The Potential for Biological Control of Eurasian
Watermilfoil (Myriophyllum spicatum): Results of the
Research Programs Conducted in 1991.
Year 2
Interim Progress Report
April 1, 1992
by
Robert P. Creed Jr. and Sallle P. Sheldon
Department of Biology
Middlebury College
Middlebury, Vermont 05753 USA
Prepared for
Region 1
U.S. Environmental Protection Agency
Boston, Massachusetts
Warren Howard
Project Officer
-------
Preface
While the final version of this report was complied and
edited by Robert Creed and Sallie Sheldon, the research
presented here and some of the writing is due to the efforts
of several people. The research at Brownington Pond was
supervised by Robert Creed, who also wrote the Brownington
section. Kathy Newbrough was in charge of the yellow perch
diet study at Brownlngton Pond and wrote that section of the
report. The research at Middlebury and Lake Bomoseen was
supervised by Sallie Sheldon. Sallie Sheldon and Robert
Creed wrote the section on the results of harvesting on
invertebrate abundance. Kathy Newbrough wrote the section
on the fish enclusion experiment and associated laboratory
experiments. Kristin Henshaw wrote the section on weevil
and watermilfoil culture, and weevil lifehistory. Robert
Creed wrote the section on the weevil choice test conducted
at Middlebury. Linda OBryan wrote the multi-lake section
of the report. The research at Castleton State College was
directed by Anne Hampton. She was assisted by Michael
Alfieri and Brian Marsh. The results presented here are
excerpts from a longer report and were edited by Sallie
Sheldon. Robert Creed and Sallie Sheldon take full
responsibility for any mistakes made in the editing the
sections written by other researchers.
2
-------
Table of Contents
Introduction . . . . . . . . . . . . . . . . . . . .
Research at Brownington
Introduction
Study Site...........
Materials and Methods
Surveys . . . . . . . . . . .
Experiments
Results
Surveys . . . . . . . . . . .
Experiments
Discussion
...... .........
.5
. . . . 8
8
. . . . 8
9
I S 9
17
33
33
4 3
50
Research at Lake Bomoseen and Middlebury.
.61
Research at Lake Bomoseen..............................61
Introduction 61
The Effect of Fish on Herbivores
Introduction . . . . . . . . . .
Materials and Methods
Results ......... . .
Discussion . . . . .
The Effect of Mechanical Harvesting
Abundance . . . . . . . . . .
Introduction
Study Sites . . . . . . . .
Materials and Methods
Results
Discussion . . . . . .
..... S..
. I SI
Research at Middlebury 83
Weevil Life History Studies.
Introduction..............
Materials and Methods
Results . . .
Discussion . . . . . . .
Pond
. .
. . .
. .
.
I...
II S I
5
...................I
I
III
III.III.I.I
S.
I
ISSSSSSSIII
.
SI.
IS.
II.
I..
SI
SI
SI
SI
62
62
62
.69
.71
72
72
72
73
75
.79
on Herbivore
5551111 ISSI
S..SI..
Establishment and Maintenance of Weevil and
Watermilfoil Cultures
Introduction . . . . . . . . . . .
Materials and Methods . . . . . . . . .
Results
Discussion .
. . 83
. . . . 83
83
85
. . . 86
IIIIII.II.
I S
IIIII....III
II5...SII I
5S5I
I
S.SIeI
. . 87
87
87
. 88
. . 89
3
-------
The Effect of Weevils from Native Watermilfoil on
Eurasian Watermi 1 foil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1
I ntroduct ion 91
Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
Results 92
D i.scussi.on . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Multilake Survey 95
Introduction 95
Materials and Methods .... . . . ........ . . ......... . . .95
Results . . . . . . . 97
D i.scussion . . . . . . . . . . . . . . . . . . . . . . . . . . 98
Research at Castleton State College . 100
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
Synopsis of Research Results 100
Communication of Research Results ........................104
Suinntar r Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 106
I.. iterature Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
Tables and Figures . . . . . . . . . . 113
4
-------
INTRODUCTION
Eurasian watermilfoil ( Myriophyllum spicatum L.) was
accidentally introduced into North America sometime between
the late 1800s and the 1940s (Bayley et al. 1968, Reed
1977, Aiken et al. 1979, Couch and Nelson 1986). Since its
introduction it has spread over much of North America (Aiken
et al. 1979, Couch and Nelson 1986, Nichols and Shaw 1986,
Painter and McCabe 1988). It was first reported in Vermont
in 1962 in Lake Champlain (Holly Crosson, Vermont Agency of
Natural Resources (VtANR), pers. comm.). A number of
methods, many of them quite costly (Freshwater Foundation
1990), have been employed to control watermilfoil in Vermont
and elsewhere, including use of drawdowns, herbicides,
bottom barriers, and mechanical harvesting. In general,
while these control methods may result in short-term
reductions in watermilfoil abundance (Bayley et al. 1968,
Nichols and Cottam 1972, Aiken et al. 1979) they do not
appear to have proven satisfactory for long-term control of
this introduced, aquatic weed (Bayley et al. 1968, Spencer
and Lekic 1974, Aiken et al. 1979).
Recently, attention has focused on the potential for
biological control of Myriophyllum spicatum . Aquatic
herbivores such as the caterpillar Acentria nivea
( =Acentropus niveus)(Lepidoptera; Pyralidae) and the weevil
Euhrychiopsis lecontei (Coleoptera; Curculionidae), have
been found associated with declining populations of
5
-------
watermilfoil in northeastern North America (Painter and
McCabe 1988, Sheldon and Creed, pers. obs.), including
Brownington Pond, Vermont. However, the exact role played
by these herbivores in bringing about these declines remains
undetermined.
We are currently evaluating the potential for insect
herbivores to act as biological control agents for Eurasian
watermilfoll. There are six main objectives to this
research:
1) Determine the probable cause(s) of the Eurasian
watermilfoil decline In Brownlngton Pond (see Figures 1-3,
Creed and Sheldon 1991).
2) Examine the grazing/boring effects of all major
herbivores on Eurasian watermilfoil and native aquatic plant
species.
3) Determine the feasibility of herbivore introductions
into other milfoil-infested lakes in Vermont.
4) Determine if Lake Bomoseen is a suitable site for
herbivore introductions/collect pre-introduction base-line
data.
5) If determined to be feasible and appropriate based on
previous research (a high-likelihood of success and
relatively free from causing negative impacts to non-target
species), use herbivorous insects to control Eurasian
watermilfoil in Lake Bomoseen.
6) Develop a public education program to keep Vermonts
citizens abreast of the results of the research.
6
-------
In addition to conducting research relating to these six
objectives we have also conducted a qualitative (1990) and a
quantitative (1991) survey of watermilfoil and associated
herbivores in several lakes in Vermont, New Hampshire and
Massachusetts. This survey (the multi-lake, multi-state
survey) was funded by the Army Corps of Engineers but the
data are relevant to the six objectives listed above. The
research described in this document is from the 1991 field
season. This is the second progress report from this five
year study.
7
-------
RESEARCH AT BROWNINGTON POND
Introduction
One of the few declines in a watermilfoil population in
N. America occurred at Brownington Pond in northeastern
Vermont. While the cause of this decline has yet to be
determined, initial samples of the watermilfoil found three
insect herbivores associated with this pest macrophyte. The
goal of the Brownington Pond research is to determine the
cause of the watermilfoil decline and ascertain the role the
insect herbivores may have played in the decline. In 1990,
we Initiated research at Brownington Pond. We monitored the
abundance of watermilfoil and its associated invertebrates.
