Influences of Acid Mine Drainage and Thermal Enrichment on Stream Fish Reproduction and Larval Survival
Andrew W. Hafs, Christopher D. Horn, Patricia M. Mazik, and Kyle J. Hartman
Northeastern Naturalist, Volume 17, Issue 4 (2010): 575–592
Full-text pdf (Accessible only to subscribers.To subscribe click here.)
Access Journal Content
Open access browsing of table of contents and abstract pages. Full text pdfs available for download for subscribers.
Current Issue: Vol. 30 (3)
Check out NENA's latest Monograph:
Monograph 22
2010 NORTHEASTERN NATURALIST 17(4):575–592
Influences of Acid Mine Drainage and Thermal Enrichment
on Stream Fish Reproduction and Larval Survival
Andrew W. Hafs1,*, Christopher D. Horn2, Patricia M. Mazik3,
and Kyle J. Hartman1
Abstract - Potential effects of acid mine drainage (AMD) and thermal enrichment on
the reproduction of fishes were investigated through a larval-trapping survey in the
Stony River watershed, Grant County, WV. Trapping was conducted at seven sites
from 26 March to 2 July 2004. Overall larval catch was low (379 individuals in 220
hours of trapping). More larval White Suckers were captured than all other species.
Vectors fitted to nonparametric multidimensional scaling ordinations suggested that
temperature was highly correlated to fish communities captured at our sites. Survival
of larval Fathead Minnows was examined in situ at six sites from 13 May to 11 June
2004 in the same system. Larval survival was lower, but not significantly different
between sites directly downstream of AMD-impacted tributaries (40% survival) and
non-AMD sites (52% survival). The lower survival was caused by a significant mortality
event at one site that coincided with acute pH depression in an AMD tributary
immediately upstream of the site. Results from a Cox proportional hazard test suggests
that low pH is having a significant negative influence on larval fish survival in
this system. The results from this research indicate that the combination of low pH
events and elevated temperature are negatively influencing the larval fish populations
of the Stony River watershed. Management actions that address these problems
would have the potential to substantially increase both reproduction rates and larval
survival, therefore greatly enhancing the fishery.
Introduction
Early life stages (ELS) of fishes are highly sensitive to perturbations in the
surrounding environment, both natural and anthropogenic (Henry et al. 1999,
Houde 1989b, Mion et al. 1998, Sandstrom et al. 1997). Consequently, the
strength of any year class and the overall population size are subject to events
and conditions that influence survival of those ELS. In impaired systems, survival
may be low, and entire year classes may be absent from the population
(Leis and Fox 1994). Several successive years of extremely low survival may
result in the extirpation of species from an impaired system (McCormick et al.
1989). Reduced fish populations in lotic systems may result from the inability
of ELS fish to survive anthropogenically induced stressors. When multiple
anthropogenic stressors are present in a system, the effects of each individual
stressor may be uncertain. In coal-bearing regions, coal-fired power plants are
often located close to mines to minimize coal transportation costs. As a result,
1West Virginia University, Division of Forestry and Natural Resources, Morgantown,
WV 26506. 2Montana Fish, Wildlife, and Parks, Thompson Falls, MT 59873. 3US
Geological Survey, West Virginia Cooperative Fish and Wildlife Research Unit,
WVU, Morgantown, WV 26506. *Corresponding author - ahafs@mix.wvu.edu.
576 Northeastern Naturalist Vol. 17, No. 4
in areas with bituminous coal seams, thermal enrichment from power plants
and acid mine drainage (AMD) often occur simultaneously, and the combined
disturbance on the biota may be unclear.
Thermal enrichment has been shown to negatively influence successful
spawning and recruitment in various temperate fishes through several
mechanisms. Oogenesis in fishes can occur abnormally early in thermally
enriched systems, resulting in egg re-absorption and/or degradation before
spawning occurs (Luksiene et al. 2000). At high temperatures, fertilized
eggs may metabolize energy stores before hatching can occur (Sandstrom et
al. 1997). This hyper-metabolism can also impact larvae, causing yolk-sac
absorption before external feeding components are fully developed (Houde
2002). Additionally, spring-spawning temperate fishes often spawn early in
thermally enriched systems (Cooke et al. 2003, Paller and Saul 1996, Sandstrom
et al. 1995), exposing ELS fish to environmental conditions that may
reduce survival. These include high river discharges that can scour young
fish from nursery habitats, acid pulses from snow melt, and thermal shock
caused by rapidly changing temperatures (Mion et al. 1998, Shuter et al.
1980). Also, spawning is often timed with the availability of an ephemeral
food source (i.e., plankton bloom), and larvae that hatch out of synchrony
with this food source can experience starvation (Houde 2002).
Acid mine drainage streams have three general characteristics that
can reduce ELS fish survival: increased acidity, elevated dissolved ion
concentrations (including metals), and metal precipitates (particularly
iron and aluminum hydroxides) (Gray 1996). Severity of AMD varies
widely within and among impacted streams, but all can negatively affect
fish (Gensemer and Playle 1999). Direct acid toxicity occurs by interfering
with ion regulation at the gills (McDonald et al. 1989, Verbost et al.
1995). Excess H+ ions compete with essential ion exchange (Na+ , Cl- , and
Ca2+), leading to ion imbalance and, in extreme cases, death (Gensemer
and Playle 1999, Potts and McWilliams 1989). In AMD waters, metal ion
toxicity functions similarly to, and occurs in conjunction with, acid toxicity,
causing ionoregulatory malfunctions (McDonald et al. 1989). The most
detrimental characteristic of mild AMD to fishes is the polymerization
and precipitation of aluminum onto the gills, which leads to gill necrosis
and asphyxia (Henry et al. 1999, Verbost et al. 1995, Witters et al. 1996).
