CLIMATE CHANGE, DISEASE AND THE
DYNAMICS OF A KELP-BED ECOSYSTEM IN
ROBERT E. SCHEIBLING, COLETTE J. FEEHAN, JEAN-SÉBASTIEN LAUZON-GUAY
Jose María Fernández-Palacios, Lea de Nascimento, José Carlos Hernández,
Sabrina Clemente, Albano González & Juan P. Díaz-González (eds.)
Servicio de Publicaciones, Universidad de La Laguna – 2013
CLIMATE CHANGE PERSPECTIVES FROM THE ATLANTIC:
PAS T, PRES ENT AND FUTURE
Congresuales 33.indb 361 30/01/2014 9:00:02
Congresuales 33.indb 362 30/01/2014 9:00:02
Along the Atlantic coast of Nova Scotia, cyclical alternations between kelp beds and sea
urchin barrens are driven by sporadic introductions of a pathogenic amoeba Paramoeba
invadens that causes mass mortality of green sea urchins Strongylocentrotus droebachiensis,
thereby enabling kelps to recolonize the rocky seabed. Sea urchins gradually repopulate
disease-affected areas and destructively graze the emergent kelp beds, returning the system
to the barrens state. Using a logistic regression model, we showed that the probability of
mass mortality of sea urchins is related to the intensity and proximity of tropical storms
and hurricanes, hypothesized to deliver the pathogenic agent, and post-storm ocean
temperature above a threshold (12°C) for a disease outbreak. We also showed that the
likelihood of deadly storms for sea urchins increased over a 30-year period (1980–2009),
a trend expected to accelerate with future ocean warming and increased hurricane in-
tensity. We currently are testing the validity of this model in a ﬁeld experiment designed
to compare predicted and observed disease outbreaks and extend our trend analysis over
5 more years. The experiment also provides insight into spatial and temporal patterns
of disease outbreaks in relation to coastal warming and a changing hydrodynamic and
biotic regime. If observed trends continue, the resilience of the kelp bed state will be
enhanced, and the sea urchin ﬁshery doomed, in areas potentially affected by disease.
KEYWORDS: kelp bed, sea urchin, disease, mass mortality, climate change, ecosystem dy-
namics, predictive modeling.
The notion that communities of interacting species can exist in alternative
organizational states, dominated by different groups of organisms at different
times, is well entrenched in ecology and currently is an area of active research and
synthesis (May, 1977; Beisner et al., 2003; Petraitis and Hoffman, 2010). Numer-
ous theoretical studies indicate the potential for alternative states in community
models, and alternative states have been documented in various terrestrial (for-
ests, grassland), freshwater (wetlands, lakes), and marine (rocky intertidal and
subtidal communities, soft-bottom assemblages, coral reefs) systems (Scheffer et
al., 2001; Folke et al., 2004; Knowlton, 2004). Understanding mechanisms that
determine the stability of a given community state or drive transitions (or phase
shifts) between states is crucial to assessing the consequences of anthropogenic
Congresuales 33.indb 363 30/01/2014 9:00:02
364 SCHEIBLING ET AL.
stressors and climate change on the structure and dynamics of ecosystems, and
the services that they provide to people. Disease has been identiﬁed as an impor-
tant driver of shifts in community state in marine systems (Harvell et al., 1999).
The incidence of disease in various marine organisms has increased in recent
decades, a trend that has been linked to ocean warming (Harvell et al., 2002).
Widespread mass mortality due to disease has been observed in sea grasses, cor-
als, molluscs, crustaceans, echinoderms and marine mammals (Lafferty et al.,
2004; Ward and Lafferty, 2004). Disease outbreaks that affect key producers or
consumers can lead to profound alterations in marine food webs and community
structure (Lessios, 1988; Muelhstein, 1989).
Along the Atlantic coast of Nova Scotia, Canada cyclical alternations
between kelp beds and sea urchin barrens are driven by recurrent outbreaks
of an amoebic disease (paramoebiasis) that causes mass mortality of the green
sea urchin Strongylocentrotus droebachiensis, the dominant herbivore (Scheibling,
1984a; Lauzon-Guay et al., 2009). This release from grazing pressure enables
kelps and other macroalgae to colonize the rocky seabed. Sea urchins gradually
repopulate disease-affected areas through larval recruitment or adult migration
from refuge populations in deeper waters. As sea urchins increase in density, they
form feeding aggregations (or fronts) that destructively graze the emergent kelp
beds, returning the system to the barrens state.
Here, we brieﬂy review the literature on the etiology and epizoology of
paramoebiasis in Strongylocentrotus droebachiensis and the ecological consequences
of this disease for alternative-state dynamics of the rocky subtidal ecosystem off
Nova Scotia. We examine evidence that increases in ocean surface temperature
and storm frequency and intensity may inﬂuence the frequency and severity of
epizootics resulting in sea urchin mass mortality, and present results of ongo-
ing empirical studies to test a statistical model relating mortality events to these
oceanographic/meteorologic factors. We also consider potential effects of local
hydrodynamic features, sea urchin population density and distribution, and the
physiological ecology of the pathogenic agent, on the introduction, spread and
persistence of disease. Finally, we apply trend analysis and simulation models
to project the frequency of disease events along this coast over the next three
decades, and discuss the implications of these projections for alternative-state
dynamics of the shallow subtidal ecosystem.
OUTBREAKS OF SEA URCHIN DISEASE IN NOVA SCOTIA
Recurrent outbreaks of disease causing mass mortalities of shallow (<20 m
depth) populations of Strongylocentrotus droebachiensis were ﬁrst recorded along
the Atlantic coast of Nova Scotia between 1980 and 1983 (Miller and Colodey,
1983; Scheibling, 1984a, 1986; Miller, 1985), and then sporadically over the next
3 decades (Scheibling and Hennigar, 1997; Miller and Nolan, 2000, 2008; Scheib-
Congresuales 33.indb 364 30/01/2014 9:00:02
CLIMATE CHANGE, DISEASE AND THE DYNAMICS OF A KELP-BED ECOSYSTEM 365
ling et al., 2010; Feehan et al., 2012a) (Fig. 1A). Anecdotal evidence suggests mass
mortalities have occurred in previous decades (Miller, 1985). Estimated annual
mortalities of sea urchins in the early 1980s ranged from 80 to 260 Kt fresh weight
in areas of complete die-off (Miller and Colodey, 1983; Miller, 1985b; Scheibling,
1986). Mortalities in the 1990s were of similar magnitude (Miller and Nolan,
2000), while those in the 2000s were more localized, as sea urchin populations
were depleted or eliminated along large tracts of coast (Scheibling et al., 2010;
Feehan et al., 2012a).
The disease in Strongylocentrotus droebachiensis is characterized by progres-
sive deterioration of muscle tissue of the body wall and water-vascular system,
resulting in loss of function of the tube feet, spines and mouthparts (Jones et al.,
1985). In early stages of infection, sea urchins cease feeding and movement, and
are unable to attach to the substratum. Many succumb to predators or are washed
ashore (Miller and Colodey, 1983; Scheibling 1984b; Feehan et al., 2012a); mori-
bund sea urchins and tests accumulate on the seabed and on beaches (Fig. 2A,
B). Spine loss and epidermal necrosis occur as the disease progresses (Scheibling
Figure 1. A) Range of mass mortalities of Strongylocentrotus droebachiensis during
widespread outbreaks of disease in the shallow subtidal zone along the Atlantic coast of
Nova Scotia, in the early 1980s (Source: Miller, 1985; Scheibllng, 1986) and the mid
to late 1990s (Source: Scheibling and Hennigar, 1997; Miller and Nolan, 2000):
black bars, near-complete mortality; grey bars, partial mortality. Also shown
are St. Margarets Bay (SMB) and other locations mentioned in the text.
B) Schematic of transitions between alternative community states (kelp beds and barrens)
following mass mortality of sea urchins and destructive grazing of kelp.