We also conducted field and laboratory experiments. In
1991, we continued to monitor the watermilfoil population
and the abundances of the associated herbivores. We also
conducted additional experiments which evaluated the effects
of various herbivores on watermilfoil in both lab and field
settings.
Study Site
Brownington Pond is a small, mesotrophic lake in
northeastern Vermont (Brownington and Derby Townships,
44°53N, 72°09W). Total surface area of the pond Is 64
hectares; maximum depth is 10.7 m with an average depth of
8
-------
5.5 m (Figure 1). There are two inlets, one on the north
shore and one on the east side, and a single outlet, Day
Brook. Less than one quarter of the shoreline has been
developed with summer camps, most of which are located along
the northeastern shore. There is a public boating access on
the west side of the pond.
Materials and Methods
Surveys
Pond Survey
The survey determining the distribution and abundance of
watermilfoil in Brownington Pond was less extensive than the
one conducted in 1990. During the first week of June, we
surveyed the pond by boat noting the location of
watermilfoil beds and also any changes in the distribution
and abundance of two of the more comon, large, native
macrophytes ( Potamopeton amplifolius, Heteranthera dubia) .
Over the course of the summer we kept track of any large-
scale changes in macrophyte distribution.
Water Chemistry
Three stations were established in the pond at which
weekly measurements of dissolved oxygen (DO) and temperature
9
-------
were made. One station was on the east side in an area with
mixed H. dubia and P. amplifolius cover. The other two
stations were located in the South and West watermilfoil
beds. At each station, DO readings were taken at 0.5 m
below the surface and at 0.5 m above the bottom.
Temperature was read from pairs of maximum/minimum
thermometers suspended from buoys. One thermometer was 0.5
m below the surface and the second was 0.5 m above the
bottom. Thermometers were reset after each weekly reading.
A survey of nutrients (nitrate, nitrite and
orthophosphate) in the water column was made on 25 June
1991. Samples were collected from the same three sites
described above for the temperature and DO readings.
However, instead of sampling a fixed point, three or more
locations were chosen to sample a broader array of potential
microhabitats within a site. Water samples were collected
using a Kemmerer sampler. Pairs of samples, one shallow and
one deep were taken at each point. Three pairs of samples
were taken on the east side, six were taken in the South Bed
and four were taken in the West Bed. We had initially
planned on taking nutrient samples on two dates but due to
the low values and lack of variation in the 25 June samples
(see results) we decided not to take the second set of
samples.
10
-------
Sediment Chemistry
Five sets of sediment samples (3 samples per site) were
taken in Brownington Pond. Samples were taken in: 1) the
West Bed, 2) a watermilfoil-free area south of the West Bed,
3) in the South Bed, 4) in a watermilfoil-free area south of
the South Bed and 5) on the east side of the pond in an area
dominated by Heteranthera dubla . Pond sediment was
collected by a SCUBA diver using a PVC tube (0.5 m long X
0.1 m diameter). The open tube was pushed through the
sediment, approximately 5-10 cm below the sediment surface.
When the tube was filled, PVC stoppers were inserted at each
end. The sealed tube was then taken to the surface and the
sediment was emptied into a 2 1 bucket and sealed. All
samples were kept cool, bubbled with nitrogen gas, sealed
and sent to the Army Corp of Engineers Waterways Experiment
Station (Vicksburg, Mississippi) for nutrient (N and P) and
organic matter analysis. Samples were sent within 48 hrs of
collection.
Plant Transects
Watermilf oil appears to be restricted to water between
2.0 - 3.5 m deep (Creed and Sheldon 1991). To see if this
distribution pattern persisted in 1991 we established three
permanent transects through both of the main beds. An
attempt was made to space the transects across the beds.
11
-------
Along each transect, locations were selected at half meter
depth intervals ranging from 0.5 m - 3.5 m deep, for a total
of twenty one sample points for each bed. At each sample
point two PVC pipe Ts were pushed into the sediment at
right angles to one another to form a cross. The four ends
of the Ts were numbered from one to four.
Transects were sampled on three dates during the growing
season. For each point to be sampled, one of the four
numbers from the Ts was selected at random from the
remaining possible numbers prior to sampling. The points
were sampled by SCUBA divers. The divers inserted a 2 m
long piece of PVC pipe into the appropriate numbered
opening. A 0.25 x 0.25 m quadrat was then placed on the
bottom at the end of the pipe and the sample taken. All
above sediment plant biomass was clipped and placed into a
numbered, plastic bag. Upon returning to the lab, plants
from each sample were sorted to species and dried in a
drying oven at 80° C. Plants were weighed after drying to a
constant weight. For clarity of data presentation, dry
weights for native species were lumped together in the
category Other.
Permanent Grids
In addition to determining the location of watermilfoil
beds in the littoral zone, we have initiated a program to
record finer scale expansions and contractions of M.
12
-------
spicatum beds using permanent grids. Four grids were
established in the pond in 1990, two In each bed. The grids
cover an area of 8 x 6 m with buoys placed every 2 m in a 4
x 5 array. Percent cover of watermilfoil was determined by
a diver using a 0.5 x 0.5 m quadrat subdivided into 25
subunits. Placement of the quadrat across the bed was
determined using a transect line made of PVC pipe with
openings placed every 0.5 m into which the quadrat was
inserted. Percent cover was evaluated along four transects
for each grid. The position of the transects corresponds to
the four lines of buoys that run along the longer dimension
of each grid, i.e. A - E (see Figure 6). The number of
quadrat subunits lying over watermilfoil plants was then
recorded. This technique generates percent cover values
ranging from 0-100%. For clarity of data presentation, we
grouped the percent cover values Into four categories - 1)
less than 25%, 2) 2550%, 3) 50-75% and 4) greater than 75%.
The grids set out in 1990 were placed on the ends and
nearshore edges of the beds as watermilfoil will be more
likely to spread laterally and Into shallow water. The
grids did not extend into deep water as watermilfoil
abundance is probably limited on the deep edge of beds by
light availability. The grids were swum in mid June (18 and
21 June), late July (24 - 26 July) and late August (26
August) during the 1991 growing season.
A new grid was established in Lake Memphremagog in a bed
of watermilfoil just north of the Whipple Bay boat access.
13
-------
No readings were taken at this grid during the 1991 growing
season.
Invertebrate Samples
To describe the watermilfoil invertebrate assemblage
quantitative samples of watermilfoil and the associated
invertebrates were taken in the South and West Beds. In
addition, samples of two abundant native macrophytes
( Potamopeton amplifolius and Heteranthera dubia ) were taken
to compare their invertebrate assemblages with those of
watermilfoil. Samples were collected using two sizes of the
Mobile Invertebrate Sampler (MIS) developed by Smith and
Sheldon (unpublished manuscript). The larger sampler (the
Super Sampler), used for both watermilf oil and the two
native macrophytes, samples an area of 0.18 m 2 ; the smaller
version (the Minisampler) was designed for sampling a single
stem of watermilfoil. Both samplers were employed by a
SCUBA diver. An area or a plant to be sampled was chosen
haphazardly. The sampler tube was then slid over the
plant(s) as the diver descended. Plants were cut near the
sediment surface, the opening of the sampler was then
covered with a 500 urn mesh sieve and then the sample was
returned to the surface. All samples were placed in
sealable, plastic bags. Super samples were preserved in 70%
ETOH; minisamples were picked soon after sampling while the
animals were still alive. Invertebrates were identified to
14
-------
the lowest taxonomic level. Dry weights were recorded for
the plants after the invertebrates were removed. Mini-
samples were taken weekly from 18 June to 27 August for a
total of 11 sample dates. Super samples were taken in June
(12, 13 and 14 June), on 3 July, on 5 August and 26
September. Only the data from the June and the August 1991
watermilfoil super samples will be presented. Also, we will
present the results of the 1990 super samples which were
collected in a similar fashion.
