Adaptive reversals in acid tolerance in copepods from lakes recovering from historical stress.

Abstract

Anthropogenic habitat disturbance can often lead to rapid evolution of environmental tolerances in taxa that are able to withstand the stressor. What we do not understand, however, is how species respond when the stressor no longer exists, especially across landscapes and over a considerable length of time. Once anthropogenic disturbance is removed and if there is an ecological trade-off associated with local adaptation to such an historical stressor, then evolutionary theory would predict evolutionary reversals. On the Boreal Shield, tens of thousands of lakes acidified as a result of SO2 emissions, but many of these lakes are undergoing chemical recovery as a consequence of reduced emissions. We investigated the adaptive consequences of disturbance and recovery to zooplankton living in these lakes by asking (1) if contemporary evolution of acid tolerance had arisen among Leptodiaptomus minutus copepod populations in multiple circum-neutral lakes with and without historical acidification, (2) if L. minutus populations were adaptively responding to reversals in selection in historically acidified lakes that had recovered to pH 6.0 for at least 6-8 years, and (3) if there was a fitness trade-off for L. minutus individuals with high acid tolerance at circum-neutral pH. L. minutus populations had higher acid tolerances in circum-neutral lakes with a history of acidification than in local and distant lakes that were never acidified. However, copepods in circum-neutral acid-recovering lakes were less acid-tolerant than were copepods in lakes with longer recovery time. This adaptive reversal in acid tolerance of L. minutus populations following lake recovery was supported by the results of a laboratory experiment that indicated a fitness trade-off in copepods with high acid tolerances at circum-neutral pH. These responses appear to have a genetic basis and suggest that L. minutus is highly adaptive to changes in environmental conditions. Therefore, restoration managers should focus on removing environmental stressors, and adaptable species will be able to reverse evolutionary responses to environmental disturbance in the years following recovery.

Full text

Page 1

Ecological Applications, 17(4), 2007, pp. 1116–1126
? 2007 by the Ecological Society of America
ADAPTIVE REVERSALS IN ACID TOLERANCE IN COPEPODS
FROM LAKES RECOVERING FROM HISTORICAL STRESS
ALISON M. DERRY AND SHELLEY E. ARNOTT
Department of Biology, Queen’s University, Kingston, Ontario K7L 3N6 Canada
Abstract.
environmental tolerances in taxa that are able to withstand the stressor. What we do not
understand, however, is how species respond when the stressor no longer exists, especially
across landscapes and over a considerable length of time. Once anthropogenic disturbance is
removed and if there is an ecological trade-off associated with local adaptation to such an
historical stressor, then evolutionary theory would predict evolutionary reversals. On the
Boreal Shield, tens of thousands of lakes acidified as a result of SO2emissions, but many of
these lakes are undergoing chemical recovery as a consequence of reduced emissions. We
investigated the adaptive consequences of disturbance and recovery to zooplankton living in
these lakes by asking (1) if contemporary evolution of acid tolerance had arisen among
Leptodiaptomus minutus copepod populations in multiple circum-neutral lakes with and
without historical acidification, (2) if L. minutus populations were adaptively responding to
reversals in selection in historically acidified lakes that had recovered to pH 6.0 for at least 6–8
years, and (3) if there was a fitness trade-off for L. minutus individuals with high acid tolerance
at circum-neutral pH.
L. minutus populations had higher acid tolerances in circum-neutral lakes with a history of
acidification than in local and distant lakes that were never acidified. However, copepods in
circum-neutral acid-recovering lakes were less acid-tolerant than were copepods in lakes with
longer recovery time. This adaptive reversal in acid tolerance of L. minutus populations
following lake recovery was supported by the results of a laboratory experiment that indicated
a fitness trade-off in copepods with high acid tolerances at circum-neutral pH. These responses
appear to have a genetic basis and suggest that L. minutus is highly adaptive to changes in
environmental conditions. Therefore, restoration managers should focus on removing
environmental stressors, and adaptable species will be able to reverse evolutionary responses
to environmental disturbance in the years following recovery.
Anthropogenic habitat disturbance can often lead to rapid evolution of
Key words:
copepods; ecosystem recovery; environmental tolerance; fitness trade-off; rapid evolution; local adaptation;
pH; zooplankton.
acidification history; adaptive reversal; anthropogenic disturbance; Boreal Shield lakes;
INTRODUCTION
Anthropogenic impacts can cause rapid evolutionary
change, which in turn can alter genetic and adaptive
properties of populations over a scale of decades
(Palumbi 2001). There are numerous examples of
adaptively diverged populations associated with anthro-
pogenic stressors (Reznik and Ghalambor 2001) that
have potential ecological consequences for populations,
communities, and ecosystems (Hairston et al. 2005).
Restoration managers need to consider contemporary
evolution in degraded and recovering habitats because
application of artificially augmented gene flow from
other habitats to enhance biological recovery can
disrupt the ability of existing populations to adaptively
track environmental changes (Stockwell et al. 2003).
A limited number of studies on recovering ecosystems
have shown evolutionary reversals following decreased
pollution: loss of metal resistance in oligochaetes after
environmental remediation (Levinton et al. 2003) and a
decline in the melanic frequency of moths in post-
industrial England (Cook et al. 2005). However, it is
unclear how populations locally adapted to a particular
stressor will respond once the disturbance is removed
because a genetically fixed trait resulting from local
adaptation to a historical stressor may continue to be
expressed even after an environmental change causes it
to lose its advantage (Stearns 1994). Also, reversals in
the direction of selection may not necessarily return a
population to its earlier state (Grant and Grant 2002).
Although certain species may be able to cope adaptively,
population declines associated with environmental
change are most often associated with local extinction
(Ricciardi and Rasmussen 1999).
Acid deposition and its interactions with climate
change, land use, and stratospheric ozone depletion have
Manuscript received 15 August 2006; revised 3 November
2006; accepted 14 November 2006. Corresponding Editor: J. S.
Baron.
1116

