 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from you: tocord.
TO AVOID FINES Mum on or baton duo duo.

DATE DUE DATE DUE DATE DUE

. ll
IL: 2—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

msu In An Afﬁrmuﬁve ActioNEqual Oppomnhy Institution
7 7 cMMpma-pi

—— ,7- ,

 

IRE-3C1
'1. 1'"
“‘AL

m9:?'V‘
‘h.htugb‘:

DIRECT AND INDIRECT EFFECTS OF THE BLACK FLY (DIPTERA:

SIMULIIDAE) LARVICIDE, W W VAR.
ISRAELEHSIS, ON SELECTED NON-TARGET AQUATIC INSECTS AND

TROUT

By

Mark Steven Wipﬂi

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Entomology

1992

6525‘ * 0 7.3" 7

ABSTRACT

DIRECT AND INDIRECT EFFECTS OF THE BLACK FLY (DIPTERA:

SIMULIIDAE) LARVICIDE. BADJLLIIS WIS VAR.
ISRAELEHSIS, ON SELECTED NON-TARGET AQUATIC INSECTS AND

TROUT

By

Mark Steven Wipﬂi

Field and laboratory studies were conducted to investigate the direct
and indirect effects of the microbial insecticide, Ms W var.
W de Barjac (Teknar®) (ELL), on non-target aquatic insects and
trout. Streamside experiments conducted in artiﬁcial channels showed
that two of the 16 species tested were sensitive to BALL applied at unusually
high dosages (> 100 ppm ELL). No insect species were killed when ELL
was applied at the recommended 22 ppm @ 1 min dosage. Predators and
detritivores consumed ELL-contaminated black ﬂy larvae without
observable adverse effects on survivorship, growth, adult emergence
success, and adult emergence phenology. Some detritivores
(Ephemeropteira) attained greater growth rates on diets containing dead
black ﬂy larvae than those without dead larvae. Trout were not sensitive to
ELL, except at rates >100,000 times the labeled rates, and trout mortality
was attributed to formulation components. In-stream studies revealed that
the diets of predatory stoneﬂies were qualitatively and quantitatively
affected after ELL applications, with one perlid species switching to

alternate F
choice trid-
biasli ﬂies

tween b?

‘ mayﬂies 0
target org:

affected 0'

 

alternate prey after black ﬂies were removed from the stream system. Prey
choice trials with predators indicated that perlodid stoneﬂies relied more on
black ﬂies than mayﬂy prey. Perlid stoneﬂies showed no preference
between black ﬂies and mayﬂies, and predatory odonates consumed more
mayﬂies than black ﬂy larvae. These studies indicated that, overall, non-
target organisms were not directly sensitive to ELL, but were indirectly
affected through loss of black ﬂy biomass following ELL application.

 

I express 5:
humored sup;
much indepen
Miller, Jean S
all aspects of
John Giesy, R
Roger Wotton
Procedures.

MCCaﬂmy. a

  

“MmTaa
teChIlicall and
mm for g:-
exCEllent 59C.
Valuable hel;

Sllppon t-llI‘C

Students for

Ol'er the pa“

 

ACKNOWLEDGMENTS

I express sincere thanks to Dr. Richard W. Merritt for his good-
humored support and help throughout this endeavor, and for allowing me
much independence. Many thanks to Drs. Bill Cooper, Don Hall, Jim
Miller, Jean Stout, and Matt Zabik for all of their valuable assistance with
all aspects of this research. I thank Drs. David Allan, Jan Ciborowski,
John Giesy, Ron Hall, Barbara Peckarsky, Bill Taylor, Ned Walker, and
Roger Wotton for many helpful suggestions on experimental design and
procedures. Thanks to Drs. Peter Adler, Bill Hilsenhoﬁ', Patrick
McCaﬂ‘erty, and Stan Szczytko for insect identiﬁcations. A sincere thanks
to Alan Tessier for his advice on statistics, and Bill Morgan for all of his
technical and computer advice. Sincere thanks to Russ Edens and Bill
LaVoie for good-natured and tireless ﬁeld assistance. I also thank the
excellent secretarial staﬁ' in the MSU Entomology Dept. for all of their
valuable help. I also want to thank Kaja Brix for her unsurpassed patient
support throughout this endeavor. I also thank numerous graduate
students for their support, thought provoking discussions, and friendship

over the past several years.

iv

 

TABLE OF CONTENTS

LIST OF TABLES ......................................................................
LIST OF FIGURES ...................................................................
INTRODUCTION .....................................................................

CHAPTER 1
PORTABLE ARTIFICIAL STREAMS FOR FIELD AND
LABORATORY EXPERIMENTS WITH AQUATIC
MACROINVERTEBRATES. .........................................
ABSTRACT .........................................................................
INTRODUCTION ................................................................
MATERIALS AND METHODS ..............................................
RESULTS ............................................................................
DISCUSSION ......................................................................
ACKNOWLEDGMENTS .......................................................
LITERATURE CITED ..........................................................

CHAPTER 2
NON -TARGET IMPACT OF EAQILLLIS W VAR.
W IN A LOTIC SYSTEM: DIRECT LETHAL
AND SUBLETHAL FOOD-CHAIN EFFECTS ON
SELECTED AQUATIC INSECTS, AND FATE OF
INTOXICATED BLACK FLY (DIPTERA: SIMULIIDAE)
LARVAE. ..................................................................
ABSTRACT .........................................................................
INTRODUCTION .............
MATERIALS AND METHODS ..............................................
RESULTS ............................................................................
DISCUSSION ......................................................................
ACKNOWLEDGMENTS .......................................................
LITERATURE CITED ..........................................................

ix

238388885653

assesses

11

Ca.
DIRI

.ABS'
1):?
BLAT

“‘V‘

DIS(
ACPE
LITE

 

CHAJU:
POP”

.ABS‘
IDFIE
BL¥T
RES:
DISC
.ACI:
LITE

R3003;

CHAPTER 3
DIRECT AND INDIRECT EFFECTS OF THE MICROBIAL
INSECTICIDE EAQ ILLUS W VAR.
WIS ON BROOK, SALVELIE US WALLS;
BROWN, SALMQ IRMA; AND STEELHEAD,
W MIXES TROUT. ...........................
ABSTRACT .........................................................................
INTRODUCTION ................................................................
MATERIALS AND METHODS ..............................................
RESULTS ............................................................................
DISCUSSION ......................................................................
ACKNOWLEDGMENTS .......................................................
LITERATURE CITED ..........................................................

CHAPTER 4
POPULATION LEVEL DISTURBAN CE FROM EAQILLHS
WELLS VAR. W IN STREAMS:
PREDATION RESPONSES OF MACROINVERTEBRATE
PREDATORS. .............................................................
ABSTRACT .........................................................................
INTRODUCTION ................................................................
MATERIALS AND METHODS ..............................................
RESULTS ............................................................................
DISCUSSION ......................................................................
ACKNOWLEDGMENTS .......................................................
LITERATURE CITED ..........................................................

RECOMMENDATIONS ............................................................

15)

. LIST OF TABLES

CHAPTER 1

Table 1. Predation experiments conducted to test the performance of
the 40 channel artiﬁcial stream unit. ......................................

Table 2. Experiments conducted to test the performance of the 40
channel artiﬁcial stream unit. ...............................................

Table 3. Predation and predator mortality rates during predation
experiments using the 40 channel artiﬁcial stream unit. ..........

Table 4. Mortality rates, cumulative molt, and emergence of aquatic
insects during collector-shredder-grazer experiments using the
40 channel artiﬁcial stream unit. ...........................................

CHAPTER 2

Table 1. Insect taxa used to test the impact of ELL on non-target
aquatic insects. ....................................................................

Table 2. Experimental variables and parameters of streamside
studies examining the impact of ELL on predatory
macroinvertebrates, conducted in artiﬁcial channels. ..............

Table 3. Experimental variables and parameters of streamside
studies examining the impact of ELL on stream
macroinvertebrates from selected ﬁmctional feeding groups
(shredders, grazers, ﬁlter-feeders), conducted in artiﬁcial _
channels. ............................................................................

Table 4. Experimental variables and parameters of streamside
studies examining drift frequency during and predation rates of
three predatory stoneﬂies following ELL exposure, conducted in
artiﬁcial channels. ...............................................................

Table 5. Predation, percent mortality, percent adult emergence
success, adult emergence phenology, and growth of predatory
aquatic insects exposed to ELL directly or indirectly through
ingesting contaminated black ﬂies. ........................................ 67

Table 6. Consumption of black ﬂies, percent mortality, percent adult
emergence success, adult emergence phenology, and growth of
selected shredders, grazers, and ﬁlter-feeders exposed to ELL

directly or indirectly through ingesting contaminated black ﬂies. 70

Table 7. Time spent drifting during ELL exposure and subsequent
predation on black ﬂy larvae after ELL exposure, with three
predatory stoneﬂies. ............................................................. 74

CHAPTER 3

Table 1. Trout species and length, ELL concentration, nature
(viable or nonviable) and exposure period, test duration, and
number of replicates per treatment used for trout toxicity tests. . 98

Table 2. L050 values for 2.2 cm long brown and 1.8 cm long brook
trout exposed to ELL for 48 hr. ............................................... 110

CHAPTER 4

Table 1 . Density estimates of macroinvertebrate prey in Medora and
Manganese Rivers 1 d before starting population level disturbance
experiments (11 = 5). .............................................................. 139

Table 2. Mean number of macroinvertebrate prey items in foreguts of
Amanda Mas collected from Medora River, and
Melina media collected from Manganese River, 1 d before
starting Medora-Manganese Rivers population level disturbance
experiments (n = 10). ............................................................ 141

Table 3. Mean mass gain and their corresponding instantaneous

growth rates (IGR) of m lyegzjgs nymphs fed ﬁve
Simlimn- Egg-Lia ratios over 6 d (n: 4). ................................... 149

 

 

 

 

CPA? l E

Figure 1
ﬁn'a:
three

pl}v

Figure 1
Clair

Figure

aqin
CB)

stre

LIST OF FIGURES

CHAPTER 1

Figure 1. Diagrammatic representation of the artiﬁcial stream unit
for aquatic macroinvertebrate studies. (D) delivery hose, (S)
threaded sleeve, (H) holding pipe, (I) input tube, (C) channel, (P)
plywood platform, (E) exit tube, (0) outﬂow pipe, (S.F.) steel frame.

Figure 2. Artiﬁcial stream unit (A), and channels with (top two
channels) and without (bottom two channels) screen collars (B).

Figure 3. Diagrammatic representation of an artiﬁcial channel for
aquatic macroinvertebrate studies (A) without screen collar, and
(B) with screen collar. ...........................................................

Figure 4. Gravity-feed pipes (A), 2 water ﬁltration barrels (B),
dispensing bucket (C), delivery hoses (D), and two artiﬁcial
stream units (E) during streamside experiments. .....................

CHAPTER 2

Figure 1. Diagrammatic representation of experimental design used
to measure in-stream mortality of Chironomidae following
Bacillus thunneisnsis var. israelensis application. AP = 13.1.1.

application point. .................................................................

Figure 2.1n-stream Chironomidae densities (A.) 'ABOVE vs.
BELOW‘ and (B. ) 'BEFORE vs. AFTER' Bacillus Lhmjngiensis
var. W applications. [ns= not signiﬁcant @ p= 0.05].

Figure 3. Ma abﬁgminalis mortality 'ABOVE vs. BELOW‘ during
in-stream ELL applications. Error bars represent S.E. [ns = not

signiﬁcant @ p = 0.05]. ..........................................................

Figure 4. Percent larval black ﬂy decomposition and release from

their attachment sites following Bacillus W var.
15mm application during (A.) May, and (3.) July, 1990.
Error bars represent S.E. ......................................................

75

78

yl

 

Figure 5. Pt
their at‘u
icfaslsrj 2
Error be

Figure 6. P4
exposure
depositio
reaches.

Clitl’l'ER E

Figure 1. P4
hr expos:
different
(p<0.05 ,1,

Figure 2. Pt
erposure
viable an
l€tters {5
(Trial 2).
belWeen

(Trials 3

“We 3. P
Exposure
represer
baI’E re;

Figure 4. F
expose]
Signiﬁe.
Signiﬁg'

I

 

 

Figure 5. Percent larval black ﬂy decomposition and release from
their attachment sites following Emma W var.
W5 application during (A. ) April, and (B. ) June, 1991.
Error bars represent S. E. ......................................................

Figure 6. Percent black ﬂy mortality, in artiﬁcial streams, from
exposure to 'ELL-tainted' and clean sediments collected from
depositional zones within ELL-treated and untreated river
reaches. Error bars represent S.E. .........................................

CHAPTER 3

Figure 1. Percent brown trout egg emergence over 26 (1 following 48
hr exposure to varying ELL concentrations. Bars headed by
different letters represent signiﬁcant difference between means
(p<0.05). Error bars represent S.E. .........................................

Figure 2. Percent brown trout mortality over 7 d following 48 hr
exposure to varying ELL concentrations (Trials 2-5), comparing
viable and nonviable ELL (Trials 3-5). Bars headed by different
letters represent signiﬁcant difference between means (p<0.05)
(Trial 2). Bar pairs headed by * signiﬁes statistical difference
between means (p<0.05), ns = not signiﬁcant, nc = not calculated
(Trials 3-5). Error bars represent S.E. .....................................

Figure 3. Percent brown trout mortality over 7 d at 0, 12, 36, and 48 hr
exposure to 2000 ppm ELL Bars headed by different letters
represent signiﬁcant difference between means (p<0.05). Error
bars represent S.E. ...............................................................

Figure 4. Percent steelhead trout mortality over 5 d following 24 hr
exposure to varying ELL concentrations. Bar pairs headed by *
signiﬁes statistical difference between means (p<0.05), ns = not
signiﬁcant, nc = not calculated. Error bars represent S.E. ........

104

105

106

108

 

Figure 5. PE
exposure
viable an
letters re
: DOI Cal'
statisticz
nc = not

Figure 6. E
steelhea
gills vn't
Exposed
shoving
particul

Figure 7. I
blood ta
hr. (A.

Figure 8. '
in three
were tn
Exposed
no Bil
differer

CRAVE};

Flgllre 1.
‘0 mea

AP=E

Figum 2.
the Me
Signiﬁ C

Ower

Figure 5. Percent brook trout mortality over 7 d following 48 hr
exposure to varying ELL concentrations (Trials 11-13), comparing
viable and nonviable ELL (Trials 12-13). Bars headed by different
letters represent signiﬁcant difference between means (p<0.05), nc
= not calculated (Trials 11&13). Bar pairs headed by * signiﬁes
statistical difference between means (p<0.05), ns = not signiﬁcant,
nc = not calculated (Trial 12). Error bars represent S.E. ...........

Figure 6. Scanning electron micrographs of gills dissected from
steelhead trout exposed to 2000 ppm ELL for 4 hr. A. Exposed
gills with no particulate and mucous accumulation (2000X). B.
Exposed gills showing particulates (2000X). C. Exposed gills
showing mucous layer (2000X). D. Exposed gills showing no
particulates or mucous (2000X). .............................................

Figure 7. Percent 02 saturation (A), and 002 partial pressure (B). of

blood taken from steelhead trout exposed to 4000 ppm ELL for 4
hr. (A. 0.05<p<0.10; B. p>0.10). Error bars represent S.E. .........

Figure 8. Thirty day steelhead trout growth (increase in body length)
in three treatments, 1) trout fed excess black ﬂy larvae for 5 d that
were treated with 30 ppm ELL for 24 hr, 2) same but trout also
exposed to 30 ppm ELL for 24 hr, 3) control treatment where trout
were fed excess live black ﬂy larvae over 5 d and water contained
no ELL Bars headed by the same letters represent no signiﬁcant
difference between means (p>0.10). Error bars represent S.E.

CHAPTER 4

Figure 1. Diagrammatic representation of experimental design used

to measure diet changes of Ammua lycgzias following Eamllns

Lhmjngiensjﬁ var. imelensis disturbance, in Manganese River.
AP= ELL application point. ..................................................

Figure 2. Macroinvertebrate content in the foreguts of AM
119911135 nymphs BEFORE and AFTER disturbance from ELL in
the Manganese River disturbance experiment (11 = 6). [ns = not
signiﬁcant @ p = 0.05, "' = p < 0.05, ** = p < 0.01, *** = p < 0.001]
[Power (= 1 - beta); A. = 0.95, B. = 0.10, C. = 0.79, D. = 0.94]. ........

109

111

113

114

130

137

 

 

Figure 3. .\
the Man
signiﬁes
[Power (

Figure 4. 1‘
MR
Mangan
0.05, "
10d : 0.6
= 0.98 8;
0.28].

Figure 5. l
the Med
(115 = n01
0.001] [F
10d : 0.

Figure 6. F
stream
the 5am.
P=005

and BL,
GXpex-im
(HS = I10
0.0011 {2'
0.41, U!

 

Figure 3. Macroinvertebrate content in the foreguts of Annamaria
magma nymphs ABOVE and BELOW disturbance from ELL in
the Manganese River disturbance experiment (11 = 6). [ns = not
signiﬁcant @ p = 0.05, * = p < 0.05, ** = p < 0.01, *** = p < 0.001]
[Power (= 1 - beta); A. = 0.52, B. = 0.08, C. = 0.42, D. = 0.80]. ........

Figure 4. Macroinvertebrate content in the foreguts of Acmnaazia
Wand EaLagnaLina madja nymphs BEFORE (= 0d) and
AFTER (= 5d & 10d) disturbance from ELL in the Medora and
Manganese Rivers (n=10). [ns = not signiﬁcant @ p = 0.05, * = p <
0.05, ** = p < 0.01, *** = p < 0.001] [Power ( = 1 - beta); A. 5d = 0.62 &
10d = 0.67, B. 5d = 0.27 & 10d = 0.27, C. 5d = 0.57 & 10d = 0.67, D. 5d
= 0.98 & 10d = 0.94, E. 5d = 0.57 & 10d = 0.52, F. 5d = 0.27 & 10d =
0.28]. ...................................................................................

Figure 5. Macroinvertebrate content in the foreguts of Ma
Matias nymphs ABOVE and BELOW disturbance from ELL in
the Medora-Manganese Rivers disturbance experiment (n = 10).
[ns = not signiﬁcant @ p = 0.05, * = p < 0.05, ** = p < 0.01, *** = p <
0.001] [Power (= 1 - beta); A. 5d = 0.89 & 10d = 0.53, B. 5d = 0.05 &
10d = 0.05, C. 5d = 0.89 & 10d = 0.53]. ........................................

Figure 6. Predation on three prey taxa by A. Ismla signata and B.
Earagnatina madja during prey choice experiments conducted
streamside in artiﬁcial channels (n=10). [prey taxa followed by
the same letter within each graph are not statistically different @
p = 0.05] [Power (= 1 - beta); A. = 0.86, B. = 0.11]. ........................

Figure 7. Mean number of prey consumed between M m
and Eamgnatjaa madja predators during prey choice
experiments conducted streamside in artiﬁcial channels (n=10).
[ns = not signiﬁcant @ p = 0.05, * = p < 0.05, ** = p < 0.01, *** = p <
0.001] [PoWer (= 1 - beta); Simaliam = 0.07, Eaeﬁa = 0.47, Em:
0.41, total = 0.24. ...................................................................

138

143

 

(Power =
signiﬁca
PWFP”
emiron.

 

 

 

Figure 8. Feeding response curves for A. mm B.
Impedadicala. and C. Manages. over a range ofBaetis and
Sinnﬂjam prey compositions. [Difference between slopes of the
two regression lines tested at p = 0.05; A. no signiﬁcant difference
(Power = 0.84), B. signiﬁcantly different (Power = 0.98), C.
signiﬁcantly different (Power = 0.98)]. Lines represent a constant
prey proportion consumed by predators over increasing % prey in
environment.

xiii

INTRODUCTION

Eacﬂlaa maringiansja var. imalansia de Barjac (ELL) is a bacterium
applied to lotic systems for reducing black ﬂy (Diptera: Simuliidae)
populations both experimentally and in some large scale control programs.
Discovered in the 19708 (Goldberg and Margalit 1977), ELL use has
increased over the years in North America and world-wide (WHO 1979,
Gaugler and Finney 1982, Lacey and Undeen 1987). It targets the larval
stages of black ﬂies to ultimately reduce adult populations.

Reducing adult populations of black ﬂies is important for medical and
economic reasons. Some species of black ﬂies carry and transmit
Onchocerciasis (or "river blindness"), a disease caused by the nematode
Qnghacama Was, that affects humans in much of tropical Africa and
South America (Philippon 1987). Over 20 million people are estimated to
suffer some degree of vision impairment due to this disease (Molloy 1984).
Black ﬂies also cause losses in animal agriculture throughout the world
(Fredeen 1985, Cupp 1987). These effects include death or morbidity,
reduced milk and meat production, and reproductive dysfunction. In
addition, black ﬂies are commonly a nuisance and interfere with many
human outdoor activities including recreation, lumbering, mining and
building activities during spring and summer (J amnback 1969, Fredeen
1977, Newson 1977, Merritt and Newson 1978, Kim and Merritt 1987).

ELL is a desirable black ﬂy control agent, due to its efﬁcacy and

apparent minimal non-target effects (Molloy and J amnback 1981, Lacey et

aL 1981C
Dejoux et
Undeen 15
are susce}
accumula'
have an a
proteolyti
inclusion
remove <
Merritt e'
Bil. is a
results if
death (D1
Gill et al
Unlliﬁes
1990).
The e
thoroug}

hi‘Sting i;

2

al. 1982, Colbo and O'Brien 1984, Pistrang and Burger 1984, Back et al. 1985,
Dejoux et al. 1985, De Moor and Car 1986, Gibbs et al. 1986, Lacey and
Undeen 1987, Merritt et al. 1989, Lacey and Mulla 1990). Black ﬂy larvae
are susceptible to ELL for three primary reasons; they physically
accumulate and ingest the bacterial particles via their ﬁlter feeding, they
have an alkaline midgut, and they have the proper complement of
proteolytic enzymes that dissolve and active the toxin-containing parasporal
inclusion bodies (Lacey and Mulla 1990). Feeding black ﬂy larvae passively
remove < 200 um sized particulates from the water column (Wotton 1977,
Merritt et al. 1978, Ross and Merritt 1987), thus ingest ELL particles, when
ELL is applied to streams. Their alkaline midgut and associated enzymes
results in binding of the toxic crystal to the larval gut wall, leading to larval
death (Dubois and Lewis 1981, Aronson et al. 1986, Lacey and Mulla 1990,
Gill et a1. 1992). The absence of any of these three factors reduces or
nulliﬁes the susceptibility of a given species to ELL (Lacey and Mulla
1990).

The environmental impact of ELL for black ﬂy control should be studied
thoroughly before large-scale control programs are implemented. Rigorous
testing is necessary to measure potential detrimental effects on lotic
systems, including the effects on non-target organisms. Some of these
effects may include 1) direct toxicity to non-target organisms, 2) indirect
lethal and sublethal impacts on non-target organisms (i.e., consumers of
ELL-contaminated larvae), and 3) food web disturbance aﬁer black ﬂy prey
biomass removal.

Several investigators have studied some aspects of ELL effects on non-

 

 

target strea
and investig
mortality 01
and dovmst
the literatu
reductions r
Mirna et a1.
Car and De
al 1986, Me
pseudorepli.
studies to d

Some ne
susceptible
1981, Gaug
1986,1353:

et al. 1989

 

”DOM
Stream.

PiStrar,

 

Ephemera
De Moor (ll
atmbute
”Matte

target stream organisms. Most of these studies have been ﬁeld oriented
and investigated invertebrate drift, population ﬂuctuations, or direct
mortality of non-target organisms either 'above and below' (= upstream
and downstream), or 'before and after' a ELL-application . The majority of
the literature reports little if any direct mortality on and population
reductions with non-target organisms (Colbo and Undeen 1980, Ali 1981,
Miura et al. 1980, Molloy and Jamnback 1981, Lacey et al. 1982, Burton 1984,
Car and De Moor 1984, Pistrang and Burger 1984, Back at al. 1985, Gibbs et
a1. 1986, Merritt et al. 1989, Xenon 1990, Lacey and Mulla 1990). However,
pseudoreplication (Hurlbert 1984) has plagued interpretations of many ﬁeld
studies to date.

Some nematoceran Diptera (other than black ﬂies and mosquitoes) are
susceptible to the toxin, especially Chironomidae (Miura et al. 1980, Ali
1981, Gaugler and Finney 1982, Car and De Moor 1984, De Moor and Car
1986, Lacey and Undeen 1987). Mortality of Chironomidae genera;
Enkiaﬂjemna , EQlypadjhm (Back et al. 1985) and Wm (Merritt
et al. 1989) was observed following ELL-applications. Back et al. (1985)
reported mortality in Blephariceridae after applying ELL to a Canadian
stream.

Pistrang and Burger (1984) observed increased drift of some
Ephemeroptera and Trichoptera in response to ELL-applications. Car and
De Moor (1984) found increased mortality in some Ephemeroptera, but
attributed mortath to the handling of these organisms. Apparent

population increases of certain Plecoptera, Coleoptera, non-target Diptera,

 

31—:—

I

[in-m

TrichoptE!
in some or:
increases

A majr
changes 0
precision
may be ur
nature of

Few it
on fish, a
oOlIlponer
Occur in r
Amine 1
ﬁngering
(Leonard
°f the die

Lebm
these ﬁsl
mortaliq
Was app]
madam
the pres,
nOt 341.1.
mortalit

Littl£

 

4

Trichoptera and Ephemeroptera were observed following ELL-application
in some cases (Molloy and J amnback 1981), but they attributed these
increases to sampling error.

A major problem with stream ecology studies measuring population
changes or mortality in macroinvertebrate populations, is the lack of
precision when sampling. Subtle changes (e.g., in population densities)
may be undetectable, or appear insigniﬁcant, given the heterogenous
nature of stream environments.

Few investigations have addressed the direct or indirect effects of ELL
on ﬁsh, even though black ﬂy larvae and adults are common ﬁsh diet
components. Black’ﬂies are an important ﬁsh food whenever these insects
occur in signiﬁcant numbers (Mann and Orr 1969, Smith and Page 1969,
Amrine 1984, Sato 1987 , Davies 1991). The stomach contents of brook trout
ﬁngerlings in a Michigan river contained 85% midges and black ﬂies
(Leonard 1941). Hess (1983) reported that black ﬂies were a signiﬁcant part
of the diets of 24 ﬁshes in a West Virginia river.

Lebrun and Vlayen (1981) reported 50% mortality in mania spp. when
these ﬁsh were exposed to unusually high doses of ELL, but found no
mortality with these ﬁsh, and sticklebacks and mosquitoﬁsh, when ELL
was applied at labeled rates. Fortin et al. (1986) measured up to 86%
mortality in brook trout fry following a ELL-application, but explained that
the preservative agent xylene was the factor responsible for mortality and
not ELL itself. Merritt et al. (1989) observed no short term growth effects or
mortality from ELL-application on rock bass in a Michigan stream.