Congresuales 33.indb 365 30/01/2014 9:00:03
366 SCHEIBLING ET AL.
and Stephenson, 1984; Roberts-Regan et al., 1988). At peak water temperatures
(see below), the disease terminates in death within weeks (Scheibling and Ste-
A previously undescribed marine amoeba, Paramoeba invadens, isolated
from tissues of diseased sea urchins was identiﬁed as the causal agent of mass
mortality of S. droebachiensis (Jones, 1985; Jones and Scheibling, 1985; reviewed by
Scheibling, 1988) (Fig. 2C). Recent genetic analysis of amoebae similarly isolated
from infected sea urchins during an epizootic in 2011 conﬁrm the identity of P.
invadens as a unique species, most closely related to Neoparamoeba branchiphila,
a parasite of other sea urchins and salmonid ﬁsh in aquaculture (Feehan et al.,
2013). Amoebae of the genera Paramoeba and Neoparamoeba are associated with
disease in sea urchins, decapod crustaceans, and ﬁsh worldwide. These infections
typically exhibit strong temperature-dependence with threshold dynamics. Para-
moeba invadens is waterborne and can be cultured on marine bacteria, indicating
a free-living existence and facultative parasitism (Jones and Scheibling, 1985).
The inability of P. invadens to survive in culture at temperatures (2 °C) above
the typical winter minimum (~ 0 °C) in coastal waters of Nova Scotia suggests
it is periodically introduced from warmer regions ( Jellett and Scheibling, 1988;
Scheibling and Hennigar, 1997). These epizootics are highly speciﬁc to S. droe-
bachiensis (Scheibling and Stephenson, 1984; Jellett et al., 1988) and sea urchin
density-dependent (Scheibling, 1984a, 1988).
Temperature is a key factor regulating the transmission and progression
of paramoebiasis in S. droebachiensis (Scheibling and Stephenson, 1984; Jellett
and Scheibling, 1988). Epizootics occur during the peak in sea temperature in
late summer or early fall, often in unusually warm years, and the rate and extent
of mortality is directly related to the magnitude and duration of peak tempera-
Figure 2. A) Moribund and dead Strongylocentrotus droebachiensis accumulating on the sea
bed, and B) a Jonah crab (Cancer borealis) scavenging a moribund sea urchin, during a disea-
se outbreak in St. Margarets Bay in 2009 (photographs by R. E. Scheibling). C) Paramoeba
invadens isolated from a moribund sea urchin (photograph by J. Johnson-MacKinnon).
Congresuales 33.indb 366 30/01/2014 9:00:03
CLIMATE CHANGE, DISEASE AND THE DYNAMICS OF A KELP-BED ECOSYSTEM 367
tures (Miller and Colodey, 1983; Scheibling and Stephenson, 1984; Miller, 1985;
Scheibling, 1986; Scheibling and Hennigar, 1997; Brady and Scheibling, 2005;
Feehan et al., 2012a). In the laboratory, the disease progresses exponentially
between 12 and 20°C and is arrested at 10 to 12°C (Scheibling and Stephenson,
1984) (Fig. 3). The time to morbidity and mortality of sea urchins infected at
different temperatures is consistent with the temperature-dependent growth rate
of Paramoeba invadens in monoexnic culture, which is zero at 5°C and greatest at
15–20°C (Jellett and Scheibling, 1988) (Fig. 3). Infected individuals recover within
30 days by lowering temperature to at least 8°C. In nature, disease outbreaks are
terminated by declining temperatures in the late fall and surviving sea urchins
recover over winter (Scheibling and Stephenson, 1984; Scheibling and Hennigar,
1997). Sea urchins in deeper, colder waters below the seasonal thermocline have
a refuge from disease (Brady and Scheibling, 2005; Scheibling et al., 2010; Feehan
et al., 2012a). Nutritional condition does not affect the rate of mortality from
paramoebiasis, and both juveniles and adults of S. droebachiensis are susceptible
(Scheibling and Stephenson, 1984). Extensive mortalities of S. droebachiensis, on
the scale of those recorded in Nova Scotia, have not been observed elsewhere in
the Northwest Atlantic, although there are reports of disease and localized die-offs
of sea urchins in Newfoundland (Hooper, 1980) and in the St. Lawrence Estuary
Figure 3. Median time to 50% morbidity and percent mortality of Strongylocentrotus
droebachiensis exposed to diseased conspeciﬁcs in water-borne transmission experiments
in the laboratory (from Scheibling, 1984a), and speciﬁc growth rate of
Paramoeba invadens in monoxenic culture (from Jellett and Scheibling, 1988).
Congresuales 33.indb 367 30/01/2014 9:00:03
368 SCHEIBLING ET AL.
and Gulf of St. Lawrence (Himmelman et al., 1983; Dumont et al., 2004). These
mortalities were associated with colder waters, and not likely due to paramoebiasis.
In the Gulf of Maine however, sea urchin mortalities in fall 1999 coincided with
an outbreak of disease in lobsters that was caused by Neoparamoeba pemaquidensis
(Mullen et al., 2004; 2005) and Caraguel et al. (2007) isolated N. pemaquidensis
from moribund sea urchins in fall 2002.
SPATIAL AND TEMPORAL VARIATION IN DISEASE OUTBREAKS AND
TRANSITIONS BETWEEN COMMUNITY STATES
Ocean currents and local hydrodynamic features presumably inﬂuence the
introduction and spread of the water-borne pathogenic agent along the coast of
Nova Scotia (Scheibling and Hennigar, 1997). Free-living Paramoeba, morpho-
logically indistinct from laboratory cultures of P. invadens, were detected in the
seawater and sediment samples near the outfall of Dalhousie University’s seawater
facility, which released untreated efﬂuent from ﬂow-through tanks containing
infected sea urchins at the time (Jellett et al., 1989). However, there has been no
attempt to sample amoebae in the water-column or sediments along the Atlantic
coast of Nova Scotia. Although the residual surface current ﬂows southwestward,
outbreaks of disease and mass mortality of sea urchins appear to have spread
both southwest and northeast along this coast in the early 1980s (Scheibling and
Stephenson, 1984; Miller, 1985; Scheibling, 1986) and mid to late 1990s (Scheib-
ling and Hennigar, 1997; Brady and Scheibling, 2005) (Fig. 1A). Detailed records
of the temporal sequence of mortalities are lacking and it is possible that the
disease spread from various foci rather than from a single origin (Scheibling,
1988). River outﬂow may have limited the spread of disease in 1981, resulting in
a discrete boundary to the mass mortality along the southwestern coast of Nova
Scotia (Miller and Colodey, 1983; Scheibling and Stephenson, 1984).
The population density of Strongylocentrotus droebachiensis affects the rate of
propagation and transmission of paramoebiasis and the range of epizootics along
the coast (Scheibling and Stephenson, 1984). In areas and years of partial die-offs
of sea urchins, morbidity and mortality were highest at shallow depths (< 5 m)
where sea urchins were most dense (Scheibling and Stephenson, 1984; Scheibling,
1988; Miller, 1985). Extreme densities of sea urchins, particularly along grazing
fronts, likely accelerate the propagation of the disease (Scheibling and Hennigar,
1997). Prior to the ﬁrst documented outbreak of disease in Halifax Harbour in
1980 (Miller and Colodey, 1983), sea urchin barrens extended along the entire
Atlantic coast of Nova Scotia (Wharton and Mann, 1981), providing an extensive
host population for recurrent epizootics, which spread along ~ 600 km (linear
distance) of coast between 1981 and 1983 (Miller, 1985; Scheibling, 1986) (Fig. 1A).
The depth range of sea urchin mass mortality is limited by the thermal
threshold (10–12°C) for transmission and propagation of paramoebiasis (Scheib-
Congresuales 33.indb 368 30/01/2014 9:00:03
CLIMATE CHANGE, DISEASE AND THE DYNAMICS OF A KELP-BED ECOSYSTEM 369
ling and Stephenson, 1984). This threshold usually occurs at 20 to 25 m depth and
varies with changes in thermal structure of the water column due to wave-driven
vertical mixing (upwelling and downwelling) during summer and fall (Scheibling,
1984a; Brady and Scheibling, 2005). The depth range of mass mortality also is
limited by the extent of rocky habitat in areas where the substratum grades to
sand at depths above the thermal threshold, such as in large embayments (Filbee-
Dexter and Scheibling, 2012). Sea urchins surviving below the threshold depth
form the source population for recolonization of areas affected by disease. Where
sea urchins occur immediately below the zone of mortality, random foraging
movements extend the population to shallower depths, as kelps and other mac-
roalgae begin to recolonize the former barrens (Lauzon-Guay et al., 2008, 2009).
