9780521861724htl 1..2

assembling natural food webs into linear chains (Poliset al., 2000), but does
illustrate that the rate of input of organic matter into the system appears to
influence the structure of the web and the dynamics of the species within it.
Unproductive systems appear to have highly connected food-web structures
(Belgranoet al., 2005) that are resistant to changes in predation pressure, either
natural (Woodward & Hildrew,2001) or deliberate (Hildrewet al., 2004). Hence,
it has been suggested that trophic cascades are restricted to productive systems
(Poliset al., 2000). The strength of the cascade in lakes increases with total
phosphorus concentrations (TP) (Jeppesenet al., 2003a). Furthermore, trophic
cascades appear to be lacking in pelagic communities from unproductive environ-
ments such as the lakes of Greenland, even though food-chain length is as
predicted by the Oksanenet al.(1981) model, body size well structured among
trophic levels, and there are pronounced effects of predators on herbivores
(Box 7.1 ).A similarpattern was found inoligotrophicalpinelakes where complete

Box 7.1 Greenland – a natural experiment on the effects

of fish in lakes

Greenland lakes are excellent for studying the cascading effects of fish on

lake ecosystems as a large proportion of the lakes are naturally without fish.

Lakes that have not been in contact with the sea at any stage since the last

glaciation are nearly all fishless. However, many lakes that were previously

connected to the sea, but are currently isolated, host land-locked populations

of Arctic char (Salvelinus alpinus, up to three morphs) and/or three-spined

sticklebacks (Gasterosteus aculeatus). Those lakes still connected to the sea

also contain anadromous specimens of char. Eighty-two lakes in the high

Arctic North-East Greenland (NEG) and near the Arctic circle in West

Greenland (WG) have recently been subject to intensive studies (Lauridsen

et al., 2001; Jeppesenet al., 2001, 2003a,b; Christoffersen, 2001 ; Brodersen &

Anderson,2002). In NEG, Arctic char is the only fish species present, while

the stickleback is present in some of the lakes in WG. Neo- and palaeolimno-

logical analyses of these lakes have shown clear differences in cladoceran

community structure among lakes with and without fish, large pelagic and

benthic forms being dominant in fishless lakes (Lauridsenet al., 2001;

Jeppesenet al., 2003a,b). Accordingly, the body size varies substantially

among the two lake types (Fig.7.2). The body mass of all zooplankton,

cladocerans and cyclopoid copepods was much larger in the fishless lakes

in both regions, while no differences were found for calanoid copepods (present

only in WG), rotifers and ciliates. The ratio of zooplankton to phytoplankton

biomass (calculated from chlorophyll a) is much higher overall in fishless

lakes. Nevertheless, no differences were seen in chlorophyll a (Fig.7.2)orin

BODY SIZE AND TROPHIC CASCADES IN LAKES 123