Aluminum precipitation is particularly lethal in mixing zones of acidic
and circumneutral waters, and is often more toxic than just high acidity in
streams alone (Henry et al. 1999, Poleo et al. 1994, Witters et al. 1996).
In systems with both thermal and AMD pollutions, synergism may reduce
larval fish survival. Aluminum toxicity can increase with increasing
temperature as elevated rates of aluminum speciation and polymerization
occur at higher temperatures (Gensemer and Playle 1999, Lydersen et al.
1990, Poleo et al. 1991). In systems influenced by both AMD and thermal
elevation, mortality events may occur more quickly and at lower aluminum
and acid concentrations than they would under non-elevated temperatures.
2010 A.W. Hafs, C.D. Horn, P.M. Mazik, and K.J. Hartman 577
The objective of this study was to quantify and assess the influences of
AMD and thermal enrichment on the stream reproduction of fishes and larval
survival. We hypothesized that the effects of multiple stressors (AMD and
temperature) would be greater than individual effects on larval fish survival.
To evaluate this hypothesis, we conducted a larval survey and in situ larval
fish assays along an environmental gradient representative of both thermal
and AMD disturbances in the same watershed.
Field-site Description
The Stony River located in Grant County, WV (Fig. 1) is a high-elevation
(650–1200 m), high-gradient (≈1.4%) system that drains into the North
Branch Potomac River. Both thermal and AMD influences are present in the
watershed (Horn 2005). Recent surveys in the system have shown that fish
populations are small and species diversity is low (Hoar 2005), likely due to
anthropogenic disturbances. Species captured during electrofishing surveys
by Hoar (2005) included Campostoma anomalum (Rafinesque) (Central
Figure 1. Sites in the Stony River watershed utilized during larval fish studies. All
sites except MSR4 were used in both trapping and in situ bioassays. MSR4 was used
only for trapping.
578 Northeastern Naturalist Vol. 17, No. 4
Stoneroller), Cyprinella spiloptera (Cope) (Spotfin Shiner), Semotilus atromaculatus
(Mitchill) (Creek Chub), Cottus bairdi Girard (Mottled Sculpin),
Etheostoma flabellare Rafinesque (Fantail Darter), Ictalurus punctatus
(Rafinesque) (Channel Catfish), Lepomis cyanellus Rafinesque (Green Sunfish), Micropterus dolomieu Lacepède (Smallmouth Bass), and Micropterus
salmoides (Lacepède) (Largemouth Bass).
A coal-fired electric generating station obtains cooling water from an impoundment
of the river (Mt. Storm Lake), and heated discharge from the lake
elevates the thermal regime in the Stony River year-round. Lake discharges
are highly variable and cease periodically because of low water levels,
thereby decreasing river temperatures. Subsequent precipitation events can
increase lake levels, inducing a thermally elevated discharge, resulting in
temperature increases in the Stony River. Moving downstream, the severity
of thermal discharge lessens with the entry of groundwater and surfacewater
tributaries, but water temperatures do remain elevated throughout the
river (≈20 km) to the confluence with the North Branch Potomac River on
the West Virginia and Maryland border.
Three tributaries downstream of the dam are AMD influenced, with two
of the three being treated prior to confluence with the Stony River. For the
sake of simplicity, all mine-influenced tributaries (treated or not) are termed
AMD from here on, even though chemical conditions may not be similar
to untreated AMD water. The first tributary to enter is the Laurel Run
mine outfall. This water is treated to reduce acidity and dissolved metals,
and treatment is effective and consistent. The next downstream tributary,
Fourmile Run, undergoes AMD treatment, but treatment is highly variable,
and the tributary is prone to large fluctuations in pH and dissolved metal
concentrations. Fourmile Run creates the most noticeable and severe AMD
conditions of any tributary to the Stony River. The third AMD tributary,
Laurel Run, is not treated and remains acidic (pH 4.0–6.0) year round, but
is relatively small and has less influence on the water quality of the Stony
River than does Fourmile Run.
Methods
The following description of study sites is in upstream to downstream
order (Fig. 1). Sites MSR0 and MSR1 are nearest the dam, and are influenced
by thermal enrichment only. Site MSR2 is downstream of the first
AMD input (Laurel Run mine outfall). The next site, 4M2, is downstream of
the AMD tributary Fourmile Run. Continuing downstream, LR2 is directly
downstream of the small AMD tributary, Laurel Run. The remaining sites,
MSR3 and MSR4, represent diminishing thermal and AMD influences along
the river continuum.
Continuous data-collection units (Datasonde 4 multi-probe hydrolab,
Hach Environmental, Loveland, CO) were maintained at all seven of the
sites mentioned above by Dominion Environmental, the environmental
quality subsidiary of Dominion Resources, Inc., Richmond, VA, which
2010 A.W. Hafs, C.D. Horn, P.M. Mazik, and K.J. Hartman 579
owns the power plant. These units recorded in-stream temperature, pH,
and specific conductivity hourly. Our experiments were conducted within
200 m of these data-collection units to allow comparison of results with
water quality parameters.
Larval trapping survey
A larval trapping survey was undertaken in 2004 to determine the timing
and relative reproductive success of fishes in the Stony River (below
Mt. Storm Lake). Larvae were captured with shallow-water quatrefoil light
traps (Aquatic Research Instruments, Lemhi, ID) using a green, six-inch
Cyalume® chemical lightstick (Omniglow Corp., West Springfield, MA) as
a larval attractant during nighttime (Kissick 1993). Previous research by
Gerhke (1994) showed that green light sticks glowed brighter during the first
hour of sampling than other colored light sticks and produced sufficient light
to attract fish larvae. Chemical light traps have been reported to sample larval
fishes as efficiently as electric light traps (Gerhke 1994, Kissick 1993).