Page 2

had huge regional effects, particularly on boreal lakes
(Schindler 1998) and oceans (Orr et al. 2005). Lake
acidification caused losses of hundreds of thousands of
fish and invertebrate populations (Minns et al. 1990).
One exception was a copepod species, Leptodiaptomus
minutus Lilljeborg, one of the most common crustacean
zooplankton on the Boreal Shield (Carter et al. 1980; see
Plate 1). This species can be found in shield lakes across
a broad pH range (pH 4–7) and tends to dominate
zooplankton communities in acid-damaged lakes
(Sprules 1975). There is evidence from an experimentally
acidified lake in Wisconsin, USA, that some populations
of L. minutus may evolve increased tolerance to low pH
conditions (Fischer et al. 2001). However, we do not
know if this is a response that can be generalized to the
thousands of lakes that have been damaged by regional
acidification.
As lakes damaged by acid precipitation chemically
recover with declining SO2 emissions across North
America and Europe (Stoddard et al. 1999, Jeffries et
al. 2000), we do not know if there will be evolutionary
responses to reversals in selection forces, i.e., will
organisms lose their tolerance to low pH? Copepod
populations may have undergone rapid local adaptation
following atmospheric acidification of lakes on the
Canadian Shield, as they did in Little Rock Lake,
Wisconsin (Fischer et al. 2001), but they may now be
facing reverse selection as lakes recover to circum-
neutral pH, especially if there is a cost associated with
increased acid tolerance. Although biological recovery is
one of the ultimate goals for SO2 emission control
programs (Meteorological Service of Canada 2004),
knowledge about genetic aspects of acidification recov-
ery and its potential significance is scarce.
The objectives of this study were to examine patterns
of acid tolerance in L. minutus in lakes that differ in
acidification history and explore evolutionary reversals
in recovering lakes. First, we tested whether or not rapid
evolution of acid tolerance had arisen among L. minutus
populations in multiple circum-neutral lakes with and
without historical acidification in a boreal landscape by
using field transplant experiments. We included replicate
lakes from both local and regional spatial scales to
determine if dispersal among local lakes had homoge-
nized acid tolerance adaptation across the local land-
scape. Theoretical work suggests that dispersal among
habitats promotes plasticity over local adaptation, by
mixing gene pools and broadening environmental
tolerances of locally specialized genotypes (Sultan and
Spencer 2002), which would result in a uniform response
from different populations among lakes.
Next, we asked if there have been evolutionary
reversals in acid tolerance after several years of lake-
water recovery to pH 6.0. We used a second field
transplant experiment to test if there are differences in
acid tolerances among L. minutus populations between
circum-neutral acid-recovering lakes and anthropogenic
and naturally acidic lakes. The circum-neutral lakes used
in this experiment had varying numbers of years since
recovery to pH 6.0, allowing us to investigate the
relationship between acid tolerance and years since
recovery. If reversals in the direction of evolution of acid
tolerance had occurred, we expected to see a decreasing
acid tolerance associated with longer times since
recovery to pH 6.0.
Finally, we investigated if there is a fitness trade-off
associated with adaptation of high acid tolerance as
acidified lakes recover to pH 6.0. We conducted a
laboratory experiment that compared survival in cope-
pods with high acid tolerance to those with low acid
tolerance, at both circum-neutral and low pH.
METHODS
Study site
Killarney Provincial Park, Ontario, Canada (468010
N, 818240W) contains .600 lakes with pH ranging from
4.3 to 7.6 (Snucins et al. 2001). This wide range in pH is
the result of both natural pH gradients and anthro-
pogenically induced pH gradients associated with SO2
emissions from nearby metal smelters during the mid-
1900s and the long range transport of acid deposition.
The circum-neutral buffered lakes contain limestone and
sandstone in their watersheds (Debicki 1982), and
therefore, high concentrations of base cations that are
able to neutralize acid deposition. Bog lakes have low
pH (?5.3) resulting from organic acids from associated
wetlands. Anthropogenically acidified lakes were histor-
ically circum-neutral lakes (Dixit et al. 2002) not
associated with limestone and sandstone and therefore,
were highly susceptible to acidification (Sprules 1975).
Many lakes in Killarney Park are starting to recover to
circum-neutral pH (pH ? 6.0) with declining SO2
emissions, but recovery rates are dependent on water-
shed properties, hydrology, and probably the degree of
acidification (Snucins et al. 2001, Keller et al. 2003). We
chose to use pH 6.0 as a threshold for recovery of L.
minutus because it is a widely used measure for
biological recovery of acid-sensitive species in the
acidification recovery literature (Keller et al. 2002). In
Killarney Provincial Park, therefore, there are a mosaic
of lakes with different acid-stress histories that we
classified as: (1) circum-neutral lakes that did not acidify
with regional acid deposition; (2) circum-neutral, acid-
recovering lakes that had a pH ? 5.2 during the 1970s
but have had a pH . 6.0 for at least six years, (3) acid-
damaged lakes that were acidic during the 1970s because
of regional acid deposition and continue to have a pH ?
5.2; and (4) bog lakes with pH ? 5.3 (Sprules 1975, Dixit
et al. 2002).
Experiment 1: field transplant with circum-neutral lakes
We hypothesized that greater acid tolerance had
arisen in copepod populations from circum-neutral
lakes with a history of acidification than in nearby and
regionally distant (.200 km, near Dorset, Ontario) un-
acidified lakes. To test this, we conducted a factorial-
June 2007 1117 ADAPTIVE REVERSALS IN ACID TOLERANCE