In 1990 we discovered that weevils lay their eggs on the
apical meristems of watermilfoil and that the early instar
larvae burrow into the meristem upon hatching (Creed and
Sheldon 1991). We initiated meristem transects across
both watermilf oil beds in 1990 to determine the density of
eggs and early instar larvae in the beds. In 1991 we
continued taking these stem transects but we sampled larger
pieces of stem (approximatley 50 cm long) in order to
collect late instar weevil larvae and pupae. Snorkelers
collected 16 stems along a transect, eight stems with intact
apical meristems and eight stems without apical meristems.
While it is possible to find all life stages on both stem
types (especially as weevils also lay their eggs on lateral
meristems), we believed that stems with intact meristems had
a greater probability of containing eggs and first instar
larvae. We believed that stems without apical meristems
were more likely to contain late instar larvae and pupae.
These two stem types were collected in pairs haphazardly by
15
-------
snorkelers swimming across the bed. Three such transects
were sampled for each bed on each sample date. Samples were
collected weekly for a total of Ii sample dates. Stems were
examined under dissecting microscopes and all lifestages of
weevils plus Acentria larvae were recorded for each stem.
As crayfish might influence the distribution and
abundance of M. spicatum (e.g., see Lodge and Lorman 1987),
we planned to assess crayfish distribution and abundance in
Brownington Pond. Our plan was to sample crayfish with
grids of traps and also to conduct nighttime SCUBA surveys
of established plots. However, due to the low densities of
crayfish encountered in our initial sampling attempts (both
with traps and SCUBA surveys) we discontinued these efforts.
What few crayfish we did collect were all identif led as
Orconectes virilis .
Fish Samples
Only five species of fish have been collected from
Brownington Pond (Unpubi. State Fisheries Survey 1980).
These include yellow perch ( Perca flavescens) , smalimouth
bass ( Micropterus dolomleui) , chain pickerel ( Esox niger),
white sucker ( Catostomus commersonhi ) and brown bullhead
( Ictalurus nebulosus) . The state survey data indicated that
the yellow perch is by far the most abundant species
numerically in this pond. Because of the abundance of
yellow perch and the fact that it is the species most likely
16
-------
to consume macrophyte associated invertebrates, our fish
survey focuses on this species.
Gill nets for large perch (>150 mm) were deployed in
Brownington Pond on three dates in 1991: 26 June, 3 August
and 27 September. A single net was deployed approximately
15 m past the western edge of the south watermi].foil bed,
perpendicular to shore, for all surveys. On 26 June a 6.4
cm (2.5) stretch mesh net was used; on 3 August and 27
September the net used had four panels with the following
mesh sizes: 3.8 cm (1.5), 5.1 cm (2), 7.6 cm (3) and 8.9
cm (3.5). The nets were deployed for approximately one
hour at dawn on all three dates. Captured fish were
measured (total length), weighed and scale samples were
removed for aging. The stomachs were then removed and
preserved in 10% forma].in. In the lab, stomach contents
were examined under a dissecting microscope and Identified
to the lowest possible taxonomic level. Scales were pressed
and used for determining the age of the fish.
Experiments
The Effect of Acentria on Watermllfoil Growth
Several small watermilfoil plants were collected from
Brownington Pond. Plants were first checked for herbivore
damage. Damaged plants (e.g., with missing meristems,
meristem damage, or significant stem damage) were rejected.
17
-------
We selected eighteen of the intact plants which were the
most similar in size. All obvious invertebrates and weevil
eggs were removed from these plants. These eighteen plants
were then weighed (blotted wet weight). We tied a marker
around the stem at the base of the plant and the length of
the stem from the marker to the tip of the apical meristem
was determined. We also counted the number of whorls on
each stem above the marker. The initial lengths of the
watermilfoil plants ranged from 160-203 mm; initial weights
ranged from 0.33-0.75 g. Much of the variation in weight
was attributable to differences in root biomass and not
above ground biomass (Creed, pers. obs.)
After processing, each watermilfoil plant was planted in
a numbered chamber. The chambers consisted of clear plastic
tubes (42 mm inside diameter) set in a PVC pipe base. We
first placed aquarium gravel in the bases to weight them
down. We then filled the remainder of each base with
strained pond sediments taken from one of the watermilfoil
beds in Brownington Pond. A tight-fitting cap covered with
500 micron, Nitex mesh was then placed on the top of the
tube. These are the same type of chambers described in
Creed and Sheldon (1991). Plants were planted in the
sediment up to the tag on the stem. The chambers were then
placed in a large wading pool (Wading Pool 1) set out of
doors in an unshaded area. The chambers were aerated with a
slow trickle of air bubbles to prevent stagnation. Plants
18
-------
were allowed to acclimate to the chambers for one day before
the Acentria larvae were added.
The experimental design was a randomized complete block
design with three treatments per row and six replicates per
treatment. The plants were randomly assigned to rows in the
wading pool. The determination of treatment within rows was
also determined using a random number table. The treatments
were 0 (control), 2 or 4 Acentrla larvae per tube. Two or
four Acentria larvae were then added to the appropriate
chamber in each row. The Acentria larvae came from a single
batch of eggs that we had collected from Brownington Pond.
The larvae had hatched and had been feeding on watermilfoil
in an aquarium for about two weeks. All larvae used in the
experiment were very similar in their initial size (mean
length ± 1 S.E. was 2.8 ± 0.13 mm, based on extra larvae not
used in the experiment). Water temperature in the pool was
monitored using a max/mm thermometer during the experiment.
Water temperatures ranged from 16.1_32.20 C during the
experiment (mean minimum temperature was 18.80 C; mean
maximum temperature was 25.80 C).
The experiment lasted for 22 days. Plants and Acentria
larvae were then removed from each chamber. After removing
the Acentria , the watermilfoll plants were measured (length
from tag to tip of rooted stem) and weighed (blotted wet
weight). Any plant material not attached to the rooted stem
was not included in the final plant weight. We al5o counted
the number of whorls of leaves remaining on each stem.
19
-------
Treatment effects were compared using an ANOVA with planned,
orthogonal contrasts (Sokal and Rolhf 1981). The recovered
Acentria larvae were preserved in 70% ETOH.
The Effect of E. lecontei Larvae on Watermilfoil Growth
The design of this experiment, the collection of plants
and the statistical analyses were the same as that discussed
above for the Acentria experiment. The experiment was
conducted in a wading pool (Wading Pool 2) adjacent to the
one described above. The initial length of the plants
ranged from 158 to 223 mm; initial weights ranged from 0.33
to 0.74 g. Late instar larvae (approximately 3-4 mm long)
were collected from watermilfoil plants in Brownington Pond
the day the experiment was initiated. Treatments consisted
of a control (0 larvae), 1 and 2 late instar larvae per
plant. The experiment lasted 9 days. Water temperatures
during the experiment ranged from 110 C to 270 C (mean
minimum temperature was 15.60 C; mean maximum temperature
was 22.5° C). We quantified change in plant length and
weight in this experiment. We also measured the amount of
stem that had been burrowed by the larvae. As weevil larvae
do not appear to feed extensively on leaves, we did not
quantify changes in number of leaves. Weevil larvae did not
feed on the plants in one replicate from both the one and
two larvae treatments so n = 5 for these two treatments; n =
6 for the control.