Trapping was conducted from 26 March to 2 July 2004 at all seven sites
mentioned previously. Over this period, ten trapping events occurred. During
each trapping event, two traps were set at each site in areas of low water
velocity and depth of >0.5 m (pool-type habitats). Traps were activated sequentially
(beginning at 2200 hours), remained active for approximately two
hours (range = 1.83–2.20 hours), and were retrieved in the same sequence
as set. Captured larvae were preserved in 10% buffered formalin (Floyd
et al. 1984) and later measured (total length) and identified to species in
the laboratory using the methods described by Auer (1982). Because light
emitted from chemical light sticks diminishes over time, capture efficiency
bias was minimized by keeping the trapping-effort times consistent. Other
factors, such as water turbidity, could also affect catch efficiency of light
traps. However, turbidity was similar across sites on each sample date, so
turbidity effects on capture efficiency would be shared across sites for a trapping
event. Temperature and water quality data for the trapping period were
obtained from Dominion Environmental’s data loggers.
The structure of the larval fish community data for all seven sites was
analyzed by a nonparametric multidimensional scaling (NMDS) ordination
technique (McCune and Grace 2002). Because effort was similar across all
sites (range = 31.57–32.09 hrs), total catch was used for NMDS ordination. The
vegan package (Oksanen et al. 2008) in program R was used to run the NMDS.
Sites that plot close together in the ordination space are more similar than sites
that plot farther apart (Merovich and Petty 2007). To assess the influences of
thermal and AMD enrichment on reproduction of stream fishes, function Envfit
(part of the vegan package in program R) was used to fit vectors for summary
statistics (minimum, maximum, standard deviation, and average) of each environmental
variable (pH, temperature, and conductivity) and plot them onto
the NMDS ordination. The number of permutations was set to 10,000, and only
vectors with P-values < 0.05 were plotted. Bray-Curtis distance metrics were
used for all NMDS ordinations (Hawkins and Norris 2000).
580 Northeastern Naturalist Vol. 17, No. 4
In situ bioassay
To examine larval fish survival, a four-week in situ bioassay was conducted
from 13 May to 11 June 2004. Six of the seven described sites were
chosen for bioassays: MSR0, MSR1, MSR2, 4M2, LR2, and MSR3. Each of
these sites represents a point of change in the environmental gradient along
the Stony River continuum and has an associated in-stream water quality
data-collection unit.
Pimephales promelas Rafinesque (Fathead Minnows) were chosen as
a representative warmwater fish for the bioassays because they are a standardized
test organism that is endorsed by the US EPA (Milam et al. 2000).
Fertilized Fathead Minnow eggs were obtained from the US EPA laboratory
in Cincinnati, OH. Upon arrival, eggs were thermally acclimated, then transferred
to water from the site MSR1 (temperature = 24 °C, pH = 7.5, specific
conductivity = 160 μS/cm3) for hatching in the laboratory. After hatch, larvae
(<24 hours old) were transported in aerated, insulated coolers to test sites.
Larvae were acclimated to river temperatures at each site by immersing
transport containers into the river until thermal equilibrium was reached.
Twenty-five larvae were placed into each of 14 test containers at each site,
for a total of 350 larvae per site. One test container from site MSR1 was
omitted from statistical analysis, as handling errors reduced the survival of
that replicate.
Larval containers were constructed from 3.7-L clear plastic jars (model
3314, Rubbermaid Home Products, Fairlawn, OH). Three 7.5-cm diameter
holes were drilled into the sides of each container along the median horizontal
axis, and covered with 0.4-mm mesh nitex screen (Sefar Filtration Inc.,
Dewey, NY). Each of these 14 larval containers was placed into a holding
rack made from untreated pine wood, which was housed in one of two 85-L
plastic tubs (model 25487, Sterilite Corporation, Townsend, MA). Water
depth in each tub was maintained to immerse each larval container >95%.
Tubs were partially buried and anchored in the stream channel, near the
stream bank. Water was gravity-fed to the tubs via a pair of 15-mm diameter
rubber hoses from 10–30 m upstream. The intake ends of hoses were
covered in coarse-mesh aluminum screen to prevent clogging. Flow through
each tub was adjusted daily to ≈5 L/min (Hulsman et al. 1983) via a valve
fitted to each feeder hose. Within each tub, a HOBO® Water Temp Pro (Onset
Computer Corp., Bourne, MA) temperature-collection unit was set to record
temperature every hour for the duration of the test.
Containers were checked daily to monitor larval survival and water
quality, and for maintenance. Living larvae in each container were
counted to determine survival. Larvae were considered dead when no
movement was observed when prodded, and deceased individuals were
removed daily following the methods of Rickwood et al. (2006). Larvae
in each of the 3.7-L jars received 2.5 mL of a concentrated suspension
of live brine shrimp nauplii (<24 hours after hatch) daily to supplement
natural food sources (Stewart et al. 1990). Each day, the screens on feeder
2010 A.W. Hafs, C.D. Horn, P.M. Mazik, and K.J. Hartman 581
hoses were cleaned of algae and detritus to minimize clogging, and flow
rates were adjusted to ≈5 L/min.
Temperature (°C), pH, and specific conductivity (μS/cm3) were measured
daily (YSI multimeter model 650, Yellow Springs, OH) within tubs at each
site. These measures were used to verify continuous data from data-collection
units maintained by Dominion Environmental near each test site. Mean
temperatures were calculated from HOBO® temperature loggers within each
tub, while mean pH and specific conductivity were calculated from the instream
data-collection units for each site.
Package survival (Therneau and Lumley 2008) in program R was used
to estimate survival along with 95% confidence intervals for each site. The
function survdiff (part of the package survival in R) was used to test for
significant differences (P ≤ 0.05) in survival curves between sites directly
downstream of AMD-impacted tributaries and non-AMD sites.
