304
Carol T. STUART et al.
orders of magnitude compared to surrounding areas
(Van Dover, 1990, 2000; Tunnicliffe, 1991). Levin et al.
(1994) noted that organic-rich zones in the deep sea
bear a striking resemblance to the frequently observed
effects of moderate organic pollution in coastal marine
environments: higher standing stock, fewer species and
numerical dominance of a few species leading to low
evenness (Pearson and Rosenberg, 1978).
We stress that patterns of species diversity on these
very large scales cannot represent controlled withinhabitat comparisons. For example, upper bathyal diversity gradients in the western North Atlantic are correlated with the variability of sediment particle size (Etter
and Grassle, 1992) as well as animal abundance. Sites
of localized organic enrichment can also experience
physical disturbance from sediment resuspension by
strong near-bottom currents (Aller, 1997), or occasional
catastrophic sediment displacement from mass-wasting
on steep slopes (Levin et al., 1994). Hydrothermal
vents occur in hard substrata, in contrast to the soft
sediments that characterize most deep-sea habitats,
and vent communities are distinctly different in being
fueled by bacterial chemosynthesis rather than by
sinking phytodetritus. Species composition and the
life-history characteristics of the faunas at enriched
sites also differ from those in neighboring habitats.
A variety of biotic and abiotic environmental causes is
undoubtedly, to some extent, responsible for geographic
variation in deep-sea species diversity (Rex et al.,
1997; Levin et al., 2001). However, the most consistent
ecological factor associated with the combination of
elevated standing stock and depressed diversity in
deep-sea communities is some form of pulsed nutrient
loading. The relative importance of the rate of nutrient
input and its variability is unclear.
TEMPORAL PATTERNS OVER GEOLOGICAL
TIME
Until quite recently, knowledge of temporal variation
in deep-sea communities was limited to seasonal
and interannual changes in reproductive pattern and
population density linked to short-term cycles of
sinking phytodetritus (Gage and Tyler, 1991; Young
and Eckelbarger, 1994). The extraordinary advances
in paleoceanography during the 1990s are beginning
to provide, for the first time, a historical dimension
to life in the deep sea over geological time-scales.
Exposed fossil assemblages of deep-sea organisms are
rare. The new understanding comes from seabed cores
taken by the Deep-Sea Drilling Project. The cores
provide a continuous sedimentary record, which can be
dated using accurate chronometric methods. Analyses
of elemental make-up and stable-isotope ratios in
microfossils and lithic materials from the cores provide
climate-proxy data for reconstructing the physical
oceanography of past oceans, including temperature
gradients, water chemistry, the deep thermohaline
circulation, surface ice volume and iceberg discharges
(Shackleton et al., 1984; McManus et al., 1994; Kennett
and Ingram, 1995; Dwyer et al., 1995; Adkins et al.,
1997; Marchitto et al., 1998; Oppo et al., 1998; Raymo
et al., 1998). Coincident climate shifts are also recorded
in the skeletons of deep-sea corals (Smith et al., 1997;
Adkins et al., 1998). These environmental changes
can be associated with the diversity and composition
of microfossil assemblages, and with the geographic
distribution and morphological evolution of individual
species (Cronin and Schneider, 1990).
While catastrophic biological crises of the remote
past have been attributed to major changes in the deep
circulation, such as Permo-Triassic mass extinction
(Knoll et al., 1996; Wignall and Twitchett, 1996;
Isozaki, 1997), the most precise record exists for
the Cenozoic, particularly the last several million
years. During this era there were glaciation cycles
in the Northern Hemisphere on time scales of 10
4 –
10
5 years, which now seem linked to variation in the
Earth’s orbit (the Milankovitch cycles). The advance
and retreat of polar ice sheets and the attendant
changes in global climate were not gradual regular
transitions. They were accompanied by briefer episodic
climate changes on millennial time scales (Dansgaard–
Oeschger events) during which temperature changes
could be surprisingly abrupt, sometimes occurring
over decades or centuries (Severinghaus et al., 1998).
Many of these climate changes appear to have been
global in scale (Kotilainen and Shackleton, 1995;
Steig et al., 1998). The causes of these short-term
fluctuations remain uncertain, but probably involved
complex interactions between the atmosphere, ocean
currents and ice coverage which may be intrinsic to
Earth’s climate, or interactions among orbital forcing
parameters and internal climate factors (Driscoll and
Haug, 1998; Marchitto et al., 1998; Raymo et al.,
1998; Weaver et al., 1998). Shifts in the ice-age climate
affected the deep-sea environment through reorganization of the thermohaline circulation and the potential
for surface–benthic coupling. Their discovery requires
Carol T. STUART et al.
orders of magnitude compared to surrounding areas
(Van Dover, 1990, 2000; Tunnicliffe, 1991). Levin et al.