Little research has been conducted on 'secondary' or 'indirect' effects

 

 

(food-ch21?
such as 6
success. 1
consumer
Marysko
predator
with EL
predator
fecundit;
nymph ]
mosqm
Paramet

Tooc
hank
high mt
01388110;
wOlgan
biologjC
the San
biochen
and thi
black ﬂ
insmac
times a

5

(food-chain toxicity) of ELL on non-target organisms. Life history aspects
such as developmental rates, molting success, emergence time and
success, and fecundity of non-target organisms may be altered when
consumers ingest ELL-contaminated black ﬂy larvae. Olejnicek and
Maryskova (1986) reported no marked increase in mortality with the lentic
predator mm glam when they were fed mosquito larvae intoxicated
with ELL. Aly and Mulla (1987) obtained similar results with three lentic
predators of mosquito larvae. They studied mortality, emergence time,
fecundity and consumption rates with the backswimmer Manama
nndalata, the dragonﬂy nymph W W, and damselﬂy
nymph Enallagma giyjla, after these predators consumed ELL-intoxicated
mosquito larvae. They found no signiﬁcant changes in these life history
parameters, except for a 50% reduction in consumption by H. nmlnlata.

'Food-chain' toxicity has been reported for predatory mites feeding on
insecticide-treated prey (McClanahan 1967). These predators experienced
high mortality after preying on mites fed host plants treated with
organophosphate and carbamate insecticides (Binns 1971, Lindquist and
Wolgamott 1980). ELL diﬂ‘ers from these conventional insecticides; it is a
biological rather than a chemical control agent. Therefore it may not have
the same toxic effects as do chemical insecticides. There may also be a
biochemical change taking place in the black ﬂy prey after they ingest ELL,
and this, in turn, may secondarily affect predators that consume treated
black ﬂies. Aly (1985) demonstrated that ELL grows and sporulates inside
infected mosquito hosts. Spore counts in infected cadavers increased to 700

times and toxicity associated with sporulation continued to increase for

several days following ELL treatment (Aly et al. 1985). Depending on
consumption rates, consumers could indirectly ingest (via contaminated
hosts) ELL at concentrations several thousand times greater than that
which is lethal to target insects.

Studies are urgently needed on (1) adverse effects on consumers of ELL-
intoxicated black ﬂy larvae, and (2) loss or alteration of a signiﬁcant
component of the food web. Any intervention disrupting or restructuring
the aquatic community could (negatively or positively) affect individuals at
other trophic levels. Relevant questions regarding ELL use in streams
include: What happens to the ELL-treated black ﬂy larvae? What percent of
the dead larvae remain attached to their substrate, being available to their
natural predators, as opposed to releasing and drifting in the water
column? Do dead, treated black ﬂy larvae decompose or are they eaten by
predators and detritivores? If black ﬂies are an important component of
insect predator diets, with numerous predatory insect species representing
over twenty aquatic families (in Coleoptera, Diptera, Hemiptera,
Megaloptera, Plecoptera, and Trichoptera Orders) (See Davies 1981 and 1991
for a complete list), do predators continue consuming black ﬂies after the
black ﬂies have been killed by ELL? If detritivores and predators consume
ELL-contaminated black ﬂies, are they secondarily affected by ingesting
those contaminated larvae? Do predators switch to alternate prey types
after black ﬂies are removed from a stream?

This research was designed to investigate the consequences, to selected

aquatic macroinvertebrates and trout, of treating stream ecosystems with

 

 

33L (or lad
macroinvert
contaminatr
features (e
after they

contamina
larvae dec

macrroinve

7
ELL, for larval black ﬂy control. Objectives included determining: 1) if
macroinvertebrate predators and detritivores, and trout consume ELL-
contaminated black ﬂy larvae, 2) if ELL indirectly affects life history
features (e.g.: mortality, growth, emergence) of these macroinvertebrates
after they consume ELL-intoxicated larvae, 3) the duration that ELL-
contaminated larvae remain "available" to consumers before the dead
larvae decompose or drift downstream, and 4) if feeding habits of these
macroinvertebrate predators change following ELL-application.

 

 

Ali, A. l
again
J. Inv

Aly, C. l

in the
45:18

Amnsoa,

 

relate

Back, 0,,

dosag

LITERATURE CITED

Ali, A. 1981. Eaglllna Lharingianaia serovar. israalansis (ABG-6108)

against chrironomids and some non-target aquatic invertebrates.

J. Invertebr. Pathol. 38:264-272

Aly, C. 1985. Germination of Eagillaa Lhmjnglanaia var. iaraalanaja spores
in the gut of Aadaa larvae (Diptera: Culicidae). J. Invert. Path.

45:1-8.

Aly, C., & M.S. Mulla. 1987. Effect of two microbial insecticides on
aquatic predators of mosquitos. J. Appl. Ent. 103:113-118.

Aly, C., M.S. Mulla, & B.A. Federici. 1985. Sporulation and toxin
production by Bacillus thminaiensis var. ismelensis in cadavers of
mosquito larvae (Diptera: Culicidae). J. Invert. Path 46:251-258.

Amrine, J .W. 1983. New River black ﬂy food habits study. Unpubl. data.

Aronson, A.I., W. Beckman, & P. Dunn. 1986. Eacillna Lhmjngianaia and
related insect pathogens. Microbial. Rev. 50:1-24.

Back, C., J. Boisvert, J .O. Lacoursiere, & G. Charpentier. 1985. High-
dosage treatment of a Quebec stream with ELL: efﬁcacy against

 

 

 

 

black 1

insect.

Binns. E.
to 521

(Acari

Burton, E
dosage

non~ta

Thesis

Car. M, 5
bentho

51:15.3

Colbo, M.‘

contro

 

Newfc

 

COIb0. M

Simu‘.

9

black ﬂy larvae (Diptera: Simuliidae) and impact on non-target
insects. Can. Entomol. 117:1523-1534.

Binns, ES. 1971. The toxicity of some soil-applied systemic insecticides

*0 Anhis m (Homoptera: Aphididae) and W Deraimilis
(Acarina: Ph3’1308eiidae) on cucumbers. Ann. Appl. Biol. 67:211-222.

Burton, D.K. 1984. Impact of Eaaillug Lhuﬁugiaugh var. hraalaugh in
dosages used for black ﬂy (Simuliidae) control, against target and

non-target organisms in the Torch River, Saskatchewan. Masters

Thesis, Univ. Manitoba, Winnipeg. Manitoba, Canada.

Car, M., & F.C. De Moor. 1984. The response of Vaal River driﬁ: and
benthos to Sjmuljum (Diptera: Nematocera) control using Eagillug

ﬂuuiugiaugh var. mm (H- 14). Onderstepoort J. Vet. Res.
51:155-160.

Colbo, M.H., & H. O'Brien. 1984. A pilot black ﬂy (Diptera: Simuliidae)
control program using Bacillus thudnaiansis var. ismelensis in
Newfoundland. Can. Entomol. 116: 1085-1096.

Colbo, M.H., & A.H. Undeen. 1980. Effect of Eaaillug thuriugiaugh var.
hraalaugh on non-target insects in stream trials for control of
Simuliidae. Mosq. News 40:368-371.

 

 

Cupp, 15-

8 SSOC

8.
m
L;

world

Davies, 1
the ft
Acade

Davies, I

(Simu

Dejoux, C
cible de
tropical
Hydrob

10

Cupp, E.W. 1987. The epizootiology of livestock and poultry diseases
associated with black ﬂies, pp. 387-395. In Kim, KC. & R.W. Merritt
(eds.), Black ﬂies: ecology, population management, and annotated

world list. Penn. State Univ. Press, University Park, PA.

Davies, D.M. 1981. Predators upon black ﬂies, pp. 139-158. In Black ﬂies:
the future of biological methods in integrated control, M. Laird (ed.).

Academic Press Inc., London.

Davies, D.M. 1991. Additional records of predators upon black ﬂies
(Simuliidae: Diptera). Bull. Soc. Vector Ecol. 16: 256-268.

Dejoux, C., F.M. Gibon, & L. Yaméogo. 1985. Toxicité pour la faune non-
cible de quelques insecticides nouveaux utilises en milieu aquatique

tmpical- IV. Le Bacillus thunneicnsis var. ismclcnsis H- 14. Rev.
Hydrobiol. Trop. 18: 31-49.

De Moor, F.C., & M. Car. 1986. A ﬁeld evaluation of Eagillug W

var. hmalangh as a biological control agent for Sjmnlium ahuLtau
(Diptera: Nematocera) in the Middle Orange River. Onderstepoort J.
Vet. Res. 53:43-50.

Dubois, N.R., & F.B. Lewis. 1981. What is Eagillug Lhmjngiaugh? J.
Arboriculture. 7: 233-240.

11

Fortin, C., D. Lapointe, 8: G. Charpentier. 1986. Susceptibility of brook

trout (Salmliuug fanﬁualh) fry to a liquid formulation of Eag'llug
Lhuriugiaugh var. hmalaugh (Teknar) used for black ﬂy control. Can.
J. Fish. Aquat. Sci. 43:1667-1670.

Fredeen, F.J.H. 1977. A review of the economic importance of black ﬂies
(Simuliidae) in Canada. Quaest. Entomol. 13: 219-229.

Fredeen, F.J.H. 1985. Some economic effects of outbreaks of black ﬂies
(Simuliumluggazi (Nicholson an Mickel)) in Saskatchewan. Quaest.
Entomol. 21: 175-208.

Gaugler, R., & J .R. Finney. 1982. A review of Eaaillug Lhmuglangh var.
hmalangh (serotype 14) as a biological control agent of black ﬂies
(Simuliidae). In: D. Molloy, (ed), Biological control of black flies
(Diptera: Simuliidae) with Eaaillug Llnuingiaugh var. hmdangh
(serotype 14): a review with recommendations for laboratory and

ﬁeld protocol. Misc. Pub. Entomol. Soc. Amer. 12:1-18.

Gibbs, KE., F.C. Brautigam, C.S. Stubbs, & L.M. Zibilske. 1986.
Experimental applications of Eaaillug Llnuiugigmgh var. hmalaugh
for larval black ﬂy control: persistence and downstrem carry,
efficacy, impact on non-target invertebrates and ﬁsh feeding.

Maine Life Sci. Agric. Exp. Stn. Tech. Bull. 123:1-25.

12

Gill, 8.8., EA. Cowles, & P.V. Pietrantonio. 1992. The mode of action of
Eaaillua Lhuungiangh endotoxins. Annu. Rev. Entomol. 37: 615-636.

Goldberg, L.J., & J. Margalit. 1977. A bacterial spore demonstrating rapid
larvicidal activity against Anopheles scrgcntii. Hrsnctacnia
unmmrlata' .Qulcxummattus.’ ' AcdcsaagntiandQulcxmmc' ' ng.
Mosq. News. 37:355-358.

Hess, L. 1983. Preliminary analysis of the food habits of some New River
ﬁshes with emphasis on black ﬂy utilization. pp. 15-21. In: W. E.
Cox and M. Kegley (Chairpersons), New River Symposium, Virginia
Polytechnic Institute and State University, Oak Hill, West Virginia.

Hurlbert, SH. 1984. Pseudoreplication and the design of ecological ﬁeld
experiments. Ecological Monographs. 54: 187-211.

J amnback, H. 1969. Bloodsucking ﬂies and outdoor nuisance arthropods of
New York State. Mem. 19. New York State Mus. and Sci. Serv. Albany,
New York.

Kennen, J .G. 1990. Effects of a larvicide Eacillug ﬂuujugiangh var.

hmalaugh on community structure of macroinvertebrates in
streams of the central Adirondack region. Masters Thesis, SUNY
College of Environmental Science and Forestry, Syracuse, N. Y.

13

Kim, KC., & R.W. Merritt (eds.). 1987. Black ﬂies: ecology, population
management, and annotated world list. Penn. State Univ. Press,
University Park, PA.

Lacey, L.A., & M.S. Mulla. 1990. Safety of Eaaillug thuﬁug‘laugh var.
hraalangh and Eaaillug gnhaaziaug to non-target organisms in the
aquatic environment (in press). In: M. Laird, L. A. Lacey, and E. W.
Davidson (eds). Safety of microbial insecticides. C. R. C. Press, Boca
Raton, FL.

Lacey, L.A., & A.H. Undeen. 1987. The biological control potential of
pathogens and parasites of black ﬂies, pp. 327-340. In K. C. Kim and
R. W. Merritt (eds). Black ﬂies: ecology, population management, and
annotated world list. Penn State Univ. Press, University Park, PA.

Lacey, L.A., A.H. Undeen, & M.M. Chance. 1982. Laboratory procedures
for the bioassay and comparative efficacy evaluation of Eacillug

Lhuringiangh var. israclcnsis (serotype 14) against black ﬂies
(Simuliidae). In: D. Molloy, (ed), Biological control of black ﬂies

(Diptera: Simuliidae) with Bacillus thminsrcnsis var. israclcnsis
(serotype 14): a review with recommendations for laboratory and

ﬁeld protocol. Misc. Pub. Entomol. Soc. Amer. 12:19-23.

Lacey, L.A., H. Escaffre, B. Philippon, A. Seketeli, & P. Guillet. 1982.

14

Large river treatment with Bacillus thurinaicnsis var. isrzaclcnsis (H-
14) for the control of Simuljum damuagum s.I. in the Onchoceriasis
Control Programme. Z. Tropenmed. Parasitol. 33:97-101.

Lebrun, P., & P. Vlayen. 1981. Etude de la bioactivite compares et des
eﬁ‘ects secondaries de Eaaillug Lhuzlugiaugh H-14. Z. Agnew. Entomol.
91: 15-25.

Leonard, J .W. 1941. Some observations on the winter feeding habits of
brook trout ﬁngerlings in relation to natural food organisms present.

Trans. Am. Fish. Soc. 71:219-227.

Lindquist, R.K., & M.L. Wolgamott. 1980. Toxicity of acephate to

thtcseiulus mrzsimﬂis and Ietranxchus uriicac. Environ. Entomol.
9:389-392.

Mann, R., & D. Orr. 1969. A preliminary study of the feeding relationships
of ﬁsh in a hard-water and a soft-water stream in Southern England. J.
Fish. Biol. 1: 31-44.

McClanahan, R.J. 1967. Food-chain toxicity of some systemic acaricides to
predaceous mites. Nature. 215: 1001-1009.

Merritt, R.W., & H.D. Newson. 1978. Ecology and management of
arthropod populations in recreational lands, pp. 125-162. In G.W.

15

Frankie and 0.8. Koehler (eds.). Perspectives in urban entomology,

Academic Press, New York.

Merritt, R.W., M.M. Mortland, E.F. Gersabeck, & D.H. Ross. 1978. X-ray
defraction analysis of particles ingested by ﬁlter-feeding animals.
Entomol. Exp. Appl. 24: 27-34.

Merritt, R.W., E.D. Walker, M.A. Wilzbach, KW. Cummins, 8: W.T.
Morgan. 1989.. A broad evaluation of ELL for black ﬂy (Diptera:
Simuliidae) control in a Michigan river: efﬁcacy, carry and non-target
effects on invertebrates and ﬁsh. J. Amer. Mosq. Cont. Assoc. 5:397-
415.

Miura, T., R.M. Takahashi, & F.S. Mulligan, III. 1980. Effects of the

bacterial mosquito larvicide, Eacillug Lhuungiangh serotype H- 14 on
selected aquatic organisms. Mosq. News 40:619-622.

Molloy, D. 1984. The black ﬂy debate. New York State Mus., State Educ.
Dept. 17: 7-10.

Molloy, D.P., & H. J amnback. 1981. Field evaluation of Eaalllug

thmiugiaugig var. hraalaugh as a black ﬂy biocontrol agent and its
effect on non-target stream insects. J. Econ. Entomol. 74:314-318.

16

Newson, H.D. 1977. Arthropod problems in recreational areas. Ann. Rev.
Entomol. 22: 333-353.

Olejnicek, J ., & B. Maryskova. 1986. The inﬂuence of Eaaillug

thun'naicnsis var. ismclcnsis on the mosquito predator Nutcnccta
glance. Folia Parasitol. 33:270-230.

Philippon, B. 1987. Problems in epidemiology and control of West African
onchcerciasis, pp. 363-373. In Kim, K.C. & R.W. Merritt (eds.), Black
ﬂies: ecology, population management, and annotated world list. Penn.

State Univ. Press, University Park, PA.

Ross, D.H., & R.W. Merritt. 1987. Factors affecting larval black ﬂy
distributions and population dynamics, pp. 90-108. In Kim, KC. & R.W.
Merritt (eds.), Black ﬂies: ecology, population management, and

annotated world list. Penn. State Univ. Press, University Park, PA.

Pistrang, L.A., & J .F. Burger. 1984. Effect of Eaaillug muduglangh var.
hraalangh on a genetically-deﬁned population of black ﬂies

(Diptera: Simuliidae) and associated insects in a montane New

Hampshire stream. Can. Entomol. 116:975-981.

Sato, G. 1987. Identiﬁcation of ﬁsh predators of simuliid larvae in
Joinville/SC. Cienc. Cult. (Sao Paulo). 39: 962-966.

17

Smith, P., & L. Page. 1969. The food of spotted bass in streams of the
Wabash River drainage. Trans. Am. Fish. Soc. 98:647-651.

World Health Organization. 1979. Data sheet on the biological control

agent Eacillug thudngiangh serotype H-14 (de Barjac 1978). Wld.
Hlth. Org. mimeo. doc. WHONBC/79.750 Rev. 1. VBC/BCDS/79.01, 46

PP-

Wotton, R. S. 1977. Sampling moorland stream black ﬂy larvae (Diptera:
Simuliidae). Arch. Hydrobiol. 79:404-412

CHAPTER 1

PORTABLE ARTIFICIAL STREAMS FOR FIELD AND LABORATORY
EXPERIMENTS WITH AQUATIC MACROINVERTEBRATES

ABSTRACT

An artiﬁcial stream unit that houses 40 replicated 480 ml circular
channels was designed for conducting aquatic macroinvertebrate
laboratory and ﬁeld studies. The unit measured 135 cm long x 91 cm wide x
80 cm high. Water was delivered to the unit with electric water pumps, in
the laboratory, and by gravity in the ﬁeld. Water current velocity was
precisely regulated in channels and the stream unit accommodated
current velocities from ==0 to >30 cm/s.

The stream unit was laboratory and ﬁeld tested using insects from
various functional feeding groups, including predators, (Plecoptera,
Odonata, Trichoptera), shredders and collectors (Diptera and
Ephemeroptera), and grazers (Ephemeroptei‘a). The unit proved to be a
useful design for predation, growth and development, emergence, and
mortality studies. It could also be used for single- and few-species toxicity
testing, insecticide efﬁcacy trials, herbivory studies, along with other
studies using aquatic invertebrates. It could be easily modiﬁed for other
applications. Construction cost was $500 and required about 80 hrs to build.

18

19

The unit was simple to transport and operate, and was efficiently operated

by one person.

2)

INTRODUCTION

A major problem inherent in stream ecology research has been the
inability to incorporate true replication of treatments, otherwise known as
pseudoreplication (Hurlbert 1984). A common approach for studying the
impact of a stress agent (e.g., predator, pesticide) on one or several
populations of aquatic invertebrates, is to 'treat' a section of river, and
sample 'above and below' or 'before and after' treatments are applied. The
sampling has typically involved repeated samples within each singly
replicated treatment.

To overcome this dilemma, stream ecologists have employed artiﬁcial
streams or channels, encompassing single species to community level
studies, thus avoiding pseudoreplication. Several variations and sizes of
artiﬁcial streams (and ponds) have been developed and used for studying
components of aquatic systems, including food limitation with
macroinvertebrates (Richardson 1991), competition and predator-prey
interactions (W alde and Davies 1980), plant-herbivore interactions
(Lamberti et al. 1989), algal community dynamics (McCormick and
Stevenson 1991), various abiotic factors (Ciborowski et al. 197 7, Steinman
and McIntire 1986, Clements et al. 1989), insecticide efﬁcacy (Gaugler et al.
1980), and single and multiple species toxicity tests (Pontash and Cairns
1989). [See Giesy (1980) and Cuffney (1990) for comprehensive reviews on
experimental ecosystem types and their applications]

Some artiﬁcial streams have incorporated air pumps (Mackay 1981), but

21

more typically, water pumps (Ryer et al. 1979, Matthews at al. 1990) to
generate water current, and have been constructed over a wide range of
sizes. Problems with most designs include high construction costs, the
inability to easily transport the units from laboratory to ﬁeld, or lack
sufﬁcient replicability.

Our objectives were to design artiﬁcial streams that (1) incorporated
high replication, (2) were compact and portable for laboratory and ﬁeld use,
(3) were cost eﬂicient to construct, maintain and operate, (4) were time
efﬁcient to set-up and operate, and (5) accommodated single species to

several species (ca. 6) macroinvertebrate studies.

MATERIALS AND METHODS

I 5 1'5 . l | 1 .
The entire stream unit (Figures 1 & 2A) consisted of 40 individual

channels (Figures 2B, 3A & B). Channels were constructed from 480 ml
plastic food storage containers (Sweetheart® Plastics, Wilmington, MA).
One cm holes were drilled through the bottom centers of each container,
and the internal surfaces of the container roughened with sand paper. A
3.0 cm long x 3.2 cm wide piece of PVC pipe was glued upright to the center
bottom inside each container (Figure 3A), and a circular 3.2 cm diameter
250 um mesh stainless steel screen was glued to the upright and of each
pipe, and served as the water drain (Figures 3A & B). Channels were
secured by placing them inside the bottom portion cut from another plastic
container (Figure 3A) that was glued to a 122 cm long x 91 cm wide x 1.6 cm

 

683m 33m $me .oaﬁ Romano HOV .33
as 3 sense Rein Q .855 Q .33 sag 6 ....aa uses: a: .883 e883... @ .33 Ease
RC .3653 Socneﬁgﬁoboﬁ caucus com ﬁns Sachem 3855.8 one no 5335838 238.9%ch .H 053m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A :5 3: V
A :5 «a: V
H..... ....8
So......:.... 3.8
"balls! I I H H H H I I H H H HHI H H I H H HﬁHFI F I I I I I H I H H H FF
mwb(3 {Ir/awn I I I I I I H I H H HI I I I H I H HI 4 I I I I I I I I I I I I H I I l I
4/, - - - ....t t - , t , , .UHD
n1
...
so»J a. $3.8
m2 ....1
X: ......
“n
25 08 n”
1
3 .n
m>._<> & .h 83
case: a E3322 2: ..UKIurIII'lI”
4 3 .u m>._<>

.o
Bodam>o 4(2dm5. "$.23

 

23A

Figure 2. Artiﬁcial stream unit (A), and channels with (top two channels)
and without (bottom two channels) screen collars (B).

23B

 

61131111915)

 

 

 

 

   

- v.; V.“-

out

 

B 11.6 cm
' ‘ *7

3.2 cm

 

 

 

, 9.0 cr_n

Figure 3. Diagramatic representation of an artiﬁcial channel for aquatic
macroinvertebrate studies (A) without, and (B) with screen collar.

25

thick piece of painted plywood (Figures 1 & 2A). Holes, 2.5 cm diameter,
were drilled through the plastic holder and plywood platform, and 2.5 cm
diameter plastic funnels were ﬁtted through these holes and glued in place.
This served as the drain tube for water exiting the channel (Figure 3A).
Channels were occasionally ﬁtted with a screen collar drain (250 um mesh)
(Figure 3B) that provided increased drain surface area to retard drain
clogging.

The stream unit was designed with four rows of ten channels (Figures 1
& 2A). Water was delivered to the stream unit using four 1.9 cm II) rubber
hoses (Figures 1 & 4) that each brought water to one of four 5.1 cm 11) x 135
cm long holding pipes (Figures 1 & 2A). Water was delivered from the
holding pipes to each channel via individual 4.2 mm ID Tygon® tubes
(Figures 1, 2A, & 3), circulated inside the channels at a desired velocity,
exited through the channel center-drain into 8.0 mm ID Tygon® tubes
(Figure 3A), and ﬁnally into one of four 5.1 cm ID x 135 cm long PVC
outflow pipes that discharged the water (Figure 1).

Water current velocity was precisely controlled in channels by Tee-
valves (D) 'water in' seen in Figure 1. Each one of four valves regulated
current velocity for each ten channel set. Each holding pipe (H) (Figures 1
& 2A) was equipped with threaded sleeves (S) that unscrewed to be taken
apart, allowing the holding pipes to be cleaned internally when necessary.
Each holding pipe was also equipped with an internal overﬂow (located ca.
2 cm from the top inside of each pipe) and an air-bleed valve to release air in

the holding pipes to maintain a high, constant water level in each pipe

 

26A

Figure 4. Gravity-feed pipes (A), 2 water ﬁltration barrels (B), dispensing
bucket (C), delivery hoses (D), and two artiﬁcial stream units (E) during
streamside experiments.

26B

 

 

27

(Figure 1). All working parts were supported by a metal support frame
(S.F.) constructed out of angle-iron (Figures 1 & 2A).

IIEI'ﬁ'll I' 'Ilﬁll

The stream unit was placed in-stream and streamside (Figure 4) during
operation. Water was gravity fed to the unit using two 2.5 cm ID plastic
plumbing pipes (A) (Figure 4) each ﬁtted with gate valves, and into one or
two 122 L plastic water ﬁltration barrels (constructed from Rubbermaid®
refuse containers) (Figure 4) to ﬁlter out large (ca. > 200 um) suspended
particulates. Filters consisted of 2.5 cm thick bonded polyester 'polyﬁl'
sheets, cut and stapled, to form 61 cm wide x 91 cm long ﬁlter bags. The
bags were secured to the previously described gravity-feed pipes.

The water volume entering ﬁltration barrels was controlled by each gate
valve, and water ﬂow was shut off when replacing ﬁlters (every 2-14 d
depending on particulate volume in the water and water volume delivered
to the ﬁlters). Filtered water then ﬂowed through a 3.2 cm opening, ca. 10
cm from the top of each ﬁltration barrel, into a 19 L dispensing bucket, that
ultimately delivered the water to each one of four water holding pipes via 1.9
cm ID rubber hoses (Figure 4). The 1.9 cm rubber hoses passed through
holes drilled 5-10 cm below the upper edge of the 19 L dispensing bucket and
delivered water into the stream units (Figure 4). After passing through the
channels, water was then returned back to the stream.

A critical factor determining successful ﬁeld operation was adequate

local vertical drop from the point where water was taken from the stream to

28

the ﬁltration bucket, and to the stream unit. I estimated 5 m vertical drop
over a 100 m stream reach at our ﬁeld site.