A Cox proportional hazard analysis (Kleinbaum and Klein 2005) was
performed to assess the effects of pH, conductivity, and temperature on hazard
rates during in situ bioassays. The mean number of fish alive on each day
among the 14 jars at each site was used for this statistical analysis. Covariates
were measured on each day throughout the study, so they were treated
as time dependent. We checked the proportional hazard assumption by correlating
scaled Schoenfeld residuals to the estimated survival function and
computing a chi-square significance test for the resulting correlation (Fox
2002). A test of proportionality was then obtained using the cox.zph function
(part of the package survival in program R). The martingale residuals were
plotted against covariates to detect nonlinearity (Fox 2002).
Results
Larval trapping survey
Larval fishes were collected in the Stony River during 2004, but overall
catch was low (Fig. 2). A total of 379 larvae were captured from 26 March to
1 June 2004. No larvae were captured during the 24 June and 2 July trapping
attempts, and no larvae were captured at MSR0 throughout the duration of
the study. Catch peaked in April, and was dominated by Catostomus commersoni
(Lacepède) (White Sucker) larvae, which comprised 90% of all
larvae captured. Other species captured were Creek Chub, Central Stoneroller,
and Smallmouth Bass. Water quality measurements recorded during
the trapping period are summarized in Table 1.
Because no larvae were captured at MSR0, a small number (1.00E-30)
was added to the White Sucker catch for that site. This adjustment was
necessary for the NMDS ordination to work properly. The NMDS ordination
with two dimensions was selected because the ease of interpretation
and the stress was low (7.28). Two convergent solutions were found after
four tries. Maximum temperature, average May temperature, average June
temperature, and average temperature all had P-values < 0.05, suggesting
correlation, and were plotted on the NMDS ordination (Table 2).
582 Northeastern Naturalist Vol. 17, No. 4
Table 1. Data collected by Dominion Environmental, Richmond, VA from the Stony River, 2002 to 2004. Data are summarized to represent major differences
among sites along the river continuum during spring months. Numbers in parentheses denote standard deviation. SpC = specific conductivity (uhmos).
Mean temp. Mean temp. Associated AMD Mean AlC Mean AlE Mean hardness
(ºC) (ºC) with AMD precipitate Max/min in main in AMD as
Site early springA late springB tributary present Mean pHC Mean SpCD SpCD stem (ppb) tributary (ppb) CaCO3 (ppm) F
MSR0 15.7 (1.1) 23.2 (2.0) No No 7.4 (0.2) 157 (5) 181 / 145 17.0 (13.6) 74 (14)
MSR1 14.7 (1.6) 21.8 (1.1) No No 7.0 (0.3) 192 (130) 572 / 125 31.2 (17.7) 68 (15)
MSR2 14.2 (1.5) 20.6 (1.8) Yes No 7.7 (0.3) 442 (305) 1214 / 145 30.3 (15.6) 31.3 (8.8) 372 (382)
4M2 13.8 (1.8) 20.6 (1.8) Yes Heavy 7.8 (0.9) 231 (358) 1373 / 167 33.5 (15.5) 239.0 (78.0) 309 (232)
LR2 12.9 (2.2) 20.1 (1.2) Yes Light 7.4 (0.4) 407 (281) 1223 / 137 34.3 (16.7) 107.0 (81.2) 310 (191)
MSR3 12.5 (2.4) 20.3 (1.5) No No 7.2 (0.9) 257 (243) 1009 / 41 31.3 (17.3) 258 (164)
MSR4 10.6 (3.0) 18.7 (0.7) No No 7.3 (0.4) 282 (120) 688 / 127 18.8 (12.9) 180 (57)
ATemperature values calculated from hourly in-stream measurements from March to April 2004.
BTemperature values calculated from hourly in-stream measurements from May to June 2004.
CpH values calculated from hourly in-stream measurements from March to June 2003 and 2004.
DSpecific conductivity values from hourly in-stream data collected from March to June 2004.
EAluminum values (total Al) calculated from four measurements from 2002 to 2004.
FHardness values calculated from five measurements from 2002 to 2004.
2010 A.W. Hafs, C.D. Horn, P.M. Mazik, and K.J. Hartman 583
In situ bioassay
Mean survival in the field bioassay ranged from 0–64% across all
sites (Table 3). Survival was lower, but not significantly different among
Table 2. Vectors for summary statistics of environmental variables collected at sites MSR0,
MSR1, MSR2, 4M2, LR2, and MSR3 along the Stony River, WV. The number of permutations
was set to 10,000, and only vectors with P-values < 0.05 (those in bold) were plotted in the
NMDS ordination.
Vectors NMDS1 NMDS2 r2 Pr(>r)
Average temperature -0.42 0.91 0.87 0.02
Standard deviation temperature -0.74 0.67 0.18 0.71
Maximum temperature -0.50 0.86 0.85 < 0.01
Minimum temperature -0.21 0.98 0.60 0.15
March average temperature -0.35 0.94 0.70 0.09
April average temperature -0.38 0.93 0.73 0.07
May average temperature -0.46 0.89 0.89 0.01
June average temperature -0.54 0.84 0.78 0.03
Average pH -0.61 -0.79 0.09 0.86
Standard deviation pH -0.91 0.41 0.22 0.60
Maximum pH -0.70 -0.71 0.16 0.75
Minimum pH 0.94 0.35 0.46 0.28
Range pH -0.82 -0.58 0.38 0.34
Average conductivity 0.90 -0.43 0.19 0.62
Standard deviation conductivity 0.55 -0.83 0.21 0.58
Maximum conductivity 0.39 -0.92 0.18 0.64
Minimum conductivity -0.02 1.00 0.18 0.73
Figure 2. Total number of larval fishes caught with light traps during each sampling
period.
