation fishery and popula-
tion dynamics of spiny lobsters, Panulirus argus in Florida
Bay, Everglades National Park, 1977–1980. Bull Mar Sci
44:78–88
Davis MW (2002) Key principles for understanding fish by-
catch discard mortality. Can J Fish Aquat Sci 59:1834–1843
Diggles BK, Ernst I (1997) Hooking mortality of two species of
shallow-water reef fish caught by recreational angling
methods. Mar Freshw Res 48:479–483
Eggleston DB, Dahlgren CP (2001) Distribution and abundance
of Caribbean spiny lobsters in the Key West National
Wildlife refuge: relationship to habitat features and im-
pact of an intensive recreational fishery. Mar Freshw Res
52:1567–1576
Eggleston DB, Lipcius RN (1992) Shelter selection by spiny
lobster under variable predation risk, social conditions,
and shelter size. Ecology 73:992–1011
Eggleston DB, Lipcius RN, Miller DL, Coba-Cetina L (1990)
Shelter scaling regulates survival of juvenile Caribbean
spiny lobster Panulirus argus. Mar Ecol Prog Ser 62:79–88
Eggleston DB, Johnson EG, Kellison GT, Nadeau DA (2003)
Intense removal and non-saturating functional responses
by recreational divers on spiny lobster prey. Mar Ecol Prog
Ser 257:197–207
Hamlett WC (1999) Sharks, skates, and rays: the biology
of elasmobranch fishes, Vol I. Johns Hopkins University
Press, Baltimore, MD
Harrington FH, Veitch AM (1992) Calving success of wood-
land caribou exposed to low-level jet fighter overflights.
Arctic 45:213–218
Herrnkind WF, Vanderwalker J, Barr L (1975) Population
dynamics, ecology and behavior of spiny lobsters off St.
Johns, U.S. V.I. Habitation, patterns of movement, and
general behavior. Sci Bull Nat Hist Mus Los Angel Cty 20:
31–45
Herrnkind WF, Childress MJ, Lavalli KL (2001) Cooperative
defence and other benefits among exposed spiny lobsters:
inferences from group size and behaviour. Mar Freshw
Res 52:1113–1124
Hunt JH (2000) Status of the fishery for Panulirus argus in
Florida. In: Phillips BF, Kittaka J (eds) Spiny lobsters:
fisheries and culture. Blackwell Science, Oxford,
p 189–199
Hunt JH, Lyons WG, Kennedy FS Jr (1986) Effects of exposure
and confinement on spiny lobsters, Panulirus argus, used
as attractants in the Florida trap fishery. Fish Bull (Seattle)
84:69–76
Juanes F, Smith LD (1995) The ecological consequences of
limb damage and loss in decapod crustaceans: a review
and prospectus. J Exp Mar Biol Ecol 193:197–223
Kaplan EL, Meier P (1958) Nonparametric estimation from
incomplete observations. J Am Stat Assoc 53:457–481
Kilgo JC, Labisky RF, Fritzen DE (1998) Influences of hunting
on the behavior of white-tailed deer: implications for
conservation of the Florida panther. Conserv Ecol 12:
1359–1364
Lima SL (1998) Stress and decision making under the risk of
predation: recent developments from behavioral repro-
ductive and ecological perspectives. Adv Stud Behav 27:
215–290
Lipcius RN, Eggleston DB (2000) Introduction: ecology and
fishery biology of spiny lobsters. In: Phillips BF, Cobb JS,
Kittaka J (eds) Spiny lobster management. Blackwell
Science, Oxford
Lipcius RN, Eggleston DB, Miller DL, Luhrs TC (1998) The
habitat–survival function for Caribbean spiny lobster: an
inverted size effect and non-linearity in mixed algal and
seagrass habitats. Mar Freshw Res 49:807–816
Loftus AJ, Taylor WW, Keller M (1988) An evaluation of lake
trout (Salvelnius namaycacush) hooking mortality in the
upper great lakes. Can J Fish Aquat Sci 45:1473–1479
Madsen J (1994) Impacts of disturbance on migratory water-
fowl. Ibis 137:567–574
Matthews TR (2001) Trap-induced mortality of the spiny
lobster, Panulirus argus, in Florida, USA. Mar Freshw Res
52:1509–1516
Melville-Smith R, Anderton SM, Caputi N (2001) Predicting
the recreational western rock lobster (Panulirus cygnus) in
Western Australia. Mar Freshw Res 52:1527–1534
Peterson CH, Black R (1994) An experimentalist’s challenge:
when artefacts of intervention interact with treatments.