Page 3

design field transplant experiment in circum-neutral
Carlyle Lake, Killarney Park, Ontario (see Plate 1). We
manipulated the following factors: zooplankton acidifi-
cation history (three levels: buffered, acid-recovering,
and un-acidified, regionally distant Dorset), water
source (two levels: lake water from Carlyle Lake and
water from each of the source lakes from which the
zooplankton were collected [‘‘home lake water’’]), and
pH treatment (three levels: circum-neutral pH 6.2–7.6,
5.4, and 4.7). This experimental design was not
completely balanced: n ¼ 3 buffered Killarney lakes, n
¼ 4 acid-recovering Killarney lakes, and n ¼ 3 un-
acidified and regionally distant lakes near Dorset,
Ontario. These lakes have a current and historical
presence of the copepod L. minutus (Sprules 1975, Keller
et al. 2002). The physicochemical properties of these
habitats are described in Table 1. All of the lakes had
long-term records of circum-neutral lake-water pH ’
6.0, except for the acid-recovering lakes, which had been
circum-neutral for at least six years previously: Bell
Lake (since 1996), George Lake (since 1998), Johnnie
Lake (since 1998), and Carlyle Lake (since 1998) (Keller
et al. 2003).
Zooplankton from each of the 10 study lakes were
stocked at 23 ambient lake densities in 66 closed,
translucent 10-L polyethylene containers. A mixed
zooplankton assemblage was stocked into these con-
tainers to minimize logistical complications, and was not
expected to influence direct effects of acid addition on L.
minutus response since indirect community effects
require longer than a week (Webster et al. 1992).
Zooplankton were collected with a 15 cm diameter
closing plankton net (80-lm mesh) at a 1.5-m depth
interval just below 1% UVB attenuation in each source
lake. One-percent UVB-attenuation depths (Morris et
al. 1995, Table 1) were calculated from dissolved organic
carbon (DOC) concentrations for each of the study lakes
to ensure that (1) zooplankton were collected from a
sufficient depth interval as to account for vertical
migration that might result from UVB avoidance, and
(2) containers were incubated at sufficient depth to
exclude UVB effects in Carlyle Lake. One zooplankton
haul from each lake was preserved (5% sugar-formalin)
to estimate initial zooplankton densities stocked into
containers. Containers were suspended from floating
frames in a 10 m deep basin in Carlyle Lake and
incubated at 1.5–2.0 m depths at 208C (YSI 95
O2/temperature meter; YSI, Yellow Springs, Ohio,
USA) for the seven-day period from 13 August to 20
August 2004.
While 33 of the containers were filled with water from
Carlyle Lake, the remaining 33 containers were filled
with home lake water. Water was collected from the top
