102
Verena TUNNICLIFFE et al.
vent species, examination of reproductive features to
date reveals no unusual or universal dispersal mode
adopted by vent invertebrates (Tyler and Young, 1999).
Direct-development, short-dispersing (lecithotrophic)
larvae and long-dispersing (planktotrophic) larvae are
all known among vent animals – and different modes
may occur in closely related species (Mullineaux and
France, 1995). It is unlikely that species range is
related to the mode of larval development (Jollivet
et al., 1998). The image of vents as highly unstable
habitats has led to the expectation of a ‘weed’ strategy
among these animals. The lack of support for this
concept points to other directions for research. One
is that adaptation to instability may not lie only in
dispersability – other ecologic adaptations are also
important (McHugh and Tunnicliffe, 1994). Another is
that the basic hypothesis may be flawed: many vent
species may have evolved in relatively stable vent
conditions.
Short-range dispersal certainly must affect withinregion distributions and gene flow. Tiny larvae must
rely on transport by ocean currents which often run
parallel to the ridge crest. Peak velocities of tidal flows
could move larvae over a kilometre in a few hours
(Mullineaux and France, 1995). Another important
mechanism may be the hydrothermal plumes that
rise some 300 m above the seafloor (Kim et al.,
1994). ‘Megaplumes’ associated with eruption push
hydrothermal fluids up over 1000 m and may provide a rare but effective mechanism of long-distance
transport. Gene flow along hundreds of kilometres of
ridge crest tends to be relatively high, and only large
offsetting transform faults induce lower population
communication (Vrijenhoek, 1997). The species noted
above as initial colonizers in succession may be particularly suited to local dispersal and establishment. Future
studies will hopefully show if later species are less
efficient in dispersal or whether habitat conditioning is
more important.
ISSUES PECULIAR TO SEEPS
Research to understand cold-seep ecosystem processes
is still in early stages but observational data indicate
that seepage of pore fluids can occur under a very broad
range of environmental conditions in the context of
flow rates, temporal stability and geological settings.
Such variability promotes the question: does species
composition, density and biomass (and production)
of the associated fauna reflect the different habitat
conditions?
Environmental conditions and habitat longevity
Depth of known seep communities ranges from 300
to 6000 m (Sibuet and Olu, 1998). Bathymetric effects
may influence species composition, as noted on the
Peruvian margin (Olu et al., 1996a); the geographic
distribution of species may be secondarily affected.
Geological features such as manganese encrustation
and depth of sediment cover also control species
composition and density (Juniper and Sibuet, 1987;
Sibuet et al., 1988). To date, fluid discharge rate is
the best-understood local environmental variable. The
relationship between discharge rates and species composition, density and general shape of clam colonies
is direct enough to be quantitatively modelled (Henry
et al., 1992). The effects of flow rate may determine
forms of symbiosis (MacDonald et al., 1990a; Olu
et al., 1997) and trophic pathways. Just as weak
flow rates do not support colonization by some seep
species, extremely rapid expulsions with fluid mud
can exclude all fauna, probably because of substratum
instability; the latter condition can occur on active mud
volcanoes and diatremes
1 (Sassen et al., 1994; Olu
et al., 1997). Faunal exclusion by extreme chemical
conditions may occur, as at sites of intense discharge
of hydrocarbons (Roberts and Carney, 1997), including
perhaps methane.
In the absence of long-term studies, comparative
data provide a first indication of the scales and
characteristics of temporal evolution at seeps. Locally,
the frequent juxtaposition of living and dead bivalves at
several geographic locations (Juniper and Sibuet, 1987;
Rosman et al., 1987; Olu et al., 1996b) indicates a relatively rapid dynamic and spatial shifting of fluid flow.
At a larger scale, surface renewal events (rapid mud
extrusion) for Barbados mud volcanoes are estimated
to occur at intervals of 100 to 1000 years. Between
events, substrata and fluid discharge properties evolve,
and these processes appear to be reflected in the composition of seep communities, with subsurface sulphide
oxidation and dependent symbioses appearing in later
stages (Olu et al., 1997). Similarly, at hydrocarbon
seeps in the Gulf of Mexico, the evolution from rapid
1 A diatreme is a volcanic intrusion emitting much carbon dioxide into the sediment. See Glossary, p. 477.
Verena TUNNICLIFFE et al.
vent species, examination of reproductive features to
date reveals no unusual or universal dispersal mode
adopted by vent invertebrates (Tyler and Young, 1999).