20
-------
The Effect of Adult E. lecontel on M. sibiricum Growth
The design of this experiment, the collection of plants
and the statistical analyses were also the same as that
discussed above f or the Acentria experiment. The .
sibiricum plants were collected from Beebe Pond, Hubbarton,
Vt, on 30 July. The initial length of the plants ranged
from 168 to 217 mm; initial weights ranged from 0.57 to 1.97
g. The adult weevils (both males and females) were
collected from watermilfoil plants in Brownington Pond on 31
July, the day before the experiment was initiated.
Treatments consisted of 0, 2 and 4 adult weevils per plant.
The experiment was run in Wading Pool 1. The experiment
lasted 13 days. Water temperatures during the experiment
ranged from 15.60 C to 26.70 C (mean minimum temperature was
17.00 C; mean maximum temperature was 22.90 C). We
quantified change in plant length and weight in this
experiment. We also recorded the number and position of all
leaves lost from the plants during the experiment and the
distribution of stem bites. We also recorded the number of
weevil eggs and larvae found on the plants.
The Effect of Herbivores on Watermilfoil Bouyancy
In late July we noticed that the tops of watermilfoil
plants in the west bed at Brownington Pond had fallen over.
The top of the bed was now approximately one meter below the
21
-------
surface of the pond. Inspection of the plants showed that
many of them had been hollowed out by weevil larvae. Many
of the collapsed plants showed little or no weevil damage,
however, and appeared to have been pulled down by nearby,
damaged plants. This observation suggested that weevils
could have an additional, negative effect on watermilfoil by
reducing the buoyancy of the plants. The following
experiment was designed to quantify this effect.
We collected adult weevils and undamaged apical portions
of watermilfoil stems on 29 July 1991. The length of the
stems was standardized and then they were sorted into ten
groups of six stems each. We then determined the blotted
wet weight for each of the groups. The mean wet weight (± 1
S.E.) of the groups of stems was 5.08 ± 0.15 g. The weevils
were sexed using a dissecting scope and then sorted into
five groups of four weevils (each with 3 females and 1
male). Watermilfoil stems were then placed into ten, 38
liter aquaria that had previously been filled with well
water on 30 July. The aquaria were placed in a line on the
ground in an unshaded area on the east side of our research
building at Brownington Pond. The aquaria did not receive
late afternoon sun. Each aquarium was aerated with a single
airstone. Weevils were added to five of the ten aquaria,
the remaining five aquaria served as controls. Assignment
of treatments to aquaria and watermilfoil and weevils to
aquaria was determined using a random number table. Water
temperature was monitored using floating thermometers in
22
-------
three of the control aquaria. Temperatures were recorded in
the morning and evening. Temperatures ranged from 160 - 29
C in the aquaria during the experiment (mean morning
temperature was 18.70 C; mean evening temperature was 23.80
C). All aquaria were covered with tight-fitting lids to
prevent the escape of the weevils. The lid consisted of a
wooden frame covered with translucent plastic to allow
transmission of light. One section of the plastic was
removed and replaced with a piece of 500 micron mesh Nitex
that was sealed in place to allow for air exchange and also
aid in temperature regulation of the aquaria.
The experiment ran for 21 days. At the end of the
experiment all watermilfoil that was not resting on the
bottom was considered as floating. We should note that all
watermilfoil settled to the bottom on cool, overcast days.
We ended the experiment on a sunny day. Floating
watermilfoil was separated from that which had settled to
the bottom of the aquaria and placed in separate, labeled
plastic bags. All herbivores were removed from the aquaria
and from the plant material. While this experiment had
initially been designed to examine the effects of weevils on
buoyancy we had contamination of four of the weevil tanks
with Acentria larvae. Therefore we will refer to the effect
on buoyancy as an herbivore effect and not simply a weevil
effect. All watermilf oil was then weighed (blotted wet
weight). Most of the watermilfol]. in one of the control
aquaria had settled to the bottom. The watermilfoil in this
23
-------
aquarium was encrusted with what appeared to be iron
precipitation. Using Dixons test (Sokal and Rohif 1981) we
determined that this replicate was a statistical outlier and
removed it from the analysis. Treatment effects were
compared using an ANOVA (Sokal and Rohif 1981). Separate
ANOVAS were performed on the weight of floating and sunk
plant material. Weight data were log transformed for the
ANOVAs.
Fish Exclusion Experiment
Predation by insectivorous fishes may influence either
the establishment of a watermilfoil herbivore population in
a lake or their distribution and abundance in the system.
The effect of fish may be through direct predation.
Alternatively, fish may indirectly influence the
distribution and abundance of herbivores through their
influence on herbivore predators and/or competitors.
To assess the impact of fish on herbivores we conducted
a fish exclusion experiment in Browningtori Pond. There were
three phases of the experiment. The first set of cages
(Phase 1) was placed in the pond in the summer of 1990.
Half of those cages (18) were sampled at the end of the 1990
growing season. The remaining eighteen cages (Phase 2) were
left in place through the winter, spring and summer. We had
hypothesized that by having cages present for an additional
year might influence invertebrate abundance, i.e., any
24
-------
potential, negative fish effect that might occur early in
the second season might be prevented. This second set of
cages was sampled in 1991. Finally, a third set of cages
(Phase 3) were placed in the pond at the beginning of the
1991 growing season. The same treatments used in Phases 1
and 2 were used in Phase 3. These cages were also sampled
at the end of the 1991 growing season.
The experimental design for all phases of the experiment
includes three treatments; a complete exclusion cage, a cage
control and an uncaged control. Complete cages and cage
controls were constructed by making cylinders of 0.6 cm mesh
that were open on one end. Cylinder ends were held open by
wire rings. Four cork floats were attached to the top of
each cage to suspend them in the water. Cage controls
differed from cages only in that large slits were cut in the
sides of the cylinder to permit access to fish. Open
controls were simply areas of the milfoil bed demarcated by
a single buoy. Placement of cages and cage controls
involved sliding the cylinder over the watermilfoil. Cages
were held in place by both pinning the lower ring into the
sediment and placing bricks on top of the ring (four pins
and bricks per cage).
Twelve rows containing each of the three treatments
(Phases 1 and 2) were set out in the south watermilfoil bed
on 20 June 1990. Four watermilfoil samples were taken with
the large MIS on 27 June to serve as initial samples of
invertebrate abundance in the bed. Cages were checked once
25
-------
a week over the course of the summer and cleaned with a
brush approximately every third week. Six rows (Phase 1)
were sampled on 5 September 1990; the remaining six rows
(Phase 2) were sampled on 28 and 29 August 1991. Phase 3
cages were placed in the pond on 12 June 1991 and were
sampled on 27 August 1991.
All three treatments for all Phases were sampled using
the large MIS sampler. For cages and cage controls this
entailed skin divers removing the top of the cage.
