302
Carol T. STUART et al.
among species. Gage and May (1993) cautioned that
the deep-sea latitudinal gradients shown by Rex et al.
(1993) might simply reflect shifts in the evenness of the
relative abundance distribution rather than richness.
Rex et al. (2000) examined latitudinal patterns
for isopods, gastropods and bivalves in the North
Atlantic by using the Shannon–Wiener Information
Function (H
) as a diversity measure (see Magurran,
1988). H
is highly correlated with E(S n ), the most
commonly used expression of deep-sea diversity (see
also Grassle and Maciolek, 1992). Like E(S n ), H
is
affected both by the number of species (S) and by the
evenness (J ). Evenness is uncorrelated with latitude in
bivalves, and more weakly correlated than are S and H
for gastropods and isopods (Rex et al., 2000). Partial
correlations between S and latitude remain significant
when the effects of J are removed; but J is uncorrelated
or more weakly correlated with latitude when S is held
constant.
The thorough systematic revision of the gastropod
family Turridae in the eastern North Atlantic by
Bouchet and War´ en (1980) permits latitudinal patterns
to be documented for this family using simple species
richness. The number of bathyal turrid species that
coexist in latitudinal bands from 36ºN to 72ºN shows
a significant poleward decrease in diversity (Rex et al.,
2000). Thus, it seems likely that latitudinal gradients in
species richness exist in the deep sea, just as in other
major ecosystems. Nevertheless, it should be borne in
mind that known macrofaunal data still represent only
three major groups in one ocean, and that estimates of
deep-sea diversity are based exclusively on scattered
remote-sampling programs rather than on the intensive
survey data available for more accessible environments.
As with terrestrial and coastal environments, it seems
reasonable to expect a variety of large-scale patterns to
emerge for different taxa and in different regions of the
World Ocean.
Potential causes of latitudinal gradients in
species diversity
It is likely that large-scale patterns of diversity in the
deep sea are shaped by both ecological and historical
factors. Historical causes involving speciation and the
radiation and geographic spread of taxa have only
recently been investigated. Hypothetical evolutionary
scenarios have been discussed by Wilson and Hessler
(1987), Stuart and Rex (1994) and Allen and Sanders
(1996). The most specific case study is Wilson’s
(1998) analysis of deep-sea isopod diversity on a
Pan-Atlantic basis. Deep-sea isopods are composed of
two distinct taxonomic components that have quite
different evolutionary and geographic origins. Asellotes
are the most diverse, and include many families
that are endemic to the deep sea. They are the
dominant isopod taxon in the deepest regions of the
oceans. Flabelliferans are a more recent clade with
no endemic deep-sea families. Their representation
decreases with depth, and they show a significant
decrease in diversity from 60º South to 50º North
(Fig. 10.6). Wilson (1998) was able to partition diversity statistically between asellotes and flabelliferans
to examine independently their contributions to overall
isopod diversity.
Wilson (1998) suggested that asellotes colonized
the deep sea first, possibly in the late Paleozoic or
early Mesozoic, and radiated there extensively. The
deep-sea flabelliferans represent a more recent invasion
from coastal habitats, probably after the mid-Tertiary,
which has not yet produced endemic higher taxa.
Their prevalence in the South Atlantic may reflect
its openness to immigration from the Indo-Pacific
area. Since the North Atlantic Basin is more isolated
from the World Ocean, lower isopod diversity in the
North Atlantic and the poleward decline in diversity
observed there could be partly a historical consequence
of the recent northward spread of flabelliferans in the
deep Atlantic. Future research on spatially-referenced
phylogenies may help to reveal the influence of
speciation and geographic spread of taxa on large-scale
patterns of biodiversity in the deep sea.
Recently, much attention has been focused on the
potential importance of Rapoport’s Rule – the tendency
for geographic ranges of species to expand at higher
latitudes as an adaptive response to greater climatic
variation [Rapoport (1982); see also Stevens (1989,
1992); on the other hand, see Gaston et al. (1998)].
For the deep-sea Turridae of the eastern North Atlantic
discussed above, latitudinal range of species is not
correlated with either latitude (n = 12, r s = 0.055, n.s.),
or the number of species (n = 12, r s = −0.050, n.s.).
One reason why Rapoport’s Rule might not obtain in
the deep sea is that below the permanent thermocline,
which occurs at upper bathyal depths, the environment
is very cold and nearly isothermal. The latitudinal
gradients in physical climatic effects, like seasonal
temperature variation, freezing, and precipitation, that
are thought to be responsible for latitudinal gradients
in the geographic ranges of terrestrial species (Stevens,
Carol T. STUART et al.
among species. Gage and May (1993) cautioned that
the deep-sea latitudinal gradients shown by Rex et al.