A method was developed to apply treatments of desired liquids to any
number of randomly chosen channels. A 23 L solution reservoir held a
known concentration of solution. The 2.6 L solution dispenser delivered
solution to the channels via 1.3 mm ID 'l‘ygon® tubes that passed through
holes drilled 2.0 cm above the bottom of the dispenser, and into the
channels. This method provided treatments of known and constant
solution concentrations through time because ﬂow rates were dependant on
the vertical drop between the solution level in the dispenser and the distal
end of the outﬂow tube connected to each channel (e.g.; vertical head). The
solution level in the dispenser remained constant through the application
period and was controlled by the height of the downspout of the solution
reservoir. As solution ﬂowed into the channels, the solution level in the
dispenser dropped, allowing air to enter the solution reservoir via its spout,
resulting in additional solution entering the dispenser. In addition to
controlling ﬂow rates, this percolation acted as a stirring mechanism that

kept the solution adequately mixed in the reservoir.

Artiﬁcial stream operation in the laboratory was similar to that in the
ﬁeld, except water was recirculated through a 600 L cooling and ﬁltration
tank using one 1.4 amperes electric water pump per each set of ten
channels. Water ﬂow into the charmels and subsequent current velocity
was regulated with the Tee-valves as described for ﬁeld operation. The

a)

outﬂow pipe however, was mounted facing the opposite direction to what it
was in the ﬁeld, thus returning water directly back to the cooling tank.

Water was ﬁltered, after returning from the channels, using an activated
charcoal/cloth mesh ﬁlter as the water ﬂowed back into the tank.

 

Stream units were ﬁeld and laboratory tested using insects from selected
functional feeding groups (predators, collectors, grazers, and shredders).
These included Plecoptera (Amuaurja sp., Eamgnatiua sp., Igauexla sp.,
Bmstcia sp.), EphemerOptera (Bactis sp.. Qacnis sp.. Ems sp.,
Ephcmcrclla sp-.Bar:a1cptouhlcbia sp., Binhlcnurus sp.). Trichoptera
(Gazamugxaha sp.), Diptera (Simuljum sp., Iiuula sp.), and Odonata
(ngazla sp.), to measure mortality and predation rates, and development in
experiments lasting up to 22 d.

Water velocity was estimated in each channel by recording the time it
took a neutrally-buoyant ﬂoat to complete 10 revolutions. Measures were
replicated 5 times.

Predators were each provided with a 25 x 25 x 1 mm piece of aluminum
placed on the bottom of each channel, providing a thin crevice between it
and the channel bottom, which served as a refuge. Pebble or leaf substrate
was provided during some experiments. Channels were ﬁtted with lids

For ﬁeld experiments, experimental animals were collected, using a D-

frame kick net, from the study site and immediately placed in channels.

30
For predation studies, predators were placed one or ﬁve (depending on
species) to a channel with up to 30 prey that were counted and replaced,
along with predator mortality measured, every 24 hr. See Tables 1 and 2 for
complete experimental details.

Laboratory tests involved using dechlorinated, carbon-ﬁltered tap water
maintained at 8° C. Insects for predator-prey studies were collected from
nearby streams, and transported in coolers to the laboratory. Surplus prey
were maintained within screened enclosures placed in the 600 L tank.
Mortality and consumption rates were recorded daily as described for the
ﬁeld experiments (Tables 1 & 2).

RESULTS

0 Water velocities were uniform across 40 channels (at mean velocity =
21.3 cm/s; SEM = 1.5, and range = 19.2-23.4 cm/s). For experiments
conducted with riﬁle-collected insects (Exps. 1-7), mean water velocity was
21 cm/s, with leaf pack-collected insects (Exps. 8-11), water velocity was
maintained near 5 cm/s.

Mortality rates were low for all species (Tables 3 & 4). Predator mortality
was zero for all species except anatanmha sp. (Table 3). Consumption
rates indicated that all predators actively fed through the experiments, with
consumption rates ranging from 1.36 prey/d for Catamnmha sp. to 8.25
prey/d for Eoyaﬁa sp. (Table 3). Insect mortality rates were also low for the
collectors, shredders, and grazers. (Table 4). Percent daily mortality ranged

from 0.0% (for Impala sp.) to 2.2% (for Simuljmn sp.) (Table 4). Caauh sp.

 

 

 

mom—w.” coﬁaﬁsmcoo GHHHV g
a assess saves s H s was s . . has «nomad Sam m.
mega» :oBQEDmcoo
cm $23.88 Succeed v N. m omum 83385 g 3 32m A.
cash
money :oBQEsmcoo Homuauv g
a ﬁnance 8383 e m S 2533 a ... . mamas. snag 2% m
and
mean» cowaaﬁsmcoo Homﬁuv g
a basses 3385 e m 2 2.2.52 8 ... a... ﬁnd 33 Sam m
moans :oﬁaﬁsmcoo
s asses sees e a a a: sedan than massage as H
@3338 359:6 a camp .85 «53 no.5 36QO 833.5 nmdgmrm a dam
mcouoamamm .anm ”833an 52.884

 

 

...:c: 833m ﬁghts Banana cs 5:... mo mocwacomcoa 2: $3 8 vmcozccoo madmaﬂonxo cosmvmcm .H 2nt

 

mm>mm~

 

 

3:388 3 m 2 888888 8 _ 8 23,: 8
mocmmpmam :33
338 w 323:
3:388 8 mm m o: 883888 38an 33,: 2
mm>mo_
3:388 8 u m m 888888 Ilqlss a. 33: m
mzuhuauﬂamm
GE 883 3%
3:388 8 8 o: 38 883888 23,: w
Smuv
3:388 8 a m 282: 83:88 333% 23: 8
8:388 8 2 S. 8 88:88 : H :3 o
@8338 55326 8 358% Sn 33:: voom @393 «:3 835 833833 a: dxm
mpgmﬁmpmm dxm 30me 8 ”8038.5

 

 

£85 8883 Emacs»: 358:0 ow 2: mo 8:888th 25 $8 3 33358 mummﬁtmmxm

.N 2nt

33

Table 3. Predation and predator mortality rates during predation
experiments using the 40 channel artiﬁcial stream unit.

 

 

 

Prey eaten per % predator
predator per d mortality per d

Exp. # Predator species 1: 3; SEM x :1: SEM

1 Agrgnegria lyggrias 2.0 :t 0.3 0.0%

2 Paramgtina media 7.0 :t 1.9 0.0%

3 1.52% m 3.1 d: 0.5 0.0%

4 ﬂeratgpsyche sparna 1.4 i 0.1 4.6 :t 1.1%

5 Mia Vin 8.3 :I: 2.5 0.0%

 

34

Table 4. Mortality rates, cummulative molt, and emergence of aquatic
insects during collector-shredder-grazer experiments using the 40 channel
artiﬁcial stream unit.

 

 

 

% insect % cummulative molt
mortality per d (or) emergence
Exp. # Insect taxa tested x j; SEM x 1: SEM
6 i 1i . 1.1 d: 0.7% ---
7 Simulium 2.2 :t 0.5% ---
Em 0.4 i 0.4%
Baetis 0.9 i 0.4%
8 r 1 hl ia 0.6 :t 0.4% ---
Prgstoia 1.4 i 0.9% ---
E h m r 11a 0.6 :t 0.4%
9 m 0.0%
10 a ni 0.8 i 0.0% 66.0 :1; 8.7% (molt)

60.0 i 7.8% (emerg.)

11 Sipllimm 1.4 i 1.4%

 

 

35

demonstrated high molt (66.0%) and adult emergence (60.0%) levels (Table
4). '

DISCUSSION

The artiﬁcial stream unit performed well under both laboratory and
ﬁeld settings. During laboratory operation, the unit ran continuously
without ﬂaw for 8 months, and for a continuous 3 months during ﬁeld
experiments. Aquatic insects appeared to behave 'normally', exhibiting no
unusual aggressive behavior during the duration of the experiments.
Predators immediately began to feed when placed in channels containing
prey. Most predators used the refuge provided them when they were
inactive.

When placed in channels, black ﬂies initially moved around on
substrate until they found suitable microhabitats, then remained stationary
until disturbed by other insects. Grazers fed on microbes that colonized
substrates within the channels as evidenced by 'trails' of grazed, microbe-
free patches.

Low mortality rates indicated that the channels were suitable for aquatic
insect studies using insects within these size classes. Although the
channels worked well for predation and insect developmental studies, they
would likely be too small for behavioral-type studies, or studies that used
larger animals. However, larger channels could be used in place of these
480 m1 channels.

36

One of the beneﬁts of this unit was the ease of transporting channels
from the laboratory to the ﬁeld, and back. The units were easily handled by
one person, and were time efﬁcient to operate. Counting and adding
insects to the channels was simple due to their small size, dimensions, and
light color. Another beneﬁt was the precise control and between-channel
uniformity of water velocity, with velocities from 0-30 cm/s or more easily
obtained by simply regulating the Tee-valves.

The units were highly suitable for ﬁeld operation provided that there
was sufﬁcient local vertical drop to create the desired water pressure and
ﬂow velocity. Insufﬁcient vertical drop would prevent stream unit
operation at the higher water velocities, but could be circumvented by
incorporating electric water pumps.

Occasional channel cleaning was necessary and was accomplished by
using small brushes. The channels needed more frequent cleaning in the
ﬁeld than the lab due to suspended sediments in stream water. The water
holding pipes also needed to be cleaned and ﬂushed ca. monthly during
ﬁeld use, which was quickly accomplished by unscrewing the threaded
sleeves and brushing the inside of each PVC pipe with a long handled
brush. The water ﬁlters also needed replacing every two to 14 d.

In hindsight, recommended modiﬁcations would include a single 15 cm
ID water holding pipe (as opposed to four 5.1 cm ID pipes) delivering water
to all 40 channels. Other modiﬁcations include channels constructed from
higher quality plastic. The channels held up for two years under heavy use
(being used in experiments in addition to those described here) but began to
break at the end of the second year.

37

In conclusion, the 40 channel artiﬁcial stream unit met our objectives.
It was inexpensive (ca. $500) and time efﬁcient to build, set up, and operate,
it was compact and portable for laboratory and ﬁeld experiments, and
incorporated high replication. It appears to be a good design for single- and
few-species toxicity testing, predation, herbivory, and other feeding studies,

as well as competition studies.

38

ACKNOWLEDGMENTS

I thank G. Arntsen for offering access to his property allowing me to
conduct this study. I also thank B. Peckarsky and D. Allen for their helpful
assistance and suggestions on channel design, and J. Velarde for
assistance with channel construction. Channels were a modiﬁcation of
those originally designed by S. Walde. This study was supported, in part, by
Sport Fishing Institute Fund grant SFRP-90-26 awarded to RWM and MSW,
Northeast Regional Black Fly Project NE-118, and Michigan State
University Cooperative Extension Service and Agricultural Experiment
Station.

I... 1......1ﬂur

39

LITERATURE CITED

Ciborowski, J .H., P.J. Pointing, & L.D. Corkum. 1977. The effect of current

velocity and sediment on the drift of the mayﬂy Ephememlla maria
Mcdunnough. Freshwater Biology. 7: 567-572.

Clements, W.H., J .L. Farris, D.S. Cherry, & J. Cairns. 1989. The inﬂuence
of water quality on macroinvertebrate community responses to copper

in outdoor experimental streams. Aquatic Toxicol. 14: 249-262.

Cuﬁ'ney, TE 1990. Experimental ecosystems: Application to ecoton'cology.
North American Benthological Society, Technical Information
Workshop Proceedings, May 1990, VPI&SU, Blacksburg, VA. 128 pp.

Gaugler, R., D. Molloy, T. Haskins, & G. Rider. 1980. A bioassay system for
the evaluation of black ﬂy (Diptera: Simuliidae) control agents under
simulated stream conditions. Can. Ent. 112: 1271-1276.

Giesy, J .P. 1980. Microcosms in Ecological Research. National Technical
Information Center, Springﬁeld, VA. 1110 pp.

Hurlbert, SH. 1984. Pseudoreplication and the design of ecological ﬁeld
experiments. Ecological Monographs. 54: 187-211.

40

Lamberti, G.A., S.V. Gregory, L.R. Ashkenas, A.D. Steinman, & C.D.
McIntire. 1989. Productive capacity of periphyton as a determinant of
plant-herbivore interactions in streams. Ecology 70: 1840-1856.

Mackay, R.J. 1981. A miniature laboratory stream powered by air bubbles.
Hydrobiologia. 83: 383-385.

Matthews, W.J., F.P. Gelwick, & T.J. Gardner. 1990. A simple system of
replicated recirculating experimental streams. J. Freshwater Ecol. 5:

437-443.

McCormick, P.V., & R.J. Stevenson. 1991. Mechanisms of benthic algal

succession in lotic environments. Ecology. 72: 1835-1848.

Pontasch, K.W., & J. Cairns. 1989. Establishing and maintaining
laboratory-based microcosms of rifﬂe insect communities: their

potential for multispecies toxicity tests. Hydrobiologia. 175: 49-60.

Richardson, J .S. 1991. Seasonal food limitation of detritivores in a montane

stream: an experimental test. Ecology. 72: 873-887.

Ryer, C.H., J .A. Wetmore, & J .L. Gooch. 1979. An artiﬁcial stream design
for lotic invertebrates. Amer. Midland Nat. 101: 447-449.

Steinman, A.D., & C.D. McIntire. 1986. Effects of current velocity and light

41

energy on the structure of periphyton assemblages in laboratory
streams. J. Phycol. 22: 352-361.

Walde, S.J., & R.W. Davies. 1984. The effect of intraspeciﬁc interference on
W mums (Plecoptera) foraging behavior. Can. J. Zool. 62: 2221-
2226.

CHAPTER 2

NON-TARGET IMPACT OF BAQELILS IHLIBINGIENSIS VAR.
ISBAELEHSIS IN A LOTIC SYSTEM: DIRECT LETHAL AND

SUBLETHAL FOOD-CHAIN EFFECTS ON SELECTED AQUATIC
INSECTS, AND FATE OF INTOXICATED BLACK FLY (DIPTERA:
SIMULIIDAE) LARVAE.

ABSTRACT

A ﬁeld study was conducted to investigate direct and indirect (food-
chain) effects of Bacillus W var. 181381311815 (ELL) on selected
non-target aquatic insects, fate of ELL-contaminated black ﬂy larvae, and
ELL persistence in stream sediments. Experiments were conducted in-
stream, and streamside using highly replicated artiﬁcial channels, to
measure lethal and sublethal (e.g., drift, feeding, growth, developmental
rates, and adult emergence success and phenology) changes in
Ephemeroptera, Plecoptera, Trichoptera, and Diptera exposed to BALL and
given ELL-contaminated food (black ﬂies and conditioned leaves). Larval
black ﬂy decomposition and release rates were measured in artiﬁcial
channels following BAA. application.

Direct mortality was seen with two species, Ma Mominalis
(Diptera: Tipulidae) and mm binnnctata (Ephemeroptera:

43

Heptageniidae), when exposed to BALL at rates above (but not at)
recommended ﬁeld rates. No signiﬁcant mortality was recorded with the
remaining 14 species tested in artiﬁcial streams. Chironomidae did not
experience higher mortality in m treated stream reaches. Amanda
ms (Plecoptera: Perlidae) nymphs exposed to a high (100 ppm 6 120
mins) 3.“. dosage drifted at a higher frequency than nymphs not exposed.

Predatory stoneﬂies consumed dead (BLL- and heat-killed) and live
black ﬂy larvae at similar rates for all trials, with no measurable adverse
effects, except one where Ismda signata consumed signiﬁcantly fewer
dead than live black ﬂies. Qemtopsxchg spams. predators consumed
signiﬁcantly more ELL-killed than live black ﬂy larvae, apparently
resulting from shorter handling time, without observable ill effects.
Detritivorous stoneﬂies and mayﬂies switched to consuming signiﬁcant
amounts of black ﬂy larvae once the larvae were killed with ELL, also
showing no negative effects. Siphlgnnms mpidns nymphs, however, grew
at higher rates when given ELL-contaminated black ﬂies. This was
attributable to increased food (dead black ﬂy larvae) quality and/or quantity.

EAL-contaminated black ﬂy larvae remained in the stream for up to 16 d
before decomposing beyond recognition. Bioassays using black ﬂy larvae
showed that pool sediments contained viable m for up to 11 d.

m appears harmless to most non-target insects when directly or
indirectly exposed even at high application rates (2 100 ppm m. m
applied at operating ﬁeld dosages poses little direct toxic threat to non-
target species. Also, predators and detritivores that consumed ELL-
contaminated black ﬂies showed no observable adverse affects.

44

INTRODUCTION

Bacillus Wm var. ismelensis de Barjac (ELL) is an effective
biocontml agent used for black ﬂies and mosquitoes, and is targeted at their
larval stages to ultimately suppress adult populations. Its use has
increased nationally and world-wide (WHO 197 9, Gaugler and Finney 1982,
Lacey and Undeen 1987) since its discovery (Goldberg and Margalit 1977).

ELL is a gut poison, thus must be ingested by potential hosts to be
effective. It contains parasporal inclusion bodies that dissolve in the insect
gut, giving rise to proteinaceous toxins that are activated in the host's gut.
The toxin binds to then solubilizes the gut cell wall, and quickly kills the
targeted larvae (Dubois and Lewis 1981, Aronson et a1. 1986). The
combination of ﬁlter feeding behavior of the target Simuliidae and Culicidae
larvae, their relatively high gut pH, and the proper complement of
proteolytic enzymes enables capture, dissolution and activation of the toxin-
containing parasporal inclusion bodies (Lacey and Mulla 1990).

Widespread acceptance of ELL use has been slow, especially for use in
lotic habitats against black ﬂies, because of possible deleterious effects on
beneﬁcial species, and community food web structure, especially ﬁsheries
(Lacey and Mulla 1990). The majority of the literature reports little effect of
3.1.1, on non-target species (Colbo and Undeen 1980, Ali 1981, Molloy and
J amnback 1981, Lacey et al. 1982, Chilcott et al. 1983, Pistrang and Burger
1984, Gibbs et al. 1986, Merritt et al. 1989), including trout (Wipﬂi 1992),
although most studies have investigated only short term (days to months)

45

population changes and macroinvertebrate driﬂ: (reviewed by Molloy 1982).
Many previously published studies on non-target B__LL impacts are difﬁcult
to evaluate because of pseudoreplication problems (Hurlbert 1984),
sampling problems and methods (Merritt and Cummins 1984), adult
emergence, low densities (Merritt et al. 1989), and patchy spatial
distributions of macroinvertebrates in aquatic habitats. These problems
may potentially mask subtle population changes with non-target species,
and population changes with species at low densities following ELL
application.

In spite of these problems, some studies have indicated that certain
non-target species may be sensitive to ELL. Experimental ELL
applications caused increased mortality in certain Chironomidae (Pistrang
and Burger 1984, Rutschke and Grunewald 1984, De Moor and Car 1986),
Blephariceridae (Back et al. 1985), and Ephemeroptera (Car and De Moor
1984). Increased drift with certain Ephemeroptera and Trichoptera
(Pistrang and Burger 1984, Dejoux et al. 1985, De Moor and Car 1986) was
also seen following ELL application. [See Lacey and Mulla 1990 for a
further review of the short term impacts of ELL on non-target species].

Few studies have investigated indirect (food-chain) lethal and sublethal,
and long term (chronic) impacts with non-target species directly or
indirectly linked to black ﬂies (e.g., predators, competitors, and other
members of the food web). There have been some food-chain toxicity studies
and predator-prey studies dealing with ELL use in lentic systems (Lacey
and Mulla 1990), but I know of no published studies lotic systems. Lacey
and Mulla (1990) stated one reason that certain lotic insects are not affected

46

by ELL is their non-ﬁltering mode of feeding (e.g., predator, scraper,
gatherer, detritivore). However, non-target insects may ingest the toxin by
consuming ELL-intoxicated black ﬂies. Aly et al. (1985) found that ELL
grows and sporulates inside dead hosts, and becomes more toxic following
ELL treatment. Little is understood about aquatic insect digestion in many
taxa. Thus, these non-target insects could potentially consume ELL, and
subsequently be lethally or sublethally affected. Lacey (1983) reported
mortality in certain species of predatory mosquitoes when these and their
mosquito prey were simultaneously exposed to varying ELL
concentrations, however prey consumption was not mentioned. Sebastien
and Brust (1981) observed predation on ELL-intoxicated (dead) mosquito
larvae by selected Odonata and Hemiptera without increased predator
mortality, but statistical comparisons were not made. No adverse effects
were seen on Odonata predators after they ingested ELL-contaminated
mosquito larvae (Aly and Mulla 1987 ). Wipﬂi (1992) reported no food-chain
toxicity with brown trout after ﬁngerlings consumed ELL-intoxicated black
ﬂy larvae.

To my knowledge, no published studies have examined functional and
ecological changes in lotic ecosystems associated with transforming live
black ﬂy larvae to dead larvae. Wipﬂi (1992) found signiﬁcant diet changes
With Amnemia 11:21:13.5 (Newman) and Eametina media (Walker)
following ELL applications, where A, 13mins did not switch, and where E
media switched to alternate prey.

The objectives of this study were to evaluate short term direct and

indirect impact of ELL in lotic systems by investigating (1) non-target

47

species mortality resulting from direct ELL exposure, (2) consumption
rates of ELL-contaminated black ﬂies by predators and detritivores, (3)
changes in life history parameters (nymphal mortality and growth rates,
and adult mass, emergence success and phenology) of selected non-target
insect species resulting from indirect ELL exposure through consuming
ELL-contaminated black ﬂy larvae, (4) the fate of ELL-contaminated black
ﬂy larvae (excluding loss due to benthic macroinvertebrate consumers),
and (5) persistence of ELL toxin in stream sediments. Components of this
study utilized streamside artiﬁcial channels (Wipﬂi 1992) speciﬁcally
designed to alleviate the problems of pseudoreplication, sampling, adult
emergence, and inherently low densities and contagious spatial -
distributions of macroinvertebrates in lotic systems; problems that have
plagued many previous studies. This study focused on macroinvertebrate
species known or suspected to be directly sensitive (Diptera,
Ephemeroptera, Trichoptera), or indirectly sensitive (Plecoptera predators
via consuming ELL-contaminated black ﬂy larvae) to ELL.

MATERIALS AND METHODS

L_S.tndlaitc

The ﬁeld studies were conducted April-July during 1990 and 1991 in
Medora River, Aetna Creek, and Manganese River in Keweenaw County,
Michigan, USA. Medora and Manganese Rivers are second order
streams that drain Lakes Medora and Manganese, respectively. Aetna

48

Creek is a ﬁrst order stream draining a small unnamed lake. Laurentian
shield predominates the regional geology giving rise to lotic systems
characterized by bedrock erosional areas, some containing sediments
within a wide particle size range (pebble through boulder), broken up by
small pools. These sites were chosen based on certain criteria; high black
ﬂy density, wide diversity and abundance of other macroinvertebrate fauna,
and appropriate site speciﬁc vertical drop for artiﬁcial stream use.

Experiments were carried out using artiﬁcial streams (Wipﬂi 1992) for
all except the Chironomidae and one Tipulidae trial (which were conducted
in-stream). The artiﬁcial streams consisted of 80 circular channels that
utilized gravity delivered stream water, and were operated at the Medora
River site. Insects were collected at or near the ﬁeld site (immediately
before experiments began), except for Atheﬁx sp. and Qaenis sp. which
were transported from nearby sites. ELL (Teknar® HP-D, #9602579)
product used in this study was supplied by Zoecon Corp., Dallas, Texas,
U.S.A.
“MW

Three water current velocities were maintained in the channels,
depending on experimental insects under investigation. In experiments
with riﬂe-collected insects (predators and baetid mayﬂies) water velocity
averaged 21 cm/s. With leaf pack-collected and pool-collected insects, water
velocity average 11 and 0-5 cm/s, respectively. All channels were ﬁtted with

clear lids during experiments.

: . “w: -; ...: u. ‘ up; .. ‘.:- : H. a... . ”a...“
channels.

Selected aquatic insect species (Table 1) were directly and/or indirectly
exposed to ELL to investigate speciﬁc changes associated with these insects
during and after exposure to this toxin: drift behavior, feeding habits,
survival, growth, and adult emergence success, and adult emergence
phenology. Speciﬁc experimental insect species were chosen based on their
suspected potential ELL sensitivity [e.g.; Diptera, Ephemeroptera,
Trichoptera (Lacey and Mulla 1990)], their functional feeding group [e.g.;
predators, Plecoptera (Merritt and Cummins 1984)], and their local
availability.

Direct ELL exposure.

Insects were exposed to ELL concentrations from 15 to 100 ppm for
periods up to 120 mins, depending on the speciﬁc experiment or taxa being
tested (Tables 2 & 3). Some exposure trials involved initially ‘pulsing'
channel water (and insects) with 10,000 ppm ELL to measure mortality
following an "extreme overdose". Label recommended rates range up to 22
ppm for 1 min. All treatments were replicated 4 to 15 times, depending on
experiment.

For mortahty experiments, treatments included exposing insects to
ELL in the absence or presence of ELL-contaminated food (as will be
further discussed) at various ELL concentrations and exposure periods,
with mortality being recorded throughout the experiment. Controls were
handled in the same, but with no ELL. Experimental variables are

50

Table 1. Insect taxa used to test the impact of B.t.i, on non—target aquatic

insects.

 

 

Taxon

Functional feeding groupa

 

Diptera
Athericidae
Atherix 181mm Walker
Tipulidae
Ma abdominalis (Say)

Ephemeroptera

Baetidae

Baetis ﬂavieriga McDunnough
Caenidae

(laenis amica Hagen
Ephemerellidae

Ephemerella subvaria McDunnough
Heptageniidae

Arthrgplea bipgngtata (McDunnough)
Leptophlebiidae

W adenine (McDunnough)

Siphlonuridae
Sinhlnnums Lanidiis McDunnough

Plecoptera
Nemouridae
mm complega (Walker)
Perlidae
mm 13882118: (Newman)
Beaming media (Walker)
Perlodidae
1391181118 51198118 Frison
15222118 m (Banks)

Trichoptera
Hydropsychidae

9mm: mm (Ross)
Philopotamidae

thmarra aLerrima Hagen

predator

shredder-detritivore

collector-gatherer, scraper
collector-gatherer, scraper
collector-gatherer, scraper
collector-ﬁlterer

collector-gatherer,
shredder-detritivore

collector-gatherer,
shredder-detritivore
shredder-detritivore

predator
predator

predator, collector-gatherer
predator, collector-gatherer

predator, collector-ﬁlterer

collector-ﬁlterer

 

8‘ Based on genus, taken from Merritt and Cummins (1984).