Coastal bathymetry and substratum type will determine the resilience of sea
urchin barrens after a mass mortality event, and hence the potential for the kelp-
bed state to establish and persist. On steeply sloping shores, where the offshore
(along-bottom) extent of urchin mortality (which deﬁnes the potential transition
zone for alternative community states) may be only 10s of meters, foraging sea
urchins can repopulate the barrens habitat within 1 to 2 years, preempting the
establishment of kelp beds and maintaining a barrens state (Brady and Scheibling,
2005). Where the seabed slopes more gradually to the threshold depth, the zone
of mass mortality may extend 100s of meters to more than a kilometer offshore
(Moore and Miller, 1983). A prolonged release from sea urchin grazing enables
kelps and other macroalgae to colonize these extensive barrens via propagules
dispersing from shallow fringe populations, in a wave-mediated refuge from sea
urchin grazing, leading to the establishment of luxuriant algal beds within 2 to
3 years (Miller, 1985; Scheibling, 1986; Johnson and Mann, 1988). Sea urchins
grazing along the offshore extent of these developing beds form aggregations
(or fronts) that gradually advance onshore to reestablish the barrens state. Deep-
dwelling sea urchins on rock or sedimentary bottoms also contribute to a larval
pool that provides recruits to the expanding barrens (Brady and Scheibling, 2005).
Recruitment to established kelp beds also may enable sea urchins to increase in
abundance within beds to the point at which they form destructive grazing ag-
gregations (Hart and Scheibling, 1988), although direct evidence for this is lack-
ing (Lauzon-Guay and Scheibling, 2010; Feehan et al., 2012b). In areas where sea
urchins are eliminated throughout the rocky subtidal zone (i.e., where the lower
boundary is above the threshold depth for the spread of paramoebiasis), kelp beds
reestablish and can persist for more than a decade before migrating sea urchins
in offshore sandy habitats encounter the beds and form grazing fronts along their
deep edge (Scheibling et al., 1999; Filbee-Dexter and Scheibling, 2012).
Disease drives the alternative-state system in the rocky subtidal zone off Nova
Scotia (Fig. 1B). Epizootics terminate the barrens phase in the transitional zone
and subsequent alternation between the kelp-bed and barrens state is depend-
ent upon the distance to physically-mediated spatial refuges for the dominant
species of each state: wave-swept shallows for kelp, colder deeper waters for sea
Congresuales 33.indb 369 30/01/2014 9:00:03
370 SCHEIBLING ET AL.
urchins. As kelps colonize offshore areas after release from grazing pressure, sea
urchin fronts gradually migrate onshore to consume them. The introduction of
P. invadens to the shallow subtidal zone, and hence the potential for an epizootic
that causes a shift to the kelp-bed state, may be largely determined by stochastic
processes associated with broad-scale meteorological and hydrodynamic condi-
tions that are poorly understood (Scheibling and Hennigar, 1997). On the other
hand, the reverse shift to the barrens state occurs through a series of determinis-
tic processes (e.g., algal succession, sea urchin foraging behaviour, kelp and sea
urchin population dynamics) that are well documented (Johnson and Mann,
1988). Mathematical models that describe the alternation between states follow-
ing a disease event demonstrate a high degree of concordance between predicted
and observed dynamics (Lauzon-Guay et al., 2008, 2009).
The timing of introduction of P. invadens to the coast of Nova Scotia, in rela-
tion to the prevailing ocean temperature and state of the ecosystem, determines
the likelihood of an epizootic and the extent of sea urchin mass mortality. If the
pathogen were introduced when the system is in the kelp-bed state, sea urchin
density likely would be too low to trigger an epizootic (even when temperatures are
conducive to the spread of disease) and the kelp-bed state would be maintained.
On the other hand, when the pathogen is introduced in the barrens state and
ocean temperatures are around the seasonal maximum (16–20 °C), the disease
propagates rapidly leading to widespread mass mortality and a shift to the kelp-
bed state (Scheibling and Hennigar, 1997; Brady and Scheibling, 2005). However,
when the introduction occurs later in the year, as falling temperatures approach
the threshold level, the disease is arrested before extensive mortality can occur
and the barrens state is maintained (Scheibling and Stephenson, 1984). Recurrent
introductions of P. invadens, depending on the frequency, can result in reversals
in the transition between states. For example, an epizootic that eliminates sea
urchins and enables kelps to expand their offshore range can be followed by a
gradual return to the barren state as a sea urchin front forms along the deep
margin of the bed and advances shoreward; a subsequent epizootic can arrest this
transition and once again return the system to the kelp-bed state (Scheibling et
al., 1999). Thus, frequent introductions of P. invadens stabilize the kelp bed state,
whereas rare introductions enable the persistence of barrens.
KILLER STORMS: HURRICANES AND OUTBREAKS OF PARAMOEBIASIS
Using a logistic regression model (see below), Scheibling and Lauzon-
Guay (2010) showed that the probability of mass mortality of sea urchins can be
predicted by the intensity and proximity of tropical storms and hurricanes, hy-
pothesized to deliver P. invadens to the coast, and post-storm ocean temperature
above a threshold (12°C) for a disease outbreak. They showed that the likelihood
of deadly storms for sea urchins increased over a 30-year period (1980–2009), a
Congresuales 33.indb 370 30/01/2014 9:00:03
CLIMATE CHANGE, DISEASE AND THE DYNAMICS OF A KELP-BED ECOSYSTEM 371
trend expected to accelerate in the near future with ocean warming and increased
hurricane intensity (Bender et al., 2010). These storm events could transport P.
invadens to coastal areas by horizontal advection from distant source populations
or vertically mix amoebae residing locally in deep basins (Scheibling and Hen-
nigar, 1997; Scheibling et al., 2010).
We currently are testing the validity of this model in a ﬁeld experiment
designed to compare predicted and observed disease outbreaks over a 4- to 6-year
period. Beginning in 2010, we have transplanted S. droebachiensis in July or early
August (before the start of the hurricane season off Nova Scotia) from a head-
land (Splitnose Point) near Halifax, Nova Scotia, where sea urchin populations
have persisted in the shallow subtidal zone for at least a decade, into cages in kelp
beds at 8 m depth in and around St. Margarets Bay, where localized outbreaks
of disease were observed following hurricanes in recent years (Scheibling et al.,
2010; Feehan et al., 2012a). Throughout the hurricane season (August to Novem-
ber), we measured sea urchin morbidity and mortality in these cages at weekly to
biweekly intervals, while monitoring sea temperature and hurricane activity. For
site descriptions and details of the experimental design and sampling procedures
in 2010, see Feehan et al. (2012a). In 2011–2012, 3 of the 6 sites within the bay
in 2010 were retained and 2 new sites were added, one at the headland on each
side of the bay; the number of replicate cages at each site was 2 in 2010 and 4 in
2011 and 2012. At one site, we also placed sea urchins in cages at 18 m depth,
below the seasonal thermocline where temperatures generally were below the
threshold for paramoebiasis.
In September of 2010 and 2011, we observed morbidity of caged sea ur-
chins at 8 m depth following the passage of a hurricane with a high probability
of association with an urchin mass mortality (Pm), according to the model: Hur-
ricane Earl, Pm = 43%, and Hurricane Katia, Pm = 70% (Fig. 4). Moribund sea
urchins displayed overt signs of paramoebiasis and transmitted the disease via a
water-borne route to healthy conspeciﬁcs in the laboratory (Feehan et al., 2012a,
Feehan et al., 2013). Paramoeba invadens was isolated from tissues of ﬁeld-collected
moribund sea urchins and cross-infected individuals in controlled laboratory
experiments, satisfying Koch’s postulates. In each year, the Pm of the hurricane
was not signiﬁcantly different (t-test) from the mean Pm of “candidate storms”
in 1980–2009 (57 ± 28% SD, n = 12), based on the model (Hurricane Earl: p =
0.11; Hurricane Katia: p = 0.14). For each storm, we predicted the time to 50%
morbidity of sea urchins with paramoebiasis (t50, d) using a relationship based
on water-borne disease transmission experiments in the laboratory:
t50 = 23492T -2.7476
where T (°C) is water temperature during the period of infection (Scheibling et
al., 2010). We estimated t50 following a hurricane from temperature records at
8 m depth in St. Margarets Bay during the ﬁeld experiment. In 2010 and 2011,
Congresuales 33.indb 371 30/01/2014 9:00:03
372 SCHEIBLING ET AL.
the observed t50 of sea urchins in cages at 8 m, pooled across all sites, closely ap-
proximated the predicted t50 based on water temperature following the hurricane:
the difference between observed and predicted values ranged from 2 to 3 days.