5 m of the water column using a battery-operated pump
TABLE 1.
experiments.
Selected physicochemical characteristics of Canadian study lakes included in field transplant and laboratory
Lake history Lake name
Lat.
(N)
Long.
(W)
Max.
depth (m)
Volume
3 104(m3)
Watershed
area (ha)pH
DOC
(mg/L)
TP
(lg/L)
1% UVB
attenuation
depth (m)
Regional
Un-acidified
(Dorset)
Local
Buffered
(Killarney)
Bigwind?
Blue Chalk?
Red Chalk?
Helen
Ishmael
Low
Teardrop§
Bell (C1, H1)
Carlyle (C1)
Davidjj (C2)
George (H1, C2)
Johnniejj
(H1, C1, C2)
Little Superior}
OSAjj
Proulx}
Heaven??
Lake of Woods#
Turbid}
458030
458110
458110
468060
468060
46806’
468020
468080
468050
468080
468010
468050
788040
788560
788560
818330
818350
818330
818240
818110
818140
818170
818240
818140
32.0
23.0
38.0
41.2
19.8
28.4
16.6
26.8
14.6
24.4
36.6
33.5
1180
447
811
1692
820
483
507
158
589
883
1243
1000
13
8596
1059
1915
5717
12952
6.8
6.9
6.7
6.9
7.0
7.6
6.8
6.3
6.3
5.8
6.2
6.1
4.0
2.4
4.6
4.1
4.1
3.0
1.1
5.5
4.1
1.9
2.4
3.6
6.5
3.1
3.2
7.0
5.8
3.4
8.0
5.4
5.2
3.0
5.0
5.0
0.4
0.7
0.4
0.4
0.4
0.6
1.5
0.3
0.4
0.9
0.7
0.5
32
Local
Acid-recovering
(Killarney)
2827
891
2857
3086
3418
Local
Acid-damaged
(Killarney)
Local
Bog
(Killarney)
468040
468030
468040
468040
468060
468060
818190
818240
818190
818170
818120
818110
33.5
39.7
28.7
17.8
6.0
9.4
179
3341
119
35
879
42
14
67
1136
4.2
4.7
4.4
4.6
5.3
5.2
0.2
0.3
0.2
3.1
3.3
3.2
4.0
2.4
3.6
5.4
4.1
5.4
0.6
0.5
0.5
9
,6.0
,8.0
16.0
24
69
Notes: Lake-water pH was measured at the time of experiments; 2004 midsummer dissolved organic carbon (DOC) and total
phosphorus (TP) data were measured by authors for all lakes except where indicated. In parentheses after acid-recovering lake
names, letters represent water source treatment (C, Carlyle water; H, home water), and numbers represent experiment (experiment
1 vs. 2). Physical characteristics of study lakes are from Reid et al. (1987) for Blue Chalk and Red Chalk lakes, from Girard et al. (in
press) for Bigwind Lake, and from Snucins and Gunn (1998) for the Killarney lakes.
? 2004 midsummer chemistry data (Ontario Ministry of the Environment, unpublished data).
? 2003 midsummer chemistry data (Ontario Ministry of the Environment, unpublished data).
§ 2001 midsummer DOC and TP (Ontario Ministry of the Environment, unpublished data).
jj 2004 midsummer pH data (Ontario Ministry of the Environment, unpublished data).
} 2000 midsummer DOC and TP data (Holt and Yan 2003).
# 1997 winter chemistry data (Snucins and Gunn 1998).
?? 2003 winter DOC (Ontario Ministry of the Environment, unpublished data) and 1997 winter TP (Snucins and Gunn 1998).
ALISON M. DERRY AND SHELLEY E. ARNOTT1118
Ecological Applications
Vol. 17, No. 4