Immediately upon removal of the cage top a SCUBA diver
descended into the cage with the MIS sampler. Upon reaching
the bottom all plants were clipped at the sediment surface,
the sieve was placed over the bottom of the sampler and the
sampler was returned to the surface. Samples were removed
from the sampler and placed into labeled, sealable plastic
bags. Phase 1 samples were preserved in 70% ETOH. Phase 2
and 3 samples were placed in a sieve stack (3 sieves with 8
mm, 2 mm and 0.425 mm openings for the top, middle and
bottom sieves, respectively) and sprayed with a jet of water
to separate the invertebrates from the larger plant pieces.
Each sample fraction was preserved in 70% ETOH.
Invertebrates were separated from macrophytes in the
laboratory and identified to the lowest feasible taxonomic
level. The macrophytes were then dried and weighed.
Invertebrate abundance was standardized by watermilfoil
biomass for statistical analysis. The data were analyzed
using an ANOVA with planned, orthogonal contrasts which
26
-------
compared 1) the full cage to the cage control and no cage to
determine a fish effect and 2) the cage control and no cage
treatments to determine if there was a cage effect. As
weevils were more abundant in 1991 and thus more likely to
show a response to fish predation, only the results of Phase
3 are presented.
Herbivore Enclosure Experiment in the Pond
The results from the wading pool experiments in 1990
demonstrated that adult and larval E. lecontei can have
strong negative effects on ! . spicatum . However, we still
need to determine if this weevil species can have a negative
effect on watermilfoil growth when feeding on larger plants
in the lake. To address this question, we conducted a large
enclosure experiment in Brownington Pond. The enclosures
were three meter tall plexiglass tubes (20.5 cm or 8 in
O.D.) which were composed of two parts. The bottom section
(1 m tall) was driven into the sediment. The upper portion
of the chamber (2 m tall) was then bolted to the bottom
section. Along the sides of the upper portion were four
pairs of ports covered with 500 um mesh which allowed for
water exchange between the enclosures and the water column.
A lid also covered with 500 urn mesh was then placed on top
of each tube.
The bottom sections of the enclosures were placed in the
pond along the nearshore edge of the South Bed by a SCUBA
27
-------
diver on 16 July. Due to the depth of the water, the tops
of the enclosure bases were not flush with the sediment
surface. Thus extra sediment which was free of other plants
was added to each base. We collected a number of small (25-
35 cm above sediment length) watermilfoll plants from the
South Bed on 16 July. The plants were cleaned of obvious
invertebrates and any weevil eggs. The plants were then
sorted into groups of six and weighed (blotted wet weight).
Six plants were to be placed in each enclosure which is
equivalent to 181 plants/rn 2 . This value is well within the
range of densities determined by surveys of watermilfoil in
the two beds during 1990. The initial mean wet weight (± 1
S.E.) of plants placed in the tubes was 4.5 (± 0.26) g.
On 17 July the plants were planted in the tube bottoms
by a SCUBA diver. Plants were gently pushed down into the
sediments until the roots were buried. The upper portion of
the tube was then bolted to the bottom. To prevent
epiphytic algae from from becoming too abundant on the
watermilfoil and the enclosure wall we added approximately
75 snails (mostly Amnicola , the dominant snail in
Brownington Pond) to each enclosure. The lids were then
bolted onto each enclosure top. Three PVC pipes which were
roped together were stuck into the sediment around each
enclosure to prevent them from swaying too much in the wind
and waves. Damaged plants were replaced In two of the
enclosures prior to weevil addition. All plants in
enclosure 10 were replaced on 23 July after seeing an adult
28
-------
. lecontel inside the enclosure. After allowing the plants
to grow for 14 days (31 July) we added four weevils (2 males
and 2 females) to half of the enclosures. Weevils were
allowed to grasp onto a watermilfoil leaf and then the leaf
was dropped into the enclosure. Assignment of enclosures to
treatments was determined using a random number table.
Enclosures were cleaned of external periphyton once a week.
Enclosures were sampled on 28 August. First, the upper
portion of the enclosure was removed from the base. Then
all plants and plant fragments remaining in the enclosure
were removed by a SCUBA diver and placed in a labeled,
sealable plastic bag. No attempt was made to recover
weevils from the weevil treatment enclosures. Plant
material was transported to the laboratory where all plants
were divided into shoots and roots and then weighed (blotted
wet weight). Shoots and roots were dried to a constant
weight at 800 C. Treatment effects were analyzed using an
ANOVA with shoot, root and total plant weight being the
response variable.
The Effect of Crayfish on Watermilfoil Distribution
Eurasian watermilfoil exhibits an interesting
distribution pattern in Brownington Pond. The dense
watermilfoil beds are located in the deeper (2.0 - 3.5 m)
regions of the littoral zone (see Creed and Sheldon 1991).
This pattern has been apparent since 1986. This absence of
29
-------
watermilfoil from shallow water does not concur with what is
known about the competitive ability of this species.
Normally, upon introduction into a body of water,
watermilfoil expands to occupy much of the littoral zone up
to the depth at which light is limiting (Carpenter 1980).
There are two main factors which might account for this
pattern. These include differences in the texture and/or
nutrient concentration of nearshore vs deep water sediments.
The alternative is that an herbivore which is more abundant
in shallow water might be excluding watermilfoil from this
habitat. While conducting watermilfoil surveys we noticed
that Brownington Pond supports a population of the crayfish
Orconectes virilis . Crayfish burrows were most commonly
seen in shallow water where hard substrata needed for refuge
were most abundant. Recent research has shown that crayfish
can have a strong negative effect on the abundance of
vegetation in lakes (e.g., Abrahamsson 1966, Flint and
Goldman 1975, Lorman and Magnuson 1978, Lodge and Lorman
1987). This experiment was designed to determine if
sediment differences and/or crayfish herbivory were limiting
the abundance of watermilfoil In the shallows.
The design of this experiment Involved watermilfoil
being planted in pots containing the two sediment types
(nearshore and deep water sediment) with pots of both
sediment types either being exposed to crayfish or not. Six
plastic flower pots were bolted to each of five racks.
Strained sediment from one of the watermilfoil beds was
30
-------
placed into half of the pots; strained nearshore sediment
was placed into the other three pots. Nearshore sediments
tended to be coarser than the watermilfoil sediments.
We then collected 30 similar-sized watermilfoil plants
from the pond. Plants were cleaned of any obvious
invertebrates and . lecontel eggs. The plants were then
measured (from the tip of the apical meristem to a tag
attached to the base of the stem) and weighed (blotted wet
weight). Plants were then grouped by length so that
similar-sized plants would be placed on the same rack.
Plants were then randomly assigned to treatments within a
rack.
The racks were set out in the pond on 23 July. The
racks were placed in a line along the south shore of the
pond in approximately one meter of water. A SCUBA diver
then planted all of the watermllf oil stems Into the
appropriate pots on each rack. Then two pots (one of each
sediment type) were enclosed in a cylindrical, plastic mesh
cage, two (again, one of each sediment type) were enclosed
in cage controls (cages with holes cut in the sides at pot
level), and the remaining two pots were left uncovered to
serve as controls. Cages were checked periodically to
ensure that crayfish had not managed to enter the exclusion
cages. The racks were sampled on 15 August. A diver
removed the cages and then placed any remaining plant
material in a labeled, plastic bag. Plants were returned to
the laboratory and measured and weighed. Treatment effects
31
-------
were to be compared using ANOVA with planned, orthogonal
contrasts.