Although a strong tropical cyclone did not pass close to the coast of Nova
Scotia during the 2012 hurricane season, a mass mortality of S. droebachiensis due
to paramoebiasis occurred in August/September. Following a report of large
numbers of dead and dying sea urchins washing onshore in Halifax Harbour on
12 August, we conducted diving and towed-video surveys in this area over the next
2 weeks. Paramoeba invadens was isolated from tissues of diseased sea urchins, and
we estimated that > 65% of the population on cobble and boulder barrens up to
700 m offshore, and across a depth range from 2 to 10 m, was moribund or dead
Figure 4. Signiﬁcant wave height (SWH) at Halifax Harbour, and temperature and the
proportion of dead or moribund Strongylocentrotus droebachiensis in experimental cages
at 8-m depth at 3 (2011–2012) or 6 (2010) sites within St. Margarets Bay (red lines),
at 2 sites (2011–2012) at the headlands on either side of the bay (black lines), at 1 site
(Splitnose Point) where sea urchins were collected for all cages (procedural control;
green lines), and at 18-m depth at 1 site within the bay (blue lines), before and after a
hurricane or tropical storm that preceded an outbreak of paramoebiasis in sea urchins.
Dotted horizontal line is the lower temperature threshold (10°C) at which sea urchins
do not exhibit signs of morbidity due to paramoebiasis. Dashed horizontal line is the
time to 50 % morbidity (t50) of sea urchins. Arrows indicate the date at which each storm
passed closest to the study area. ND: No data.
Congresuales 33.indb 372 30/01/2014 9:00:04
CLIMATE CHANGE, DISEASE AND THE DYNAMICS OF A KELP-BED ECOSYSTEM 373
at this time. The mass mortality in Halifax Harbour was preceded by a sharp
peak in signiﬁcant wave height (2.7 m) on 1 August, followed by a 4-week period
of unusually warm sea surface temperatures (mean ± SD: 20 ± 1 °C). The high
incidence of morbidity/mortality by mid-August is consistent with an introduction
of P. invadens around the time of the wave event, given a t50 of 11.5 days based on
an estimated mean bottom temperature of 16 °C during the intervening 2-week
period (estimate based on mean temperature at 2 – 10 m depth in Halifax Har-
bour). Interestingly, a similar peak in signiﬁcant wave height (2.9 m) occurred on
27 June, followed by a 10-day period of similarly warm sea surface temperatures
(19 ± 1 °C), but there was no report of unusual mortality of sea urchins in July.
These wave events were due to local weather disturbances and not associated
with passing tropical storms. Such wave events during summer are infrequent
along the coast of Nova Scotia. For example, a signiﬁcant wave height > 2.6 m in
August was only observed in 5 out of 34 years (1976–2009) in the Halifax region.
On 24 August 2012, 60 and 100 % of caged sea urchins were dead or dying
at our two sites on the headlands on either side of St. Margarets Bay, compared
to 20 % at one site on the eastern shore of the bay (Luke Island) and none at
the remaining two sites along the western shore (Fig. 4). Mortality due to dis-
ease gradually increased at Luke Island, exceeding 50 % by 26 September and
reaching 90 % by 19 October. No further mortality was observed over the next
18 days, while bottom temperatures remained around the 10 – 12 °C threshold
for paramoebiasis. Sea urchins at the other two sites remained asymptomatic
throughout the experimental period with minimal mortality by 6 November. At
Splitnose Point, the control site and source of experimental animals, morbidity
and mortality ﬁrst was observed in cages (and on the surrounding seabed at 8
m) on 24 September (~ 40 days after the mass mortality in Halifax Harbour),
accounting for 55 % of the caged sea urchins (Fig. 4). At this time scattered
pockets of diseased and dying sea urchins and their tests also were evident at
shallower depths (3 – 4 m). Disease was not present at this site or at two nearby
sites (Duncan’s Cove and Bear Cove) 4 km N and 7 km NNW (linear distance)
from Splitnose Point, during towed-video surveys on 27 August. By 23 October,
mortality of caged urchins at Splitnose Point had reached 90 %.
In each year, rates of morbidity and mortality were much lower for sea urchins
in cages at 18 m at a site (The Lodge) on the western shore of St. Margarets Bay, in
accordance with lower water temperatures at this depth (rarely above 12 °C) compared
to 8 m (Fig. 4). However, when surviving asymptomatic sea urchins were collected at
the end of the experiment in 2010 and maintained at 16 °C in the laboratory, they
succumbed to paramoebiasis and transferred the disease to healthy conspeciﬁcs
through a waterborne route (Feehan et al., 2012a). This suggests that sea urchins
in deeper, colder waters can harbor the pathogen for some period after a disease
outbreak in shallower water. A similar effect of depth on the propagation of disease
was observed in Halifax Harbour in August 2012. The rate of morbidity and mortal-
ity decreased linearly with depth from 2 to 10 m, where temperature dropped from
Congresuales 33.indb 373 30/01/2014 9:00:04
374 SCHEIBLING ET AL.
20 to 13 °C respectively; moribund sea urchins or tests were not observed below 10
m, where temperature continued to decrease to 11 °C at 18 m depth, 1 km offshore.
Thus far, the results from our ﬁeld experiment provide equivocal support for
the “killer storm” hypothesis and the efﬁcacy of the statistical model of Scheibling
and Lauzon-Guay (2010) to predict disease outbreaks in sea urchins based on
hurricane or tropical storm activity and post-storm seawater temperature. The
occurrence of a mass mortality in 2012, without a prior storm that passed near to
the coast of Nova Scotia (only tropical storm Chris with Pm = 0.1 % on 19 June was
within the spatial range considered in the model prior to the disease outbreak),
indicates that other advective mechanisms may account for the introduction of
P. invadens to shallow coastal areas. This is not inconsistent with previous records
off Nova Scotia. A widespread mass mortality of S. droebachiensis occurred in 1983,
also a year without signiﬁcant tropical storm or hurricane activity (Scheibling
and Hennigar, 1997). In fact, this was the only year (out of 13) in which a mass
mortality event was not preceded by a tropical cyclone in the 30-year record
used in the statistical model (Scheibling and Lauzon-Guay, 2010). Scheibling
and Lauzon-Guay (2010) suggest that experiments that incidentally or deliber-
ately released laboratory-cultured P. invadens along the coast may have caused
the mortality event in 1983. This is unlikely to have been the cause of the 2012
event because infected sea urchins were maintained in ﬂow-through aquaria in
a quarantine facility at Dalhousie University, in which all efﬂuent seawater was
treated with strong bleach before being released into the ocean.
Our experiment and observations of sea urchin mortality in surrounding
areas provide insight into spatial and temporal patterns of disease outbreaks along
the coast of Nova Scotia. In 2010, the time to morbidity or mortality of trans-
planted sea urchins varied widely among sites and 20 – 40 % survived at 3 of the
6 sites within the bay, once water temperatures fell below the 12 °C threshold in
November (Fig. 4). Importantly, there was no evidence of paramoebiasis prior to
Hurricane Earl, although temperatures at 8 m depth averaged 19 °C for a 15-day
period in August. The predicted t50 at this temperature is 7 days, indicating sea
urchins likely were not infected before the storm. We did not observe diseased
sea urchins or unusual mortality among our source population at Splitnose Point
or along headlands adjacent to Halifax Harbour (~ 40 km linear distance from
St. Margarets Bay) in 2010, suggesting the disease outbreak was a localized event.
Paramoeba invadens may have been locally introduced, perhaps by vertical mixing
within the bay, or more broadly distributed along the coast but disease may not
have propagated outside of the bay. Large semi-protected embayments such as St.