584 Northeastern Naturalist Vol. 17, No. 4
sites directly downstream of AMD-impacted tributaries and non-AMD sites
(χ2 = 1.7, df = 1, P = 0.20, Fig. 3). Most mortality observed occurred during
the first fifteen days of the test. Survival declined steadily at all sites except
below Fourmile Run (4M2), which experienced an acute mortality event
between the 21st and 22nd of May 2004. This mortality event coincided with
Table 3. Physiochemical conditions at in situ bioassay sites. Data for site 4M2 encompasses
only the days up to the major mortality event (day 10 of test), while data for other sites covers
the entire test period. Numbers in parentheses denote standard deviation.
Mean specific Specific
Mean Temperature conductivity conductivity Mean
Site temp. (ºC) range (ºC) Mean pH pH range (μS/cm3) range (μS/cm3) survival
MSR0 23.8 (3.1) 17.5–27.8 7.7 (0.2) 7.3–7.9 185 (40) 159–278 0.44 (0.33)
MSR1 22.8 (2.6) 15.5–26.5 7.7 (0.2) 7.5–8.1 173 (22) 149–232 0.60 (0.38)
MSR2 21.9 (3.4) 14.7–26.7 7.9 (0.1) 7.7–8.1 558 (400) 104–1223 0.64 (0.38)
4M2 22.3 (2.0) 18.7–24.3 7.6 (0.5) 6.3–8.1 490 (256) 331–1170 0 (NA)
LR2 20.4 (3.2) 13.7–24.9 7.6 (0.1) 7.3–7.9 509 (320) 201–1170 0.56 (0.37)
MSR3 19.8 (2.6) 14.3–23.5 7.6 (0.1) 7.5–8.0 473 (276) 201–1042 0.52 (0.36)
Figure 3. Comparison of daily mean survival for larval Fathead Minnows at sites directly
downstream of AMD impacted tributaries and non AMD sites during in situ bioassays.
2010 A.W. Hafs, C.D. Horn, P.M. Mazik, and K.J. Hartman 585
an 18-hour depression in pH to less than 5.0 in the tributary Fourmile Run (Fig. 4).
Water quality measurements recorded at in situ bioassay locations are summarized
in Table 3.
Cox proportional hazard test revealed that low pH had a strong negative
influence on survival of larval Fathead Minnows in our in situ bioassay
(Table 4). The log likelihood ratio test for the model including temperature,
conductivity, and pH was significant (log likelihood = 45.8, df = 3,
P < 0.001). Figure 5 summarizes predicted survival at the levels of pH,
conductivity, and temperature that were present during this study. Both the
proportional hazard (global model: χ2 = 5.86, df = 3, P = 0.12) and the nonlinearity
assumptions were met.
Discussion
Larval trapping survey
Trapping suggests that abundance of fish larvae in the Stony River is low.
Floyd et al. (1984) captured 4549 larvae in 96 hours of light trapping out of the
Table 4. Results from the Cox proportional hazard test (n = 150; 25/site).
Coefficient se(Coefficient) z P-value
Conductivity 0.001 0.001 1.08 0.28
pH -2.580 0.399 -6.47 <0.01
Temperature 0.163 0.104 1.57 0.12
Figure 4. In situ survival of larval Fathead Minnows at site 4M2 and pH flux in Fourmile
Run, which enters the Stony River directly upstream of 4M2.
586 Northeastern Naturalist Vol. 17, No. 4
Figure 5. Cox proportional hazard survivorship curves for Fathead Minnows at the
varying levels of pH, conductivity (μS/cm3), and temperature (°C). Graphs A and B
show estimated survival at different temperatures while holding conductivity and
pH constant at mean values (368 and 7.72, respectively; A) and constant at values of
600 and 7.72, respectively (B). Graphs C and D show estimated survival at different
conductivities while holding temperature and pH constant at mean values (21.85 and
7.72, respectively; C) and constant at values of 25 and 7.5, respectively (D). Graphs
E and F show estimated survival at different pH values while holding temperature
and conductivity constant at mean values (E) and constant at values of 25 and 600
(F), respectively.
2010 A.W. Hafs, C.D. Horn, P.M. Mazik, and K.J. Hartman 587
Middle Fork of Drake's Creek, KY. Marchetti et al. (2004) captured 4672 larval
fish from the upper mainstem Sacramento River, CA, in 120 hours of light
trapping. Niles and Hartman (2007) captured 9221 larvae from the Kanawha
River, WV, in 378 hours of light trapping. Using the same methods as Floyd
et al. (1984), Marchetti et al. (2004), and Niles and Hartman (2007), we only
captured 379 larvae in 220 hours of trapping in this study, providing evidence
to suggest that overall stream fish reproduction in the Stony River is low.
A possible explanation for our low capture rates is that the spring spawning
event did not entirely overlap with our trapping period (26 March to
2 July 2004). However, the peak capture rates for all four species we captured
were between 22 April and 23 May. Catch rates decreased with every
trapping event prior to 22 April. Catch rates for all species declined after 23
May, and no larvae were captured in the two trapping periods after 1 June.
Thus, the timing of our trapping seems to have encompassed the major
spring spawning event, and the low larval fish capture rates are not likely an
artifact of our sampling design, but rather are most probably explained by
some other factor.
Fitted vectors on the NMDS ordination provided strong evidence that
high water temperatures were having a significant influence on the larval
fish communities at our sites in the Stony River, WV. Early life stages of
fishes have been found to be very sensitive to temperature conditions (Houde
1989a, Pepin 1991). Thus, it is likely that thermally induced mortality during
the ELS of fishes in the Stony River is a major factor contributing to low
catch rates.