. lecontei Oviposition Experiment
Late in the summer, some E. lecontei were found on .
sibiricum collected at Lake Iroquois. We were unclear
whether these weevils were already present on . sibiricum
when the . s icatum was introduced into the lake or if they
came in on the introduced M. spicatum . Regardless of their
origin, we were curious to see if weevils collected from M.
sibiricum would display a preference for the native
watermilfoil when presented with both species.
The experiment was conducted in wading pool 1 (see
above). We placed stems of both watermilfoil species into
3.875 1 plastic bags filled with pool water. Each stem had
a single apical meristem. The stems were checked to make
sure that the meristem was intact. Any weevil eggs present
on the meristem were removed. Six of the bags received a
pair of weevils (one male, one female) collected from M.
s icatum in Brownington Pond. The remaining six bags
contained a pair of weevils collected from ! . sibiricum .
Stems were checked either once a day or every other day for
the number of eggs on the apical meristem. Whenever eggs
were found in a bag both watermilfoil species were replaced
with new stems. The experiment ran for six days and the
meristems were checked for eggs four times.
32
-------
Results
Surveys
Pond Survey
Eurasian watermilfoil appears to have increased slightly
in abundance since the survey of 1990 (Figure 1). The main
areas of expansion are 1) the small bed which is northeast
of the South Bed and 2) the three small clumps along the
2.0 - 3.0 depth interval between the South Bed and the West
Bed. While scattered stems of watermilfoil were observed
throughout the pond no other large clumps were found. P.
amplifolius distribution and abundance generally appeared
similar to that observed in 1990. The only change appeared
to be an increase in abundance in the southwestern bay by
the outlet. The abundance of H. dubia also appeared
unchanged. This species is most abundant in the deeper
water (1.5 - 3.0 m) along the east side of the pond.
Water Chemistry
Dissolved oxygen concentrations ranged from 8.0 to 10.0
mg/i during the summer (Figure 2A&B). For the temperatures
recorded on these dates the pond water had oxygen
saturations of approximately 100%. There was little
difference in dissolved oxygen concentration between the
33
-------
surface and the bottom of the pond. Dissolved oxygen
concentrations had increased to a little over 16.0 mg/i by
late October. Surface and bottom water temperatures also
remained fairly constant during the summer (Figure 3A&B).
Temperatures ranged from 180 - 270 C.
Very little variation was found in orthophosphate
concentration in Brownington Pond (Table 1). Concentration
of orthophosphate was slightly higher in the native plant
beds on the east side of the pond with shallow water having
a slightly higher concentration. This may due to the fact
that this site is near the main inlet stream. No variation
in orthophosphate concentration was observed in either of
the watermilfoil beds. The concentration of nitrite and
nitrate did not vary at all around the pond. Due to this
lack of variation in nutrient concentration we decided not
to continue the water chemistry survey.
Sediment Chemistry
Unfortunately, our sampling technique did not produce
samples that could be used for sediment analysis by the Army
Corps of Engineers laboratory at the Waterways Experiment
Station. Apparently, we need to be able to remove more
water from the samples. We will attempt to get better
samples in the following year.
34
-------
Plant Transects
Eurasian watermilfoil dislayed a similar pattern of
distribution to that observed in the pond in 1990. Eurasian
watermilfoil was abundant in the 2.0-3.0 m depth interval
(Figures 4 and 5). Occasionally, plants were found at a
depth of 3.5 m. No significant watermilf oil biomass was
sampled in water less than 2.0 m deep. Watermilfoil biomass
increased over the course of the summer (West Bed: from a
maximum of approximately 32 g/m 2 to 96 g/m 2 ; South Bed: from
68 g/m 2 to 90 g/m 2 . These values are comparable to those
observed on the transects collected in Brownington Pond in
1990 (Creed and Sheldon 1990).
Eurasian watermilfoil clearly dominated the 2.0-3.0 m
depth interval. Native macrophytes, which were abundant in
the shallower water (and occasionally in the deeper water),
were extremely uncommon within the watermifoil beds (Figures
4 and 5). This pattern was also similar to that observed in
1990. Native macrophyte taxa commonly collected in the
shallows included P. amplifolius , P. praminius, Heteranthera
dubla, Chara sp., Isoetes sp., Saaittaria sp. and Nalas
flexilis. Nitella sp. was the macrophyte commonly found on
the deep side of the watermilfoil beds.
35
-------
Permanent Grids
Substantial increases in watermilfoil abundance were
observed on one of the permanent grids and modest expansion
was observed on the three other grids (Figures 6 - 9). The
grid where the greatest expansion was observed was the East
Grid on the South Bed (Figure 9). Only 25% of the grid area
displayed heavy (>75% cover) cover at the end of the 1990
growing season. Expansion at this bed was already apparent
in June of 1991 and by the end of the summer the grid was
filled with dense watermilfoil. Moderate expansion was
observed at both grids on the West Bed and at the West Grid
on the South Bed (Figures 6 - 8). Dense watermilfoil cover
increased by about 100% on both the West Grid, South Bed and
the North Grid, West Bed from the end of 1990 to 1991.
Much of these grids still had less than 25% cover. A 75%
increase was observed at the Middle Grid on the West Bed
(Figure 6). Thus, expansion was observed at all four grids
compared to 1990. However, only one grid (East Grid, South
Bed) appears to have a rapidly expanding margin.
Invertebrate Samples
Stem Transects . In the South Bed there was a steady
increase in the mean number of eggs per meristem (1.0 - 3.0
eggs/meristem) for plants with intact apical meristems
(Figure bA). On plants without intact apical meristems,
36
-------
there were more eggs found on the lateral meristems at the
end of the summer but a steady increase in egg number such
as that observed f or the intact stems was not observed
(Figure lOB). Larvae found in meristems (meristem
larvae=first instar larvae) tended to increase in abundance
over the course of the summer regardless of the presence of
an intact meristem (Figure 11A&B). Larvae found in the
stems (stem larvae) displayed peaks in abundance in mid-
summer and late summer for plants with intact meristems;
only one obvious peak was observed for the plants without an
apical meristem (Figure 12A&B).
In the West Bed, egg numbers on plants with intact
meristems fluctuated between 1.0 and 1.75-2.0 over the
summer (Figure 13A). Egg numbers were higher in the West
Bed than in the South Bed at the beginning of the summer; no
pattern of steady increase was seen as was observed in the
South Bed. Egg numbers declined on plants without intact
meristems by midsummer, picking up again late in the season.
Numbers of meristem larvae remained consistently high on
both intact plants and plants without apical meristems
(Figure 14MB). Stem larvae peaked around mid-summer on
intact plants (Figure 15A). Three peaks in the abundance of
stem larvae were observed on plants without apical meristems
(Figure 15B). The position of these peaks is very similar
to those observed in the South Bed for the minisamples (see
below). Eggs and meristem larvae were rarely found in the
collections made on 14 September. Only stem larvae and
37
-------
pupae were still obvious. By 31 October, all lifestages of
the weevil were no longer collected from the watermilf oil.
Minisamples . There was little change in the dominant
taxa found on watermilfoil over the course of the summer in
the minisamples (Tables 2A&B, 3A&B). Typically, the most
abundant animals on the plants were oligochaets and snails,
primarily the genus Amnicola . Both . lecontel and Acentria
were more abundant on watermilfoil in 1991 than 1990 (see
Tables 1 and 2 in Creed and Sheldon 1991). E. lecontei were
collected from stems on every date in 1991 while they were
only collected on 50% of the dates in 1990. Maximum
abundance of . lecontei in 1990 was 0.8 weevils per stem as
compared to 3.2 for 1991. Three peaks in abundance for .
lecontei were observed during 1991 in the South Bed. These
were on July 2, July 23 and August 20, 1991 (Figure 16A).