(1994) noted that organic-rich zones in the deep sea
bear a striking resemblance to the frequently observed
effects of moderate organic pollution in coastal marine
environments: higher standing stock, fewer species and
numerical dominance of a few species leading to low
evenness (Pearson and Rosenberg, 1978).
We stress that patterns of species diversity on these
very large scales cannot represent controlled withinhabitat comparisons. For example, upper bathyal diversity gradients in the western North Atlantic are correlated with the variability of sediment particle size (Etter
and Grassle, 1992) as well as animal abundance. Sites
of localized organic enrichment can also experience
physical disturbance from sediment resuspension by
strong near-bottom currents (Aller, 1997), or occasional
catastrophic sediment displacement from mass-wasting
on steep slopes (Levin et al., 1994). Hydrothermal
vents occur in hard substrata, in contrast to the soft
sediments that characterize most deep-sea habitats,
and vent communities are distinctly different in being
fueled by bacterial chemosynthesis rather than by
sinking phytodetritus. Species composition and the
life-history characteristics of the faunas at enriched
sites also differ from those in neighboring habitats.
A variety of biotic and abiotic environmental causes is
undoubtedly, to some extent, responsible for geographic
variation in deep-sea species diversity (Rex et al.,
1997; Levin et al., 2001). However, the most consistent
ecological factor associated with the combination of
elevated standing stock and depressed diversity in
deep-sea communities is some form of pulsed nutrient
loading. The relative importance of the rate of nutrient
input and its variability is unclear.
TEMPORAL PATTERNS OVER GEOLOGICAL
TIME
Until quite recently, knowledge of temporal variation
in deep-sea communities was limited to seasonal
and interannual changes in reproductive pattern and
population density linked to short-term cycles of
sinking phytodetritus (Gage and Tyler, 1991; Young
and Eckelbarger, 1994). The extraordinary advances
in paleoceanography during the 1990s are beginning
to provide, for the first time, a historical dimension
to life in the deep sea over geological time-scales.
Exposed fossil assemblages of deep-sea organisms are
rare. The new understanding comes from seabed cores
taken by the Deep-Sea Drilling Project. The cores
provide a continuous sedimentary record, which can be
dated using accurate chronometric methods. Analyses
of elemental make-up and stable-isotope ratios in
microfossils and lithic materials from the cores provide
climate-proxy data for reconstructing the physical
oceanography of past oceans, including temperature
gradients, water chemistry, the deep thermohaline
circulation, surface ice volume and iceberg discharges
(Shackleton et al., 1984; McManus et al., 1994; Kennett
and Ingram, 1995; Dwyer et al., 1995; Adkins et al.,
1997; Marchitto et al., 1998; Oppo et al., 1998; Raymo
et al., 1998). Coincident climate shifts are also recorded
in the skeletons of deep-sea corals (Smith et al., 1997;
Adkins et al., 1998). These environmental changes
can be associated with the diversity and composition
of microfossil assemblages, and with the geographic
distribution and morphological evolution of individual
species (Cronin and Schneider, 1990).
While catastrophic biological crises of the remote
past have been attributed to major changes in the deep
circulation, such as Permo-Triassic mass extinction
(Knoll et al., 1996; Wignall and Twitchett, 1996;
Isozaki, 1997), the most precise record exists for
the Cenozoic, particularly the last several million
years. During this era there were glaciation cycles
in the Northern Hemisphere on time scales of 10
4 –
10
5 years, which now seem linked to variation in the
Earth’s orbit (the Milankovitch cycles). The advance
and retreat of polar ice sheets and the attendant
changes in global climate were not gradual regular
transitions. They were accompanied by briefer episodic
climate changes on millennial time scales (Dansgaard–
Oeschger events) during which temperature changes
could be surprisingly abrupt, sometimes occurring
over decades or centuries (Severinghaus et al., 1998).
Many of these climate changes appear to have been
global in scale (Kotilainen and Shackleton, 1995;
Steig et al., 1998). The causes of these short-term
fluctuations remain uncertain, but probably involved
complex interactions between the atmosphere, ocean
currents and ice coverage which may be intrinsic to
Earth’s climate, or interactions among orbital forcing
parameters and internal climate factors (Driscoll and
Haug, 1998; Marchitto et al., 1998; Raymo et al.,
1998; Weaver et al., 1998). Shifts in the ice-age climate
affected the deep-sea environment through reorganization of the thermohaline circulation and the potential
for surface–benthic coupling. Their discovery requires