(1993) might simply reflect shifts in the evenness of the
relative abundance distribution rather than richness.
Rex et al. (2000) examined latitudinal patterns
for isopods, gastropods and bivalves in the North
Atlantic by using the Shannon–Wiener Information
Function (H
) as a diversity measure (see Magurran,
1988). H
is highly correlated with E(S n ), the most
commonly used expression of deep-sea diversity (see
also Grassle and Maciolek, 1992). Like E(S n ), H
is
affected both by the number of species (S) and by the
evenness (J ). Evenness is uncorrelated with latitude in
bivalves, and more weakly correlated than are S and H
for gastropods and isopods (Rex et al., 2000). Partial
correlations between S and latitude remain significant
when the effects of J are removed; but J is uncorrelated
or more weakly correlated with latitude when S is held
constant.
The thorough systematic revision of the gastropod
family Turridae in the eastern North Atlantic by
Bouchet and War´ en (1980) permits latitudinal patterns
to be documented for this family using simple species
richness. The number of bathyal turrid species that
coexist in latitudinal bands from 36ºN to 72ºN shows
a significant poleward decrease in diversity (Rex et al.,
2000). Thus, it seems likely that latitudinal gradients in
species richness exist in the deep sea, just as in other
major ecosystems. Nevertheless, it should be borne in
mind that known macrofaunal data still represent only
three major groups in one ocean, and that estimates of
deep-sea diversity are based exclusively on scattered
remote-sampling programs rather than on the intensive
survey data available for more accessible environments.
As with terrestrial and coastal environments, it seems
reasonable to expect a variety of large-scale patterns to
emerge for different taxa and in different regions of the
World Ocean.
Potential causes of latitudinal gradients in
species diversity
It is likely that large-scale patterns of diversity in the
deep sea are shaped by both ecological and historical
factors. Historical causes involving speciation and the
radiation and geographic spread of taxa have only
recently been investigated. Hypothetical evolutionary
scenarios have been discussed by Wilson and Hessler
(1987), Stuart and Rex (1994) and Allen and Sanders
(1996). The most specific case study is Wilson’s
(1998) analysis of deep-sea isopod diversity on a
Pan-Atlantic basis. Deep-sea isopods are composed of
two distinct taxonomic components that have quite
different evolutionary and geographic origins. Asellotes
are the most diverse, and include many families
that are endemic to the deep sea. They are the
dominant isopod taxon in the deepest regions of the
oceans. Flabelliferans are a more recent clade with
no endemic deep-sea families. Their representation
decreases with depth, and they show a significant
decrease in diversity from 60º South to 50º North
(Fig. 10.6). Wilson (1998) was able to partition diversity statistically between asellotes and flabelliferans
to examine independently their contributions to overall
isopod diversity.
Wilson (1998) suggested that asellotes colonized
the deep sea first, possibly in the late Paleozoic or
early Mesozoic, and radiated there extensively. The
deep-sea flabelliferans represent a more recent invasion
from coastal habitats, probably after the mid-Tertiary,
which has not yet produced endemic higher taxa.
Their prevalence in the South Atlantic may reflect
its openness to immigration from the Indo-Pacific
area. Since the North Atlantic Basin is more isolated
from the World Ocean, lower isopod diversity in the
North Atlantic and the poleward decline in diversity
observed there could be partly a historical consequence
of the recent northward spread of flabelliferans in the
deep Atlantic. Future research on spatially-referenced
phylogenies may help to reveal the influence of
speciation and geographic spread of taxa on large-scale
patterns of biodiversity in the deep sea.
Recently, much attention has been focused on the
potential importance of Rapoport’s Rule – the tendency
for geographic ranges of species to expand at higher
latitudes as an adaptive response to greater climatic
variation [Rapoport (1982); see also Stevens (1989,
1992); on the other hand, see Gaston et al. (1998)].
For the deep-sea Turridae of the eastern North Atlantic
discussed above, latitudinal range of species is not
correlated with either latitude (n = 12, r s = 0.055, n.s.),
or the number of species (n = 12, r s = −0.050, n.s.).
One reason why Rapoport’s Rule might not obtain in
the deep sea is that below the permanent thermocline,
which occurs at upper bathyal depths, the environment
is very cold and nearly isothermal. The latitudinal
gradients in physical climatic effects, like seasonal
temperature variation, freezing, and precipitation, that
are thought to be responsible for latitudinal gradients
in the geographic ranges of terrestrial species (Stevens,