No larval fishes were captured at site MSR0, while White Suckers and
Central Stonerollers were captured at MSR1. Although temperature is having
a significant effect on the fish communities of sites MSR0 and MSR1, the
temperature regimes at both of these sites were very similar. It is therefore
unlikely that temperature is the only reason that larval fish were not captured
at site MSR0. Bed coursing and loss of spawning gravel below dams is a
common occurrence (Kondolf 1997). All of the species captured in this study
spawn over medium- to small-sized gravels (Curry and Spacie 1984, Lukas
and Orth 1995, Miller 1962, Ross and Reed 1978). MSR0 is approximately
200 m from the dam, where loss of spawning habitat resulting from scouring
could be contributing to the zero capture rate at this site.
Temperature effects from the thermally enriched Mount Storm Lake
diminished farther downstream in the Stony River as water from coldwater
tributaries and groundwater inputs entered the system. Site MSR2 is directly
below a tributary which caused the average late spring temperature to be
2.6 °C cooler than MSR0, the farthest upstream site. Smallmouth Bass have
been classified as a coolwater fish (Eaton and Scheller 1996), and larvae
were captured in MSR2 but not in MSR1, where high temperatures were still
present. This result implies that if the temperature regime of the Stony River
was decreased, the habitat for larval Smallmouth Bass could be substantially
improved in the section downstream of the impoundment.
588 Northeastern Naturalist Vol. 17, No. 4
Because Smallmouth Bass are sensitive to acidic conditions (Snucins and
Shuter 1991), the capture of Smallmouth Bass at MSR2, an AMD-influenced
site, is evidence that MSR2 is treated effectively. Larval fish commonly
experience downstream drift (Brown and Armstrong 1985), so it is possible
that the Smallmouth Bass captured at MSR2 had drifted from upstream.
However, Smallmouth Bass were not captured at MSR1, which was only approximately
250 m upstream of MSR1. Therefore, if Smallmouth Bass were
present upstream of MSR2, they should have been captured in the trapping
that occurred at MSR1. The only other site where larval Smallmouth Bass
were captured was MSR4, which is the site farthest downstream, where AMD
effects have diminished. The capture of Smallmouth Bass at MSR2 suggests
that if all AMD tributaries within the study area were treated effectively, fish
populations could be improved and the Smallmouth Bass population would
likely increase.
Only three larval fish were captured at 4M2, a severely AMD-influenced
site. This capture rate was much lower than the capture rates at MSR1 and
MSR2, the other AMD-influenced sites. Sporadic treatment failures on 4M2,
like the event on 21–23 May, are the most probable reason for low capture
rate at 4M2. Treatment keeps pH high the majority of the time in Fourmile
Creek, but sporadic treatment failures may be enough to have a severe negative
influence on the stream fish communities.
In situ bioassay
Significant larval mortality was observed in conjunction with fluctuations
in AMD severity. The major mortality event that occurred at site 4M2
followed lowered pH (<5.0) in the AMD tributary, Fourmile Run. This event
did not create a significant difference between survival curves of sites directly
downstream of AMD-impacted tributaries and non-AMD sites because
other treated sites (MSR2) mitigated the impacts. Fluctuations in the pH
of Fourmile Run created acutely toxic conditions in the Stony River likely
related to fluctuations in metal solubility (aluminum). Mixing zones between
acidic and circumneutral streams are often highly toxic due to aluminum
polymerization and precipitation (Henry et al. 1999, Poleo et al. 1994). Even
when conditions are not overly acidic (pH > 6.0), high mortality can occur
(Verbost et al. 1995). Thus, it seems that when an AMD tributary is treated
to circumneutral pH, acute larval toxicity is reduced. However, alterations
in treatment that lead to periodic depressions of pH and increased aluminum
solubility can create acutely toxic conditions. Such conditions could have
severely negative effects on larval fish populations when larvae encounter
these mixing zones during downstream drift. Similarly, adults exposed to
mixing zones during spawning migrations may experience mortality or
avoid these areas, further reducing potential larval production. With 23,000
km of AMD-influenced streams in the United States (Sasowsky et al. 2000),
the implications for riverine fishes are significant.
The survival rates of larvae at MSR1 and MSR2 (upstream and downstream
of the Laurel Run mine outfall) suggests that effective treatment of
2010 A.W. Hafs, C.D. Horn, P.M. Mazik, and K.J. Hartman 589
mine wastes creates water quality conditions tolerable by larval fishes. Larval
survival was also similar to MSR1 and MSR2 at site LR2, even though it
is downstream of an AMD tributary that is acidic on occasion. The range in
pH (7.3–7.9) at LR2 over the course of the study suggests that treatment was
effective at creating stable pH levels, at least over the short term. The stable
pH levels at LR2 may be the reason that the survival estimates were similar
among MSR1, MSR2, and LR2. Another possible explanation is that because
of physical constraints, assays were placed on the side of the Stony River opposite
the confluence with the AMD tributary (LR2). Thus, any mixing-zone
effects that occur near the confluence may have been missed.
Results from the Cox proportional hazard test clearly show pH largely affects
survival rates of larval fishes in the Stony River. Predicted survival rates
decrease substantially at low pH values. High temperature also had major
influences on estimated survival rates from Cox proportional hazard models.
The combination of high spring temperatures and low pH from ineffectively
treated AMD tributaries is having severe negative impacts on the larval fish
communities of the Stony River. Management actions that would decrease
spring and summer water temperatures in combination with effectively treating
AMD-impacted tributaries should greatly benefit the fish populations of
the Stony River through increased larval survival and recruitment.
Acknowledgments
Thanks go to Cara Hoar and Brandon Keplinger for assistance in planning, field,
and lab work. Thanks go to George Merovich for review of statistical procedures.
Also, thanks to the Dominion biologist crew for cooperation in all phases of this investigation.
Funding was provided by Dominion Environmental, Richmond, VA and
the US Geological Survey. Use of trade names does not imply endorsement by the
US Government.
Literature Cited
Auer, N.A. 1982. Identification of larval fishes of the Great Lakes basin with emphasis
on the Lake Michigan drainage. Special Publication 82–3. Great Lakes
Fishery Commission. Ann Arbor, MI.
