Trophic Cascades in Ecosystems

advertisement
Trophic Cascades in Ecosystems
(from: http://www.inhs.uiuc.edu/chf/pub/surveyreports/jul-aug96/cascade.html
Wolves eat moose and moose eat fir needles on Isle Royale in Lake Superior. This food
chain is in some form of balance, but suppose that wolves are artificially stocked, say
doubling their numbers. In response, we expect fewer moose and more fir needles. The
stocking of wolves creates a kind of alternating ripple effect that changes the numbers of
organisms in various levels of the food chain. This is called a "trophic cascade" and has
been observed in a number of lakes and a few terrestrial systems, disturbed from the top
down. However, the magnitude--the percent change in the biomass at each trophic level-of the changes often diminishes at lower trophic levels. In addition, we can imagine
perturbing a food chain at the bottom, say by increasing light or nutrient levels. Then we
would expect more producers (green plants), more herbivores, more carnivores--all
changes of the same sign as we go up the food chain, or at least none of opposite sign.
This is also often, but not always, seen. (See figure below.)
Four-level food chain. Signs of influence are indicated. Bootom-up influences are all
positive: increasing producers tends to increase herbivores. Top-down influences are
negative: increasing carnivores tends to decrease herbivores. Cropping reduces each
respective compartment's biomass. Stocking would be represented as negative cropping.
The trophic cascade is a potential management tool (for example, for controlling aquatic
vegetation by controlling carnivorous fish), and its details reveal much about the
interactions between trophic levels and the control of ecosystems. This subject goes back
at least to a controversial 1960 paper by Hairston, Smith, and Slobodkin who, observing
that the earth is green (and not overgrazedly brown), conjectured on how producers,
herbivores, and carnivores interact...all with no equations.
Both top-down and bottom-up influences occur in the same system, and a proper theory
incorporates both mechanisms simultaneously. Figure 1 shows the interactions in a fourlevel food chain subject to cropping and to manipulation of the available resources.
Early in the discussion there was a tendency to talk of top-down and bottom-up effects as
mutually exclusive. After listening to colleagues and their strongly held opinions, I began
to feel that the problem represented by Figure 1 had strong analogies to the physics of
magnetism (a food chain was just a linear chain with nearest-neighbor interactions).
While only nearest neighbors in the food chain interact directly, all compartments interact
indirectly as they respond to changes in resources and to cropping. Rather than use the
rusty machinery from magnetism, I sought intuitive arguments of how the trophic
cascade should pop out of Figure 1. The arguments were not very convincing, and each
time a Survey colleague reviewed them, he asked for more rigor. The final stroke came
when a well-respected systems ecologist and modeller with a physics background said,
"You may not be right." Then I dusted off the mathematics and solved the problem
exactly.
The resulting approach combines all interactions shown by arrows in Figure 1. The crux
of how one level impacts another is the degree to which a preyed-upon level can
compensate its reduced biomass by capturing more of its prey. With high compensation,
top-down effects diminish rapidly at lower trophic levels, which is often seen
experimentally. For the particular type of predator-prey interaction known as ratio
dependence (which is part of a continuing debate), the trophic cascade should produce
changes that alternate in sign and diminish by a factor of roughly ten at each successively
lower level. In that case a 10% increase in the top carnivore biomass will produce
changes of -1%, +0.1%, and -0.01% in carnivores, herbivores, and producers,
respectively. The last of these is likely undetectable over experimental noise, and hence
consistent with some observations that the trophic cascade disappears around the level of
zooplankton. On the other hand, the same system shows changes of the same sign and
magnitude for a bottom-up perturbation, which is sometimes observed.
If instead "prey-dependent" predation (the basis of the Lotka-Volterra predator-prey
equations) is assumed, the trophic cascade is not seen. Rather than a pattern of (+ - + -),
we see a pattern (+ 0 + 0), which is suspected in some river systems.
The approach I have developed seems to be able to predict essentially all observed
patterns in experiments. Future work will include testing this claim more carefully. As
usual, there will be the necessary job of testing a theory that is logically satisfying, but
based on many simplifications, against real biological mechanisms and data.
Complicating factors include:
1. the fact that food chains are often actually webs, in which case the theory here breaks
down,
2. the difficulty of applying a theory based on steady state to real dynamic systems,
3. the complexity of real predator-prey interactions, including prey shifts at different life
stages and under different environmental conditions,
4. the possiblity that the predicted effects are too small to detect experimentally.
Reference:
Herendeen, R. 1995. A unified quantitative approach to bottom-up:top-down and trophic
cascade hypotheses. Journal of Theoretical Biology 176:13-26.
Robert Herendeen, Center for Aquatic Ecology
Download