Margarets Bay may serve as “incubation” sites for P. invadens, with warmer surface
temperatures and longer water residence times (Heath, 1973), compared to the
outer coast. In 2011, variation among sites in the rate of morbidity was much
lower following Hurricane Katia and mortality was complete at all experimental
sites within the bay and at adjacent headlands (Fig. 4). There also was a partial
die-off of sea urchins at Splitnose Point in late October and sea urchin ﬁshers
Congresuales 33.indb 374 30/01/2014 9:00:04
CLIMATE CHANGE, DISEASE AND THE DYNAMICS OF A KELP-BED ECOSYSTEM 375
observed mass mortalities around this time at sites along the southwestern shore
(Lockport and Shelburne) and southern tip of Nova Scotia (Barrington Bay),
spanning 40 km (linear distance) of coast (C. Hopkins, sea urchin harvester,
Shelburne, Nova Scotia, personal communication).
In 2012, the initial observation of paramoebiasis in Halifax Harbour in
early August, followed by subsequent outbreaks of disease at our experimental
sites around the mouth of St. Margarets Bay, suggests a southwesterly spread of
disease, consistent with the residual current direction. Alternatively, there may have
been separate introductions in these two areas separated by several weeks. In the
absence of a major storm event, the introduction of P. invadens may have been as-
sociated with warm water intrusions from across the shelf. Satellite imagery shows
a broad band of warm (18 – 19 °C) surface water impinging on the central part
of the mainland coast near Halifax in late July and early August (Fig. 5), which
preceded the disease outbreak in Halifax Harbour around mid-August. A second
Figure 5. Sea surface temperature (SST) along the Atlantic coast of Nova Scotia (see in-
set map to locate measurement grid) averaged over 8-day intervals from 1 June to 27 Au-
gust 2012. These images were generated from global Level-3 standard mapped images
(SMIs) of 4 x 4 km spatial resolution MODIS/Aqua 11 µm daytime SST. The SMIs were
obtained from the NASA Goddard Space Flight Center Ocean Color Web (Feldman and
Congresuales 33.indb 375 30/01/2014 9:00:04
376 SCHEIBLING ET AL.
intrusion of even warmer water (20 – 21 °C) moved into this area and along shore
to the northeast in mid to late August, which may have contributed to a secondary
outbreak of disease at the mouth of St. Margarets Bay. Entrainment of large volumes
of slope and shelf water by mesoscale cyclonic eddies (warm-core rings), emanat-
ing from the Gulf Stream, results in mixing and advective transport of planktonic
organisms, such as ﬁsh larvae (Flierl and Wroeblewski, 1985; Myers and Drinkwater,
1989), across the shelf. Scheibling and Hennigar (1997) suggest that these warm
water intrusions also may contribute to the transport of P. invadens to coastal areas.
The widespread mass mortality of sea urchins in 1983 also was associated
with unusually warm sea surface temperatures along the coast in a year without
a major tropical cyclone (Scheibling and Lauzon-Guay, 2010). These sporadic
occurrences indicate that large storm events, such as hurricanes, do not neces-
sarily precede mass mortalities of sea urchins and therefore are not the only
mechanism of introduction of P. invadens to coastal areas. This does not preclude
the possibility of a synergistic interaction between severe storms and warm-water
intrusions that may accelerate advection and mixing of surface waters (Crad-
dock et al., 1992) and the transport of the planktonic pathogen. In 2012, we ﬁrst
observed paramoebiasis at Splitnose Point 14 days after tropical storm Leslie
(Pm = 0.325) had passed close to the coast of Nova Scotia on 10 September. The
temperature at 8 m depth at this site over the 2-week period following Leslie was
~ 13 °C. This gives a t50 of 20 days, which exceeds the observed time to > 50 %
morbidity by almost a week. However, a peak in signiﬁcant wave height, possibly
associated with Leslie, occurred on 5 September, 21 days before we observed >
50 % morbidity, which is consistent with the predicted t50 . Mixing due to this
relatively weak tropical storm may have locally introduced the pathogen to this
area after the outbreaks of disease in Halifax and St. Margarets Bay. It did not
appear sufﬁcient, however, to spread P. invadens throughout St. Margarets Bay,
as evidenced by the survival of asymptomatic sea urchins at our two sites along
the western shore during the experimental period in 2012.
The occurrence of subtropical and tropical ﬁsh species in coastal waters
of Nova Scotia in late summer and fall is further evidence of cross-shelf advec-
tion that could transport a planktonic pathogen. In 1995, there were reports of
tropical ﬁsh and sea turtles in the shallow nearshore waters immediately preced-
ing a sea urchin epizootic (Scheibling and Hennigar, 1997). We observed grey
triggerﬁsh (Balistes capriscus) actively swimming near the mouth of St. Margarets
Bay on 22 July and 22 August 2012, during the warm-water events that coincided
with outbreaks of paramoebiasis in the region. The local media also reported
triggerﬁsh and other subtropical and tropical species, including ﬂying ﬁsh and
seahorses, in the region around this time (http://www.cbc.ca/player/News/
Canada/NS/ID/2270769998/). These observations of exotic species in coastal
waters provide circumstantial evidence for a mechanistic link between outbreaks
of sea urchin disease and meteorologic and oceanographic features (Scheibling
and Hennigar, 1997).
Congresuales 33.indb 376 30/01/2014 9:00:04
CLIMATE CHANGE, DISEASE AND THE DYNAMICS OF A KELP-BED ECOSYSTEM 377
Interestingly, the disease outbreaks that we observed in transplanted sea
urchins between 2010 and 2012, and during another experiment in St. Margarets
Bay in 2009 (Feehan et al., 2012b), occurred among artiﬁcially generated aggrega-
tions of sea urchins in cages or experimental plots, at sites where naturally occur-
ring adults were rare. This may reﬂect previous situations when P. invadens was
introduced to coastal areas in a kelp-bed state, but the disease passed unnoticed
because of a low density of adult sea urchins in kelp beds.
CLIMATE CHANGE AND ECOSYSTEM EFFECTS OF DISEASE IN
SEA URCHIN POPULATIONS
Scheibling and Lauzon-Guay (2010) found that the strongest storms pass-
ing along the coast of Nova Scotia tended to increase in intensity and proximity
to the coast over a 30-year period (1980–2009). Sea temperatures during the
hurricane season also increased during this period. These trends were associ-
ated with an increase in the frequency of hurricanes and tropical storms with
the greatest potential for sea urchin mortality. If these trends continue due to a
changing ocean climate, this could limit the resilience of sea urchin populations
and maintain the ecosystem in the kelp-bed state.
To explore the possible effects of climate change, we used quantile regres-
sion to conduct a trend analysis of storm characteristics (wind speed, distance
from the coast) and ocean temperature that affect the probability of sea urchin
mass mortality (Pm), using procedures described by Scheibling and Lauzon-Guay
(2010). We ﬁrst performed a separate regression for every 5th quant ile from the 5 th
to the 95th quantile for each of these factors over a 30-year period (1980–2009). To
predict the distribution of each factor over a 100-year period, we then extended
the regression equation for each quantile backwards to 1940 and forwards to
2040. For each year from 1940 to 2040, we randomly picked a number of storms
(n) from the empirical distribution of the number of storms observed annually
between 1980 and 2009. For each of the n storms in a given year, we obtained
a maximum sustained wind speed (W, km h-1) when the hurricane was closest
to the coast (within our study grid between 35°N and the coast of Nova Scotia,
and between 55°W and 70°W) by randomly selecting one of the 19 wind-speed
quantiles for that year, and followed the same procedure for distance of a storm
from the coast (D, 100s km) and water temperature (T), a dummy variable for the
temperature threshold based on the mean temperature (Tm) at 0–10 m depth in
the 2-week period following each storm (T = 1 if Tm > 12.2 °C, T = 0 if Tm < 12.2 °C).
Based on these environmental factors, we calculated Pm for ea ch o f the n storms in
a year using the logistic regression model of Scheibling and Lauzon-Guay (2010):
Pm = 1/(1 + e-z)
z = -14.352 + 0.082W - 0.069D2 + 4.966T
Congresuales 33.indb 377 30/01/2014 9:00:04
378 SCHEIBLING ET AL.
We then calculated the cumulative probability of having a mass mortality
(cPm) in that year based on the Pm for each storm:
cPm = 1- ((1- Pm1 ) • (1- Pm2 ) •...• (1- Pmn))
We repeated this procedure 1000 times and calculated the mean cPm and
95% CI of these iterations for each year. For comparison with predicted values
of cPm, we calculated the observed cPm for each year from 1980 to 2011 using Pm
values based on the actual storms occurring in those years.