Mar Ecol Prog Ser 111:289–297
Phillips BF, Cobb JS, Kittaka J (1994) Spiny lobster manage-
ment. Blackwell Scientific, Cambridge, MA
Pollock KH, Winterstein SR, Bunck CM, Curtis PD (1989)
Survival analysis in telemetry studies: the staggered entry
design. J Wildl Manag 53:7–15
Ratchford SG (1999) The influence of chemical Communica-
tion on shelter selection, shelter sharing, and aggregation
among spiny lobsters, Panulirus argus. PhD dissertation
North Carolina State University, Raleigh, NC
Ratchford SG, Eggleston DB (1998) Size- and scale-depen-
dent chemical attraction contribute to an ontogenetic shift
in sociality. Anim Behav 56:1027–1034
Ratchford SG, Eggleston DB (2000) Temporal shift in the
presence of a chemical cue contributes to a diel shift in
sociality. Anim Behav 59:793–799
Ryer CH (2002) Trawl stress and escapee vulnerability to
predation in juvenile walleye pollock: Is there an un-
observed bycatch of behaviorally impaired escapees?
Mar Ecol Prog Ser 232:269–279
Ryer CH, Ottmar ML, Sturm EA (2004) Behavioral impair-
ment after escape from trawl codends may not be limited
to fragile fish species. Fish Res 66:261–269
Sharp WC, Bertelsen RD, Leeworthy VR (2005) Long-term
trends in the recreational lobster fishery of Florida, United
States: landings, effort, and implications for management.
NZ J Mar Freshw Res 39:733–747
Sinclair ARE, Arcese P (1995) Population consequences of
predation-sensitive foraging: the Serengeti wildebeest.
Ecology 76:882–891
Thrush SF, Hewitt JE, Funnell GA, Cummings VJ, Ellis J,
Schultz D, Talley D, Norkko A (2001) Fishing disturbance
and marine biodiversity: the role of habitat structure in
simple soft-sediment systems. Mar Ecol Prog Ser 223:
277–286
243
Mar Ecol Prog Ser 303: 235–244, 2005
Veale LO, Hill AS, Brand AR (2000) An in situ study of preda-
tor aggregations on scallop (Pecten maximus (L.)) dredge
discards using a static time-lapse camera system. J Exp
Mar Biol Ecol 255:111–229
Warner K (1978) Hooking mortality of lake dwelling land-
locked Atlantic salmon, Salmo salar. Trans Am Fish Soc
107:518–522
Yarmoloy C, Bayer M, Geist V (1988) Behavior responses and
reproduction of mule deer, Odocoileus hemionus, does
following experimental harassment with an all-terrain
vehicle. Can Field-Nat 102:425–429
Zar JH (1984) Biostatistical analysis, 2nd edn. Prentice-Hall,
Englewood Cliffs, NJ
Zimmer RK, Commins JE, Kenneth AB (1999) Regulatory
effects of environmental chemical signals on search
behavior and foraging success. Ecology 80:1432–1446
244
Editorial responsibility: Otto Kinne (Editor-in-Chief),
Oldendorf/Luhe, Germany
Submitted: June 22, 2004; Accepted: May 24, 2005
Proofs received from author(s): October 18, 2005