Only two peaks appear in the West Bed data (Figure 16B).
The position of the two West Bed peaks appear roughly
similar to the first and last peaks in the South Bed.
Abundances of weevils built up quite rapidly in the West
Bed; by 2 July there were already over 3 weevils per stem
compared to only 1/stem for the South Bed.
Acentria were collected on every date in the West Bed
and eight of eleven dates in the South Bed (Figures 17 A&B).
Maximum abundance of Acentria in 1990 was 0.8 per stem
compared to 2.8 per stem in 1991. The data for 1991 suggest
that there may be two generations of Acentria in Brownington
Pond. Peaks in abundance were observed in early June and
38
-------
mid-August for both beds. Acentria females were observed in
the pond laying eggs at about the same times.
Super Samples 1990 . The same taxa which show up as the
dominant taxa on watermilfoil in the minisamples also
dominated the super samples in 1990 (Tables 4 and 5). These
include the oligochaets, the amphipod Hyallela , chironomids,
the mayfly Caenis (early in the season) and the snail
Amnicola . Few taxa showed significant increases in
abundance across dates in the South Bed. Two thirds of the
taxa (the same taxa found in the South Bed) showed
significant increases In abundance in the West Bed. This
difference between beds may simply be due to the fact that
the last sample in the West Bed was taken later than that
for the South Bed and all we are seeing is a seasonal
effect. Alternatively, there may be some Important
difference in the environment of these two beds that results
in the differential rates of increase of these invertebrate
taxa. Certainly, changes in weevil density over the summer
in 1991 were different in the two beds (see minisample
results). There also appeared to be differences In weevil
phenology between the two beds in 1991. So the difference
observed in the abundance of these taxa may be real. One
interesting observation Is that Euhrvchiopsls showed a
significant increase In the South Bed but not the West Bed.
The reverse was true for Acentria.
Super Samples 1991 . The same taxa that dominated the
1990 super samples were also the dominant taxa in the 1991
39
-------
samples (Table 6A). As in 1990, these taxa most frequently
showed significant differences between dates. Of the taxa
that showed a significant date effect most increased in
abundance from June to August (Oligochaets, Hyracarina,
Euhrvchiopsis, Acentria , Anisoptera, Oxvethira , Planorbid
snails, Ainnicola and Phvsa) . Only the Chironomidae and the
mayfly Caenis decreased in abundance from June to August.
Six taxa displayed significant differences by site (Table
6A). Oligochaets, chironomids, Caenis , and the Anisoptera
were more abundant in the South Bed. The planorbids and
Amnicola (both gastropods) were more abundant In the West
Bed. Significant date by site Interactions were observed In
five taxa. These Interactions mean that the abundances of
these taxa did not change in the same fashion in the two
beds over the two dates. For four of the taxa ( Caenis,
Acentria, Enallagma , and planorbids), the abundance In the
West Bed decreased while abundances In the South Bed
increased. For Amnicola , abundances increased In both beds
but they increased much more sharply In the South Bed.
The density of . lecontei ranged from 2-6 times higher
in the 1991 super samples compared to the 1990 samples
(Compare values in Table 6B with those in Tables 4 and 5).
Acentria , on the other hand, had similar densities to those
recorded in 1990. The highest value for Acentria density
was actually recorded In 1990.
The native plant super samples from 1990 and 1991 have
not yet been processed.
40
-------
Fish Samples
The number of yellow perch sampled in June, August and
September was 29, 28 and 13, respectively, for a total of 70
perch. In June, 25 of the sampled perch contained prey; the
number of perch containing prey in August and September was
26 and 10, respectively. Length, weight and age data are
summarized in Table 7. Only those prey taxa which which
were found in more than 20% of the perch stomachs (i.e.,
frequency of occurrence (FO) greater than 20%) are included
in Figures 18-21.
The dominant prey taxa (dominance determined as any prey
taxon which had >20% FO when fish from all dates were
combined) in the yellow perch stomachs collected in June
were chironomids, amphipods, mayf lies (both Caenis and
Baetis ) and odonates (both dragonflies and
damselflies)(Figure 18). The frequency of occurrence for
all of these prey taxa had decreased by the August
collection with cladocerans being the dominant prey.
Odonates (just dragonflies) and cladocerans were the
dominant prey found In perch stomachs in September.
Chironomids had declined still further to a FO of 20% and
amphipods and mayf lies were not found in the perch stomachs.
Prey groups that had a frequency of occurrence less than 20%
when all dates were combined, but greater than 20% for a
given date included: Coleopterans (Gyrinidae) and Orconectes
(crayfish) in June (Figure 19), Diptera pupae and larvae
41
-------
( Chaoborus ) in August (Figure 20), and Diptera pupae and
Trichoptera in September (Figure 21). No Acentria larvae or
Euhrychiopsis (adults or larvae) were found in yellow perch
stomachs.
At first glance, these data suggest a different pattern
of feeding by yellow perch between 1990 and 1991. In 1990,
cladocerans dominated the midsummer collections (26 July)
and were codominant prey items in late August with
chironomids and Chaoborus (see Figure 22 which is Figure 11
from Creed and Sheldon 1991). However, qualitative
inspections of perch guts in June of 1990 showed that the
perch were consuming considerable quantities of benthic and
macrophyte-associated invertebrates (Creed, pers. obs.).
Thus, the pattern of heavy consumption of benthic/macrophyte
invertebrates in early summer, followed by a midsummer diet
dominated by cladocerans, followed by a late summer/early
fall diet with cladocerans and some macroinvertebrates as
codominant prey did occur in both years. No young-of-the-
year perch were collected in 1991 so it was not possible to
compare the diet of this group between years.
The perch collected in 1991 were more evenly divided
among age classes than in 1990 (Figure 23). For example, in
1991 the percent of fish in the age classes 3, 4 and 5 was
24%, 31%, and 37%, respectively. In 1990, the corresponding
age class percentages were 56%, 21% and 9% (Creed and
Sheldon 1990). The mean total length for these three age
classes was also smaller in 1991 (Figure 24).
42
-------
Experiments
The Effect of Acentria on Watermilfoil Growth
Acentria larvae had a strong effect on change in
watermilfoil length (Figure 25A). The mean change in length
of the control plants was 68.5 mm compared to -7.5 mm for
the 2-Acentria treatment and -37.5 mm for the 4-Acentria
treatment. The contrast between the 0-Acentria treatment
(control) versus the two Acentria treatments was highly
significant. There was no significant difference between
the 2- and 4-Acentria treatments.
The change in the number of leaf whorls on the stems was
similar to the change in length response (Figure 25B).
Control plants added an average of 10 new leaf whorls. The
contrast between the controls and the Acentria treatments
was highly significant. Watermilfoil plants with Acentria
showed either little change in the number of leaf whorls (2-
Acentria treatment) or a loss of leaf whorls (4-Acentria
treatment). The two Acentria treatments were not
significantly different from one another.
Acentria larvae also had a significant effect on change
in plant weight (Figure 25C). Control plants gained the
most weight and were significantly different from the two
Acentria treatments. While watermilf oil plants with two
Acentria larvae gained slightly more weight than plants with
four larvae the difference was not statistically
43
-------
significant. The observed increase In weight in the two
Acentria treatments appeared to be almost entirely due to
increases in root biomass as length of these plants either
did not change or decreased.
