eLearning

are an important step in the extrapolation of ecologi-
cal understanding to the understanding andmanage-
ment of whole ecosystems. Carpenter (1996) pro-
vides additional arguments for experiments at
ecosystem scales.
VERY SMALL SCALES:
BOTTLE TECHNIQUES
A classic use of bottle experiments is to measure
photosynthesis in lakes. When the results of such
experiments were applied to whole ecosystems, the
predictive power was poor (Schindler and Comita
1972; Bower and others 1987). A closer look at the
methods revealed a number of logical flaws. For
example, traditional ways of measuring photosyn-
thesis and nutrient limitation often employ strings
of small, clear glass bottles, filled with lake water
containing natural plankton communities and incu-
bated at several depths for several hours. In most
older experiments, including my own, the question
of transparency of containers to ultraviolet (UV)
radiation was not addressed.
We now know that the interpretation of such
experiments hadmany flaws. Communities in bottles
incubated near the lake surface are usually less
productive than those at slightly deeper depths. This
was commonly assumed to be ‘‘photoinhibition,’’
caused by exposure to bright sunlight (Vincent and
others 1984). Some early experiments where the
depth of incubation was varied during the experi-
ment (Marra 1978) gave higher results than did
static exposures, but did not trigger widespread
attention to a critical question: how natural were
the exposures? However, when a whole lake at ELA
was spiked with 14C-labeled dissolved inorganic
carbon, and uptake by phytoplankton was com-
paredwith bottles over several days, itwas clear that
photoinhibition of themagnitude indicated by bottle
experiments did not correlate with whole-lake re-
sults (Bower and others 1987).
Epilimnetic mixing, and the resulting exposure of
plankton communities to both photosynthetically
active and UV solar radiation, is a complex process
poorlymimicked by static incubations. For example,
detailed temperature profiles indicate that on days
with moderate to brisk winds, the epilimnions of
lakes at ELA are mixed rather rapidly, so that
exposure to high UV radiation would be brief.
However, consideration of several thousand tem-
perature profiles spanning several orders of magni-
tude in lake area indicates that winds are light
enough to allow near-surface microstratification to
occur on 25%–70% of midsummer days (M. A.
Xenopoulos and D. W. Schindler unpublished).