Direct-development, short-dispersing (lecithotrophic)
larvae and long-dispersing (planktotrophic) larvae are
all known among vent animals – and different modes
may occur in closely related species (Mullineaux and
France, 1995). It is unlikely that species range is
related to the mode of larval development (Jollivet
et al., 1998). The image of vents as highly unstable
habitats has led to the expectation of a ‘weed’ strategy
among these animals. The lack of support for this
concept points to other directions for research. One
is that adaptation to instability may not lie only in
dispersability – other ecologic adaptations are also
important (McHugh and Tunnicliffe, 1994). Another is
that the basic hypothesis may be flawed: many vent
species may have evolved in relatively stable vent
conditions.
Short-range dispersal certainly must affect withinregion distributions and gene flow. Tiny larvae must
rely on transport by ocean currents which often run
parallel to the ridge crest. Peak velocities of tidal flows
could move larvae over a kilometre in a few hours
(Mullineaux and France, 1995). Another important
mechanism may be the hydrothermal plumes that
rise some 300 m above the seafloor (Kim et al.,
1994). ‘Megaplumes’ associated with eruption push
hydrothermal fluids up over 1000 m and may provide a rare but effective mechanism of long-distance
transport. Gene flow along hundreds of kilometres of
ridge crest tends to be relatively high, and only large
offsetting transform faults induce lower population
communication (Vrijenhoek, 1997). The species noted
above as initial colonizers in succession may be particularly suited to local dispersal and establishment. Future
studies will hopefully show if later species are less
efficient in dispersal or whether habitat conditioning is
more important.
ISSUES PECULIAR TO SEEPS
Research to understand cold-seep ecosystem processes
is still in early stages but observational data indicate
that seepage of pore fluids can occur under a very broad
range of environmental conditions in the context of
flow rates, temporal stability and geological settings.
Such variability promotes the question: does species
composition, density and biomass (and production)
of the associated fauna reflect the different habitat
conditions?
Environmental conditions and habitat longevity
Depth of known seep communities ranges from 300
to 6000 m (Sibuet and Olu, 1998). Bathymetric effects
may influence species composition, as noted on the
Peruvian margin (Olu et al., 1996a); the geographic
distribution of species may be secondarily affected.
Geological features such as manganese encrustation
and depth of sediment cover also control species
composition and density (Juniper and Sibuet, 1987;
Sibuet et al., 1988). To date, fluid discharge rate is
the best-understood local environmental variable. The
relationship between discharge rates and species composition, density and general shape of clam colonies
is direct enough to be quantitatively modelled (Henry
et al., 1992). The effects of flow rate may determine
forms of symbiosis (MacDonald et al., 1990a; Olu
et al., 1997) and trophic pathways. Just as weak
flow rates do not support colonization by some seep
species, extremely rapid expulsions with fluid mud
can exclude all fauna, probably because of substratum
instability; the latter condition can occur on active mud
volcanoes and diatremes
1 (Sassen et al., 1994; Olu
et al., 1997). Faunal exclusion by extreme chemical
conditions may occur, as at sites of intense discharge
of hydrocarbons (Roberts and Carney, 1997), including
perhaps methane.
In the absence of long-term studies, comparative
data provide a first indication of the scales and
characteristics of temporal evolution at seeps. Locally,
the frequent juxtaposition of living and dead bivalves at
several geographic locations (Juniper and Sibuet, 1987;
Rosman et al., 1987; Olu et al., 1996b) indicates a relatively rapid dynamic and spatial shifting of fluid flow.
At a larger scale, surface renewal events (rapid mud
extrusion) for Barbados mud volcanoes are estimated
to occur at intervals of 100 to 1000 years. Between
events, substrata and fluid discharge properties evolve,
and these processes appear to be reflected in the composition of seep communities, with subsurface sulphide
oxidation and dependent symbioses appearing in later
stages (Olu et al., 1997). Similarly, at hydrocarbon
seeps in the Gulf of Mexico, the evolution from rapid
1 A diatreme is a volcanic intrusion emitting much carbon dioxide into the sediment. See Glossary, p. 477.