8:: 8H @ 88 8H 8 1le .m 3
@838 85 8383 am. 83383

:vbﬁoov

 

338.88 .88 83:85.8 8 H E om; 8388 8385-434m Am .25 3 g 59 m
388 .8888 8 88 8H 8 8.8 2 8 Add 3
8825 8883888 :59... 3898 $5 5385 8% 838883 3833
328.88 .38 8383.8 8 E S om; 8382 383...-ﬂ Am .9»: 3 «Inga 89 8
H88 om © 8%: ma 888 g
8883 8383-3 8 mg
328.88 .38 858898 c 3 ca om; €385.88: AN .3: G 3 89 m
H88 om © 8mg 3 © Add 8
@8888 $5 #885 w. 838883 a
328.88 .88 83888 p m OH cm; 8.382 63895.43 8 .95 2 g 89 N
88 cm 8 8.8 8H 88: 3.8
8825 853888 :38: 88th @8838dﬂ8 a
3:888 .33 8:898 HV m 2 8H 8383.38: 8 .8: :H 38% 8: H
03.8838 833:6 Z :058 «33. Add. .8588 .86QO 88% .8
8388888 dxm .3 8083 829688. mm M35 .8888& dxm
.88me 8335an 8888888

 

 

838838 88mg: 8 “882688 8383882888888 .888on
no Add mo 838m 2: “5.88888 8:88 83.88888 mo 8308888 98 8388‘, H3808toaxm .m 3nt

$58 888035 0

.85 x83 28. 38898 .8 328838 .58 555 n

.8888 833m 23 8 Add 3 889$
633m .5 208me 8:8 @385 .5 585 u 8885-43 ”88898 .8: 8.8 22:88 :85 8888 @88me
583 83 cm a 8m 8883 O omv n 8 8898 8.5: .8 583 u 838.5me 50.580 nv .85 x83 95 u 834 a

 

Ea 8; @ 8.5 2:. @ dd 3
8898 85 583 mm 888623 @838»

 

3:388 8 N. v 3 8&2 8885-43 a .25 a 83.8 89 w
38 oﬁ © 8% 2: © ﬂ 3
mm 3898 85 583 d £38883 3%
35388 .88 838va v N. m omum 88,85 vmummbﬁﬂﬂ Am .95 Q 334mg 39 N.
is oﬁ © 8% 2: © Add 3
3898 .85 583 an 3889:: a
323.88 .82 5888 8 H a om; vats 8885.3 8 .3: a g 82 8
98.2888 aoﬁmhé Z nosmu «885 Add. 8E8 .88QO 80% 6:
28888me .95 a: 583 5835288. .3 825 . 888.5 .935
888.5 ”888.5 $388889

 

 

.8288 .N 3nt

«ﬂag a + an acmaamob
mm 888 a .88 8H 8 Add 8.8 2:

Q5883

 

8 3898 $033 $3 + 88 #083 3
$23.88 w 3 m @885-§ + 853-0 8 8963.0 3 g 89 E

88 02 © 83 8% 2: 88888 d

3 3898 $82: $8 + 35 x33 mung M
$23.88 v N. mumum $835-43 + $263-0 Am £9636 3 g .M 89 Q

88 2 @ 8% 82 a .88 3 © 888%
3:888 8 m m. 8% a a .888 o 2 8 Add «gag 82 a

88 B 8 888 82 a .88 3 © 888888
323.88 w m m 888 2 a .888 o 3 @ Add 3 89 a

B $825 85325 :35 ~58 om © Ema ooH Mamet 1343 3.6%
$23.88 w w m ”4.3 + 8283 AN £950? 3 gm 8% 2

$835 moammSEm ”:35 H58 om @ Ema ooH memow .H A a

3:388 68th vﬁmobﬁdd cm 8 ~33.qu
83 80395588 w w m .wﬁmmbémoa om Am .3: cm Q 33 :3 m 89 m
owmusmmma somumpsv Z @3885 ammov Add was @393 23% .o:
988:8qu .me \ .30me 358329 38QO dxm

 

 

22:85 Hmmommﬁm
E 38358 .Amumwoﬂéﬁa .98qu $8388an 35on mqmvmmm 355083 voaomdom 89¢ 33588266888
885m no Add mo 8893 33 9:58:38 3653 3.68333 mo $338qu 98 3:32.? ﬁguoﬁtmnxm .m mime

ApuGOUV

 

88 cm © Add 888 om o..— 8398

 

 

 

3:388 v N. m m 3888 $3 + 883-8 8 .8828 3 dm 3 89 9
88 oma @ Add 89m 03 3 @8898
3:888 v u m m 3838 88... + .8826 Am .8826 3 .8 3 $9 ma
88 ca © 4.34M
8am ooH 8% 828” Add 3 8398
3:888 @ u m m 8888 $3 + 8283 Am .8828 3 dm 3 89 S
88 84 8 Add 88 OS
88 £308 3 @8898 $888 83 + 85 :85 a
3:888 w u m 2 "8885-141: m + 8283 Am .8283 3 3833 $9 2
328828 883888 8% 8888 Add + aw 88888.5
8888 888888 :88 mm. 8888 Am .88 ca © 8.9m 88m ooH
$38 #838»: o... 8888 $888 .83 + 85 883 «Hand
$28.88 v S m 0H @8885-§ + 8883 Am .8283 S g .89 9
08.8888 aomumgsw Z £8888 mmmow Add E8 “883 88% 68
98888.8m dxm \ 8035 388839 883m .qu

 

 

.9808 .m 8389

88838
8 m - m 8 @3883 83 33. 3:388 88 .3838 8883 8.83 38 388.888 388 38 833 dosagmqoo 0

83383 88 8: .8883
88 @3888 8.83 3888 88: :33 :88 .8 880 8:3 8 3088 83888888 .8 8:88: 8:» 888.83 833/ a

38888 83 :wseﬁ 8:8: 888: 88.8: 838: Add. 88: coed: 338808 m 3.8 3 :33 :08 .8

M88888: 8:..— 8 38888 888883 3 838 8538 883-43 .8 838:8 :8 m 88 n 88: Add .8838 388
8:3 8 8:88 .83 8.888 80.: 888:8 883 8:33:88 u 883-0 3:38 8833 8.83 388883 :8 8 85
83m .85 83: :388838883 68: 8 3835.88 .8980: @338 8:... 8 Add 3 8898 8338 .8 88888
:83 833 3: 83: u 8883-4.qu x8898 88 88 38888 89: 8883 3088.8 :85 833 838

8 8.8358.“ 88 cm n 8: .883 O omv n 3 8888 833 3: 83: u 8383-88: ”3888 "V 8:: 83: 8...: u 884 m

 

88 .3 © .38 8.8 OS 88

 

3:388 .8388. coed: 3 8898 3888 83 + gm
33 8388588 v u m m 83 :03: 88833.3 om Am .8380 2 888m 5m: 8
88 03 © 11.38 8.8 OS 88
3:388 .8358. coed: 3 8898 3088 83 + 88333
32 8888888 8 N. ... m 88 883 8888.8 om a .888 a 8888 82 8
".88888 8:88: Z £88888 8838 841ml: :8 . @383 .88.». .8
338833: .83: \ 388: 38883:. 882mm 8::

 

 

.8288 .m 388:.

Table 4. Experimental variables and parameters of streamside studies
examining drift frequency during and predation rates of three predatory
stoneflies following B.t,i, exposure, conducted in artiﬁcial channels.

 

 

 

Predator sp. Treatments N Parameters measureda
mm 1) 0 ppm, 2) 100 ppm 10 time spent drifting
lyggrias B,t.i, @ 120 min predation rate
Baragnetina 1) 0 ppm, 2) 100 ppm 10 time spent drifting
media ELL @ 120 min predation rate

M 1) 0 ppm, 2) 100 ppm 10 time spent drifting
Sim B,t.i= @ 120 min predation rate

 

a Time spent drifting was estimated by counting the number of revolutions
that each animal drifted and converting to time based on known water
current velocity. Drift data were taken during the last 60 min of 13,101,

exposure for each treatment pair.

57

presented in Tables 2 & 3. Dead experimental insects were counted and
removed at least once every two days throughout the experiment.

For behavioral drift experiments (conducted only with predators),
insects were exposed to 100 ppm BALL for 120 mins (Table 4), with drift
measured in each of ﬁve replicates over 10 mins during the last 60 mins of
BAA. exposure. The number of revolutions individual insects drifted was
recorded. These data were converted to time spent drifting based on the
known distance of one revolution (21 cm) and water current velocity.
Controls contained insects exposed to no ELL-

Experiments were also conducted to investigate changes in predator
feeding habits following ELL exposure. Individual plecopteran predators
were exposed to ELL (0 or 100 ppm @ 120 mins) and then given a known
number of live black ﬂy larvae 6 hrs later. Number of black ﬂies consumed
over 24 hrs was determined the following day by counting remaining larvae
and subtracting this number from the original number. Controls were
handled in the same fashion, but contained no predators to measure non-
predator prey loss.

Indirect (food-chain) ELL exposure.

Experiments on food-chain exposure to BALL were carried out with
predators, detritivores, collectors, and scrapers by supplying these with
excess ELL-killed or live black ﬂy larvae throughout the experiment. Some
experiments included 'heat-killed' black ﬂy larvae as a third treatment.
For experiments 1-7 (Table 2), 'm-killed' larvae were added to channels
as live larvae, then treated with ELL along with experimental insects.

Remaining experiments involved adding 'pre-ELL-killed' larvae to
channels, then exposing both these and experimental insects to ELL. The
'heat-killed' black ﬂy larvae were exposed to 45 °C water for ca. 20 secs
either directly in the artiﬁcial channels (with experimental insects
removed), or prior to being placed in the channels. These experiments
were often coupled with the direct exposure experiments described above
(Tables 2 & 3).

For experiments investigating consumption rates of non-target insects
on live and dead (ELL- and heat-treated) black ﬂies, experimental insects
were provisioned with a known number of black ﬂy larvae prior to the
experiment (Experiments 1-7, 9, 20, 21; Tables 2 & 3). Consumed black ﬂies
were counted and replaced every 24 hrs. To estimate non-consumptive
black ﬂy 'loss', controls were included that contained black ﬂies with no
experimental insects.

For survival experiments using Plecoptera and Diptera predators
(Experiments 2, 3, 4, 8; Table 2), one predator was placed per channel and
replaced every 24 to 48 hrs following death or adult emergence. Trichoptera
predator experiments started with 5 predators per channel, and proceeded
without predator replacement (Experiment 7; Table 2). All other survival
experiments incorporated ﬁve or ten experimental insects per channel
(supplied with excess black ﬂies), also without replacement. Dead insects
were counted, removed, (and replaced depending on experiment) on a daily
basis.

Two experiments investigated the inﬂuence of ELL on insect growth

59
rate (Tables 2 & 3; Experiments 4, 16), both measured diﬁ‘erently. For
Experiment 4 (Impala sp.), similar sized nymphs were placed in
channels, treatments applied, and following adult emergence, adult wet
mass was measured using a Sartorius‘” 1207 MPZ pan balance. Growth
was also estimated for a mayﬂy. Sinhlnnums sp. (Experiment 16), by
collecting nymphs within a narrow size class (middle instars), placing 10
nymphs per channel for 10 channels, applying treatments, and measuring
increased body mass over 7d. A subsample of 10 nymphs was taken at the
onset of the experiment and placed in Kahle's solution (Borror et al. 1989),
oven dried at 35 °C for 48 hrs, and weighed on a Calm® 27 electrobalance to
the nearest 0.01 mg to establish average initial nymphal mass. At the end
of the experiment, live nymphs were collected and placed into vials
containing Kahle's solution for each treatment and replicate, and nymphal
mass estimated as above. Growth was calculated by subtracting estimated
initial dry mass from ﬁnal dry mass for each nymph per replicate per
treatment. Nymphal mortality was negligible for both treatments so data
were not weighted to account for unequal nymph number/channel at the
end of the experiment.

Two experiments (#4, 15) were designed to investigate adult emergence
phenology, and four experiments (#4, 9, 10, 15) involved studying adult
emergence success with four insect species following indirect (coupled with
direct) 3.12.1, exposure. Phenology was measured by recording the day
individual adults emerged throughout the experiment (up to 21 d). Time
period required for adult emergence was calculated as the number of days

from the onset of the experiment to adult emergence, and was averaged

6)

across all individuals per channel (=replicate). Adult emergence success
was measured by counting adults that successfully emerged throughout
the experiment. Cumulative emergent adults were recorded for each
replicate for each treatment.

During experiments with leaf pack-collected insects, stream
conditioned leaves were placed in channels, and replenished as they were
consumed by experimental insects. Conditioned leaves consisting of
approximately equal ratios of speckled alder (Alma mm (Du Roi)
Sprengel), sugar maple (Am sagghamm Marshall), paper birch (3311.113
9312mm Marshall), and northern red oak (Qnmns mbxa Linnaeus),
were collected from the river, fragmented into small (ca. 2 x 2 cm) pieces,
and placed in channels prior to experiments. The equivalent of two leaves
per channel was added in experiments with detritivorous mayﬂies and
stoneﬂies, and ﬁve leaves per channel with I, abdominalis.

ELL applications were made using an inverted 23 L plastic container
that delivered a ELi—water solution of known and constant ELL
concentration to pre-speciﬁed channels. The 23 L container delivered ELL
solution to a 2.6 L bucket which was ﬁtted with ten 1.3 mm ID tubes at its
base, which ultimately delivered the ELL solution to respective channels
(Wipﬂi 1992). Treatments consisted of no ELL and ELL dosages from 15-
100 ppm applied up to 120 mins, depending on experiment. A sudden 10,000
ppm ELL 'pulse' was administered with a 5 ml pipette at the onset of some
experiments (Tables 2 & 3), that remained for several seconds before being
ﬂushed through the channels. If a species was found to be ELL sensitive,

61

additional experiments followed at progressively lower ELL concentrations
(and without the 'pulse') to ﬁnd a 'critical' point beyond which signiﬁcant

mortality did not occur.

 

ELL induced mortality was investigated with L ahdgminalis during
in-stream experiments. Treatments were blocked across three streams to
reduce pseudoreplication problems. This design allowed for true
replication of treatments, but did not avoid the problem of non-random
application of treatments (i.e., the control treatment was the upstream
reach, and the ELL treatment was the downstream reach, across all
blocks). Five I. ahdgminaljs larvae were placed in each of six 2 mm mesh
galvanized metal screen bags (15 x 15 cm) along with approximately 20
stream-collected conditioned leaves. With insects and leaves freshly added,
bag tops were stapled closed and immediately placed two per site in Medora
and Manganese rivers, and Aetna Creek. Bags were completely
submerged and placed against natural leaf packs; microhabitats where I,
ahdgminalis larvae were collected. Bags were longitudinally distanced
from each other by ca. 50 m in each river. ELL was calibrated, based on
speciﬁc stream discharge, and applied the following day at 22.5 ppm over 1
min (recommended ﬁeld rates) at a point midway between bags in each
stream. Thus, the upstream bag served as a control while the downstream
bag was ELi-exposed. Treatments were blocked across rivers giving three
blocks of two treatments. One week later, bags were collected and I.
ahdnminalis larvae removed to record mortality.

An experiment designed to measure Chironomidae mortality resulting

62

from ELL applications was carried out in Manganese River during May
and June, 1991. To eliminate psuedoreplication problems, treatments were
longitudinally blocked upstream through time. This design did not,
however, incorporate interdispersion of treatments. A 1 km stretch of river
was longitudinally sectioned into 5 overlapping blocks such that sequential
blocks overlapped each previous block by ca. 50% (Figure 1) as one
progressed upstream. Block 1 was farthest downstream, and block 5
farthest upstream, ending approximately 100 m below Lake Manganese
outlet. This design allowed treatment comparisons 'above and below'
(across ﬁve blocks) and 'before and after' (across four blocks) ELL.

ELL was applied (100 ppm @ 1 min) weekly to each successive block,
beginning with block 1 during week 1, and ending with block 5 during week
5. To estimate Chironomidae densities, three rocks (surface area; mean =
457, range = 287-941) were randomly chosen and collected from erosional
zones in each 'plot' within each respective block, one week following ELL
application to that block. Macroinvertebrates were immediately brushed off
each set of three rocks into a sorting tray. All visible Chironomidae (down
to ca. 0.5 mm) were counted and placed in a vial containing 70% ETOH for
each treatment per block.

A randomized complete block design (RCB) was used for in-stream

exposure studies with both Tipulidae and Chironomidae.

 

Larval black ﬂy decomposition and release rates following ELL
application were investigated in the artiﬁcial channels. Four trials were

conducted over two years at different mean water temperatures. One trial

    
     

Lake Manganese

 

 

 

 

 

 

 

 

 

ABOVE/BELOW
~>
BEFORE/AFTER
Block 5 ?
Block 4
> p
Block 4 J Block 3
' r
Block 3
J Block 2
' P
Block 2
J Block I
9
Block I
b

 

 

Figure 1. Diagramatic representation of experimental design used to
measure in-stream mortality of Chironomidae following Bacillus
Lhmingjensia var. was application. AP = ELL application point.

64

ran in May and another in July during 1990. Third and forth trials were
conducted in April and June 1991.

Ten black ﬂy larvae were placed in each of ten channels at 21 cm/s mean
water current velocity. After acclimating for 2 hrs, larvae were treated
with 20 ppm ELL for 60 mins. Release rates were evaluated by counting
the number of larvae that released from their attachment site at 24 hr
intervals until no larvae were attached. Decomposition rates were
estimated by rating attached larvae on a 1-4 scale based on the level of bodily
loss; owing either to larval fragmentation or fungal encasement: 1 = intact
larvae, no encasement, 2 = individual larvae showing an estimated 1-33%
bodily loss or encasement, 3 = individual larvae showing an estimated 34-
66% bodily loss or encasement, 4 = individual larvae showing an estimated
67-99% bodily loss or encasement.

I B | . . | . I 1° |

A small braided section of the Medora River was chosen to study ELL
persistence in stream sediments. ELL was calibrated (based on calculated
discharge of this braided section) and applied at an overdose of 1000 ppm for
10 mins.

Every 1-2 d, black ﬂy larvae were collected from Medora River and placed
10 per channel in 20 channels, and allowed to acclimate for 2 hrs. Water
velocity in channels was maintained near 21 cm/s. About 300 ml of stream
sediments were collected at each of two pool sites within the braided section
of river by skimming a 1 cm layer from pool sediment surfaces. Sediments

were collected approximately 5 m upstream (=control, no ELL) and

65
downstream (=ELL-tainted) from the ELL application point.

‘Control' and 'ELL-tainted' sediments were separately mixed with 23 L
of stream water into each of two 23 L applicators, and applied over 1 hr to
respective channels. The following day, larval black ﬂy mortality was
recorded by counting the number of live larvae (determined by observing
movement presence or absence when prodded) and subtracting this value
ﬁom 10. This experiment proceeded until ELL-containing sediments no
longer. killed larvae.

III SI |° |° ] 1

All experiments conducted in artiﬁcial channels employed a complete
randomized design (CRD). Data on prey consumption, percent mortality,
growth rates, percent adult emergence, and adult emergence phenology
from these experiments were analyzed using one-way AN OVA. Treatment
means were compared to control means using Dunnett's test (two-sided), or
using a T-test at p = 0.05 (Steel and Torrie 1980). Prey consumption data
were analyzed based on daily means. Mortality, growth, and adult
emergence success data were analyzed using cumulative values. Adult
emergence phenology was recorded as the number of days from the onset of
the experiment until 60% of adults emerged, for each treatment.
Individual treatments were compared to control treatments (Ho: control
mean = 'treatment' means; HA: means at), for all experiments.

Data from in-stream Tipulidae and Chironomidae experiments were
analyzed with two-way ANOVA (treatment x block) at p = 0.05 (RCB). The
effects of time (BEFORE and AFTER), and space (ABOVE and BELOW),

66

could not be separated out, and therefore, became part of the blocking eﬁ‘ect.
Hypotheses tested were; Ho: control mean = 'treatment' means; HA: means
at. Data from ELL and ELL-contaminated black ﬂy larvae persistence
trials were recorded through time and SE. calculated for graphical
representation.

Normality and variance heterogeneity assumptions were checked and
data transformed using log10 [x + 1], when necessary. Sample and
treatment independence was assumed based on experimental design and

protocol employed.

RESULTS

: . Hm: -; _.'-: u. . “’3 .. .N. : H. “I. . ”0...”:
channels.

Predators consumed a signiﬁcant quantity of dead (ELL and heat-
treated) black ﬂies, without observable treatment effects, relative to controls
(Table 5). Ismrh signnm nymphs consumed signiﬁcantly fewer dead
than live larvae during Experiment 2 (p < 0.05). All other experiments with
this predator indicated no reduced predation on dead (ELL heat-killed) vs.
live larvae (Table 5). Resulting predator mortality after consuming ELL-
treated larvae was not statistically signiﬁcant. In addition, adult
emergence phenology and success, and adult mass of L signal; was not

affected for nymphs that consumed ELL-intoxicated black ﬂy larvae.
Amanda homing and E media both consumed approximately an

Aﬁbﬁoov

 

o.o c 3 H 3 Samara a 338
--- ed a md H m6 3: 2 Huang 89 m
ed a we H ed 585.13 a «quad
o.o e .3 H 2: as: a 333 89 m
a 3A H 3.9 m wd H 3: m mam: H odw 0.0 a HA H wdH weasepu-g AN am
a m? H 3.2 a we H w: c. 3:. H 0.8 ed a E H mm 2,: : Hague...— 82 e
o.o a o4 H mac. 335.43 a
..-- --- ed a NA H Qm 3395.36: 8 g
od m ad H Wm 2,: 3 g 89 m
od c 3 H 3 385.43 a maﬁa
o.o c 9o H Hm 2.: a a 89 m
m hm H 93 a 5m H km a cg H Wm @335- .n q n 8
a Hm H a: e «a H an a 3 H hm 385.32 a ”maﬁa
a hm H mm a hm H mm a 3 H «a can a a 82 H
38V mmmE monomcmﬁo $835 $23.88 .Qﬁsmaoo nmmow 3 2:98 $6on 58% .o:
£st :35 mocewcoﬁm 12.38%: mm xoﬂm .Gomﬁdcoo. .3 x33 83¢on dam
3:5 3:: mama 25pm cs es ”massages

 

 

«.35 x33 veamﬁﬁmucoo mcsmmwﬁ nudes“ 38865 .8 £88m. Add 3 @3098 38%: cassava bopmcma mo
53on was $33323 oecewcmﬁo :35 .8825 85325 :35 £623 £53.88 33.89 6333.5 .m 2nt

.413.

B vmmogxm Ampopmvma «.88 323 ha M33 .1. w3m2ﬁﬂ x3398 Ho: 98 22:85 8on @9688 €onva
ﬁtsv 8mm om n .8.“ .833 O 09» n 3 @0898 max/.81..“ «623 n uSmmSé—«mn ”2838 "V 85 x35 9,: n 35 a

.mmcmu wmuﬁcmvgm mummxse EB <>OZ< .8 $8-9 @3me
was: .mod n a an HGBMEG 238$:me no: v.8 Amazoﬁtonxm 3:35 .833 8:3 ms... 3 3333“ .m@ H £832 a

 

 

o.o 585.3 a ﬂag
3 2.: s 3H 52 w
a mm H o? o. No H mm c385.ﬁ1m.a Hg
m. 3 H qua a no H 3 a»: a «mag 82 H
3:: mmmS moammumam 3825 3:388 @8538 ammow 134M Shim momuomm 23% .0:
£35 23%» oucmmgmﬁm 3:993 ha Mona .cosmvcoo. ma x33 833.5 AVE
:35 EH5 989 :35 05 Q» $308389

 

 

.5308 .m 3nt

w

equal amount of live versus dead black ﬂy larvae, without increased
predator mortality (Table 5).

mm mm, on the other hand, did increase feeding rates on
black ﬂies alter the black ﬂy larvae were killed with ELL (p < 0.01) (Table 5)
without increased predator mortality. Athezix W predators, also did
not suffer increased mortality after being fed ELL-contaminated black ﬂies,
although predation on the black ﬂies was not veriﬁed.

No direct macroinvertebrate mortality resulted following 3.1.1. exposure
with all detritivore, grazer, and collector species tested except for a ﬁlter-
feeding mayﬂy, and a detritivorous crane ﬂy (Table 6). Anthmplea
bjpnnmm nymphs exposed to 10,000 ppm 3.1.1, 'pulse' followed by 100 ppm
3.12.1. over 120 mins resulted in 24% mortality, signiﬁcantly greater than the
control (p < 0.05). 1591113 ahdnminalis larvae exposed to BALL at 10,000 ppm
'pulse' plus 100 ppm for 120 mins, and 100 and 20 ppm ELL for 120 and 60
mins, resulted in 100, 92 and 36% mortality for Exps. 17, 18, and 19,
respectively, and had signiﬁcantly greater mortality than did the controls (p
< 0.001) (Table 6).

Detritivores that were fed black ﬂy larvae consumed dead (ELL-killed
and heat-killed) but not live black ﬂies. Siphlnnma mpjdns consumed a
signiﬁcant number of dead versus live black ﬂy larvae (p < 0.001) (Table 6).
During experiments with Ldicala. menleta. Ladentixa. Em
and Q, amica, nymphs were observed handling and appeared to be feeding
on dead black ﬂy larvae, but feeding was not quantiﬁed. In addition, m-
treated larvae had to be replenished regularly throughout experiments in

70

3-3. H ”83

odTogv ”@588

Amyuﬁoov

«mania
dd 8% 2: 6. «338 d

 

--- --- cam: ”dam m --- 35:8 Q SHEEN 89 mm
a q: H eon Had 8% 82 a
m m6 H v.3 Add Sam 3 a Magda
m 3 H 3:. secs 2 3% 82 a
a S H o8 dd 8% 82 a
a 3 H emu dd 8% 3 a snag
--- --- a m6 H oém 35:8 3 3 89 2
a 3: H 3a a 3 H Q8 g can 2: Am aqua
w GS H 0.8 H.. mm H oém 35:8 3 ”2:28:35 89 2
m wdm H 9Q. m. ms H odm n md H Qmm @385- .. H m 8
m mm: H Be a 3 H ow a E H mam 38.5-32 a 3%
m mm; H 04% a QB H 93 m «6 H md 9,: Q mandazqmm 89 m
336 u .8 38V 3328 $836 $23.88 .Qﬁsmcoo ammoc Add use v33» 53% 6:
5mm mama 33cm. meow oocowumﬁm :39qu mm xoﬂm 358329 36QO dxm
Rana? ES. 99% H :33 e e

 

 

«.85 x33 wouwcmESQoo waﬁmowﬁ nmsopﬁ

38865 .8 383% Add 3 3898 mumcoﬂémﬁa can 68an .mumcwmuﬁ @3023 no £3on van .3332.“
85325 “Eve .9383 monomumﬁm :33 “:8ng $33.88 “smegma .85 #33 mo :oﬁagmcoo .w mime

ﬂ

 

 

6.82;

 

a Q... H Q8 Hem 8% 2: a wlzmclqmm:
m Qm H Qm 3888 2 Emma 82 S
8 Q2: 8+4; 8% 8H 3. Has
a Qo 3.588 a ﬁg 82 S
a oQo H 88 a E H Q3 in 8% 82 a :2 2
a 88 H wQH m 3. H QS 888 3 88:83am 8% a
as: Cam quE
8 .8 H Q8 8 m H 2 H Qm H QE 8 Qm H Q2 $13: 8% 82 8
H. Qm H QE 8 m H m: a 5. H Q8 a «.8 H Q8 181m: 8% 2: a 888
m S H Q8 H. m H 8 a E H Q8 a 8.8 H Q3 3888 3 H883 $9. a
£08 88
H. Qw H Qm 8+ij m 8% 2: a
8 0.8 H Q8 14mm 8% 8H 8 £88
8 3. H Qﬁ 3.588 a 8% $9 8
:08 Q8 .8 Rs: wmmeoﬁm $838 3:388 @8228 ammow lqlm. .m 98 63mm: 88% 8:
88m. $88 8:38 088 8889888 Hanna»: mm #85 mucmqswourr $8QO .axm
88882 :8: 988 H :38 Q0 8 .

