Community Ecology Processes, Models, and Applications

environmental stochasticity, and may also lower
their resilience to demographic perturbation. Sec-
ond, harvesting disproportionately targets large an-
imals near the top of marine food chains (Botsford
et al. 1997; Paulyet al. 1998; Jacksonet al. 2001).
Because most ecosystems support few species of
apex predators, this combination of demographic
vulnerability and targeted human pressure means
that marine (and other) ecosystems under human
influence are inherently vulnerable to loss of an
entire functional group or trophic levels at the top
of the food web (Jacksonet al. 2001; Duffy 2002;
Dobsonet al. 2006). The predicted pattern that re-
sults is ‘trophic skew’, a vertical compaction and

blunting of the trophic pyramid due to proportion-

ally greater losses of higher level species (Duffy

2003).

8.2.2 Empirical evidence for trophic skew in the ocean

Several lines of evidence illustrate that human im-

pacts cause predictable changes in the functional

structure of marine communities. The importance

of life history traits in mediating responses is

illustrated by data from the North Sea. Sustained

size-selective fishing during the late 20th century

shifted the demersal fish community toward

increased aggregate growth rate and decreased

average age and length at maturity as smaller,

faster maturing species gained in relative abun-

dance, while larger, slower growing species declined

(Jenningset al. 1999a). A similar pattern was found

on fished coral reefs (Jenningset al. 1999b).

The decline of apex predators in the oceans has

taken on iconic status since Paulyet al. (1998) pre-

sented evidence for ‘fishing down the food web’,

i.e. a worldwide decline in the mean trophic level of

fishery landings, which they suggested resulted

from sequential depletion of large-bodied preda-

tors. Although this pattern results in part from ad-

dition of new fisheries at lower trophic levels

(‘fishing through the food web’; Essingtonet al.

2006), there is abundant evidence that extinction

and depletion of marine animals are consistently

biased toward loss of large animals at high trophic

levels (Paulyet al. 1998; Jacksonet al. 2001; Dulvy

et al. 2003; Byrneset al. 2007), resulting in broad-

scale declines in both abundance and body size of

marine predators (Baumet al. 2003; Myers and

Worm 2003; Hsiehet al. 2006).

But marine ecosystems are also increasingly af-

fected by exotic invasions, which could in principle

counteract the loss or depletion of species (Sax and

Gaines 2003). Byrneset al. (2007) explored this pos-

sibility by synthesizing data from global tallies of

marine extinctions (Dulvyet al. 2003) and marine

invasions in four well-documented areas, and as-

signing each species a trophic level. They found

that the gains and losses of biodiversity did not

compensate one another functionally but instead

led to consistent directional change in the shape of

16

4

10

12

14

16

18

20

(a)

(b)

6810

log 2 maximum total mass

log 2 body mass

12 14 16

15

14

13

15

N (

0 /

00

)

15

N (

0 /

00

)

12

11

10

(^9051015)
Figure 8.2Among North Sea fishes, trophic level (as
indexed byd^15 N signature) is unrelated to maximum body
mass of the species (a) but closely related to individual
body mass (b). Thus, trophic level is a property of
individuals, not species, and increases through ontogeny.
Reproduced with permission from Jenningset al. (2001).
STRUCTURE AND FUNCTIONING OF EMERGING MARINE COMMUNITIES 99