via intense predation by humans and invasion
of non-indigenous species. The studies reviewed
here suggest that these emerging marine ecosys-
tems show several recurrent patterns that raise
the following intriguing questions for future re-
search.
The decline in large animals through overhar-
vesting constitutes an important loss of functional
diversity, which is eroding complex food webs into
topologically simpler, and probably more strongly
linked, food chains. This simplification may lead to
fundamental change in community and ecosystem
dynamics with several potential consequences. Are
these simpler food chains indeed less stable gener-
ally and, specifically, more vulnerable to dramatic
regime shifts than naturally diverse communities
(Folkeet al. 2004)?
There is some evidence that intense exploitation
may change not only the strength of interactions
but also the mode of control, from bottom-up in
lightly exploited, naturally diverse systems to top-
down in heavily exploited systems, as in the north-
west Atlantic (Franket al. 2006). Indeed, this re-
view suggests that trophic cascades may be more
common in marine systems than concluded previ-
ously (Jennings and Kaiser 1998). The different
conclusions may stem in part from newer, fine-
scale data (e.g. Franket al. 2006); but it is also
conceivable that marine ecosystems have changed
even in the last decade toward states more vulner-
able to perturbation. Does heavy exploitation gen-
erally shift marine communities towards stronger
top-down control?
Accelerating invasions of non-indigenous species
are changing marine communities worldwide (Ruiz
et al. 2000). Do these emerging communities interact
in different ways as a result of the lack of shared
evolutionary history among species? For example,
do marine consumer–prey interactions involving
non-indigenous species differ systematically from
those involving only native species, as they do on
land where non-indigenous consumers promote
‘invasional meltdown’ (Parkeret al. 2006)?
Finally, it is becoming increasingly clear that evo-
lutionary change often occurs on similar timescales
to ecological interactions among species, and can be
critical to understanding the dynamics of those in-
teractions (Thompson 1998; Hairstonet al.2005).
This is especially true of human predation on marine
fishes, which generally targets larger, more econom-
ically valuable individuals, and accordingly has pro-
duced declines in average body size in many
exploited marine fish species over recent decades
(Hsiehet al. 2006). If length and age at maturity are
at least partially heritable, the resultant size- and
age-selective mortality means that rapidly maturing
genotypes will be favoured under fishing mortality
(Law 2000) and this truncation in size structure will
produce not only ecological ramifications through
the ecosystem but also evolutionary change. In par-
ticular, length and age at sexual maturity are key life
history traits affecting fitness. Controlled experi-
ments in both laboratory (Conover and Munch
2002) and field (Reznick and Ghalambor 2005) con-
firm that size-selective mortality can produce sub-
stantial genetically based changes in age and size at
maturity within a few generations. Data from com-
mercially exploited fishes also indicate that age and
size at maturity have substantial heritabilities, and
many stocks indeed have shown predicted declines
in age and size at maturity over recent decades
(Hutchings and Baum 2005). A critical question for
future research is how much of the change in life
histories of wild fish stocks results from evolution
versus other factors such as release from competi-
tion, and whether this evolution reinforces the hys-
teresis between exploited and unexploited states of
marine ecosystems, as suggested for cod-dominated
ecosystems (Olsenet al. 2004; de Rooset al. 2006).Acknowledgments
I am grateful to James Douglass, Peter Morin and
Herman Verhoef for comments that improved the
manuscript, and to the National Science Founda-
tion for support (OCE-0623874).114 APPLICATIONS