Our simulations yield a sigmoidal trajectory of mean cPm over a 100-year
period from 1940 to 2040 (Fig. 6). The mean cumulative probability of a mass
mortality event was low (< 30%) prior to the 1980s, when the ﬁrst outbreaks of
sea urchin disease were documented (Miller and Colodey, 1983), and steadily
Figure 6. Results of simulations to predict the cumulative probability of mass mortality (cP
of Strongylocentrotus droebachiensis due to paramoebiasis associated with passing
hurricanes or tropical storms along the Atlantic coast of Nova Scotia from 1940 to 2040,
based on trends in storm characteristics (wind speed, distance from the coast) and post-
storm water temperature measured from 1980 to 2009. Data are mean cP
and 95% CI for
each year over the 100-year period. Circles are the predicted cP
each year from 1980 to
2011 based on actual storm and temperature data for these years (Scheibling and Lauzon-
Guay, 2010). Solid circles are years when mass mortalities of sea urchins were recorded in
the shallow subtidal zone to 25 m depth; open circles are years when mass
mortalities were not reported (Scheibling and Lauzon-Guay, 2010; this study).
Congresuales 33.indb 378 30/01/2014 9:00:04
CLIMATE CHANGE, DISEASE AND THE DYNAMICS OF A KELP-BED ECOSYSTEM 379
increased over the next 3 decades to 90% by 2012. The projected cPm reaches
a plateau at ~ 98% by 2030. The observed cPm each year from 1980 to 2011 fell
within the 95% CI of the predicted mean cPm (Fig. 6). During this period, there is
a signiﬁcant correlation (r = 0.46, P = 0.015, n = 31) between the predicted mean
cPm ea ch y ear b ased on ou r si mula tion s and the cPm predicted using the storm and
temperature data given in Scheibling and Lauzon-Guay (2010) (Fig. 6).
Using the quantile regression procedures described above to estimate
cPm, we predicted hurricane and tropical storm characteristics and post-storm
temperatures for 10 random storms per year between 1940 and 2040. Our trend
analysis over the 100-year period shows: 1) a clear decline in the upper limit of
the distance of a storm from the coast, 2) an increase in maximum sustained wind
speed when the storm was closest to the coast, and 3) an increase in post-storm
temperatures above the threshold for an epizootic (Fig. 7). The results of these
simulations are consistent with the increase in cPm over this period, based on an
independent simulation model.
Figure 7. Results of simulations based on trend analysis of hurricane or tropical storm
characteristics and post-storm water temperature measured from 1980 to 2011
(Scheibling and Lauzon-Guay, 2010; this study), showing: 1) a decline in the upper
limit of the distance of a storm from the coast, 2) an increase in maximum sustained
wind speed when the storm was closest to the coast (relative size of bubbles), and 3) an
increase in post-storm temperatures above the threshold (12°C) for an outbreak of para-
moebiasis in Strongylocentrotus droebachiensis (blue bubbles, below threshold; red bubbles,
above threshold), over the 100-year period from 1940 to 2040.
Congresuales 33.indb 379 30/01/2014 9:00:04
380 SCHEIBLING ET AL.
A changing ocean climate also may inﬂuence the spatial range of Para-
moeba invadens and the frequency of epizootics as the winter minimum in water
temperature begins to exceed 2 °C, the predicted lower tolerance limit of the
amoeba (Jellett and Scheibling, 1988). On the Atlantic coast of Nova Scotia, water
temperature typically drops below 2 °C in February and March, the two coldest
months. Regression analysis of a 33-year temperature record (0 – 10-m depth)
along the central region of this coast shows a signiﬁcant warming trend in Febru-
ary/March (Fig. 8). During our experiment, winter temperature at 8-m depth in
St. Margarets Bay rarely dropped below the 2 °C tolerance limit for P. invadens
in 2011 (4 days) and 2012 (1 day), indicating that the pathogen may be capable
of overwintering in the shallow subtidal zone as temperature continues to rise.
Given the observed rate of warming over the past three decades (0.039 °C y-1),
the predicted average winter temperature will reach 2.0 °C by 2020 and 2.7 °C by
Figure 8. Average daily water temperature along the Atlantic coast of Nova Scotia during
the annual summer/fall peak (August, September and October; red circles) and the winter
trough (February and March; blue circles) over a 33-year period (1980 – 2012). Data for
1980–2009 are based on records at 0–10-m depth over 70 km (linear distance) of coast
(from Halifax to Lunenburg) and 5 km offshore (for data source, see Scheibling and
Lauzon-Guay, 2010). To extend these records to 2012, we used average daily temperature
based on our temperature records at 8-m depth along the western shore of St. Margarets Bay
for each period (August – October, February – March). Lines represent the linear regression
of average temperature in each year at the peak (T
) or trough (T
of the annual cycle against year (Y) based on the 33-year record
= 0.064Y - 114.4, r
= 0.175, p = 0.015); T
= 0.039Y - 77.2, r
= 0.170, p = 0.017)
Congresuales 33.indb 380 30/01/2014 9:00:04
CLIMATE CHANGE, DISEASE AND THE DYNAMICS OF A KELP-BED ECOSYSTEM 381
2040. The rate of warming for the summer/fall peak in the annual temperature
cycle is even greater (0.064 °C y-1), with predicted average peak temperatures of
14.5 °C by 2020 and 15.8 °C by 2040.
An increase in the frequency of sea urchin epizootics associated with in-
creasing hurricane intensity and increasing ocean surface temperatures off Nova
Scotia is expected to shift the rocky subtidal ecosystem to a kelp-bed state that
would remain stable for the foreseeable future (Scheibling and Lauzon-Guay,
2010). This could beneﬁt lobster and ﬁnﬁsh ﬁsheries that rely on kelp beds as
nursery habitat and a major source of primary production for benthic food webs
(Wharton and Mann, 1981; Steneck et al., 2002). On the other hand, recurrent
mass mortality of Strongylocentrotus droebachiensis has caused the collapse of the
sea urchin ﬁshery in Nova Scotia (Miller and Nolan, 2000, 2008), and there is
little prospect for recovery on a coastal scale. Changes in the shallow subtidal
ecosystem resulting from an increased frequency of sea urchin epizootics also
are likely to have profound effects on the structure and functioning of adjacent
communities that utilize kelp detritus and sea urchin feces as a food resource
(Sauchyn et al., 2011; Krumhansl and Scheibling, 2012a). This connectivity in
the ﬂux of organic matter and energy between kelp beds and offshore benthic
assemblages across a depth gradient remains largely unexplored (Krumhansl
and Scheibling, 2012b).
CONCLUSIONS AND PERSPECTIVES FOR FUTURE RESEARCH
The origin of infective populations of Paramoeba invadens and mechanisms of
dispersal, propagation and survival of amoebae in nature remain unresolved, de-
spite three decades of observation along the Atlantic coast of Nova Scotia (Scheib-
ling and Lauzon-Guay, 2010) and recently renewed research efforts (Feehan et al.,
2012a). Reports by sea urchin harvesters in the 1990s (Miller and Nolan, 2000)
and our recent ﬁeld studies (Scheibling et al., 2010; Feehan et al., 2012a, b) suggest
that outbreaks of paramoebiasis can be patchy along this coast. More extensive
monitoring of disease outbreaks across a range of habitats and spatial scales is
needed to identify potential disease hotspots and increase our understanding
of the introduction and spread of paramoebiasis. For example, sampling tissues
of deep-living sea urchins for presence of P. invadens, could establish whether
amoebae are locally present in deep basins where bottom temperatures are at or
above the lower tolerance limit (2 °C) established in laboratory experiments. The
possibility of a resistant cyst stage also warrants further research in studies aimed
at better understanding the physiological response of P. invadens to environmental
variables such as temperature and salinity, particularly as winter temperatures in
shallow coastal areas rise above the lower tolerance limit. Genetic markers of P.
invadens, isolated from naturally infected sea urchins, could be used to effectively
monitor the amoeba in water and sediment samples before, during and after a
Congresuales 33.indb 381 30/01/2014 9:00:04
382 SCHEIBLING ET AL.
disease outbreak to gain insights into host-pathogen dynamics, factors affecting
the spread of disease (e.g., sea urchin density, temperature, salinity, currents),
and the fate of amoebae as temperatures drop. These genetic tools also would
enable us to search broadly for source populations and explore dispersal mecha-
nisms, such as advection and turbulent mixing by hurricanes, which potentially
introduce P. invadens to the shallow coastal waters. Finally, a multiyear record of
hurricane activity, surface temperature and occurrence of paramoebiasis in Nova
Scotia, based on longitudinal studies and controlled ﬁeld experiments, such the
one presented here, is needed to critically test the Scheibling and Lauzon-Guay
(2010) model. Further exploration of the mechanistic links between hurricanes,
water temperature and spread of paramoebiasis will allow us to more accurately
predict the magnitude and extent of epizootics that can profoundly affect the
ecology of subtidal communities.