 

 

.8288 .m 833.

.88 08.5... :8 88 $888.95 :83qu 8288.va ammommammm oz 9
88:58.5 .8 833.88% 88388 8m m 8585 88m n

.Godvnv 88 $8855 38 <>OZ< .8 83-8 @889
mam: .mod n a 8 888ch 388.888 8: 8.8 AmEmEEonm 5583 .833 8:88 85 .3 326:8.“ .mm H 382 a

 

 

m 5 H Q8 Qo 8888-83 8 88.88%
a 3. H Q3 8.8 9,: Q g 88 8
n 3. H Q8 Qo 838.8.qu 8 88838
m 3 H QH Qo 2: a 8.8888. 8.9 8
8 SH H Q8 jam. 8% 8 a 38888.8
#8 Qo 3888 2 38a 82 9
8:88 88 .8 3:5 vmmHoEm 8828 3:388 @8888 ammov Add 98 @388 .88» 8:
88m $88 8:38 088 8283888 8:98? ma Moﬂm 388.839 88QO .me
888882 :8: 988 H :38 8 8

 

 

.8388 .m 8389

73

channels containing these detritivores. This larval loss was attributed to
detritivore consumption as complete microbial decomposition and
fragmentation took more time (one to two weeks). Ms W and A.
W nymphs did not consume dead black ﬂy larvae as did detritivores.
Nymphal molt frequency and adult emergence phenology was not affected
for Q. amica exposed to 3.1.1. and given ELL—contaminated leaves and black
ﬂies (Table 6).

Life history parameters of mortality and adult emergence were not
affected for S. m: after nymphs consumed ELL-contaminated food
(Table 6). However, S. 1391an nymphs given ELL-contaminated black ﬂies
as well as leaves grew signiﬁcantly faster than nymphs with only leaves for
food (p < 0.001).

Predators showed no behavioral changes following ELL exposure except
for A. lymn’as nymphs that increased time spent drifting at 100 ppm BALL
concentration (p < 0.05) (Table 7). None of the We predators teswd showed
immediate (over the following 24 hrs) feeding habit changes in response to
BALL application (Table 7). After being exposed to ELL. these predators
consumed equal amounts of black ﬂies as did predators not exposed to 3.1.1..

 

Although ’there was a trend towards lower Chironomidae densities in
ELL-treated stream sections, 'ABOVE and BELOW' (Figure 2A) and
'BEFORE and AFTER' (ﬁgure 2B) midge population levels were not
signiﬁcantly different between ELL-treated and untreated stream reaches,
during the 5 wk study. In-stream studies with I. abdominalis showed that

Table 7. Time spent drifting during B,t,i, exposure and subsequent
predation on black ﬂy larvae after B.t.i, exposure, with three predatory

 

 

 

stoneﬂiesa.

Time spent Subsequent
Predator sp. Treatments drifting (mins)b predationc
AW 1) control 3.6 i- 3.5 a 7.4 i 1.3 a
122m 2) 100 ppm mi, 27.0 :t 8.6 b 7.6 i 1.2 a
23mm 1) control 0.0 8.7 i 2.2 a
media 2) 100 ppm BALL 0.0 8.5 i 2.2 a
Maﬁa 1) control 8.2 i 5.3 a 10.9 :t 1.1 a
m 2) 100 ppm B,t.i, 2.6 i- 2.3 a 10.5 i 1.2 a

 

a Means :t SE. followed by the same letter (within predators) not

signiﬁcantly different at p = 0.05, using paired T-test.

b Time spent drifting was estimated by counting the number of revolutions
that each animal drifted and converting to time based on known water
current velocity. Drift data were taken during the last 60 min of 3.1.1,

exposure for each treatment pair.

c Mean number of black ﬂy larvae consumed / predator / 24 h, immediately

following 3.1.1. exposure.

75

 

250° A. —e— 'ABOVE' B.t.i. (control)
—9— 'BELOW’ B.t.i. (treated)

2000

1500

1000

500

 

 

 

 

 

0 l ' I ' 1 v 1 v I .
6 May 17 May 22 May 31 May 6June

DATE

 

MEAN NUMBER OF CHIRONOMIDAE I SQUARE METER

 

 

 

 

2500 B —9— 'BEFORE' B.t.i. (control)

' —.— 'AFI'ER' B.t.i. (treated)
2000
1500
1000
500

ns
0 l I I ' ﬂ
0 1 2 3 4 5
BLOCK

Figure 2. In-stream Chironomidae densities (A.) 'ABOVE vs. BELOW' and

(3.) 'BEFORE vs. AFTER' Bacillus thun'ngiensis var. israelensis
applications. [ns = not signiﬁcant @ p = 0.05].

76

these nematoceran Diptera were not sensitive to ELL when it was applied
at recommended rates. ELL-treated and untreated I. abdominalis larvae
showed equal levels of mortality (Figure 3).

 

Complete larval black ﬂy 'loss' from substrate (owing to larval release or
decomposition) took 4 to 16 d (Figures 4A & B, 5A & B), and appeared to be
dependent on water temperature. Higher water temperatures resulted in
greater release of larvae, and in greater decomposition rates. Black ﬂies
immediately began releasing from their attachment sites following ELL-
application, with the greatest release rates within 2 days of application. All
released larvae were dead when collected (24 hr intervals). Decomposition
rates lagged behind release rates with signiﬁcant larval 'loss' from
decomposition or fungal encasement beginning as late as 10 (1 following
larval death, recorded at the lowest water temperature (5.5 °C) (ﬁgure 5A).
I B | . . | . | 1° |

Bioassays indicated that the toxic components of ELL remained present
or viable in stream pool sediments for 11 d (Figure 6). Viability loss began 7
- 9 d following 3.1.1, application, with most rapid decline occurring 9 to 12 (1.

DISCUSSION

Exposure to ELL did not result in increased macroinvertebrate
mortality for most Diptera, Ephemeroptera, Plecoptera, and Trichoptera

tested in this study. However, two species were ELL sensitive at unusually

 

 

 

 

 

 

 

 

>- 100 '
t .
:r‘
t- 80 ‘
a:
0 cl
2
<
.J
:3
E
I-
32
z
4
Lu
2
'ABOVE' B.t.i. 'BELOW B.t.i.
TREATMENT

Figure 3. T191113 abdominalis mortality 'ABOVE vs. BELOW‘ during in-
stream B.t.j, applications. Error bars represent S.E. [ns = not
signiﬁcant @ p = 0.05].

78

A. Water temperature:
mean . 15.8° C

 
   

1oo_l’ange=11.O-19.O°C _100

1 .

80 .. —6— % Released _ 80
. —O-— % Decomposition ,.

60 - - 60

4O - BTI applied .. 40

20 - - 20

0 1 r 0

 

 

 

B Water temperature: ——e— % Released
mean-19.1°c —O— %Decomposition

MEAN % BLACK FLY LARVAE RELEASED
MEAN % LARVAL DECOMPOSITION

 

 

 

100_ranger-17.0-22.O°C {100

80-: -80

6° ‘ BTI applied :60

40- -4o

20: :20

01 . , . . . . 'o
O 1 2 3 4 5 6

DAYS FOLLOWING BTI APPLICATION

Figure 4. Percent larval black ﬂy decomposition and release from their

attachment sites following Bacillus 1212mm var. was
application during (A.) May, and (B.) July, 1990. Error bars represent

S.E.

A Water temperature:
' mean a 55° C
range . 1.0 -10.0° C

  
  
   

 

 

 

 

100- -100
. —e— % Released _
80 . —-O— % Decomposition .-. I. .- _ 80
60 - BTI applied - 60
1 1-
§ 40 - - 40
< ‘ ' Z
51 20 - - 20 2
ul . _ I:
l: a)
I“ 0 - I I 1 I ' I 1 O 2
g 1 3 5 7 9 11 13 15 5
a:
< 8
" o
>.
d B. Water temperature: a.
a: mean - 19.0° c E
2 range: 16.5 -21.5° C 5
a: 100:8Tlapplied [100 32
z
322 so - - 80 <
m
< . . E
m
E 60 ~ - 60
4o - - 4o
20 _ —6— % Released _ 20
. —O—- % Decomposition _
0 '1 0

 

 

 

13579111315

DAYS FOLLOWING BTI APPLICATION

Figure 5. Percent larval black ﬂy decomposition and release from their

attachment sites following Bacillus Lhmjngiensis var. 15139131515
application during (A.) April, and (B.) June, 1991. Error bars represent

S.E.

 

   
 

 

 

 

t -—

3 100

< ..

E

0 so -

5 l

5 60 -

l I

5 40

S . —O— BTI-tainted SEDIMENTS
m _ —O-- 'CLEAN'SEDIMENTS
32 20

z I

a O-l 1 ﬂ I ‘ I i I
s

0 2 4 6 8 10 12 14
DAYS FOLLOWING BTI APPLICATION

Figure 6. Percent black ﬂy mortality, in artiﬁcial streams, from exposure
to B.t.i.-tainted and clean sediments collected from depositional zones
within m-treated and untreated river reaches. Error bars represent
S.E.

81

high dosages; Anhmnlea hinunctata. a ﬁlter feeding mayﬂy. and I.
abdominalin, a detritivorous crane ﬂy.
Ma ahdgminalis was sensitive to B.t.i. at all concentrations above

ﬁeld operating dosages. Others have also reported nematoceran Diptera
sensitivity to B.t.i. (Back et al. 1985, Pistrang and Burger 1984, de Moor and
Car 1986). When I exposed these detritivores to B.t.i. at recommended
rates, their survival was not affected. This suggests that when treating
lotic systems with B.t.i., applicators need to accurately measure stream
discharge and carefully calculate and apply the appropriate B.t.i. dose so
that non-target species will not be killed. Shredders in stream ecosystems
play a major role converting large particulate organic matter (CPOM) such
as leaf litter and woody debris into ﬁne particulate organic matter (FPOM)
(Cummins et al. 1989). Loss of these and other Dipteran shredders (e.g.,
Bxillia and Xﬂgtms spp.) could profoundly retard CPOM processing and
could affect nutrient ﬂow in streams.

Arthmplga hinunmm nymphs exposed to high levels of 34.1, also died.
Logistical problems, however, prevented me from further testing the
sensitivity of these nymphs to lower B.t.i. concentrations. The ﬁlter-feeding
habits of this mayﬂy may have allowed a route for B.t.i. to their digestive
tracts. This, combined with the high B.t.i., application rate (10,000 ppm
'pulse' plus 100 ppm @ 120 mins) that these mayﬂies were exposed, resulted
in 25% nymphal mortality over 7 d. Survival would probably not have been
affected if A. hinunmm nymphs had been exposed to operational ﬁeld

dosages.

82

Chironomidae populations were not signiﬁcantly reduced following
B.t.i.. application. Chironomids were grouped to evaluate the impact of
B.t.i.. on the total Chironomidae population. Individual species, however,
may have been sensitive to B.t.i., but species population changes may have
been masked by taxonomically lumping the Order. Several studies have
indicated that certain chironomid species are sensitive to B.t.i. (Miura et al.
1980, de Moor and Car 1986, Merritt et al. 1989).

This study investigated sub-chronic, food-chain, and ecological impacts
of B.t.i. on non-target macroinvertebrates. Predators consumed dead and
live black ﬂies at equal rates for all predators tested with two exceptions.
1m aignm nymphs consumed fewer B.t.i.-treated than live black ﬂy
larvae during one trial in 1990, and Q, mama predators ate more dead than
live black ﬂy larvae. Peckarsky and Penton (1989) found that certain
predatory stoneﬂies rely on hydrodynamic cues from mobile prey to elicit an
attack. Predation rates by L signata, depending on their starvation levels at
the onset of individual feeding trials, may have been controlled by the
presence or absence of hydrodynamic cues of live and dead black ﬂy larvae,
respectively. Lack of consumption rate differences between the two dead
(heat- and ELL-killed) black ﬂy larval types by L W (and A. Imaging)
indicated that the presence of B.t.i. toxin in black ﬂy cadavers did not deter
feeding by these predators. ‘

Higher capture success and decreased handling times on dead versus
live black ﬂy larvae may have been causative factors for the increased
predation rate of Q, mama. Live black ﬂy larvae frequently driﬂ; or respond
with aggressive biting behavior (Hart 1979) when encountered by other

invertebrates.

Experiments designed to investigate insect drift in response to B.t.i.
showed that one plecopteran species, A. lymrias. was sensitive to the
presence of ELL, of the three species tested. This was the only species that
driMd in response to B.t.i., but did so at an usually high B.t.i. dose (100
ppm 6 120 mins). Previously published studies also have indicated that
certain Ephemeroptera (Pistrang and Burger 1984) and Trichoptera
(Dejoux et al. 1985) may show a drift response in the presence of B.t.i..

Consumption of BALL-intoxicated black ﬂy cadavers by predators and
detritivores did not result in changes in nymphal survival, development,
adult emergence success and phenology of the 15 species tested. However,
growth rates for S. rapidns, a detritivore, signiﬁcantly increased when
these nymphs ate dead black ﬂy larvae over a 7 d period. The mayﬂies
consumed the highly nutritious black ﬂy larvae even in the presence of
stream conditioned leaves, resulting in faster growth rates than those
nymphs given only conditioned leaves as natural food.

The sudden short term biomass conversion of live to dead black larvae
aﬁer B.t.i. application may have important ecological implications to
consumers (e.g., detritivores such as S, mnidna) in lotic systems.
Increased growth and body size may equate to shorter generation times,
earlier phenological events, higher fecundity, and ultimately increased
production for those insects utilizing dead black ﬂy biomass. The
availability of B.t.i.-contaminated prey following B.t.i. application was
found to be variable and depended on water temperature. Some black ﬂy

84

larvae remained intact for up to 16 din colder water, and thus were
potentially available to consumers.

Predators utilized dead black ﬂy larvae as food following B.t.i. use.
Thus, immediate ecological consequences for this trophic level may be
minimal. However, longer term impacts through loss of (black ﬂy) prey
may be important, as these predators rely on black ﬂies for food, with
overall prey consumption reduced following B.t.i. application (W ipﬂi 1992).

The short persistence (11 d) of B.t.i, in stream sediments in this study
suggests that this compartment did not serve as a long term B.t.i. source.
Physical factors and pool dwelling benthos therefore can potentially
resuspend B.t.i, up into the water column and B.t.i. will remain lethal to
sensitive species (e.g., black ﬂies) for up to 11 d. Black ﬂies can quickly
recolonize B.t.i.-treated areas by drifting from upstream sites or egg
hatching if the toxin is no longer viable or present. However, B.t.i. may be
taken up and released in other stream compartments.

In conclusion, this study showed that most taxa of aquatic insects were
not directly affected by B.t.i.. I found that I, ahdgminalia died at unusually
high B.t.i. dosages, but not at recommended ﬁeld dosages. A ﬁlter feeding
mayﬂy, A, W, also died at high B.t.i. concentrations. In addition,
all insects tested did not show changes in nymphal survival and growth,
and adult emergence success and phenology following food-chain exposure
to B.t.i., even though predators and detritivores consumed dead, B.t.i.-
contaminated black ﬂy larvae.

85

ACKNOWLEDGMENTS

I thank G. Arntsen for offering access to his property allowing me to
conduct this study. I also thank R. Edens for his tireless assistance in the
ﬁeld, R. Hall for suggestions on experimental procedures, and D. Molloy,
for helpful comments on experimental design. I also thank W. Hilsenhoff
for numerous insect identiﬁcations, W. McC aﬁ‘erty for mayﬂy
identiﬁcations, and S. Szczytko for stoneﬂy identiﬁcations. This study was
supported, in part, by Sport Fishing Institute Fund grant SFRP-90-26,
Northeast Regional Black Fly Project NE-118, and Michigan State
University Cooperative Extension Service and Agricultural Experiment
Station.

86

LITERATURE CITED

Ali. A. 1981. Bacillusthun‘nsiensis eerovar.israelensis(ABG-6108)
against chrironomids and some nontarget aquatic invertebrates. J.

Invertebr. Pathol. 38: 264—272.

Aly, C. and M.S. Mulla. 1987. Effect of two microbial insecticides on
aquatic predators of mosquitos. J. Appl. Ent. 103: 113-118.

Aly, C., M.S. Mulla, and B.A. Federici. 1985. Sporulation and toxin

production by Bacillus thunnsiensis var. ismclensis in cadavers of
mosquito larvae (Diptera: Culicidae). J. Invert. Path 46: 251-258.

Aronson, A.I., W. Beckman, & P. Dunn. 1986. Bacillus thndngiensis and
related insect pathogens. Microbiol. Rev. 50: 1-24.

Back, C., J. Boisvert, J .O. Lacoursiere and G. Charpentier. 1985. High-
dosage treatment of a Quebec stream with Bacillus mutingignsis var.
W3 efﬁcacy against black ﬂy larvae (Diptera: Simuliidae) and
impact on non-target insects. Can. Entomol. 117-: 1523-1534.

Borror, D.J., C.A. Triplehorn, & N.F. Johnson. (eds.). 1989. An
Introduction to the Study of Insects (Sixth edition). Saunders College
Publishing, Philadelphia, PA, U.S.A. 875 pp.

Car, M. & F.C. De Moor. 1984. The response of Vaal River drift and
benthos to Simullnm (Diptera: Nematocera) control using Bacillus

thnﬁngiensis var. imdﬁnaiﬁ (H-14). Onderstepoort J. Vet. Res. 51: 155-
160.

Chilcott, C.N., J .S. Pillai, & J. Kalmakoﬂ'. 1983. Efﬁcacy of W

W var. mm as a biocontrol agent against larvae of
Simuliidae (Diptera) in New Zealand. N. Z. J. Zool. 10: 319- .

Colbo, M.H. & A.H. Undeen. 1980. Effect of Bacillus tlnm'ngignsls var.
ismelensis on non-target insects in stream trials for control of
Simuliidae. Mosq. News. 40: 368-371.

Cummins, K.W., M.A. Wilzbach, D.M. Gates, J .B. Perry, & W.B.
Taliaferro. 1989. Shredders and riparian vegetation. Bioscience. 39: 24-
30.

Dejoux, C., F.M. Gibon, & L. Yaméogo. 1985. Toxicité pour la faune non-

cible de quelques insecticides nouveaux utilises en milieu aquatique

tropical. IV. Le Bacillus ﬂmzjngignsjs var. madman. Rev. Hydrobiol.
Trop. 18:31-49.

De Moor, F.C., & M. Car. 1986. A ﬁeld evaluation of M1113 thnn'ngjgn gig
var. ismelensis as a biological control agent for W M

 

88

(DipterazNematocera) in the Middle Orange River. Onderstepoort J. Vet.
Res. 53:43-50.

Dubois, N.R. & F.P. Lewis. 1981. What is Bacillus Lhuzingicnsis? J.
Arboriculture. 7: 233-240.

Gaugler, R. & J .R. Finney. 1982. A review of Bacillus Lhuziugicusis var.
ismclcnsis (serotype 14) as a biological control agent of black ﬂies
(Simuliidae), In: D. Molloy, (ed.). Biological Control of Black Flies
(Diptera: Simuliidae) with Bacillus thuzingicnsis var. isnaclcnsis
(serotype 14): A Review with Recommendations for Laboratory and Field
Protocol. Misc. Pub. Entomol. Soc. Amer. 12: 1-18.

Gibbs, KE., F.C. Brautigam, C.S. Stubbs, & L.M. Zibilske. 1986.
Experimental applications of Bacillus mutingicnsis var. ismclcnsis for
larval black ﬂy control: persistence and downstrem carry, eﬁcacy,
impact on non-target invertebrates and ﬁsh feeding. Maine Life Sci.
Agric. Exp. Stn. Tech. Bull. 123: 1-25.

Goldberg, L.J., & J. Margalit. 1977. A bacterial spore demonstrating rapid
larvicidal activity against Ancnhclcs scrgcntii, Hzanctania unguiculata,
Culcx unixitattus, Acdcs acgypji and Culcx pipicns. Mosq. News. 37:
355-358.

89

Hart, DD. 1979. Patchiness and ecological organization: experimental
studies within stream benthic communities. Ph.D. Thesis. University of

California, Davis.

Hurlbert, SH. 1984. Pseudoreplication and the design of ecological ﬁeld
experiments. Ecol. Monographs. 54: 187-211.

Lacey, L.A. 1983. Larvicidal activity of Bacillus pathogens against
W mosquitoes (Diptera: Culicidae). J. Med. Entomol. 20: ’
620-624.

 

Lacey, L.A. & M.S. Mulla. 1990. Safety of Bacillus thuﬁngicnsis var.
ismclcnsis and Bacillus sphacricus to non-target organisms in the
aquatic environment, In: M. Laird, L.A. Lacey, and E.W. Davidson
(eds.), Safety of Microbial Insecticides. C. R. C. Press, Boca Raton, FL.

Lacey, L.A., & A.H. Undeen. 1987. The biological control potential of
pathogens and parasites of black ﬂies, In: K.C. Kim and R.W. Merritt,
(eds.), Black Flies: Ecology, Population Management, and Annotated
World List. Penn. State Univ. Press, University Park, PA.

Lacey, L.A., H. Escaffre, B. Philippon, A. Seketeli, & P. Guillet. 1982.
Large river treatment with Bacillus thuringicusis var. iszaclcnsis (H-14)
for the control of Simulium damnasum s.I. in the Onchoceriasis Control
Programme. Z. Tropenmed. Parasitol. 33: 97-101.

Merritt, R.W., & K.W. Cummins (eds.). 1984. An Introduction to the
Aquatic Insects of North America (2nd ed.), Kendall/Hunt Publishing
Co., Dubuque, IA. 710 pp.

Merritt, R.W., E.D. Walker, M.A. Wilzbach, KW. Cummins, & W.T.
Morgan. 1989. A broad evaluation of B.T.I. for black ﬂy
(Diptera:Simuliidae) control in a Michigan river: efﬁcacy, Carry and
nontarget effects on invertebrates and ﬁsh. J. Amer. Mosq. Cont. Assoc.
5: 397-415.

Miura, T., R.M. Takahashi, & F.S. Mulligan. 1980. Effects of the bacterial
mosquito larvicide. Bacillus thunnsiansis var. israclcnsis serotype H-14
on selected aquatic organisms. Mosq. News 40: 619-622.

Molloy, D. (ed.). 1982. Biological control of black ﬂies (Diptera: Simuliidae)
with Bacillus thuringicnsis var. ismelﬁnsiﬁ (Serotype 14): a review with
recommendations for laboratory and ﬁeld protocol. Misc. Publ. Entomol.
Soc. Amer. 12: 1-30.

Molloy, D., & H. J amnback. 1981. Field evaluation of Bacillus
thuzingicnsis var. israclcnsis as a black ﬂy biocontrol agent and its effect

on nontarget stream insects. J. Econ. Entomol. 74: 314-318.

Peckarsky, B.L. & M.A. Penton. 1989. Mechanisms of prey selection by

 

91

stream-dwelling stoneﬂies. Ecology. 70: 1203- 1218.

Pistrang, L.A. & J .F. Burger. 1984. Effect of Bacillus thuringicusis var.
israclcnsis on a genetically-deﬁned population of black ﬂies (Diptera:
Simuliidae) and associated insects in a montane New Hampshire
stream. Can. Entomol. 116: 975-981.

Rutschke, J ., & J. Grunewald. 1984. The control of black ﬂies (Diptera:
Simuliidae) as cattle pests with Bacillus thun'ngicnsis H-14. Zentralbl.
Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Ref. 258: 413- .

Sebastien, R.J., & R.A. Brust. 1981. An evaluation of two formulations of

mm thudngicnsis var. isaaclmsis for larval mosquito control in sod-
lined simulated pools. Mosquito News. 41: 508-512.

Steel, R.G.D., & J .H. Torrie. (eds.). 1980. Principles and procedures of
statistics. Second edition. McGraw-Hill Book Co. New York, NY, U.S.A.
633 pp.

WHO. 1979.. Data sheet on the biological control agent Bacillus
thuringicnsis serotype H—14(de Barjac 1978). WHO/VBC/79.750. Rev. 1.

VBC/BCDS/7 9.01. World Health Organization.

Wipﬂi, M.S. 1992. Direct and indirect effects of Bacillus mun'ugicnsis var.

 

92

ismclcnsis on non-target aquatic insects and trout. Ph.D. dissertation.
Michigan State University, East Lansing, MI, U.S.A.

CHAPTER 3

DIRECT AND INDIRECT EFFECTS OF THE MICROBIAL INSECTICIDE
BAQILLIIS W VAR. ISBAELENSIS 0N BROOK.
SALYELINIIS EQNIIHALIS; BROWN, SALMQ mm; AND
STEELHEAD. W8 MXKISS TROUT.

ABSTRACT

Direct toxicity and indirect (food-chain) eﬂ'ects of Bacillus thuﬂngicnsis
var. ismclgmsis (B.t.i.) (Teknar® HP-D) on trout were studied in the
laboratory. Direct exposure to three size classes; egg, 1.5 - 2.4 cm long, and
7.2 - 8.2 cm long ﬁsh of brook, Salmlinus fautinalis; brown, Balms mum;

and steelhead, anhcnmchus mykiss, at B.t.i. concentrations up to 10,000
ppm was studied with both viable and non-viable (autoclaved) B.t.i.. There

was no mortality below 1000 ppm B.t.i. for all three trout species of all size
classes tested with viable and autoclaved B._t,i.. There was no difference in
toxicity between viable and non-viable B.t.i. for all three trout species at all
concentrations, except two trials where the autoclaved material was
signiﬁcantly more toxic than the viable B.t.i.. L050 values for brown and
brook trout alevins were between 1561 and 2321 ppm for both viable and
nonviable B.t.i.. Scanning electron micrographs showed particle and

mucous accumulation on ﬁsh gill surfaces, with about 50% ﬁsh that were

 

 

 

94
exposed to B.t.i, having these accumulations. Blood-gas analyses revealed
low oxygen levels in blood taken from ﬁsh exposed to 4000 ppm BLi, for 4
hr.

Food-chain toxicity experiments were conducted with brown trout (mean
length 4.3 cm) (fed excess ELL-contaminated or live black ﬂy larvae daily
for 5 d) to assess mortality and growth over 30 d. Trout ate an equal amount
(about 40 black ﬂies / trout / d) of each larval type (B.t.i.-killed or live). Trout
fed B.t.i.-contaminated black ﬂies experienced no signiﬁcant mortality or
changes in growth rates relative to trout given live (uncontaminated) black
ﬂies.

This study indicated that the B.t.i, toxin was not responsible for trout
mortality, instead, reduced survival was attributable to formulation
component(s). Trout were not affected when indirectly exposed to B.t.i. via

consuming contaminated prey.