Page 4

in Carlyle Lake and an integrated tube sampler (5 m
long; 2.5-cm internal diameter) in each of the other
study lakes. We filtered the water through an 80-lm
mesh to remove crustacean zooplankton, but allowed
rotifers, phytoplankton, and bacteria to pass through
into the containers. Incubation of L. minutus in both
Carlyle Lake water and home lake water enabled us to
differentiate between the possible effects of algal food
quality and quantity and potential effects from local
adaptation to home-water conditions.
Three acid treatments (circum-neutral pH 6.2–7.6, 5.4,
and 4.7) were imposed using 1 mol/L sulfuric acid for
each acidification history and water source treatment
combination. The pH in each container was measured
using a WTW inoLab 740 pH meter (WTW, Woburn,
Massachusetts, USA). Two replicates of the buffered
home-water treatments (Ishmael and Low lakes) and
one replicate of the acid-recovering lake history treat-
ment in Carlyle water (George Lake) were accidentally
not acidified to pH 4.7. Otherwise, pH treatments were
all within 60.2 units of the target pH. Therefore,
although there were 66 containers in the experiment,
data from 63 containers were statistically analyzed.
At the end of the experiment, 9 L of water from each
container was filtered through 80-lm mesh to collect
zooplankton that were subsequently preserved in 5%
sugar-formalin. The remaining l L of water from each of
the containers was 80-lm filtered to remove zooplank-
ton, and then a 0.4-L subsample was analyzed for pH
(WTW inoLab 740 pH meter). An estimate of total
chlorophyll a was obtained by collecting phytoplankton
from a subsample of 0.3 L on a Fisherbrand 42.5 mm
diameter, 1.2-lm G4 glass fiber filter (Fisher Scientific
Canada, Nepean, Ontario, Canada). A measure of
edible chlorophyll a was obtained by first filtering a
0.3-L subsample through a 30-lm mesh and then
collecting the phytoplankton from the filtrate on a.1.2-
lm glass fiber filter since copepods graze on highly
edible algae ,30–35 lm (Cyr and Curtis 1999).
Adult L. minutus and fifth-stage calanoid copepodites
were counted from initial zooplankton samples (Ni) and
from samples taken from each container at the end of
the experiment (Nf) under a Leica MZ16 dissecting
microscope (Leica Microsystems [Canada], Richmond
Hill, Ontario, Canada). Final abundance (Nf ) was
sometimes greater than the initial abundance (Ni) due to
maturation of nauplii and early-stage copepodites. Only
copepods that had been alive or recently alive at the time
of formalin preservation were counted. Individuals that
exhibited signs of decay or incomplete body structure
were not counted.
Experiment 2: field transplant with acidified lakes
Our second field transplant experiment compared acid
tolerance of L. minutus populations from circum-neutral
acid-recovering lakes with anthropogenic acid-damaged
and naturally acidic bog lakes in Killarney Park (Table
1). If L. minutus had responded to a reversal in selection
in acid-recovering lakes, then we expected to find a