The authors thank the Ocean Biology Processing Group (Code 614.2) at the
NASA Goddard Space Flight Center, Greenbelt, MD 20771, for the production
and distribution of the MODIS/Aqua SST data, and Michael Brown for provid-
ing the processed imagery. This research has been continuously supported since
1982 by grants from the Natural Sciences and Engineering Research Council
(NSERC) of Canada to RES.
Congresuales 33.indb 382 30/01/2014 9:00:04
BEISNER, B.E., HAYDON, D.T. an d CUDDINGTON, K. (2003) Alternative stable states in ecology.
Frontiers in Ecology and the Environment, 1, 376–382.
BENDER, M.A., KNUTSON, T.R., TULEYA, R.E., SIRUTIS, J.J., VECCHI, G.A., GARNER, S.T. and
HELD, I.M. (2010) Modeled impact of anthropogenic warming on the frequency
of intense Atlantic hurricanes. Science, 327, 454.
BRADY, S. and SCHEIBLING, R.E. (2005) Repopulation of the shallow subtidal zone by green
sea urchins (Strongylocentrotus droebachiensis) following mass mortality in Nova Scotia,
Canada. Journal of the Marine Biological Association of the United Kingdom, 85, 1511–1517.
CARAGUEL, C.G., O’KELLY, C.J., LEGENDRE, P., FRASCA, S., GAST, R.J., DESPRÉS, B.M., CAWTHORN,
R.J. and GREENWOOD, S.J. (2007) Microheterogeneity and coevolution: an exami-
nation of rDNA sequence characteristics in Neoparamoeba pemaquidensis and its
prokinetoplastid endosymbiont. Journal of Eukaryotic Microbiology, 54, 418.
DUMONT, C.P., HIMMELMAN, J.H. and RUSSELL, M.P. (2004) Size-speciﬁc movement of green
sea urchins Strongylocentrotus droebachiensis on urchin barrens in eastern Canada.
Marine Ecology Progress Series, 276, 93–101.
FEEHAN, C., SCHEIBLING, R.E. and LAUZON-GUAY, J.-S. (2012a) An outbreak of sea urchin
disease associated with a recent hurricane: Support for the «killer storm hypothe-
sis» on a local scale. Journal of Experimental Marine Biology and Ecology, 413, 159–168.
FEEHAN, C., SCHEIBLING, R.E. and LAUZON -GUAY, J.-S. (2012b) Aggregative feeding behavior
in sea urchins (Strongylocentrotus droebachiensis) leads to gap formation and destruc-
tive grazing in a Nova Scotian kelp bed. Marine Ecology Progress Series, 444, 69–83.
FEEHAN, C., JOHNSON-MACKIN NON, J., SCHEIBLING, R.E., LAUZON -GUAY, J.-S. and SIMPSON,
A.G.B. (2013) Validating the identity of Paramoeba invadens, the causative agent of
recurrent mass mortality of sea urchins in Nova Scotia, Canada. Diseases of Aquatic
Organisms, 103, 209–227.
FELDMAN, G.C. and MCCLAIN, C.R. (2012) Ocean Color Web, MODIS Reprocessing 2012.0,
NASA Goddard Space Flight Center (ed. by KURING, N. and BAILEY, S.W.). http://
FILBEE-DEXTER, K. and SCHEIBLING, R.E. (2012) Hurricane-mediated defoliation of kelp
beds and pulsed delivery of kelp detritus to offshore sedimentary habitats. Marine
Ecology Progress Series, 445, 51–64.
Congresuales 33.indb 383 30/01/2014 9:00:04
FLIERL, G.R. and WROEBLEWSKI, J.S. (1985) The possible inﬂuence of warm core Gulf
Stream rings upon shelf water larval ﬁsh distribution. Fishery Bulletin, 83, 313–330.
FOLKE, C., CARPENTER, S., WALKER, B., SCHEFFER, M., ELMQVIST, T., GUNDERSON, L. and HOLLING,
C.S. (2004) Regime shifts, resilience, and biodiversity in ecosystem management.
Annual Review of Ecology and Systematics, 35, 557–581.
HART, M.W. and SCHEIBLING, R.E. (1988) Heat waves, baby booms, and the destruction of
kelp beds by sea urchins. Marine Biology, 99, 167–176.
HARVELL, C.D., KIM, K., BURKHOLDER, J.M., COLWELL, R.R., EPSTEIN, P.R., GRIMES, D.J.,
HOFMANN, E.E., LIPP, E.K., OSTERHAUS, A.D.M.E., OVERSTREET, R.M., PORTER, J.W.,
SMITH, G.W. and VASTA, G.R (1999) Emerging marine diseases–climate links and
anthropogenic factors. Science, 285, 1505–1510.
HARVELL, C.D., MITCHELL, C.E., WARD, J.R., ALTIZER, S., DOBSON, A.P., OSTFELD, R.S. and
SAMUEL, M.D. (2002) Climate warming and disease risks for terrestrial and marine
biota. Science, 296, 2158–2162.
HEATH, R.A. (1973) Variability of water properties and circulation of St. Margaret’s Bay, Nova
Scotia. Fisheries and Research Board of Canada, Dartmouth, NS.
HIMMELMAN, J.H., CARDINAL, A. an d BOURGET, E. (1983) Community development following
removal of urchins, Strongylocentrotus droebachiensis, from the rocky subtidal zone of
the St. Lawrence Estuary, Eastern Canada. Oecologia, 59, 27–39.
HOOPER, R. (1980) Obser vations on algal-grazer interactions in Newfoundland
and Labrador. Canadian Technical Reports of Fisheries and Aquatic Sciences, 954,
JELLETT, J.F., NOVIT SKY, J.A., CANTLEY, J.A. and SCHEIBLING, R.E. (1989) Non-occurrence of
free-living Paramoeba invadens in water and sediments of Halifax Harbour, Nova
Scotia, Canada. Marine Ecology Progress Series, 56, 205–209.
JELLETT, J.F. and SCHEIBLING, R.E. (1988) Effect of temperature and prey availability on
growth of Paramoeba invadens in monoxenic culture. Applied and Environmental
Microbiolog y, 54, 1848–1854.
JELLETT, J.F., SCHEIBLING, R.E. and WARDLAW, A.C. (1988) Host speciﬁcity of Paramoeba
invadens, a sea urchin pathogen. Echinoderm biolog y (ed. by BURKE, R.D., MLADENOV,
P.V., LAMBERT, P. and PARSLEY, R.L.), pp. 755–761. Balkema, Rotterdam.
JOHNSON, C.R. and MANN, K.H. (1988) Diversity, patterns of adaptation, and stability of
Nova Scotian kelp beds. Ecological Monographs, 58, 129–154.
JONES, G.M. (1985) Paramoeba invadens n. sp. (Amoebida, Paramoebidae), a pathogenic
amoeba from the sea urchin, Strongylocentrotus droebachiensis, in eastern Canada.
Journal of Protozoology, 32, 564 –569.
JONES, G.M., HEBDA, A.J., SCHEIBLING, R.E. and MILLER, R.J. (1985) Histopathology of the
disease causing mass mortality of sea urchins (Strong ylocentrotus droebachiensis) in
Nova Scotia. Journal of Invertebrate Pathology, 71, 559–565.
Congresuales 33.indb 384 30/01/2014 9:00:04
JONES, G.M. and SCHEIBLING, R.E. (1985) Paramoeba sp. (Amoebida, Paramoebidae) as
the possible causative agent of sea urchin mass mortality in Nova Scotia. Journal of
Parasitology, 71, 559–565.
KNOWLTON, N. (2004) Multiple ‘stable’ states and the conservation of marine ecosystems.