 

 

95

INTRODUCTION '

Bacillus thudngicnsis var. iszaclcnsis (de Barjac) (serotype H-14) (B.t.i.)
is toxic to black ﬂy and mosquito larvae and is currently used in control
programs in North America and other parts of the World (WHO 1979,
Dejoux and Elouard 1990). B.t.i. contains a proteinaceous protoxin (Dubois
and Lewis 1981, Aronson et al. 1986) which is solubilized by enzymes in the
larval midgut, activating the toxin. The toxin binds to and lyses the midgut
epithelial cells, destroying the midgut cells and killing the infected larvae.
Recent published studies report minimal impact of B.t.L on nontarget
invertebrates (Car and deMoor 1984, Pistrang and Burger 1984, Gibbs et al.
1986, Merritt et al. 1989, Lacey and Mulla 1990, Wipﬂi and Merritt 1992).
Few studies have investigated toxicity of B.t.i. to ﬁsh, including both direct
and indirect (food-chain) toxicity, even though black ﬂy larvae are an
important diet component for trout (Elliott 1973, Williams 1981, Gibbs et al.
1986), salmon (Williams 1981), whiteﬁsh (Hocking and Pickering 1954), bass
(Hess 1983), creek chubs (Keast 1966), and numerous other species (Davies
1981, Hess 1983).

Brook trout, Salycliuus faminalis, were susceptible to high doses of
B.t.i., but mortality was attributable to xylene in the formulation (Fortin et
al. 1986). Mortality in the fathead minnow, Eimculmlcs nmmclas
Raﬁnesque, was attributed to dissolved oxygen depletion due to formulation
ingredients rather than to direct toxicity ﬁom the B.t.i. toxin (Snarski 1990).
Merritt et al. (1989) reported no signiﬁcant change in ﬁsh numbers

 

 

follow
term 1
site, I:

ofS.f

concex
classe:
or for:
sublet.
Contan

black ﬂ

egg to
ex1308111

“IElI' m
betwe e 17‘

 

%

following B.t.i. application in a Michigan river, although these were short
term experiments and most species were in low abundance at their study
site, making it unfeasible to detect ﬁsh population changes. Feeding habits
of S. fantinalis and the slimy sculpin, Bonus ccgnatus, were not aﬁ'ected by
B.t.i. (Gibbs et al. 1986).

The objectives of this study were to determine 1) the lethal
concentrations of B.t.i. to brook, brown, and smelhead trout with ﬁsh size i
classes from eg to 8.2 cm length, 2) if trout mortality was due to B.t.i. toxin

 

or formulation components, and 3) if trout experience lethal (mortality) or
sublethal (growth) effects after indirect B.t.i. exposure by consuming B.t.i;
contaminated black ﬂy larvae.

MATERIALS AND METHODS

This study had two primary components: direct toxicity tests at varying
B_.1,,i,(Teknar® HP-D) concentrations and exposure periods, and indirect
(food chain) toxicity tests by feeding ﬁngerling trout B.t.i.-contaminated
black ﬂy larvae. In the ﬁrst component, brook, S. fontimlis; brown, Balms
mum; and steelhead, anhcnmmus mykiss trout, ranging in size from
eg to 8.2 cm- length, were exposed to B.t.i. over a wide range of
concentrations (0 to 10,000 ppm; B.t.L volume to water volume ratio) and
exposure periods (15 min to 48 hr). The second component involved feeding
brown trout fry ELL-contaminated or live black ﬂy larvae for 5 d to compare
their mortality and growth over 30 d. All experiments were conducted
between 26 October 1990 and 7 June 1991 in the Zoology Dept. aquatics

97'

laboratory at Michigan State University. Trout were transported from state
ﬁsh hatcheries at Wolf Lake and Marquette, M1, for all but one study,
where brown trout were electroﬁshed from Hunt Creek, MI, U.S.A.

I T. . .I I I
GanemLemﬁmentaancedm B.t.i. treatments. and trout species
along with their associated size classes for toxicity trials, are given in Table

1. To measure mortality from differing concentration, a wide B.t.i.

 

concentration range at logarithmic intervals was initially tested at 48 hr
exposure to approximate 0 and 100% mortality limits. Then, concentrations
were narrowed over progressively smaller increments and ranges, to
compare viable and nonviable B.t.i, and to determine L050 values. Trout
were exposed to viable (nonautoclaved) and nonviable (autoclaved) B.t.i. to
determine if B.t.i, toxin or B.t.i, formulation was responsible for trout
mortality. Autoclaving denatures B.t.i. protein toxin and kills spores,
resulting in loss of toxicity (D. Grant, pers. comm.) without affecting
formulation stability (M. Zabik, pers. comm.). In order to verify that
autoclaved B.t.i. was non-toxic, I ran bioassays with mosquito (Acclcs
triscziatus) larvae using autoclaved and nonautoclaved B.t.i. at 0, 1, 10, 100,
and 1000 ppm. Mosquito mortality was 100% for larvae exposed to
nonautoclaved B.t.i. and negligible (< 1%) for autoclaved B.t.i, at all
dosages, showing that autoclaving B.t.i. for 45 min eliminated its viability.
Fish were maintained in a 600 L tank for a minimum of two days in

dechlorinated, carbon ﬁltered tap water (7 - 8 °C), prior to individual toxicity

Asa mﬁ how UQNVQHOSSNV _.— H m wﬂﬁwﬁwdoﬂ H z .A@m>w~00uﬁw ...on .maﬁﬂ maﬁa? H >N

.252 2532, 3 2833 £283 3 Add .858 Sn 5.8“:

 

 

S b we ZN> ooom .o o.m #85 mg
a N. we ZN> ooom dog dog .ooE .o ma M83 NH
m N. we > 83 .83 .oom 63 .o ma 838.5 S
m. m Sr. ZN> ooo.oH .ooom .83 .o ed 38588 9
Eva @ omH
m b am .m > ..Evm © 3 ..EN © 3 .o no wmmﬁmoum o
Eva © omH
m N. em .m > ..ﬁvm © 3 ..5m © 3 .o «N. 85.5 m
e N. we do .§ .§ .o > ooom .o md 55.5 b
e v we .om Jam .2 .o > ooom .o Na 58.5 o
ooom dog .83
e N. we 22¢ .83 .oomH .83 .o Na 58.5 m
e N. mv ZN> ooow .ooom .ooom .o ﬁn 550.5 a
m N. we ZN> ooo.oH .82 63 .o ﬂu 58.5 m
m N. we > ooo.oH .83 63 .3 .o ma 58.5 m
m 8 we > 88.9. .82 .03 .2 .8 meme e395 H
E2585 3.33 25 9:58: 1893 A88 $523 a
8a 858.56 voted «Add 85955858 5352 26.5 ~25
826.238 82 8.8898 Add 585 .26

 

 

.285 52x8 255 .So vows 55:58.5 pom m38232 mo .8225 was £258.56
82 .8th 8.8898 28 $3858: .8 @323 258: .8585Gouc8 Add :5ch one @28on 5328 .H 2nt

99

tests. They were fed trout pellets until experiments began. Experimental
units were 475 ml plastic drinking cups containing 20 eggs or 10 ﬁsh (1.5 -
2.3 cm length), or 4 L plastic tubs containing 5 ﬁsh (7 .2 - 8.2 cm length). At
the onset of each toxicity trial, experimental units were ﬁlled with tank
water to 75% of their capacity, and suspended in the tank water using
styrofoam rafts. This tank bath maintained a uniform and relatively
constant water temperature (7.5 :l: 0.5 °C) across experimental containers

 

during trials. All chambers were aerated with bubblers.

Water temperature, pH, and dissolved oxygen (D.O.) were measured in
experimental containers at the start and ﬁnish of each trial. Water was
changed in all containers immediately following B.t.i, exposure, and at 3 d
intervals, thereafter.

WM B.t.i. was added to experimental units at the
onset of each experiment. All 3.1.1, treatments (e.g., concentrations,
exposure duration, viability) were replicated ﬁve times in a complete
randomized design unless otherwise stated (Table 1). Controls consisted of
trout with no B.t.i.. Trout were randomly chosen and added to
experimental units.

Following exposure to B.t.i., brown trout eggs were held in their
respective containers until they hatched or died (up to 26 (1). Dead eggs,
identiﬁed by an opaque appearance and loss of eye spots, were removed
daily. For remaining trials, ﬁsh were held an additional 5 (1, except for one
steelhead trial where ﬁsh were held 4 d. During and after each trial, ﬁsh

were evaluated as live, immobilized, or dead. Live ﬁsh maintained a

100

normal upright position in the water and showed no sign of stress.
Immobilized ﬁsh were unable to maintain an upright position and rested
lateral or ventral side up, but still showed respiratory, convulsive, or
irritability movements (Sprague 1973). Dead ﬁsh lacked the above described
behavior and were removed daily.

WW AN OVAs were performed on all mortality data,
which were transformed via logarithmlo [x + 1] to normalize data, and
reduce variance heterogeneity, when necessary. Means were compared
using Tukey's test at p = 0.05 (Steel and Torrie 1980).

LC5os were determined for appropriate toxicity tests (tests that
contained B.t.i. concentrations that resulted in no 0 and 100% ﬁsh
mortality) using logarithmlo-probit transformations (Finney 1971) for B.t.i.
concentrations and percent trout mortality, respectively. Abbott's formula
(Abbott 1925) was used to correct mortality data against mortality in
controls when necessary.

WW Scanning electron micrographs (SEM)
were taken of brown trout gills from ﬁsh that were held in water containing
B.t.i. and water with no ELL to detect morphological changes, or presence
of particulates or mucous. Trout were exposed to 0 (control) and 2000 ppm
B.t.i. for 4 hr under the same experimental conditions and procedures
described for toxicity trials. The experimental unit was one trout per 4 L
tub and both treatments were replicated ﬁve times within two blocks.
Immediately after the 4 hr exposure period, trout were removed from the
water, gills dissected from the ﬁsh, and placed in ﬁxative for SEM.

Following the ﬁxation process, gills from each ﬁsh were scanned at 20X,

 

 

200]
was

fore

(mad
tnxk
(mnu
ﬁvet
expo:
abou1
mxsl
unﬂd
&nnpl
8Dace

blood

“ween

 

 

101

200x, 2000K, and 20,000X magniﬁcation and a representative micrograph
was taken for each gill sample (= one ﬁsh) from the same approximate area
for each gill at each SEM magniﬁcation.

W5. Blood from brown trout exposed and not
exposed to ELL was analyzed for blood 02 and 002 levels. Experimental
conditions and procedures for these trials were the same as the previous
toxicity studies except the experimental unit was an 18.9 L plastic bucket
containing two 17.8 - 20.3 cm long trout. Both treatments were replicated
ﬁve times. Trout were exposed to B.t.i. for 4 hr and control trout not
exposed. Blood was then drawn from the caudal artery and vein area,
about 3 cm anterior from the base of the tail, using a 5 ml B-D® syringe and
30G1/2 needle. Due to the small size of the trout, arterial or venous blood
could not taken without mixing the two blood types, so all analysed blood
samples were an arterial-venous blood mixture. Syringe and needle dead
space was ﬁlled with Neprin" salt solution to reduce coagulation. Alter
blood was drawn from each ﬁsh, syringes were placed on ice and taken to
the laboratory for blood-gas analysis. Blood-gas concentration
measurements were averaged across both ﬁsh per experimental unit for
each replicate. AN OVA was used to analyse %02 saturation and %002

concentration in the blood.

II E l l . | . 'I
This experiment involved feeding brown trout ﬁngerlings (mean length

4.3 cm) live or ELL-killed black ﬂy larvae, Qnenhia 513991311513 (Diptera:

102

Simuliidae), for ﬁve continuous days to study survival and growth over 30 d.
Three treatments were 1) trout fed excess black ﬂy larvae [that had been
treated with 30 ppm B.t.i. for 24 hr] for 5 d, 2) same as (1), with trout also
exposed to 30 ppm B.t.i. for 24 hr, and 3) control [trout fed excess live black
ﬂy larvae over 5 d with trout and black ﬂies exposed to no B.t.i.].
Treatments were replicated four times in a completely randomized design.

Experimental conditions were the same as for previously described
experiments using the 4 L plastic tub containing 10 trout each (mean length
= 4.3 cm). At the start of the experiment, trout were fed black ﬂies (live or
B.t.i.-killed depending on treatment) until they approached satiation (ca.
400 larvae / 10 trout) where feeding became suppressed. This was repeated
daily for 5 d. Trout were maintained for an additional 25 d and fed 0.3 g
trout chow / day/ 10 ﬁsh. Water was replaced and ﬁsh mortality recorded
every 2 to 4 d.

A subsample of 40 ﬁsh was taken at the onset of the experiment to
estimate initial trout length and at the end of the experiment to determine
30 d growth. Growth was determined by subtracting the initial length
estimate from the ﬁnal length for each ﬁsh, and averaging measurements
across the ﬁsh in each 4 L tub for each replicate and treatment. Total
mortality for each replicate was also recorded. AN OVA and Tukey's
Studentized range tests were used to determine treatment differences for
trout growth and mortality.

 

103

RESULTS

I I . ’I | |

Water temperature ranged 7 - 8 °C, D.0. between 82 - 100% saturation,
and pH ranged 8.4 - 8.9 in all experimental units for the duration of all
trials. Tank water used to ﬁll experimental units contained 0.68 mmho
soluble salts and 233 ppm alkalinity.

Trout generally became immobilized within 48 hr exposure, at B.t.i.
concentrations 2 1000 ppm. When immobilized (but still living) ﬁsh were
placed back into clean water, they recovered and survived the remainder of
the trial.

All three trout species were not sensitive to B.t.i. below 1000 ppm, and
species showed similar mortality rates at increased B.t.i. concentrations.
Signiﬁcant brown trout egg mortality occurred at 10,000 ppm B.t.i. with 27%
successful hatcheling emergence (Figure 1). All other concentrations of
B.t.i. resulted in insigniﬁcant mortality.

Signiﬁcant ﬁsh mortality, 51% at 1000 ppm and 100% at 10,000 ppm
B.t.i., was seen for 1.5 cm long brown trout (Figure 2, Trial 2). There were
no signiﬁcant mortality differences in brown trout exposed to viable and
nonviable B.t.i. (Figure 2, Trials 3, 4, 5) at all concentrations, with the
exception of Trial 5 at 1800 ppm B.t.i. The nonviable B.t.i. killed more trout
than viable B.t.i. at this concentration.

Brown trout mortality was directly related to B.t.i. exposure duration.
Data from Trial 6 were variable but exhibited a trend towards more
mortality at longer exposure periods (Figure 3). This experiment was

 

TRIAL 1; BROWN TROUT EGGS
a a
100- a

  
   
  

80'
60"

40'

MEAN % EGG EMERGENCE

 

0 10 100 1000 10000
B.T.l. CONCENTRATION (PPM)

Figure 1. Percent brown trout egg emergence over 26 (1 following 48 hr
exposure to varying ELL concentrations. Bars headed by different
letters represent signiﬁcant difference between means (p<0.05). Error
bars represent S.E.

Figure 2. Percent brown trout mortality over 7 (1 following 48 hr exposure to
varying B_._t_,i, concentrations (Trials 2-5), comparing viable and
nonviable B.t.i. (Trials 3-5). Bars headed by different letters represent
signiﬁcant difference between means (p<0.05) (Trial 2). Bar pairs
headed by * signiﬁes statistical difference between means (p<0.05), ns =

not signiﬁcant, nc = not calculated (Trials 3-5). Error bars represent
S.E.

MEAN % TROUT MORTALITY

 

105
TRIAL 2; 1.5 CM BROWN TROUT c
100
80
60
40

20

 

0 10 100 1000 10000

TRIAL 3; 2.1 CM BROWN TROUT no
100 -

 
 
   

80. I man.
'. nut-maxi.
60-
4o-

20-

 

0 100 1000 10000
TRIAL 4; 2.1 CM BROWN TR'O‘CUT

‘t‘i‘If'MM-Ffﬁﬁﬂn

.93

.
maria-.5

\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

 

. x»!

5715‘ 1;. .-.-.~. .-.-.~.- ,. ..-.-

0 1000 1200 1400 1600 1800 2000

B.T.I. CONCENTRATION (PPM)

106

TRIAL 6; 2.2 CM BROWN TROUT

 

 

60-

100 -
. b
so -
so -
>_ 40 -
E 20 ‘
< _
l— .
m o _ "
o
E o 12 24 36 4s
'5 TRIAL 7; 2.3 CM BROWN TROUT
o\° 30 _
z
<
I.”
E

40-

20—

,1

 

 

0 12 24 36 48

B.T.l. EXPOSURE PERIOD (HRS)

Figure 3. Percent brown trout mortality over 7 d at 0, 12, 36, and 48 hr
exposure to 2000 ppm 3.1.1, Bars headed by diﬁ'erent letters represent
signiﬁcant difference between means (p<0.05). Error bars represent S.E.

107

repeated (Trial 7). Brown trout exposed to B.t.i. for 36 and 48 hr suffered
signiﬁcantly greater mortality than ﬁsh exposed for S 24 hr (Figure 3, Trial
7).

There was no mortality with 7.2 cm long brown and 8.2 cm long
steelhead trout in Trials 8 & 9 at all concentrations (0 - 150 ppm) and
exposure periods (2 - 24 hr), therefore, the data were not presented.
Signiﬁcant mortality was recorded for 2.4 cm long steelhead trout exposed
to 10,000 ppm B.t.i. for 24 hr, but not at 5000 ppm or less (Figure 4, Trial 10).

Similar trends for brook trout were also apparent with signiﬁcant
mortality seen at 1500 ppm (Figure 5, Trial 11). There was signiﬁcantly
more mortality with ﬁsh exposed to nonviable versus viable B.t.i. at 1800
and 2000 ppm (Figure 5, Trial 12). Another trial was run at 2000 ppm B.t.i.
with 10 replicates per treatment that indicated no differences in brook trout
mortality between viable and nonviable B.t.i. at 2000 ppm (Figure 5, Trial
13).

LC50 values for brook and brown trout alevins exposed to B.t.i. over 48 hr
ranged between 1561 - 2321 ppm (Table 2). LC50 values could not be
calculated for steelhead trout owing to limited ﬁsh availability.

Wylie. SEM showed particulates and mucous
on the gills of some (about 50%) but not all trout exposed to B.t.i. (Figure 6).
The occurrence of particulates was variable among B.t.i. treated trout, with
some gills showing no (Figure 6A), and others showing some (Figure 6B)
particulate matter accumulation. Some gills from treated trout had a

mucous layer covering their surface (Figure 60), while others did not.

108

TRIAL 10; 2.4 CM STEELHEAD nc

 

 

 

100 "
E .
3’ - I viable B.t.i.
.. 80 .. . .
g . '14" nonovuable B.t.l.
5 60 .
h I
3
a: 40 .
F 1
a?
‘2: 20 "
u: no no
2 0 —, — . '_1 — .
0 1 000 5000 1 0000

B.T.I. CONCENTRATION (PPM)

Figure 4. Percent steelhead trout mortality over 5 d following 24 hr
exposure to varying B.t.i. concentrations. Bar pairs headed by *
signiﬁes statistical difference between means (p<0.05), ns = not
signiﬁcant, nc = not calculated. Error bars represent S.E.

Figure 5. Percent brook trout mortality over 7 d following 48 hr exposure to
varying B.t.i. concentrations (Trials 11-13), comparing viable and
nonviable B.t.i. (Trials 12-13). Bars headed by different letters represent
signiﬁcant difference between means (p<0.05), nc = not calculated
(Trials 11&13). Bar pairs headed by * signiﬁes statistical difference

between means (p<0.05), ns = not signiﬁcant, nc = not calculated (Trial
12). Error bars represent S.E.

MEAN % TROUT MORTALITY

109

TRIAL 1 1; 1.6 CM BROOK TROUT

 

  

 

100-
801 b
60-
4o-
20- a
0- — . T. _
0 100

500 1000 1500
B.T.I. CONCENTRATION (PPM)

TRIAL 12; 1.8 CM BROOK TROUT
100 -

1

80- . viable B.t.i.
non-viable B.t.i. .

60-

4o-

 

 

O 1400 1600 1800 2000
B.T.I. CONCEN'I’RA‘I'ION (PPM)

TRIAL 13; 2.0 CM BROOK TROUT
100 -

80~

d

60-
J

40-

20-

1 a

 

 

 

0

 

control viable nonviable
B.T.I. VIABIUTY

110

Table 2. LC50 values for 2.2 cm long brown and 1.8 cm long brook trout
exposed to B.t.i. for 48 hr.

 

 

 

 

LC501
trial trout
# species viable B.t.i. nonviable B.t.i.
5 brown 1691 1561
12 brook 2321 1792

 

1 Expressed as parts per million, B.t.i. volume to water volume ratio.

111A

Figure 6. Scanning electron micrographs of gills dissected from steelhead
trout exposed to 2000 ppm B.t.i. for 4 hr. A. Exposed gills with no
particulate and mucous accumulation (2000X). B. Exposed gills
showing particulates (2000X). C. Exposed gills showing mucous layer
(2000X). D. Exposed gills showing no particulates or mucous (2000X).

 

112

Gills taken from 'control' trout had no particulates or mucous (Figure 6D)
for all trout tested.

W Blood gas (02, 002) levels in B.t.i. exposed
and unexposed trout was highly variable. Percent 02 saturation ranged
from 19.0 - 89.0% in 'control' trout, and 2.2 - 31.6% in trout exposed to 4000
ppm B.t.i. for 4 hr (Figure 7A). Percent 02 saturation was less in blood
taken from B.t.i. exposed trout (0.05 < p < 0.10, F = 5.83) versus unexposed
trout. Carbon dioxide levels were less variable (14.8 - 27.0 mm Hg for
control trout, and 15.6 - 38.0 mm Hg with B.t.i.-treated trout). Partial
pressure 002 (mm Hg) was not signiﬁcantly different between the two
treatments (Figure 7B) (p > 0.10, F = 0.80).

11 E I l . I . .I
The ten trout in all replicates of each treatment consumed 350 - 400 live

or B.t.i.-killed black ﬂy larvae daily for ﬁve continuous days. There was no
brown trout mortality in all three treatments. Trout growth ranged 2.2 - 4.1
mm over 30 d, with no signiﬁcant differences in growth (p > 0.10, F = 1.49)
among the treatments (Figure 8).

DISCUSSION

I I . .I | |
The lack of ﬁsh mortality below 1000 ppm B.t.i. in all toxicity trials
indicated that this toxin is safe to hatchery trout when they are exposed to

113

 

 

 

 

 

 

 

 

 

5 100

"' - A.

E .

a 80 -

< 4

a)

z

m

‘5

O

32

Z

<

m

E B.t.i. treated trout untreated trout
100 - B

I? 80 -

E ‘

3; so ~

N .

0

U

z

<

m

E

 

 

 

 

B.t.i. treated trout untreated trout

TREATMENT

Figure 7. Percent 02 saturation (A), and 002 partial pressure (B), of blood

taken from steelhead trout exposured to 4000 ppm B.t.i. for 4 hr. (A.
0.05<p<0.10; B. p>0.10). Error bars represent S.E.

114

 

1°: BROWN TROUT FINGERLINGS

 

 

so 0 GROWTH (MM)

 

 

 

 

B.t.i. larvae B-t-i- larvae control
+B.t.i. water

Figure 8. Thirty day steelhead trout growth (increase in body length) in
three treatments, 1) trout fed excess black ﬂy larvae for 5 d that were
treated with 30 ppm B.t.i. for 24 hr, 2) same but trout also exposed to 30
ppm B.t.i. for 24 hr, 3) control treatment where trout were fed excess live
black ﬂy larvae over 5 d and water contained no B.t.i. Bars headed by the
same letters represent no signiﬁcant difference between means (p>0.10).
Error bars represent S.E.

115

recommended (15-22 ppm for 1 min) concentrations of B.t.i.. Exposure to
B.t.i. at 1000 ppm or less for 48 hr (= 196 X recommended exposure period)
did not induce mortality for all but one trial (Trial 2). Trial 2 included the
smallest (non-egg) size class (mean 1.5 cm length) of all ﬁsh tested.
Smaller ﬁsh may be more sensitive to B.t.i. than larger ﬁsh, although I did
not conduct side by side size-class-tests. Toxicity levels were similar for
eggs and ﬁsh.

Autoclaving B.t.i. did not reduce its toxicity, indicating that the B.t.i.
toxin was not responsible for mortality, demonstrating that formulation
component(s) were responsible for ﬁsh mortality. This result was
consistent across all trials for all three species. In fact, at 1800 ppm in
Trial 5, and 1800 and 2000 ppm in Trial 12, the autoclaved (nonviable) B.t.i.
was signiﬁcantly more toxic than viable B.t.i.. Fortin et a1. (1986) attributed
the formulation ingredient xylene in Teknar‘” HP-D to brook trout mortality
in their studies. Xylene was not in the B.t.i. formulation I used (D. Ross,
Zoecon Corp., pers. comm.). Due to proprietary reasons, I cannot report
the formulation ingredients in Teknar® .

L050 values for brown and brook trout also showed that B.t.i. was 'safe'
when these ﬁsh species were exposed to B.t.i. at recommended dosages and
exposure times. These L050 values were calculated from tests where ﬁsh
were exposed to B.t.i. over a time period 196X greater (48 hr versus 0.25 hr)
than that recommended for ﬁeld application.

ScanninLelectmericmaranhs. SEM was performed on ﬁsh gills to get
a better understanding of factors responsible for ﬁsh mortality. Fish often

116

became immobilized and appeared stressed at B.t.i. concentrations 2 1000
ppm. Particulates in the B.t.i. formulation, if adhering to the ﬁsh gill
surface, would be seen with SEM. The presence of particulates and
mucous on gills from some of the B.t.i.-treamd ﬁsh suggested that there
may be some interference with gas exchange at the gill surface. However,
results were variable. The gill ﬁxation process for SEM preparation may
wash or dissolve material adhering to the gill surface, and may have been,
in part, responsible for this variability among treawd trout. None of the
gills from the control trout had particulates or mucous attached.
Wags. Depleted oxygen levels in the blood of trout
exposed to B.t.i. agreed with our SEM ﬁndings. Apparently, ﬁsh gills lost
their capacity for oxygen uptake. The presence of particulates and mucous
on some of the 'treated' trout may be interfering with 02 uptake, but did not
affect 002 levels in the blood. The absence particulates on other 'treawd'
trout suggest that soluble components in the B.t.i. formulation may have
been responsible for suppressed blood 02 levels and eventual ﬁsh death at

high dosages.

II E I l . I . .I
Black ﬂy larvae ﬁlter, collect and ingest suspended particulates

(Wallace and Merritt 1980), including B.t.i., from the water column. Due to
their non-ﬁltering mode of feeding, ﬁsh (and other predators) probably
ingest minimal amounts of B.t.i. directly. They may however, indirectly
ingest B.t.i. through consuming B.t.i.-contaminated black ﬂy larvae. Large
masses of black ﬂy larvae drift downstream immediately following B.t.i.

