22. Assessing the Effects of Acidification on Aquatic Ecosystems: Insights from Lake Experiments
333
1997). The duration of such experiments should be
planned carefully to avoid scale-dependent artifacts
(e.g., excessive periphyton growth on mesocosm
walls). Experiments in larger (> 1000 liters) mesocosms can be run for longer periods of time (weeks
or months) while smaller mesocosms « 1000 liters) should be terminated within a couple of
weeks. Systematic considerations of the effects of
mesocosm size are worthy assessments from a variety of perspectives (Petersen et al. 1997) and may
have particular value in evaluating acidification
effects.
Previous studies have employed a variety of experimental treatments to study the effects of acid
on aquatic communities. Acid is generally added to
mesocosms in a dilute form to avoid the temporary
formation of highly acidic patches of water. Typically, sulfuric acid is used to lower pH in mesocosms (but see Barmuta et al. 1990 for an example
of a combination of sulfuric and nitric acids). In
most studies, acid is evenly mixed throughout the
mesocosm. However, Locke and Sprules (1993)
successfully manipulated pH in the epilimnion versus the whole water column of thermally stratified
mesocosms. Interestingly, zooplankton communities responded similarly in the epilimnetic and
whole water column acidification treatments suggesting that a high pH hypolimnetic refuge does not
increase survival of acid-sensitive species. While
most studies reduce pH to target levels in a single
step, Havens (1992) allowed a 7-day acclimation
period during which pH was reduced in small steps.
Fischer (1997) compared zooplankton responses to
a single sustained acidification (PRESS sensu
Bender et al. 1984) versus multiple short-term acidification events (PULSE sensu Bender et al. 1984).
Crustacean zooplankton responded similarly in
PRESS and PULSE treatments, while some shorterlived rotifer species exhibited recovery when the
pH was raised in the PULSE treatment (Fischer,
unpublished data). In all of these studies, interpretation of results was aided by the inclusion of an
unacidified control treatment.
Community responses to acidification in mesocosm experiments occur through direct and indirect
pathways. For example, several studies (Barmuta et
al. 1990; Locke and Sprules 1993; Fischer 1997)
have noted increases in acid-tolerant species such
as Bosmina longirostris following decreases in their
acid-sensitive competitors (e.g., Holopedium gibbe rum and Daphnia spp.). Several different approaches have been used to explore mechanisms
driving community responses to acidification.
Locke and Sprules (1993) used changes in body
size, egg ratio, and lipid status to make inferences
about the role of acid sensitivity and food web interactions in controlling zooplankton responses.
Fischer (1997) applied first order autoregressive
models (Ives 1995) to experimental data to differentiate the role of direct and indirect effects in zooplankton community responses to acidification.
While most mesocosm studies have focused on
community-level responses, Havens (1992) also
monitored the effects of acidification on ecosystem
parameters such as algal carbon assimilation rates
and carbon transfer efficiency from algae to
zooplankton.
Short-term experiments in smaller containers
have the potential to reveal some of the mechanisms operating in responses to acidification. In
particular, they can be used to test the direct response of smaller organisms to acid effects. Results
of such experiments must be interpreted with caution, however. Results of standard laboratory experiments may not be very effective for predicting
community responses at the whole-lake scale. In an
explicit comparison of laboratory bioassays and a
whole-lake acidification experiment, Gonzalez and
Frost (1994) found that bioassays for one rotifer
species correctly predicted the population's decrease in response to acid conditions. However, detrimental effects of acid conditions in bioassays occurred only when acid conditions were combined
with reduced food availability. Rotifers were insensitive to acid conditions if they were fed ad libitum
and incubated in the lake in 4-liter containers for 2
to 3 days, as is standard in many bioassay protocols.