Progress in Oceanography, 60, 387–396.
KRUMHANSL, K.A. and SCHEIBLING, R.E. (2012a) Detrital subsidy from subtidal kelp beds is
altered by the invasive green alga Codium fragile ssp. fragile. Marine Ecology Progress
Series, 456, 73−85.
KRUMHANSL, K.A. and SCHEIBLING, R.E. (2012b) Production and fate of kelp detritus. Marine
Ecology Progress Series, 467, 281–302.
LAFFERTY, K.D., PORTER, J.W. and FORD, S.E. (2004) Are diseases increasing in the ocean?
Annual Review of Ecology, Evolution, and Systematics, 35, 31–54.
LAUZON-GUAY, J.-S., SCHEIBLING, R.E. and BARBEAU, M.A. (2008) Formation and propaga-
tion of feeding fronts in benthic marine invertebrates: a modeling approach using
kinesis. Ecology, 89, 3150–3162.
LAUZON-GUAY, J.-S., SCHEIBLING, R.E. and BARBEAU, M.A. (2009) Modeling phase shifts in a
rocky subtidal ecosystem. Marine Ecology Progress Series, 375, 25–39.
LAUZON-GUAY, J.-S. and SCHEIBLING, R.E. (2010) Spatial dynamics, ecological thresholds
and phase shifts: modeling grazer aggregation and gap formation in kelp beds.
Marine Ecology Progress Series, 403, 29−41.
LESSIOS, H. (1988) Mass mortality of Diadema antillarum in the Caribbean: What have we
learned? Annual Review of Ecology and Systematics, 19, 371 – 393.
MAY, R.M. (1977) Thresholds and breakpoints in ecosystems with a multiplicity of stable
states. Nature, 269, 471–477.
MILLER, R.J. (1985) Succession in sea urchin and seaweed abundance in Nova Scotia,
Canada. Marine Biology, 84, 275–286.
MILLER, R.J. and COLODEY, A.G. (1983) Widespread mass mortalities of the green sea urchin
in Nova Scotia, Canada. Marine Biology, 73, 263–267.
MILLER, R.J. and NOL AN, S. (2000) Management of the Nova Scotia Sea Urchin Fishery: A Nearly
Successful Habitat Based Management Regime. Canadian Stock Assessment Secretariat
Research Document 2000/109 (41 pp.), Department of Fisheries and Oceans, Ottawa.
MILLER, R.J. and NOLAN, S.C. (2008) Management methods for a sea urchin dive ﬁshery
with individual ﬁshing zones. Journal of Shellﬁsh Research, 27, 929–938.
MOORE, D.S. and MILLER, R.J. (1983) Recovery of macroalgae following widespread sea
urchin mortality with a description of the nearshore hard bottom habitat on the
Atlantic coast of Nova Scotia. Canadian Technical Reports of Fisheries and Aquatic
Sciences, 1230, vii-94.
MUELHSTEIN, L. (1989) Perspectives on the wasting disease of eelgrass Zostera marina.
Diseases of Aquatic Organisms, 7, 211–221.
Congresuales 33.indb 385 30/01/2014 9:00:04
MULLEN, T. E., RUSSELL, S., TUCKER, M. T., MARATEA, J. L., KOERTING, K., HINCKLEY, L., DE
GUISE, S., FRASCA JR., S., FRENCH, R. A., BURRAGE, T. G. and PERKINS, C. (2004) Para-
moebiasis associated with mass mortality of American lobster Homarus americanus
in Long Island Sound, USA. Journal of Aquatic Animal Health, 16, 29–38.
MULLEN, T.E., NEVIS, K.R., O’KELLY, C.J., GA ST, R. and FRASCA, Jr. S. (2005) Nuclear small-
subunit ribosomal RNA gene-based characterization, molecular phylogeny and
PCR detection of Neoparamoeba from western Long Island Sound lobster. Journal of
Shellﬁsh Research, 24, 719–731.
MYERS, R.M. and DRINKWATER, K.F. (1989) The inﬂuence of Gulf Stream warm core rings on
recruitment of ﬁsh in the northwest Atlantic. Journal of Marine Research, 47, 635–656.
PETRA ITIS, P.S. and HOFFMAN, C. (2010) Multiple stable states and relationship between
thresholds in processes and states. Marine Ecology Progress Series, 413, 189–200.
ROBERTS-REGAN, D.L., SCHEIBLING, R.E. and JELLETT, J.F. (1988) Natural and experimentally
induced lesions of the body wall of the sea urchin Strongylocentrotus droebachiensis.
Diseases of Aquatic Organisms, 5, 51–62.
SAUCH YN, L.K., LAUZON-GUAY, J.-S. and SCHEIBLING, R.E. (2011) Sea urchin fecal production
and accumulation in a rocky subtidal ecosystem. Aquatic Biology, 13, 215–223.
SCHEFFER, M., CARPENTER, S.R., FOLEY, J.A. and WALKER, B. (2001) Catastrophic shifts in
ecosystems. Nature, 413, 591–596.
SCHEIBLING, R.E. (1984a) Echinoids, epizootics and ecological stability in the rocky sub-
tidal off Nova Scotia, Canada. Helgolander Wissenschaftliche Meeresuntersuchungen,
SCHEIBLING, R.E. (1984b) Predation by rock crabs (Cancer irroratus) on diseased sea urchins
(Strongylocentrotus droebachiensis) in Nova Scotia. Canadian Journal of Fisheries and
Aquatic Sciences, 41, 1847–1851.
SCHEIBLING, R.E. (1986) Increased macroalgal abundance following mass mortalities of
sea urchins (Strongylocentrotus droebachiensis) along the Atlantic coast of Nova Scotia.
Oecologia 68, 186–198.
SCHEIBLING, R.E. (1988) Microbial control of sea urchins: Achilles’ heel or Pandora’s box?
Echinoderm biology (ed . by BURKE, R.D., MLADENOV, P.V., LAMBERT, P. and PARSLEY, R.L.),
pp. 745–754. Balkema, Rotterdam.
SCHEIBLING, R.E., FEEHAN, C. and LAUZON-GUAY, J.-S. (2010) Disease outbreaks associated
with recent hurricanes cause mass mortality of sea urchins in Nova Scotia. Marine
Ecology Progress Series, 408, 109–116.
SCHEIBLING, R.E. and HENNIGAR, A.W. (1997) Recurrent outbreaks of disease in sea urchins
Strongylocentrotus droebachiensis in Nova Scotia: evidence for a link with large-scale
meteorologic and oceanographic events. Marine Ecology Progress Series, 152, 155–165.
SCHEIBLING, R.E., HENNIGAR, A.W. And BALCH, T. (1999) Destructive grazing, epiphytism,
and disease: the dynamics of sea urchin-kelp interactions in Nova Scotia. Canadian
Journal of Fisheries and Aquatic Sciences, 56, 2300–2314.
Congresuales 33.indb 386 30/01/2014 9:00:04
SCHEIBLING, R.E. and LAUZON-GUAY, J.-S. (2010) Killer storms: North Atlantic hurricanes
and disease outbreaks in sea urchins. Limnology and Oceanography, 55, 2331–2338.
SCHEIBLING, R.E. and STEPHENSON, R.L. (1984) Mass mortality of Strongylocentrotus droeba-
chiensis (Echinodermata: Echinoidea) off Nova Scotia, Canada. Marine Biology, 78,
STENECK, R.S., GRAHAM, M.H., BOURQUE, B.J., CORBET T, D., ERLANDSON, J.M., ESTES, J.A. and
TEGNER, M.J. (2002) Kelp forest ecosystems: biodiversity, stability, resilience and
future. Environmental Conservation, 29, 436−459.
WARD, J.R. and LAFFERTY, K.D. (2004) The elusive baseline of marine disease: Are diseases
in ocean ecosystems increasing? PLoS Biology, 2(4): e120.
WHARTON, W.G. and MANN, K.H. (1981) Relationship between destructive grazing by the
sea urchin, Strong ylocentrotus droebachiensis, and the abundance of American lobs-
ter, Homarus americanus, on the Atlantic coast of Nova Scotia. Canadian Journal of
Fisheries and Aquatic Sciences, 38, 1339–1349.
Congresuales 33.indb 387 30/01/2014 9:00:05