Brown, A.V., and M.L. Armstrong. 1985. Propensity to drift downstream among
various species of fish. Journal of Freshwater Ecology 3:3–17.
Cooke, S.J., J.F. Schreer, D.P. Philipp, and P.J. Weatherhead. 2003. Nesting activity,
parental care behavior, and reproductive success of Smallmouth Bass,
Micropterus dolomieu, in an unstable thermal environment. Journal of Thermal
Biology 28:445–456.
Curry, K.D., and A. Spacie. 1984. Differential use of stream habitat by spawning
Catostomids. American Midland Naturalist 111:267–279.
Eaton, J.G., and R.M. Scheller. 1996. Effects of climate warming on fish thermal
habitat in streams of the United States. Limnology and Oceanography
41:1109–1115.
Floyd, K.B., R.D. Hoyt, and S. Timbrook. 1984. Chronology of appearance and habitat
partitioning by stream larval fishes. Transactions of the American Fisheries
Society 113:217–223.
590 Northeastern Naturalist Vol. 17, No. 4
Fox, J. 2002. Cox proportional-hazards regression for survival data. In R. Fox
(Ed.). An R and S-PLUS Companion to Applied Regression. Sage, Thousand
Oaks, CA.
Gensemer, R.W., and R.C. Playle. 1999. The bioavailability and toxicity of aluminum
in aquatic environments. Critical Reviews in Environmental Science and
Technology 29:315–450.
Gerhke, P.C. 1994. Influence of light intensity and wavelength on phototactic behaviour
on larval Silver Perch Bidyanus bidyanus and Golden Perch Macquaria
ambigua and the effectiveness of light traps. Journal of Fish Biology 44:741–75
Gray, N.F. 1996. Field assessment of acid mine drainage contamination in surface
and ground water. Environmental Geology 27:358–361.
Hawkins, C.P., and R.H. Norris. 2000. Performance of different landscape classifications
for aquatic bioassessments: Introduction to the series. Journal of the North
American Benthological Society 19:367–369.
Henry, T.B., E.R. Irwin, J.M. Grizzle, M.L. Wildhaber, and W.G. Brumbaugh. 1999.
Acute toxicity of an acid mine drainage mixing zone to juvenile Bluegill and Largemouth
Bass. Transactions of the American Fisheries Society 128:919–928.
Hoar, C.C. 2005. Fish response to discharge events from a power plant cooling
reservoir in a river affected by acid mine drainage and thermal influences. M.Sc.
Thesis. West Virginia University, Morgantown, WV. 70 pp.
Horn, C.D. 2005. Investigation of the effects of thermal enrichment and acid mine
drainage on sensitive aquatic biota in the Stony River, Grant County, WV. M.Sc.
Thesis. West Virginia University, Morgantown, WV. 101 pp.
Houde, E.D. 1989a. Subtleties and episodes in the early life of fishes. Journal of Fish
Biology 35:29–38.
Houde, E.D. 1989b. Comparative growth, mortality, and energetics of marine fish larvae:
Temperature and implied latitudinal effects. Fisheries Bulletin 87:471–495.
Houde, E.D. 2002. Mortality. Pp. 64–87, In L.A. Fuiman and R.G. Werner (Eds.).
Fishery Science: The Unique Contributions of Early Life Stages. Blackwell Sciences
Limited, Malden, MA.
Hulsman, P.F., P.M. Powles, and J.M. Gunn. 1983. Mortality of Walleye eggs and
Rainbow Trout yolk-sac larvae in low-pH waters of the LaCloche Mountain area,
Ontario. Transactions of the American Fisheries Society 112:680–688.
Kissick, L.A. 1993. Comparison of traps lighted by photochemicals or electric bulbs
for sampling warmwater populations of young fish. North American Journal of
Fisheries Management 13:864–867.
Kleinbaum, D.G., and M. Klein. 2005. Survival Analysis: A Self-Learning Text. 2nd
Edition. Springer, NewYork, NY.
Kondolf, G.M. 1997. Hungry water: Effects of dams and gravel mining on river
channels. Environmental Management 21:533–551.
Leis, A.L., and M.G. Fox. 1994. Effect of mine tailings on the in situ survival of
Walleye (Stizostedion vitreum) eggs in a northern Ontario river. Ecoscience
1:215–222.
Lukas, J.A., and D.J. Orth. 1995. Factors affecting nesting success of Smallmouth
Bass in a regulated Virginia stream. Transactions of the American Fisheries Society
124:726–735.
Luksiene, D., O. Sandstrom, L. Lounasheimo, and J. Andersson. 2000. The effects
of thermal effluent exposure on the gametogenesis of female fish. Journal of Fish
Biology 56:37–50.
2010 A.W. Hafs, C.D. Horn, P.M. Mazik, and K.J. Hartman 591
Lydersen, E., B. Salbu, A.B.S. Poleo, and I.P. Muniz. 1990. The influence of
temperature on aqueous aluminum chemistry. Water, Air, and Soil Pollution
51:203–215.
Marchetti, M.P., E. Esteban, M. Limm, and R. Kurth. 2004. Evaluating aspects of
larval light-trap bias and specificity in the northern Sacramento River system:
Do size and color matter? Pp. 269–279, In R. Feyrer, R.L. Brown, and J.J. Orsi
(Eds.). Early Life History of Fishes in the San Francisco Estuary and Watershed.
American Fisheries Society, Bethesda, MD.
McCormick, J.H., K.M. Jensen, and L.E. Anderson. 1989. Chronic effects of low pH
and elevated aluminum on survival, maturation, spawning, and embryo-larval
development of the Fathead Minnow in soft water. Water, Air, and Soil Pollution
43:293–307.
McCune, B., and J.B. Grace. 2002. Analysis of Ecological Communities. MjM Software
Design, Gleneden Beach, OR. 300 pp.