lower acid tolerance in acid-recovering lakes than in
acid-damaged or bog lakes. To test this hypothesis, we
used a 3 3 4 factorial design manipulating zooplankton
acidification history (three levels: from acid-recovering,
acid-damaged, and bog lakes) and pH treatment (three
levels: circum-neutral ?6.0, 5.4, 4.7, and 3.8). Epilim-
netic zooplankton from three lakes in each of the
acidification history categories were collected and
stocked into 36 closed, translucent 10-L polyethylene
containers on 18 June 2005. Because of low abundances
of L. minutus in Bell and Carlyle lakes at the time of the
experiment in 2005, circum-neutral acid-recovering lakes
were comprised of George, Johnnie, and David (pH of
5.6 as of 2005) lakes. Zooplankton were incubated in
only Carlyle Lake water for this experiment.
Methods for zooplankton collection and stocking
paralleled the approach described previously for exper-
iment 1. Containers were suspended from wooden
frames and incubated at 1.5–2.0 m depths at 248C in
acid-recovering Carlyle Lake as a common garden
experiment for seven days. Four acid treatments within
each of the ‘‘acidification history’’ treatments were set
based on prior titrations with 1 mol/L sulfuric acid and
were maintained within 60.2 units of the target pH. As
with the first transplant experiment, an initial zooplank-
ton sample from each lake and a final zooplankton
sample from each container were preserved in 5% sugar-
formalin.
Experiment 3: laboratory assay for chronic pH effects
Finally, we investigated if there is a trade-off
associated with high acid tolerance in L. minutus at
circum-neutral pH. A 32-day two-factor laboratory
experiment was conducted for L. minutus individuals
collected from two side-by-side Killarney lakes with
distinct acidification histories and lake-water pH:
circum-neutral, buffered Teardrop Lake and anthropo-
genic acid-damaged OSA Lake (Table 1). Copepods in
each lake-type were subjected to two pH levels: neutral
(pH 6.7 6 0.3) and acidic (pH 4.7 6 0.2). Zooplankton
were collected from each lake using an 80-lm plankton
net, placed in 2-L Nalgene containers in coolers, and
transported to the laboratory at Queen’s University,
Kingston, Ontario. Females without eggs and males (20
adults per replicate) were placed in 100 mL of 0.2-lm-
filtered Carlyle Lake water for each replicate (n ¼ 3).
Algal food was a 1:1 mix of Chlamydomonas reinhardtii
and Cryptomonas erosa (University of Toronto Culture
Collection of Algae and Cyanobacteria, Department of
Botany, University of Toronto, Toronto, Ontario,
Canada) and concentrations ranged from 3 to 6 3 103
cells/mL. Previous feeding experiments had indicated
that survival was highest at this food concentration
under neutral pH conditions (P. Lam, unpublished data).
Water and algae were refreshed, pH was adjusted using
1 mol/L sulfuric acid, and nauplii were removed every 3
to 5 days. Copepod and algal cultures were incubated at
June 20071119ADAPTIVE REVERSALS IN ACID TOLERANCE