117

applications (Merritt et al. 1989). Fish, and other consumers, could
potentially satiate on B.t.i.-contaminated black ﬂy larvae and indirectly
ingest high amounts of B.t.i..

These results indicated that trout mortality did not increase after the
trout were starved, then satiated with B.t.i.-contaminated black ﬂy larvae
for ﬁve continuous days. Snarski (1990) showed that fathead minnows were
not affected by ingesting B.t.i., and that B.t.i. spore counts in the ﬁsh and in
their gastrointestinal tract quickly dropped following B.t.i. exposure. Trout
growth was also not affected in our experiments when trout consumed
B.t.i.-contaminated prey.

I conclude from these experiments that direct and indirect exposure to
B.t.i. did not affect brook, brown and steelhead trout when it was applied at
recommended and higher than recommended rates. Fish mortality
occurred at a concentration about 100 X greater and an exposure period
about 200 X longer than recommended for larval black ﬂy control. L050
values for B.t.i. with brown and brook trout were > 100 X above
recommended application rates for exposure periods that were 196 X
greater than recommended (e.g., margin of safety = 100 x 196 = 19,600).
Formulation ingredients, and not B.t.i. toxin, was responsible for ﬁsh
mortality.

J '

118

ACKNOWLEDGMENTS

I express sincere thanks to J. Giesy, J. Kocik and B. Crawford for their
technical help and suggestions; MSU Center of Electron Optics for SEM
work; MSU veterinary clinical pathology staff for assistance with blood-gas
analyses; Michigan DNR for supplying trout; D. Ross, Zoecon Corp. for
supplying Teknarm HP-D and helpful comments; W. Cooper, D. Hall, and
J. Stout for helpful comments and use of the aquatics laboratory; E. Walker
for supplying bioassay mosquito larvae; and S. Usiak and L. Aronoﬁ‘ for
technical help. This research was supported by Sport Fishing Institute
Fund grant SFRP-90-26 awarded to RWM and MSW, and National
Institutes of Health grant A1-21884.

119

REFERENCES

Abbott, W.S. 1925. A method of computing the eﬂ‘ectiveness of an
insecticide. J. Econ. Entomol. 18: 265-267.

Aronson, A.I., W. Beckman, & P. Dunn. 1986. W W
and related pathogens. Microbiol. Rev. 50: 1-24.

Car, M., & F.C. De Moor. 1984. The response ofVaal River drift and
benthos to Simulium (Diptera: Nematocera) control using Bacillus

W var. iﬁmﬂﬁnsiﬁ (H-14). Onderstepoort J. Vet. Res.
51: 155-160.

Davies, D.M. 1981. Predators upon black ﬂies, pp. 139-158, In M.
Laird [ed.]. Black ﬂies: the future of biological methods in integrated
control. Academic Press, Inc., London, 1981.

Dejoux, C., & J .M. Elouard. 1990. Potential impact of microbial
insecticides on the freshwater environment, with special reference to
the WHO/UNDP/World Bank Onchocerciasis control programs, pp.
65-83, In M. Laird, L. A. Lacey and E. W. Davidson, [eds.]. Safety of
microbial insecticides. CRC Press, Inc., Boca Raton, FL. 259 pp.

Dubois, N.R., & F.B. Lewis. 1981. What is mm W? J.

120

Arboriculture. 7: 233-240.

Elliott, J .M. 1973. The food of brown and rainbow trout (Salmq

mm and S. gamed) in relation to the abundance of drifting
invertebrates in a mountain stream. Oecologia. 12: 329-347.

Finney, DJ. 1971. Probit analysis. 3rd ed. Cambridge University
Press, London. 333 pp.

Fortin, C., D. Lapointe, & G. Charpentier. 1986. Susceptibility of brook
trout (Salxelinns fontinalis) fry to a liquid formulation of Bacillus

thudngiensis serovar. amalgam (Teknar®) used for black ﬂy
control. Can. J. Fish. Aquat. Sci. 43: 1667-1670.

Gibbs, KE., F.C. Brautigam, C.S. Stubbs, & L.M. Zibilske. 1986.
Experimental applications of B.t.i. for larval black ﬂy control:
persistence and downstream carry, efﬁcacy, impact on non-target
invertebrates and ﬁsh feeding. Maine Life Sci. Agric. Exp. Stn. Tech.
Bull. 123:1-25.

Hess, L. 1983. Preliminary analysis of the food habits of some New
River ﬁshes with emphasis on black ﬂy utilization. Proc. New River

Symp. VPI&SU, Roanoke, VA. pp. 15-21.

Hocking, B., & L.R. Pickering. 1954. Observations on the bionomics

 

121

of some northern species of Simuliidae (Diptera). Can. J. Zool. 32:
99-113.

Keast, A. 1966. Trophic relationships in the ﬁsh fauna of a small
stream. Great Lakes Res. Div., Univ. Michigan Pub. 15: 51-79.

Lacey, L.A., & M.S. Mulla. 199d. Safety oansillns mum var.
ismslsnsis and Bacillus snhsszisns to non-target organisms in the
aquatic environment, pp. 169-188, In M. Laird, L. A. Lacey and E.

W. Davidson, [eds.]. Safety of microbial insecticides. CRC Press, Inc.,
Boca Raton, FL. 259 pp.

Merritt, R.W., E.D. Walker, M.A. Wilzbach, KW. Cummins, & W.T.
Morgan. 1989. A broad evaluation of B.t.i. for black ﬂy (Diptera:
Simuliidae) control in a Michigan river: efﬁcacy, carry and nontarget
effects on invertebrates and ﬁsh. J. Amer. Mosq. Control Assoc. 5:
397-415.

Pistrang, L.A., 8:. J .F. Burger. 1984. Effect of Bsgjllns thmjngisnsis
var. ismslsnsis on a genetically deﬁned population of black ﬂies
(Diptera: Simuliidae) and associated insects in a montane New
Hampshire stream. Can. Entomol. 116: 975-981.

Snarski, V.M. 1990. Interactions between Basillns thnzingisnsis var.

122

ismelcnsis and fathead minnows. Eimaahalis mains
Raﬁnesque, under laboratory conditions. Appl. Environ. Microbiol.

56: 2618-2622.

Sprague, J .B. 1973. The ABCs of pollutant bioassay using ﬁsh, pp. 6-30,
In Biological methods for the assessment of water quality. ASTM
STP 528. American Society for Testing and Materials, Philadelphia,
PA.

Steel, R.G.D., & J .H. Torrie. 1980. Principles and procedures of
statistics. 2nd ed. McGraw-Hill Book Co., New York, NY. 633 pp.

Wallace, J .B., & R.W. Merritt. 1980. Filter feeding ecology of aquatic
insects. Ann. Rev. Entomol. 25: 103-132.

WHO. 1979. Data sheet on the biological control agent Basillns
thmjngisnsis serotype H-14 (de Barjac, 1978), mimeographed
document, WHO/VBC/7 9.750, Rev. 1, VBC/BCDS/7 9.01, World
Health Organization.

Williams, DD. 1981. The ﬁrst diets of postemergent brook trout
(331121111118. fontinalis) and Atlantic salmon (Salim solar) alevins in

a Quebec river. Can. J. Fish. Aquat. Sci. 38: 765-771.

Wipﬂi, M.S. 1992. Direct and indirect effects Of Bacillus thuringiensis var.

123
israelensis on non-target aquatic insects and trout. Ph.D. dissertation.

Michigan State University, East Lansing, MI, U.S.A.

CHAPTER 4

POPULATION LEVEL DISTURBAN CE FROM BAQILLIIS

W VAR. ISRAELENSIS IN STREAMS: PREDATION
RESPONSES OF MACROINVERTEBRATE PREDATORS

ABSTRACT

A ﬁeld study was conducted to 1) assess the impact of the black ﬂy
(Diptera: Simuliidae) larvicide, Bacillus thnﬁngisnsis var. ismslsnsis de
Barjac (B.t.i.), on the diets of selected predatory stoneﬂies, and 2) determine
the importance of black ﬂy larvae as prey for selected predatory Plecoptera
and Odonata. Changes in feeding habits of Amends Mas and
Eszagnstins msdis (Plec0ptera: Perlidae) were studied in response to
locally reduced larval black ﬂy populations in two Laurentian shield rivers
by collecting predators from 0 to 10 (1 following B.t.i. application, and
inspecting their foregut contents for macroinvertebrate prey. Black ﬂies
were the major diet component before B.t.i. was applied, but the number of
black ﬂies cOnsumed dropped signiﬁcantly for both stoneﬂy species
following B.t.i. application. Total number of prey consumed decreased for
A. lysmjss but not for E. msdin, and non-black ﬂy prey consumption

signiﬁcantly increased for 12. media but not for A. ms, following B.t.i.
application.

124

125

In prey choice trials conducted streamside in artiﬁcial channels, A.
lymn’ss andL msdin showed no selectivity or preference for black ﬂy over
mayﬂy prey. Issnsﬂn signstn and I. dissls (Plecoptera: Perlodidae).
however, consumed signiﬁcantly more black ﬂies than both Essﬁs
ﬂsﬂstrigs (Ephemeroptera: Baetidae) and Epssm gitrsn (Heptageniidae)
prey. mm mm (Odonata: Aeshnidae) consumed more 8. Was
than S. m prey, showing a switch to the more abundant prey taxon,
at increasing prey densities.

I concluded that B.t.i. indirectly and differentially aﬁ‘ected predators by
reducing black ﬂy biomass. Perlodidae nymphs relied more heavily on
black ﬂies and should subsequently be most affected by B.t.i. applications,
Perlidae were not selective towards black ﬂies and should be moderately
susceptible. Odonata selected other prey over black ﬂies and are probably
least susceptible to B.t.i.. Predator-prey interactions and prey community
structure may be indirectly affected when predators such as E. msdis
switch to consuming alternate prey, following B.t.i. application.

 

126

INTRODUCTION

Disturbance in stream communities varies spatially and temporally, in
magnitude, frequency, predictability, and in the time required for
disturbance to occur (Sousa 1984). Although deﬁnitional discrepancies
exist (Poff 1992), White and Pickett (1985) broadly deﬁned disturbance as
"any relatively discrete event in time that disrupts ecosystem, community,
or population structure, and changes resources, substrate availability, or
the physical environment".

The responses to disturbance of various aquatic biota vary according to
the nature of the disturbance (e.g., spate vs. organic pollution), and show
differing levels of resilience, depending on the biological, chemical, and
physical factors of the system involved (Resh et al. 1988). Anthropogenic
perturbations, such as chemical pollution, generally invoke community
level responses as opposed to population or trophic level responses, with
repercussions lasting up to many years, depending on the nature of the
disturbance (Wallace 1990).

Bacillus thnringismsis var. ismelansis de Barjac (B.t.i.) is a highly
eﬂ'ective biological larvicide used against black ﬂies (Dejoux et al. 1985, De
Moor and Car 1986, Lacey and Undeen 1987, Lacey and Mulla 1990), with
minimal direct (toxic) effects on non-target fauna (Molloy and J amnback
1981, Gibbs et al. 1986, Merritt et al. 1989, Lacey and Mulla 1990). The high
efﬁcacy and selectivity of this bacterial larvicide has prompted its increased

world-wide use for suppressing black ﬂy populations (Dejoux and Elouard

127

1990).

Information is lacking on functional and ecological impact associated
with applying B.t.i. and removing an entire population (e.g., black ﬂies)
from stream community food webs. Any intervention that disrupts aquatic
community structure could directly or indirectly affect other trophic levels.
If black ﬂies comprise a large fraction of animal standing stock in a given
system, the sudden loss of black ﬂy biomass may have negative or positive
effects on other populations, particularly on predators and competitors of
black ﬂies. The impact on predators may be minimal, however, if the
predators are generalists and switch to alternate prey, assuming suﬁcient
alternate prey are available. Specialist predators would suffer greatest
indirect consequences through prey resource loss. Competitors, on the
other hand, may beneﬁt through additional food or space resources when
black ﬂies are removed from the system.

Indirect effects also may be felt by other taxa in the same trophic level
with black ﬂies, through shared predators. Both positive and negative
indirect interactions among non-competing prey are theoretically possible
(Abrams 1987 , Holt and Kotler 1987), and have been experimentally
demonstrated (Schmitt 1987). Predators switching to alternate prey
(following black ﬂy biomass loss) may signiﬁcantly impact these alternate
prey populatiOns, leading to a restructuring of the prey community and
potentially resulting in successional repercussions throughout the food
web. Miller and Kerfoot (1987) discuss such indirect links in detail where
species A affects species C indirectly by affecting species B.

The objectives of this study were to 1) determine the relative importance

 

128

of black ﬂy larvae as prey for selected macroinvertebrate predators below
lake outlets in two Laurentian shield streams, 2) study the direct trophic
impact of prey population level disturbance on these predators by removing
black ﬂy larvae ﬁ'om these systems, using B.t.i., and 3) investigate prey
preference of these and other macroinvertebrate predators relating to black

ﬂy and alternate prey abundances.

MATERIALS AND METHODS

Field experiments investigated in-stream changes in feeding habits of
two lotic predators, Asmnsnn‘s lysmiss (Newman) and Esmgnstins msdis
(Walker) (Plecoptera: Perlidae), following B.t.i. application to remove the
black ﬂy populations. Prey choice with A. lysmiss, 2. media, Issnsﬂn
signstn (Banks), I. digsls Frison (Plecoptera: Perlodidae), and Bonds
ﬁncsa (Say) (Odonata: Aeshnidae) was also studied, in artiﬁcial streams.
Prey included black ﬂies, Emsimnlinm fnsgnm Syme and Davies and
Simulium W(Say) (Diptera: Simuliidae), and mayﬂies, Bastis
ﬂsﬁstﬁgn McDunnough (Ephemeroptera: Baetidae) and Ensgms mrss
(Walker) (Heptageniidae). These insects are common macroinvertebrates
in lotic systems (Merritt and Cummins 1984, Stewart and Stark 1988) and
were abundant at the selected ﬁeld sites.

51115113125.

These experiments were conducted during April-May 1990 at Morgan

Creek, Marquette Co., and during May-June 1991 at Medora and

129

Manganese Rivers, Keweenaw Co., Michigan, U.S.A. These locations were
chosen because macroinvertebrate predators, black ﬂy larvae and other
prey were abundant during the study periods.
I l . I l D. I 1

Three hypotheses on trophic level disturbance in these stream systems
were tested: H1: reducing the prey population would reduce predation levels
of predatory macroinvertebrates (e.g., stoneﬂies), H2: predators would
switch to consuming alternate prey taxa in the absence of black ﬂies, and
H3: removing an abundant prey taxa (e.g., black ﬂies) would indirectly
affect other prey taxa (through the shared predator link). Stomach content
analyses of nymphal predatory stoneﬂies collected from disturbed and
control habitats were used to test these hypotheses.

WWI. Changes in the feeding habits of the
predator, A. lysmjss, was studied in response to black ﬂy population
disturbance from B.t.i.. A 700 m stretch Of Manganese River was sectioned
into six overlapping (200 m long) blocks such that sequential blocks
overlapped each previous block by about 50% (Figure 1), as one progressed
upstream. Block 1 was farthest downstream, and block 6 farthest upstream,
ending approximately 200 m below Lake Manganese outlet. This design
allowed for treatment comparisons 'ABOVE and BELOW' and 'BEFORE and
AFFER' disturbance, across six blocks. A

B.t.i. (Teknar® HP-D) was applied (100 ppm @ 1 min) weekly to each
successive block, starting with block 1 during week 1, and ending with block 6
during week 6. To estimate predator diets, 10 A 1yc_or_i_as nymphs were

collected, using a D-frame kick net, from erosional zones in each 'plot' within

130

   
 

     

Lake Manganese

<—
1U6JJ03

ABOVE/BELOW

 

 

 

 

 

 

 

 

 

 

 

 

 

 

BEFORE/AFTER
Block 6 V ‘
Block 6
‘.
~> <
Block 5 5
Block S
v v 1
Block 4 Block 4
b , :
BIOCK 3 BlOCk 3
r 1
Block 2
Block 2
Le» <1-
<
Block 1
Block I
<-

 

 

 

Figure 1. Diagramatic representation of experimental design used to
measure diet changes of Amms MES following Basillns
anngjsnsis var. mslsnsis disturbance, in Manganese River. AP- =
B.t.i. application point.

 

 

 

 

 

 

131

each respective block, one hour before and one week following m
application at a midpoint within that block. These nymphs were
immediately placed in a vial containing 70% ETOH for each treatment per
each block. Digestive tracts were removed from each predator, and their
contents down to and including the foregut, were carefully teased out,
macroinvertebrate prey items were identified to lowest possible taxon and
counted, for each predator. These digestive tract contents will hereafter be
referred to as 'foregut contents' in this and following experiments.

MW A 50 m reach of river
immediately below both Medora and Manganese Lake outlets was selected
to investigate changes in feeding habits with two predators, A. hearing and
2, media, following larval black ﬂy removal from their respective
communities.

First, to determine their natural diet, 10 A, homing and 10 2. media
were collected from Medora R. and Manganese R., respectively (using a D-
frame kick net), and immediately placed in 70% ETOH. Predator digestive
tracts were examined for macroinvertebrate prey, as above.

Following the above sampling scheme, one B.t.i. application (100 ppm 0
1 min) was made to both Medora and Manganese Rivers, and three stoneﬂy
collections took place at 5 day intervals beginning the day B.t.i. was applied
(=Od). Ten predators were collected per river on each sampling day (0d, 5d,
10d), and gut contents identiﬁed as above. Thus, experimental controls
were 0d, which were compared to 5d and 10d. Stones, with attached black
ﬂies, were collected from B.t.i.-treated stream reaches 1, 24, 48, and 72 hr

132

after application, and indicated that there was nearly 100% larval black ﬂy
mortality within 1 hr, and also showed that black ﬂy larvae no longer
remained intact and attached to substrates beyond 48 hr.

A non-m-treated reach in Medora River served as an additional
control. This design provided an additional comparison, 'm-treated
habitat' (=BELOW) and 'untreated habitat' (=ABOVE). This was not
feasible in Manganese River.

Prey composition and densities were estimated on stone substrates
(predator microhabitats) at each site one day prior to the experiment. Five
stones ca. 15 cm mean diameter were selected at random within each study
reach, removed from the stream (with a D-frame net placed immediately
below the stones to collect macroinvertebrates drifting from the stones), and
all macroinvertebrates were brushed from the stones and net into white
enamel sorting trays. Macroinvertebrates detectable by the naked eye were
collected and placed in 70% ETOH, and later sorted to the family level and
counted, for each rock (n = 5) for both rivers. Stone surface area was
estimated by completely covering each stone with a single layer of
aluminum foil. The foil was then removed from each stone and placed ﬂat
on 20 cm x 28 cm (1 cm2 increment) grids, and the number of grids covered
by the foil counted for each stone. Prey densities were then calculated per
100 cm2 stone surface area.

B l | l . E . |

The hypotheses tested in these experiments were: H1: macroinvertebrate
predation is higher on black ﬂies versus baetid and/or heptageniid
mayﬂies, and H2: predator growth rates are greater on diets higher in

133

black ﬂy composition.

Two sets of prey choice trials were conducted with ﬁve predator species,
in artiﬁicial channels (Wipﬂi 1992). The ﬁrst set of choice trials involved
offering 2, media and L signata equal ratios (10:10:10) of black ﬂies
mm mm baetid mayﬂies Baatia mm and heptageniid
mayﬂies Enema yjtma, daily over nine days. These feeding trials were
conducted in streamside at the Morgan Creek ﬁeld site. Stream water was
gravity-fed to channels each containing stream-collected cobble substrate,
one predator, and 30 prey. Prey body lengths ranged 3-4 mm; size classes
co-occurring naturally with predators at this site. Control channels
contained no predator. Predator and control treatments were replicated 10
and 5 times, respectively, utilizing a complete randomized design. Controls
were used to monitor and correct for 'non-predator' prey loss. Predation
was measured by counting consumed prey every 24 hrs, and dead, injured,
and consumed prey were replaced daily, until day nine.

The second set of prey choice trials involved offering A, 1159135, L
dicala, and B, maﬁa ﬁve ratios of black ﬂies, S. W, to mayﬂies, B,
ﬂaxiatrjga. Sjmnmlm-Baetis prey ratios were 0:16, 4:12, 8:8, 12:4, and 16:0
at 16 total prey, and 0:20, 5:15, 10:10, 15:5, and 20:0, at 20 total prey.
Consumed prey were counted and replaced daily over 6, 4, and 1 d for A,
lymﬁaa, L digala, and B. 1mm, respectively. Total prey number was
maintained at 16 (for one day) and 20 (for ﬁve days) for A, lymn'aa, and at 16
for L dicala and H.. m. Concurrent controls included no predators to

evaluate and correct for non-predator prey 'loss'. All treatments were

134

replicated four times, and predation was monitored and prey replaced as in
the previous choice experiments.

Aemnemia lyemjiae growth was measured, and the corresponding
instantaneous growth rates (IGRs) were calculated for individual nymphs
in each treatment (i.e., prey ratio) over 6 d to determine if predator growth
differed on Simulium vs. Baetie diets. Waters (1977) equation, IGR = [In (Mr
/M) / t], was used to measure instantaneous growth rates, where Mi =
initial nymphal mass, Mr = ﬁnal nymphal mass, and t = time (6 d). Towel
blotted nymphal wet masses were measured with a Sartorius® 1207 MP2
balance to the nearest 0.001 mg.

A complete randomized design was employed for experiments with L
signata, L dicaladi. media, Bi yineea, and A, lyemziae. Treatments were
randomly assigned to experimental units (e.g., individual artiﬁcial
streams).

SI I' |° ] l

Treatments for the Manganese River trophic level disturbance
experiment were longitudinally blocked upstream, facilitating a RCB
design with six blocks (n = 6). Although this design provided true
replication, treatments could not be randomly assigned (interdispersed)
within blocks. Water current carries B.t.i. downstream from the point of its
application, resulting in the upstream portion of each block serving as the
untreated control.

In the Medora-Manganese Rivers experiment on trophic level
disturbance, the experimental unit was an individual predator (n = 10).

One needs to be cautious when using individual predators as replicates

135

because of pseudoreplication (Hurlbert 1984), however, treatments could not
be blocked across additional rivers because of legal restrictions.

Predator foregut contents from disturbance experiments were analysed
for three or four components (depending on experiment): number of black
ﬂies, number of chironomids, number of other prey, and number of total
prey. AN OVA was used to test treatment effects on predation within each
of these prey catagories. Predator foregut contents from the control
habitats were compared to foregut contents of predators collected from
disturbed habitats (n = 10, Medora-Manganese R. experiment; n = 6,
Manganese R. experiment).

Prey consumption data from prey choice trials with L media and L
signata were compared between prey for each predator species using a chi-
squared test. Daily prey consumption data were averaged over the nine
days for each of the ten individual predators (for both species), and
compared to expected consumption at 1:1:1 (null hypothesis). In addition,
predation on total prey (three species combined) and individual species was
compared between L dieaia and 2. media using two-way AN OVA (predator
species at 2 levels x day at 9 levels) with predators as replicates (n = 10).

In choice trials with A. lyeoijae, L dieala, and B. yingea, numbers of
prey consumed as a function of prey density were plotted (for each prey
type), data linearized (loglob' + 1]) if needed, to ﬁt a linear model, and a
linear least squares regression ﬁtted to the means. Slopes from regressions
were compared between prey types for each predator.

Amnemja lyeen'ae growth data were analysed using ANCOVA (SAS
1985) across 5 treatments with initial nymphal body mass as the covariate

 

136

(n = 4).

Data were checked for normality, variance homogeneity, additivity, and
proper model ﬁt, then transformed when appropriate. Statistical power
was calculated for all experiments (Cohen 1988).

RESULTS

I l . I l D' | I
WILL Predation levels by A, m nymphs n
was signiﬁcantly less in B.t.i. disturbed habitats (Figures 2 & 3). Predators

 

consumed signiﬁcantly fewer black ﬂies in disturbed habitats when
comparing both BEFORE and AFTER (Figure 2A), and ABOVE and
BELOW (Figure 3A) the B_,Li, disturbance (p<0.01). Signiﬁcantly fewer
Chironomidae also were consumed by these predators comparing BEFORE
and AFTER (p<0.05) (Figure 2B), but not ABOVE and BELOW (p>0.05)
(Figure 3B) the disturbance. Other prey (non-black ﬂy, non-chironomid
prey) were not differentially preyed upon by A, lymziaa between the
disturbed and non-disturbed habitats (Figures 2C & 3C) (p>0.05). Total
prey taken by these predators was signiﬁcantly different between the
disturbed and non-disturbed habitats when the data were analysed
BEFORE and AFTER (p<0.001) (ﬁgure 2D) and ABOVE and BELOW
(p<0.05) (Figure 3D).

W The macroinvertebrate diet of
A, 1192235 and 12, media closely coincided with prey composition (Table 1)

Figure 2. Macroinvertebrate content in the foreguts of Aerenegria lyeerias
nymphs BEFORE and AFTER disturbance from B.t.i. in the Manganese
River disturbance experiment (n = 6). [ns = not signiﬁcant @ p = 0.05, * =
p < 0.05, ** = p < 0.01, *** = p < 0.001] [Power (= 1 - beta); A. = 0.95, B. =
0.10, C. = 0.79, D. = 0.94].

NUMBER OF PREY ITEMS IN THE FOREGUTS OF 10 PREDATORS

137

 

40

d

30‘

q

20.

10-

1

 

Oar—v—Q—v—O—I—Q—I—Qq—Q—I—O—F‘

40

A. BLACK FLIES

—9— BEFORE
—O— AFTER

Mu

 

30-

20..

q

10~

B. CHIRONOMIDAE

—9— BEFORE
—.— AFTER

 

40

 

 

30-

20-

10a

C. OTHER PREY

—e— BEFORE
—O— AFTER

 

 

 

40

30-

20..

10-

 

D. COMBINED PREY

—e— BEFORE
—O— AFTER

 

 

 

Figure 3. Macroinvertebrate content in the foreguts of Amneeﬁa lyeqrias
nymphs ABOVE and BELOW disturbance ﬁom ELL in the Manganese
River disturbance experiment (11 = 6). [ns = not signiﬁcant @ p = 0.05, * =
p < 0.05, ** = p < 0.01, *** = p < 0.001] [Power (= 1 - beta); A. = 0.52, B. =
0.08, C. = 0.42, D. = 0.80].

NUMBER OF PREY ITEMS IN THE FOREGUTS OF 10 PREDATORS

138

 

 

A. BLACK FLIES

—e— ABOVE
—-O—- BELOW

 

 

 

B. CHIRONOMIDAE

—-6— ABOVE
—-.— BELOW

 

 

 

20-

10-

C. OTHER PREY

—e— ABOVE
—O— BELOW

"8

 

‘ I ‘7 I I V I I I U "

 

 

D. COMBINED PREY

—e— ABOVE
—O—- BELOW

 

 

 

 

 

139

Table 1. Density estimates of macroinvertebrate prey1 in Medora and
Manganese Rivers 1 d before starting population level disturbance
experiments (n = 5).