McDonald, D.G., J.P. Reader, and T.R.K. Dalziel. 1989. The combined effects of pH
and trace metals on fish ionoregulation. Pp. 221–242, In R. Morris, E.W. Taylor,
D.J.A. Brown, and J.A. Brown (Eds.). Acid Toxicity and Aquatic Animals. Cambridge
University Press, Cambridge, UK.
Merovich, G.T., and J.T. Petty. 2007. Interactive effects of multiple stressors and restoration
priorities in a mined Appalachian watershed. Hydrobiologia 575:13–31.
Milam, C.D., J.L. Farris, and J.D. Wilhide. 2000. Evaluating mosquito control pesticides
for effect on target and nontarget organisms. Archives of Environmental
Contamination and Toxicology 39:324–328.
Miller, R.J. 1962. Reproductive behavior of the Stoneroller Minnow, Campostoma
anomalum pullum. Copeia 2:407–417.
Mion, J.B., R.A. Stein, and E.A. Marschall. 1998. River discharge drives survival of
larval Walleye. Ecological Applications 8:88–103.
Niles, J.M., and K.J. Hartman. 2007. Comparison of three larval fish gears to sample
shallow water sites on a navigable river. North American Journal of Fisheries
Management 27:1126–1138.
Oksanen, J., R. Kindt, P. Legendre, B. O'Hara, G.L. Simpson, P. Solymos, M.H.H.
Stevens, and H. Wagner. 2008. Vegan: Community ecology package. R package
version 1.15-1. Available online at http://cran.r-project.org/, http://vegan.rforge.
r-project.org/. Accessed 15 October 2009.
Paller, M.H., and B.M. Saul. 1996. Effects of temperature gradients resulting from
reservoir discharge on Dorosoma cepedianum spawning in the Savannah River.
Environmental Biology of Fishes 45:151–160.
Pepin, P. 1991. Effects of temperature and size on development, mortality, and survival
rates of the pelagic early life-history stages of marine fish. Canadian Journal
of Fisheries and Aquatic Sciences 48:503–518.
Poleo, A.B.S., E. Lydersen, and I.P Muniz. 1991. The influence of temperature on
aqueous aluminium chemistry and survival of Atlantic Salmon (Salmo salar L.)
fingerlings. Aquatic Toxicology 21:267–278.
Poleo, A.B.S., E. Lydersen, B.O. Rosseland, F. Kroglund, B. Salbu, R.D. Vogt, and
A. Kvellestad. 1994. Increased mortality of fish due to changing Al-chemistry of
mixing zones between limed streams and acidic tributaries. Water, Air, and Soil
Pollution 75:339–351.
Potts, W.T.W., and T.G. McWilliams. 1989. The effects of hydrogen and aluminium
ions on fish gills. Pp. 201–220, In R. Morris, E.W. Taylor, D.J.A. Brown, and J.A.
Brown (Eds.). Acid Toxicity and Aquatic Animals. Cambridge University Press,
Cambridge, UK.
592 Northeastern Naturalist Vol. 17, No. 4
Rickwood, C.J., M.G. Dube, L.M. Hewitt, T.G. Kovacs, J.L. Parrott, and D.L.
MacLatchy. 2006. Use of paired Fathead Minnow (Pimephales promelas) reproductive
test. Part1: Assessing biological effects of final bleached kraft pulp mill
effluent using a mobile bioassay trailer system. Environmental Toxicology and
Chemistry 25:1836–1846.
Ross, M.R., and R.J. Reed. 1978. The reproductive behavior of the Fallfish Semotilus
corporalis. Copeia 2:215–221.
Sandstrom, O., E. Neuman, and G. Thoresson. 1995. Effects of temperature on life
history variables in perch. Journal of Fish Biology 47:652–670.
Sandstrom, O., I. Abrahaamsson, J. Andersson, and M. Vetemaa. 1997. Temperature
effects on spawning and egg development in Eurasian Perch. Journal of Fish
Biology 51:1015–1024.
Sasowsky, I.D., A. Foos, and C.M. Miller. 2000. Lithic controls on the removal of
iron and remediation of acidic mine drainage. Water Research 34:2742–2746.
Shuter, B.J., J.A. MacLean, F.E.J. Fry, and H.A. Regier. 1980. Stochastic simulation
of temperature effects on first-year survival of Smallmouth Bass. Transactions of
the American Fisheries Society 109:1–34.
Snucins, E.J., and B.J. Shuter. 1991. Survival of introduced Smallmouth Bass in lowpH
lakes. Transactions of the American Fisheries Society 120:209–216.
Stewart, A.J., L.A. Kszos, B.C. Harvey, L.F. Wicker, G.J. Haynes, and R.D. Bailey.
1990. Ambient toxicity dynamics: Assessments using Ceriodaphnia dubia and
Fathead Minnow (Pimephales promelas ) larvae in short-term tests. Environmental
Toxicology and Chemistry 9:367–379.
Therneau, T. (original), and T. Lumley (ported). 2008. Survival: Survival analysis,
including penalised likelihood. R package version 2.34-1.
Verbost, P.M., M.H.G. Berntssen, F. Kroglund, E. Lydersen, H.E. Witters, B.O. Rosseland,
B. Salbu, and S.E.W. Bonga. 1995. The toxic mixing zone of neutral and
acidic river water: Acute aluminium toxicity in Brown Trout (Salmo trutta L.).
Water, Air, and Soil Pollution 85:341–346.
Witters, H.E., S. Van Puymbroeck, A.J.H.X. Stouthart, and S.E.Wendelaar Bonga.
1996. Physicochemical changes of aluminium in mixing zones: Mortality and
physiological disturbances in Brown Trout (Salmo trutta L.). Environmental
Toxicology and Chemistry 15:986–996.