Page 5

208C under 16:8 h of light:dark photoperiod in a
Percival model 1-36LLVL incubator (Percival Scientific,
Perry, Iowa, USA). Surviving adults were counted on
days 5, 13, 22, and 32 of the experiment under Leica
MZ16 dissecting microscopes.
Statistical analyses
For each of experiments 1 and 2, initial (Ni) and final
(Nf) L. minutus copepod abundances were ln(x þ 1)-
transformed to normalize data distributions and ho-
mogenize variance. These data were analyzed with
factorial-design (lake history 3 water source 3 pH)
analyses of covariance (ANCOVAs) with Ni as a
covariate and Nfas a dependent response variable. A
covariate (Ni) was used to control for differences in
initial stocking densities among lakes rather than
standardization by proportions (Nf/Ni) because propor-
tions can cause changes in the covariance and correla-
tion structure of the data matrix that can result in
different statistical outcomes compared to raw abun-
dances (Jackson 1997). For experiment 1, assumptions
of normality were met, but equal variance was violated.
Although the effect of violating these ANCOVA
assumptions is a greater risk of accepting a false null
hypothesis (Type I error) (Rheinheimer and Penfield
2001), we were able to reject H0for the interaction of
interest (lake history3pH: see Results). For experiment
2, all assumptions were met except for homogeneity of
regression slopes. There was a significant effect of lake
history on the covariate Ni because initial stocking
densities (Ni) were higher in bog (27.2 6 0.5 L.
minutus/L; mean 6 SE) and acid-damaged lakes (9.9
6 0.05 L. minutus/L) than in acid-recovering lakes (3.4
6 0.13 L. minutus/L) (factorial ANOVA on Ni; lake
history main effect, F ¼ 10.49, df ¼ 2, 24, P , 0.01;
Tukey’s hsd tests, P , 0.01). Violation of homogeneity
of regression slopes in an ANCOVA can increase the
probability of accepting a false H0. We conducted a
factorial ANOVA on ln(Nf þ 1) without Ni as a
covariate and found that the ANCOVA results for
experiment 2 were supported by an analysis that was not
based on this assumption.
To test if final chlorophyll a concentrations and
conductivity measurements varied among treatments in
experiments 1 and 2, we used factorial ANOVAs. For
experiment 1, both normality and equal variance
assumptions were violated for total chlorophyll a,
despite log-transformation, but variances were equal
for log-transformed edible chlorophyll a. All ANOVA
assumptions were met for log-transformed total and
edible chlorophyll a concentrations in experiment 2.
To investigate the relationship between copepod acid
tolerance and time since recovery among acid-recovering
lakes, copepod survival (ln(Nf/Ni)) at pH 4.7 from acid-
damaged and acid-recovering lake history categories in
experiments 1 and 2 was linearly regressed on length of
time since lake recovery to circum-neutral conditions.
Regression assumptions of independence, normality,
and equal variance were met.
For experiment 3, log(x þ 1)-transformed number of
alive adult copepods were analyzed with a multivariate
repeated-measures ANOVA (RMA) to examine both
among- and within-treatment effects. These results were
presented despite the fact that most normality and
equality of variance assumptions were violated because
RMA is robust when sample sizes are equal. Sphericity
assumptions were met.
All ANCOVAs and ANOVAs were followed by
Tukey’s hsd post hoc comparisons. Statistical analyses
were conducted with Statistica 6.0 software (Statsoft,
Tulsa, Oklahoma, USA).
RESULTS
Experiment 1: field transplant with circum-neutral lakes
Leptodiaptomus minutus final abundance varied both
among lake-acidification history categories and pH
treatments (Table 2). Copepod abundance was lowest
in the pH 4.7 treatments (ANCOVA; Tukey’s hsd tests,
P , 0.01). The effect of pH treatment on copepod final
abundance, however, was dependent on lake history, as
indicated by a significant lake history 3 pH interaction
(Table 2, Fig. 1). Within the acid-recovering lake history
category, L. minutus final abundance was not lower at
pH 4.7 than at pH 5.4 or circum-neutral pH (Fig. 1).
However, there were significant declines in L. minutus
density at pH 4.7 within circum-neutral buffered and
regionally un-acidified Dorset lake history categories
compared with pH 5.4 and circum-neutral pH treat-
ments (ANCOVA; Tukey’s hsd test, P , 0.01; Table 2,
Fig. 1). At pH 4.7, L. minutus individuals had greater
final abundance in the acid-recovering lake history
treatment than in the circum-neutral buffered or the
regionally un-acidified (Dorset) lake history treatment
TABLE 2.
iments 1 and 2 for Leptodiaptomus minutus final abundance
(Nf) with initial abundance (Ni) as a covariate.
Results of ANCOVAs from field transplant exper-
Experiment and factordfFP
1) Circum-neutral lakes
Covariate (Ni)
Lake history (LH)
Water source (WS)
pH
LH 3 WS
LH 3 pH
WS 3 pH
LH 3 WS 3 pH
Error
1
2
1
2
2
4
2
4
74.27
5.56
2.11
63.66
0.70
4.55
0.83
0.60
,0.01
,0.01
0.15
,0.01
0.50
,0.01
0.44
0.67
46
2) Acidified lakes with different histories
Covariate (Ni)
Lake history (LH)
pH
LH 3 pH
Error
1
2
3
6
15.76
0.67
34.82
0.43
,0.01
0.52
,0.01
0.85
23
Note: Boldface P values indicate significant main and
interaction effects (P , 0.05).
ALISON M. DERRY AND SHELLEY E. ARNOTT1120
Ecological Applications
Vol. 17, No. 4