 

 

 

 

 

 

Medora River Manganese River

Taxa #l100 cm2 :1: sem % #/100 cm2 i sem %
Diptera

Simuliidae 249.5 2!: 36.6 92.54 360.2 :1: 42.9 91.79

Chironomidae 7.0 i 0.7 2.59 13.4 i 1.5 3.41
Trichoptera

Hydropsychidae 10.1 i 2.6 3.74 17.4 i 1 .5 4.43
Ephemeroptera

Baetidae 0.8 i 0.4 0.30 1 .2 :t 0.2 0.31

Heptageniidae 0.1 i 0.1 0.04 0.0 0.00

Trichorythidae 0.1 :t 0.1 0.04 0.0 0.00
Plecoptera

Perlidae 0.8 :l: 0.4 0.30 0.0 0.00
Acarina 1.1 i 0.5 0.41 0.2 :1: 0.2 0.06
other 0.1 :l: 0.1 0.04 0.0 0.00

 

1Prey included macroinvertebrates visible to the naked eye while in white
sorting trays, but less than half the body length of the stoneﬂy predators
under investigation.

140

in both systems (Table 2). Simuliidae prey were the most abundant

 

(numerically 92-93% of all prey) at the study sites, and were consumed most
frequently (numerically 90-93% of prey). Hydropsychidae and
Chironomidae composition at the site ranged 34%, while their composition
in predator foreguts ranged 0-6%. Remaining prey taxa were less frequent
in both stream reaches and in both predator's diets (Tables 1 & 2).

Both predator species consumed signiﬁcantly fewer black ﬂies at 5d and
10d relative to 0d post-m-disturbance (p<0.05) (Figures 4A 8: 4D).
Amnemja Medan collected from the BALL-treated reach (=BELOW) in
Medora R. also consumed signiﬁcantly fewer black ﬂies than those
collected ﬁom the untreated reach (=ABOVE) at 5d and 10d post-
disturbance (p < 0.05) (Figure 5A).

Bazagnetina media collected at 5d and 10d intervals consumed
signiﬁcantly more alternate (non-black ﬂy) prey relative to those collected at
0d (p<0.05) (Figure 4E), with hydropsychid caddisﬂies comprising the
majority of consumed alternate prey. Aemnemzia lyeeziae did not show a
signiﬁcant predation shift to alternate prey in ELL-disturbed reaches
relative to controls (p>0.05) (Figures 4B & 5B).

Total prey consumed by predators in ELL-treated habitats (AFTER) was
signiﬁcantly less for A, lyemjaa relative to those predators collected from
control (BEFORE) habitats (Figure 4C) at both 5d and 10d (p<0.05), reﬂecting
black ﬂy biomass loss. When comparing predator diets in ABOVE vs.

 

BELOW treatments, total prey consumption for A, lyceziaa nymphs was
signiﬁcantly different at 5d (p<0.05), but not 10d (p=0.06) (Figure 5C).

141

Table 2. Mean number of macroinvertebrate prey items in foreguts of

mm Macias collected from Medora River. and Baresnetina media
collected from Manganese River, 1 d before starting Medora-Manganese

Rivers population level disturbance experiments (n = 10).

 

 

 

 

 

Medora River Manganese River

Taxa #/foregut i sem % #lforegut :l: sem %
Diptera

Simuliidae 3.7 :t: 1.7 90.24 10.5 :t: 2.4 92.92

Chironomidae 0.2 i 0.1 4.88 0.7 i 0.4 6.19
Trichoptera

Hydropsychidae 0.1 :t 0.1 2.44 0.0 0.00
Ephemeroptera

Baetidae 0.0 0.00 0.0 0.00

Heptageniidae 0.0 0.00 0.0 0.00

Trichorythidae 0.0 0.00 0.0 0.00
Plecoptera

Perlidae 0.1 :1: 0.1 2.44 0.0 0.00
Acarina 0.0 0.00 0.0 0.00

other 0.0 0.00 0.1 i 0.1 0.89

 

 

142

 

 

 

 
  

 

 

 

 

 

 

 

 

 

 

 
  

 

 

 

 

 

 

Acroneurla lycarias Paragnetlna media
6 q A. BIACK FUES 16 1 D. BLACK FUES
'— [ 12 ‘
(=5 4 ‘ 4
‘a‘:‘ 8 ~
8 2 - -- l *
g: 311 APPLIED 4 ‘ BTI APPLIED t
l- v ‘ *
g 0 l o I T I if
m
E 6 q B. OTHER PREY 13 . E. OTHER may
m ‘
3f 12 -
g 4 "' 1
BTI APPLIED s .
In ~ - *
I: 2 l « BTI APPUED
>- ‘ 4 . ¢
m
E ‘ (B ns 1
u. 0 ?\4—————,— 0 "—"l' I I
0
E 6 C. rCOMBINED PREY 16 . F. COMBINED PREY
In ‘ ‘
g 1
z 4 . 12 4
Z ‘ ns ns
5 . e-
5 2 . 0 ‘ B11 APPLIED
BTI APPLIED 4 .
* I « I
o I o I 1 I
0d 5d 10d 0d 5d 10d

DAY RELATIVE TO BTI APPLICATION

 

Figure 4. Macroinvertebrate content in the foreguts of Aemnemja lyeeﬁae
and Earagnetina media nymphs BEFORE (= 0d) and AFTER (= 5d & 10d)
disturbance from B.t.i. in the Medora and Manganese Rivers (n=10). [ns
= not signiﬁcant @ p = 0.05, * = p < 0.05, ** = p < 0.01, *** = p < 0.001]
[Power ( = 1 - beta); A. 5d = 0.62 & 10d = 0.67, B. 5d = 0.27 & 10d = 0.27, C.
5d = 0.57 & 10d = 0.67, D. 5d = 0.98 & 10d = 0.94, E. 5d = 0.57 & 10d = 0.52,

F. 5d = 0.27 & 10d = 0.28].

143

 

   

Acroneurla lycorias

6 . A. BLACK FLIES
ABOVE
BELOW

4 .

2 .

BTI APPLIED
0 I

 

 

 

 

6- B. OTHER PREY

—.—- ABOVE
—9— BELOW

' BTI APPLIED

 

 

. ns ns

 

 

C. COMBINED PREY

—.— ABOVE
* —e— BELOW

MEAN NUMBER OF PREY ITEMS PER PREDATOR FOREGUT

 

 

 

 

DAY RELATIVE TO BTI APPLICATION

Figure 5. Macroinvertebrate content in the foreguts of W W
nymphs ABOVE and BELOW disturbance from B_,_t,i, in the Medora-
Manganese Rivers disturbance experiment (n = 10). [ns = not
signiﬁcant @ p = 0.05, * = p < 0.05, ** = p < 0.01, *** = p < 0.001] [Power (=
1 - beta); A. 5d = 0.89 & 10d = 0.53, B. 5d = 0.05 & 10d = 0.05, C. 5d = 0.89 &
10d = 0.53].

144

Earagneiina media, on the other hand, did not consume signiﬁcantly less
total prey following Mi, application (p>0.05) (Figure 4F), owing to
signiﬁcant increased predation on alternate prey (p<0.05) (Figure 4E).

12 I | I . E . I

Predators L aignata and 2,, media showed a signiﬁcant predation
difference among prey types. Iagperia signata nymphs consumed
signiﬁcantly more black ﬂies than mayﬂies (p<0.05) (Figure 6A). This
predator consumed 1.5-3.5 Simulium, 0.1-1.3 Baetie, and 0-0.4 Enema per
day.

Eaiagnetina media, on the other hand, did not consume signiﬁcantly
more of one prey type (p>0.05) (Figure 6B). Predation rates ranged between
0.5-5.5 prey eaten/predator/day, on the three prey types.

There were no signiﬁcant differences between the number of black ﬂies
consumed by 2.. media and L eignata (p>0.05) (Figure 7), but there were
signiﬁcant differences between combined prey (combined across the three
prey species), and individual mayﬂy taxa. Earagnetina media nymphs
consumed signiﬁcantly more combined, Baetia, and Emma prey than did
I. signata (p<0.01) (Figure 7).

Remaining prey choice trials indicated signiﬁcant differences in prey
types (Simulium vs. Baetie) consumed between two of the three predators.
Amnenzia Medan consumed similar amounts of the two prey types
(Figure 8A), with regression slopes between prey types not signiﬁcantly
different (p>0.05). Iaeaeﬂa dieaia consumed signiﬁcantly more Simulium
than Baeﬁe(p<0.001) (Figure 8B). BmLezia yineea demonstrated a different

 

 

145

 

 

 

 

6
E 5 A. Isoperla slgna ta
5 I —o— Simulium
E 4 - —o—- Baetis
B I —a— Epeorus
I: 3 -
a. .
o 2 '-
u.I . a
s 1-

1 x b

,_ CE: 4.»
2 s
E 71 B . Paragnetlna media
a. u
u. 6 - —o— Simulium a
g ——o— Baelis " a
I.”
m
E
a . a
z
z
<
m
2

 

 

 

 

Figure 6. Predation on three prey taxa by A. M sigeata and B.
Banagnﬁina media during prey choice experiments conducted
streamside in artiﬁcial channels (n=10). [prey taxa followed by the same
letter within each graph are not statistically different @ p = 0.05] [Power
(= 1 - beta); A. = 0.86, B. = 0.11].

146

 

100 ‘
‘ I Isopeda signata
80 . Paragnetina media

 

 

 

 

Simulium Baetis Epeorus Spray
PREY SPECIES combined

MEAN I PREY CONSUMED I PREDATOR / 90

Figure 7. Mean number of prey consumed between Isapeﬂa signata and
Bax‘agaetina media predators during prey choice experiments
conducted streamside in artiﬁcial channels (n=10). [ns = not signiﬁcant
@ p = 0.05, * = p < 0.05, ** = p < 0.01, *** = p < 0.001] [Power (= 1 - beta);
Simulium = 0.07, Baejzie = 0.47, ED391115 = 0.41, total = 0.24.

147

 

 

 

  

 

 

 

1:: A. Acroneuria Modes
2 4 Baetis
a: SImuIIum
2
<
G
Ll]
I:
n.
ﬁ 0 I I I 1
'u-
a 14 - B
>_ . . lsoperla dicala
D. 10 -
u_ .
O 8 - Simulium
2 4 -
D a
g 2 q 4—3 Baetis
< 0 r . L
E 14
a: 12 . C. Boyeria vinosa
P '1
< '1 .
m 839115
2
Q
'—
g Simulium
In
a:
n.

 

 

0 I I I I
O 25 50 75 1 00

 

°/o PREY IN ENVIRONMENT

Figure 8. Feeding response curves for A. We ineriae, B. legeeLla
disala. and C. Roxana mesa. over a range of Baetis and Simulium prey
compositions. [Difference between slopes of the two regression lines
tested at p = 0.05; A. no signiﬁcant difference (Power = 0.84), B.
signiﬁcantly different (Power = 0.98), C. signiﬁcantly different (Power =
0.98)]. Lines represent a constant prey proportion consumed by
predators over increasing % prey in environment.

 

 

 

 

pl‘l
pﬂ

inc

avai

148

prey density-prey type-predation relationship than both plecopteran
predators. These nymphs consumed signiﬁcantly more Baetia than
Simulium (p<0.001), with a switch (Price 1984) to the more abundant prey at
increasing prey density (Figure 8C) for both prey species.

Growth for A, lyeen'ae nymphs was not signiﬁcantly different among
the ﬁve black ﬂy—mayﬂy ratios (p>0.05) (Table 3). IGRs ranged 36-53
ug/mg/d for these predatory stoneﬂies for the ﬁve treatments during the six

day experiment.
DISCUSSION

The foregut contents of A, lyeeiiaa and 2, media closely reﬂected prey
availability. Black ﬂies were the major prey taxa available to the
macroinvertebrate predators at the study sites and likewise formed the
largest fraction of A, lyeen'ae and 2, media diets during the study.
Predatory stoneﬂies display differing levels of generalist predation, that
may or may not reﬂect prey availability (Stewart and Stark 1988).

Removing black ﬂy larvae from these aquatic systems signiﬁcantly
impacted overall predation levels and composition of prey consumed by A,
lyeeziaa and 2, media. I hypothesized that these opportunistic predators
would switch to alternate prey; this was the case for 2, media, but not A,
lyeeziae. One explanation for A, liceijaa not switching was that non-black
ﬂy prey were not sufficiently abundant, or were not within suitable size

classes typically attacked by A, lyeeziae. However, available prey seemed to

 

 

 

Ta
rat

HT:

 

1 Me:
same
0.15).

149

Table 3. Mean mass gain and their corresponding instantaneous growth
rates (IGR) of Aeneneezia lyeeziae nymphs fed ﬁve Simeiiem-Baetie ratios

over 6 d (n = 4).

 

 

Simuliamﬁaetie ratio Mass gain :I: SEM (mg)1 IGR :I: SEM (ug/mg/d)

 

0:20 130.8 i 34.8 a
5:15 139.8 :I: 29.0 a
10:10 142.8 :I: 33.7 a
15:5 125.8 :I: 32.5 a
20:0 85.0 :I: 25.7 a

49.9 :I: 8.9
48.5 :I: 7.1
53.1 :I: 10.6
48.3 :I: 11.2
35.8 :I: 13.2

 

1 Means (ug body mass gain / mg initial body mass / day) followed by the
same letter are not signiﬁcantly different (alpha = 0.05, Power [= 1 - beta] =

0.15).

150

be well within the size range that predators were consuming before
disturbance.

An alternative explanation was that ELL affected predator feeding
behavior, causing them to eat less following exposure. This was rejected
since stoneﬂy predation was not affected up to 24 hr following B.t.i.
exposure (Wipﬂi 1992). It was also possible that predatory stoneﬂies were
not generalists, but specialists and 'preferred' or attacked certain prey
types, with predation related to encounter rates, handling time, and
capture success. Prey choice trials indicated that peer predators
consumed black ﬂy and mayﬂy prey at equal or near equal frequencies,
while perlodid stoneﬂies strongly selected black ﬂies over mayﬂies.
Observations made during these experiments indicated low capture to
encounter ratios with perlodid relative to peer predators on Baetia prey
(Wipﬂi, pers. obs.). This may, in part, explain different predation levels
among prey types, and between predator species. Perlodid stoneﬂies (L
aignaia) appeared highly aggregated and were almost exclusively collected
from tree branches suspended in the water column at the Medora River
ﬁeld site, microhabitats where black ﬂy larvae were most abundant and
alternate prey nearly non-existent. Perlid stoneﬂies were never collected
from these suspended substrates, but were collected exclusively ﬁ'om
substrates on the stream bed, areas where black ﬂy larvae occurred at more
even ratios with other prey taxa.

Prey choice by the odonate predator differed markedly from the

stoneﬂies. Belezia yinesa attacked the more mobile Baetie prey, taking

 

 

 

151

fewer attached Simulium prey. These predators probably encountered
more mobile than sedentary prey owing to the 'ambush' type feeding
behavior characteristic of Odonata (Merritt and Cummins 1984). During
these experiments, B, yineaa was frequently observed taking prey while
stationary. The stoneﬂies, however, actively searched during feeding
trials. Black ﬂies generally remained stationary after they located suitable
microhabitats within the artiﬁcial streams, and stayed stationary unless
disturbed by other insects. Thus, odonate-black ﬂy encounter rates were
low. Baetie nymphs frequently swam between microhabitats during
experiments, resulting in higher predator-prey encounter rates. The
feeding response exhibited by B. m suggested that this predator may
actively switch (Price 1984) to alternate prey, when black ﬂies become less
abundant. This switching should result in this predator being less
sensitive to a B.t.i. disturbance, relative to the predatory stoneﬂies.

Some macroinvertebrate predators were more sensitive than others to
prey population level disturbances. Perlodid and odonate nymphs relied
more on certain prey taxa such as black ﬂies and mayﬂies, regardless of
overall prey availability, suggesting some degree of specialization. Perlids,
however, showed no biased selection towards speciﬁc prey taxa, reﬂecting
generalist feeding habits. Specialists like L dieala and L eignaia should be
most affected by prey loss through a trophic level disturbance that
speciﬁcally reduces the prey populations on which these predators rely.
However, selective predators such as B, yinesa, relied on prey taxa that
were not affected by the imposing disturbance. Generalist predators should
fall between the above extremes, having the ability to switch to alternate

152

prey, provided alternate prey are sufficiently abundant. This was true for
2, media in the stream disturbance experiments, and for A, lyeeriaa in the
prey choice experiments.

Aezeneezia lxeeziaa nymphs grew equally well over a complete range of
Simulium-Mejia diet compositions, suggesting that predators such as
these generalist will not be affected by removing selected prey taxa, if there
are sufﬁcient alternate prey. However, this experiment had a low power
(i.e., low probability of ﬁnding difference between treatments). There was
an interesting trend which led me to form an alternative hypothesis on prey
suitability: higher growth rates would be achieved on diets comprised of
near even compositions of prey taxa. A higher growth rate trend with
increasing prey taxa evenness (i.e., 8:8 versus 0:16, 4:12, 12:4, and 16:0 prey
taxa A to prey taxa B) was recorded for A, lyeeziaa over 6 d. This hypothesis
needs further testing.

The disturbance of removing black ﬂies quantitatively and qualitatively
changed stoneﬂy diets. Had these predators not been food limited, reducing
prey abundance would not have affected predation levels. Even though
predators may not be temporally food limited, a drastic reduction in prey
abundance (e.g., 92% during these disturbance experiments) could reduce
prey populations below that necessary to maintain predator populations at
their pre-existing levels. Alternatively, bottlenecks (Neill 1988) in the
system (i.e., low prey abundance in summer when many prey taxa have
emerged) may cause predators to be temporally food limited. Prey may be

excessively abundant at other times of the year, being under-utilized by the

153
predators.
Predation by 2, media on alternate prey species (comprised mostly of
Hydropsychidae caddisﬂies) intensiﬁed following ELL disturbance. This

 

increased predation on less 'preferred' prey could in turn lead to a
restructuring of the prey community by these macroinvertebrate predators,
especially if such a disturbance is repeated frequently enough to prevent
prey recolonization. Alternatively, the niche vacated by black ﬂies, .‘
following ELL application, may be ﬁlled by other species. Removing black I
ﬂies should provide additional food and space and should allow other ﬁlter
feeders (e.g., hydropsychid caddisﬂies, midges) to colonize habitats
previously occupied by black ﬂies.

ELL applications to aquatic communities where black ﬂies dominate
may differentially affect specialsit and generalist predators. Results of this
study indicated that specialist predators may be most affected by prey-level
disturbances (e.g., ELL applications), through loss of 'preferred' prey.
Generalist predators may be least affected by prey disturbances if alternate
prey are abundant. Black ﬂy larvae were most abundant in these streams
in spring and summer, and were not an abundant food resource
throughout the year. Thus, ELL may simply remove black ﬂies from
stream systems earlier than natural adult emergence would, resulting in
minimal impacts on predatory species. Nonetheless, long term studies,
associated with ELL use, are needed to understand the consquences to
black ﬂy predators.

 

 

Ad

ide

ex;
FI'E

 

154

ACKNOWLEDGMENTS

I thank G. Amtsen for offering access to his property allowing me to
conduct this study, and R. Edens for tireless ﬁeld assistance. Thanks to P.
Adler for identifying black ﬂies, W. Hilsenhoﬂ' for dragonﬂy and stoneﬂy
identiﬁcations, and W. McCafferty for mayﬂy identiﬁcations. I also thank
B. Peckarsky, J. Ciborowski, and R. Wotton for their suggestions on
experimental design, A. Tessier for help with statistical analyses, and S.
Fradkin, D. Herms, B. Sobczak, and E. Walker for helpful comments on
text content. This study was supported, in part, by Sport Fishing Institute
Fund grant SFRP-90-26 awarded to RWM and MSW, and Northeast
Regional Black Fly Project NE-118.

155

LITERATURE CITED

Abrams, RA. 1987. Indirect interactions between species that share a
predator: varieties of indirect effects, pp. 38-54, in WC. Kerfoot & A. Sih
[eds.], Predation: direct and indirect impacts on aquatic communities.

University Press of New England, Hanover, New Hampshire, U.S.A. 386
PP-

COhen, I. 1988. Statistical power analysis for the behavioral sciences. Second

edition. L. Erlbaum Assoc.

Dejoux, C. 8: J.M. Elouard. 1990. Potential impact of microbial insecticides on
the freshwater environment, with special reference to the
WHO/ UNDP/ World Bank, Onchocerciasis control programme, pp. 65-83,
In M. Laird, L. A. Lacey and E. W. Davidson, [eds.], Safety of microbial
insecticides. CRC press, Inc., Boca Raton, FL. 259 pp.

Dejoux, C., F.M. Gibon, & L. Yaméogo. 1985. Toxicité pour la faune non-

cible de quelques insecticides nouveaux utilises en milieu aquatique

tmpical. IV. Le Bacillus thnzinsiensis var. israelensis H-14. Rev.
Hydrobiol. Trop. 18: 31-49.

De Moor, F.C., & M. Car. 1986. A field evaluation of Baeillug tiIiergieasia

 

var. israelensis as a biological control agent for Simulium chutteri (Diptera:

156

Nematocera) in the Middle Orange River. Onderstepoort J. Vet. Res. 53:43-
50.

Gibbs, K.E., F.C. Brautigam, CS. Stubbs, 6: L.M. Zibilske. 1986. Experimental
applications of M ghuringieneie var. M5 for larval black ﬂy
control: persistence and downstream carry, efficacy, impact on non-target
invertebrates and fish feeding. Maine Life Sci. Agric. Exp. Stn. Tech. Bull.
123: 1-25.

Holt, RD., 8: HP. Kotler. 1987. Short-term apparent competition. Amer. Nat.
130: 412430. *

Hurlbert, S. H. 1984. Pseudoreplication and the design of ecological ﬁeld

experiments. Ecological Monographs. 54: 187-211.

Lacey, L. A., & M. S. Mulla. 1990. Safety of Baseline ﬂmzingieneie var.
maeleneie and Eaeillna enhaezieua to non-target organisms in the
aquatic environment, pp. 169- 188. In M. Laird, L. A. Lacey and E. W.
Davidson, [eds.], Safety of microbial insecticides. CRC press, Inc., Boca
Raton, FL. 259 pp.

Lacey, L.A., 8: AH. Undeen. 1987. The biological control potential of
pathogens and parasites of black ﬂies, I_n_ K.C. Kim and R.W. Merritt, [eds.],
Black ﬂies: ecology, population management, and annotated World list.

Penn. State Univ. Press, University Park, PA.

157

Merritt, R.W., 8: KW. Cummins (eds.). 1984. An Introduction to the Aquatic
Insects Of North America (2nd ed.), Kendall/ Hunt Publishing Co.,
Dubuque, IA. 710 pp.

Merritt, R. W., E. D. Walker, M. A. Wilzbach, K. W. Cummins, 8: W. T.
Morgan. 1989. A broad evaluation of ELL for black ﬂy (Diptera:
Simuliidae) control in a Michigan river: efﬁcacy, carry, and non-target
effects on invertebrates and ﬁsh. J. Amer. Mosq. Control Assoc. 5 : 397-
415.

Miller, T.E., & W.C. Kerfoot. 1987. Redeﬁning indirect effects, pp. 33-37, I_n_
W.C. Kerfoot 8: A. Sih [eds.], Predation: direct and indirect impacts on
aquatic communities. University Press of New England, Hanover, New

Hampshire, U.S.A. 386 pp.

Molloy, D., 8: H. Jamnback. 1981. Field evaluation of Bag‘llus thuringieeeie'
var. israelensis as a black ﬂy biocontrol agent and its effect on non-target

stream insects. J. Econ. Entomol. 74: 314-318.

Neill, W.B. 1988. Community responses to experimental nutrient
perturbations in oligotrophic lakes: the importance of bottlenecks in size-
structured populations, pp. 236-255. Ia B. Ebenman and L. Persson (eds.),

Size structured populations. Springer-Verlag, Berlin.

158

Poff, N.L. 1992. Why disturbances can be predictable: a perspective on the
disturbance in streams. J. N. Am. Benthol. Soc. 11: 86-92.

Price, P.W. 1984. Insect ecology. Second edition. John Wiley 8: Sons, New
York, NY.

Resh, V.H., A.V. Brown, A.P. Covich, M.E. Gurtz, H.W. Li, G.W.
Minshall, S.R. Reice, A.L. Sheldon, J .B. Wallace, & R.C. Wissmar. 1988.
The role of disturbance in stream ecology. J. N. Am. Benthol. Soc. 7: 433-
455.

SAS Institute Inc. 1985. SAS user's guide: statistics. Version 5 edition.
SAS Institute, Cary, North Carolina, U.S.A.

Schmitt, R.J. 1987. Indirect interactions between prey: apparent
competition, predator aggregation, and habitat segregation. Ecology 68:
1887-1897.

Sousa, WP. 1984. The role of disturbance in natural communities. Ann.
Rev. Ecol. Syst. 15: 353-391.

Stewart, K.W., & B.P. Stark. 1988. Nymphs of North American stoneﬂy
Genera (Plecoptera). The Thomas Say Foundation, Entomological Soc.
Amer. U.S.A. 460 pp.

159

Wallace, J .B. 1990. Recovery of lotic macroinvertebrate communities from
disturbance. Environ. Management. 14: 605-620.

White, RS, 8: S.T.A. Pickett. 1985. Natural disturbance and patch
dynamics: an introduction, In The ecology of natural disturbance and
patch dynamics, S.T.A. Pickett 8: RS. White [eds.]. Academic Press, New
York, N.Y.

Wipﬂi, M.S. 1992. Direct and indirect effects of Bacillus thuringiensis var.
i5raelensis on non-target aquatic insects and trout. Ph.D. dissertation.
Michigan State University, East Lansing, MI, U.S.A.

RECOMMENDATIONS

These studies indicated that, when used at labeled rates, ELL does
not pose a direct or indirect toxic threat to non-target stream insects and
trout. Some non-target aquatic insects (Ephemeroptera and Diptera) were
sensitive to ELL at unusually high dosages, but were not aﬁ‘ected when
ELL was applied at recommended ﬁeld rates. This research indicated that
when using ELL, care must be taken to accurately calibrate stream
discharge, in order to apply the correct dosage of ELL, avoiding an
overdose.

The main impact of using ELL was on predatory species, through
the loss of black ﬂy biomass. In systems where black ﬂies are the dominant
food-web component, species such as predatory insects and ﬁsh, if highly
dependent on black ﬂies for food, will probably suffer from a decreased food
resource. If black ﬂies are the dominant species, and remaining animal
species that are dependent on black ﬂy larvae are not common, the impact
on the stream community will probably be minimal.

In summary, the impact of using ELL in a given stream appears to
be system speciﬁc, dependent upon community structure and secondary
productivity. The impact of greatest concern will be the loss of black ﬂy

biomass from the stream community food-web.

160