Discrepancies between small-scale and whole-lake
responses were much more striking for a second
rotifer species. Bioassays suggested that this species would decline under acid conditions. Instead,
they increased to dominate rotifer biomass during
the most acidified stage of the experiment (Gonzalez and Frost 1994). These rotifer results were
consistent with the general observation that many
popUlation changes in the acidified lake were
driven by indirect, food-web-related mechanisms
rather than direct responses to acidification (Webster et al. 1992). Similarly, Fischer (1997) working
with other zooplankton taxa quantified population
333
1997). The duration of such experiments should be
planned carefully to avoid scale-dependent artifacts
(e.g., excessive periphyton growth on mesocosm
walls). Experiments in larger (> 1000 liters) mesocosms can be run for longer periods of time (weeks
or months) while smaller mesocosms « 1000 liters) should be terminated within a couple of
weeks. Systematic considerations of the effects of
mesocosm size are worthy assessments from a variety of perspectives (Petersen et al. 1997) and may
have particular value in evaluating acidification
effects.
Previous studies have employed a variety of experimental treatments to study the effects of acid
on aquatic communities. Acid is generally added to
mesocosms in a dilute form to avoid the temporary
formation of highly acidic patches of water. Typically, sulfuric acid is used to lower pH in mesocosms (but see Barmuta et al. 1990 for an example
of a combination of sulfuric and nitric acids). In
most studies, acid is evenly mixed throughout the
mesocosm. However, Locke and Sprules (1993)
successfully manipulated pH in the epilimnion versus the whole water column of thermally stratified
mesocosms. Interestingly, zooplankton communities responded similarly in the epilimnetic and
whole water column acidification treatments suggesting that a high pH hypolimnetic refuge does not
increase survival of acid-sensitive species. While
most studies reduce pH to target levels in a single
step, Havens (1992) allowed a 7-day acclimation
period during which pH was reduced in small steps.
Fischer (1997) compared zooplankton responses to
a single sustained acidification (PRESS sensu
Bender et al. 1984) versus multiple short-term acidification events (PULSE sensu Bender et al. 1984).
Crustacean zooplankton responded similarly in
PRESS and PULSE treatments, while some shorterlived rotifer species exhibited recovery when the
pH was raised in the PULSE treatment (Fischer,
unpublished data). In all of these studies, interpretation of results was aided by the inclusion of an
unacidified control treatment.
Community responses to acidification in mesocosm experiments occur through direct and indirect
pathways. For example, several studies (Barmuta et
al. 1990; Locke and Sprules 1993; Fischer 1997)
have noted increases in acid-tolerant species such
as Bosmina longirostris following decreases in their
acid-sensitive competitors (e.g., Holopedium gibbe rum and Daphnia spp.). Several different approaches have been used to explore mechanisms
driving community responses to acidification.
Locke and Sprules (1993) used changes in body
size, egg ratio, and lipid status to make inferences
about the role of acid sensitivity and food web interactions in controlling zooplankton responses.
Fischer (1997) applied first order autoregressive
models (Ives 1995) to experimental data to differentiate the role of direct and indirect effects in zooplankton community responses to acidification.
While most mesocosm studies have focused on
community-level responses, Havens (1992) also
monitored the effects of acidification on ecosystem
parameters such as algal carbon assimilation rates
and carbon transfer efficiency from algae to
zooplankton.
Short-term experiments in smaller containers
have the potential to reveal some of the mechanisms operating in responses to acidification. In
particular, they can be used to test the direct response of smaller organisms to acid effects. Results
of such experiments must be interpreted with caution, however. Results of standard laboratory experiments may not be very effective for predicting
community responses at the whole-lake scale. In an
explicit comparison of laboratory bioassays and a
whole-lake acidification experiment, Gonzalez and
Frost (1994) found that bioassays for one rotifer
species correctly predicted the population's decrease in response to acid conditions. However, detrimental effects of acid conditions in bioassays occurred only when acid conditions were combined
with reduced food availability. Rotifers were insensitive to acid conditions if they were fed ad libitum
and incubated in the lake in 4-liter containers for 2
to 3 days, as is standard in many bioassay protocols.
Discrepancies between small-scale and whole-lake
responses were much more striking for a second
rotifer species. Bioassays suggested that this species would decline under acid conditions. Instead,
they increased to dominate rotifer biomass during
the most acidified stage of the experiment (Gonzalez and Frost 1994). These rotifer results were
consistent with the general observation that many
popUlation changes in the acidified lake were
driven by indirect, food-web-related mechanisms
rather than direct responses to acidification (Webster et al. 1992). Similarly, Fischer (1997) working
with other zooplankton taxa quantified population