The Effect of Weevil Larvae on Watermilfoil Growth
Weevil larvae did not have a consistent effect on
watermilfoil growth (Figures 26A&B). The presence of one
larva reduced watermilfoil change in length compared to the
control but there was no difference In change in length
between the two larva treatment and the control. The result
of this varied response was that the contrast between the
control and the weevil treatments was not significant.
However, the contrast between the one larva and the two
larvae treatment was significant (Figure 26A). Weevil
larvae did have a consistent effect on change in weight,
(Figure 26B). Control plants gained about twice as much
weight as either of the weevil treatments. The contrast
between the control and the two weevil treatments was
marginally significant (p=O.08). The contrast between the
two weevil treatments was not significant. The mean (± 1
S.E.) amount of stem hollowed by weevil larvae in the two
treatments was as follows: one larva treatment, 75.4 ± 6.7
mm (range 59 - 98 mm); two larvae, 106.2 ± 18.6 mm (range 59
- 160 mm). These values translate into burrowing rates of
8.4 mm/day for single larvae and 11.8 mm/day for two larvae.
44
-------
Weevil larvae burrowed through internodes 6 and greater; no
burrowing damage was observed in internodes 1 - 5.
The Effect of Adult Weevil Feeding on Myriophyllum sibiricum
Weevil adults did feed on Z!i. sibiricum but they did not
have a significant effect on either change in plant length
or weight (Figure 27A&B). The contrast between the control
and the weevil treatments for change in plant weight was
marginally significant (p
-------
The Effect of Herbivores on Watermilfoil Buoyancy
Herbivores had a significant effect on watermilfoil
buoyancy. Significantly more watermilfoil was floating in
the control aquaria than in the aquaria with herbivores
(F=19.97, p
-------
mostly the genus Tetragoneuria) , and the pulmonate snail
Physa . All of these taxa were more abundant in the cages
than in the open watermilfoil bed. However, densities in
the full cages and cage controls were not very different for
either Hyalella or Physa ; dragonflies were twice as abundant
in cages compared to cage controls. Marginally significant
fish effects were observed for the Hydracarina, all weevils
combined, and the damselfly Enallagma . Weevils and
Enallagma tended to be more abundant in the cage controls
and the open watermilfoil (control); Hydracarina were most
abundant in the open watermilf oil. Thus the highest
abundances observed for these three taxa were in areas
accessible to fish.
Significant cage effects were observed for Oligochaets,
Hyalella , Anisoptera, immature pianorbid snails, and the
snails Amnicola and Physa . The abundances of Oligochaets,
immature planorbids and Amnicola all appear to have been
depressed by the presence of cages. On the other hand, taxa
such as Hvalella , the Anisoptera and Physa were more
abundant inside of cages. These last three taxa are known
to be vulnerable to fish which suggests that this cage
effect Is actually an expression of a fish effect (see Table
8). If the results of the fish and cage contrasts are
combined, the only taxa which showed a significant positive
response to the exclusion of fish were Hyalella , Anisoptera
and Physa .
47
-------
The three watermilfoil herbivores, Acentria,
Euhrvchiopsis and Parapovnx did not show a significant
response to the exclusion of fish. These herbivores were
either more abundant where fish were present ( Euhrvchiopsis
and Parapovnx ) or they showed no discernable distribution
pattern with respect to treatment ( Acentria) .
Herbivore Enclosure Experiment in the Pond
Unfortunately, we had some critical problems with this
experiment. First, we had contamination of some of the
tubes with first instar Acentria . Second, the snails were
introduced into the tubes too early and they appear to have
fed upon the watermilfoil meristems. Finally, we appear to
have lost some of the adult weevils from some of the weevil
treatments. All in all, we did not have a consistent
treatment effect. The mean wet weight (± 1 S.E.) of the M.
spicatum in the control was 9.008 (± 0.837) g; mean wet
weight In the weevil treatment was 9.692 (± 1.125). In a
few of the tubes where the weevil treatment appeared to have
been successfully established the plants .were collapsed on
the bottom. Thus we feel that this experiment, when
modified to take into account the above factors, should
work.
48
-------
The Effect of Crayfish on Watermilfoil Distribution
This experiment did not work either. Adult weevils fed
on watermifoil plants in all the treatments with the result
that all but one plant decreased in weight and all plants
decreased in length. These results do not rule out the
possibility that crayfish are important in generating the
watermilfoil distribution pattern in Brownington Pond.
Obviously, the importance of crayfish can not be determined
when weevil abundance is high.
. lecontei Oviposition Experiment
Eggs were recovered from four of the six . sibiricum
weevil females (females 1-4) on all four dates. One of the
. sibiricum females (female 5) only laid eggs on two dates
and the remaining female (female 6) laid no eggs at all; due
to the small number of eggs laid, these two females (5 and
6) will not be considered in the discussion. Only two of
the M. spicatum weevil females laid eggs during the
experiment. Due to the small sample size, the results for
the M. spicatum females will not be discussed in detail.
Female 1 laid eggs only on ! . sibiricum , laying as many
as 5 eggs in 24 hrs (Table 9). Females 2-4 made one change
in their oviposition patterns during the trials but did not
otherwise deviate from an egg laying pattern. Two of these
females (3 and 4) laid more eggs on 14. s icatum while
49
-------
female 2 laid almost equal numbers of eggs on both
watermilfoll species. No oviposition pattern was apparent
for the two M. spicatum females which laid eggs.
In summary, three general oviposition patterns were
observed in this small sample of weevil females collected
from . sibiricum in L. Iroquois: 1) more eggs laid on .
sibiricum , 2) more eggs laid on M. spicatum and 3) equal
numbers of eggs laid on both watermilfoil species.
Discussion
Survey Results
The abundance of Eurasian watermilfoil increased
slightly in 1991 compared to the 1990 surveys. The most
notable expansion was the appearance of the small bed to the
northeast of the South Bed. This expansion was also
apparent on a finer scale in some of the grids, most notably
the North Grid on the West Bed and the East Grid on the
South Bed. While some lateral movement of the beds was
apparent, there was no change in bed position with respect
to depth. The plant transects demonstrated that the beds
occupied the same depth zone as in 1990. The factor that
prevents the beds from moving Into shallow water Is still
unknown. Crayfish did not appear to be abundant enough to
control large plants. It is possible that small . spicatum
individuals that colonize shallow water are consumed by
50
-------
crayfish or some other shallow water herbivore. There do
not appear to be any adverse environmental conditions that
might prevent Eurasian watermilfoil from colonizing the
shallows with the exception of possible sediment
differences.
Both Euhrychiopsis and Acentria were more abundant in
Brownington Pond in 1991 compared to 1990. Euhrvchiopsis
was more abundant in both the minisamples and super samples
in 1991. Acentria only appeared to be more abundant in the
minisamples. The lower Acentria densities in the 1991 super
samples may be due to the fact that 5 August (the date of
the second super sample) was before the peak in Acentria
abundance (see minisample data, Figure 17). More
importantly, the invertebrate survey data (primarily the
minisample data) suggest that . lecontei and not Acentria
is the herbivore which is having the strongest effect on .4.
spicatum abundance. A partial collapse of the West Bed was
observed in midsummer. No collapse was observed in the
South Bed until late August. Acentria abundances in both
beds were identical during the summer (see Figure 17).
Weevil abundances, on the other hand, were dramatically
different between the two beds. Weevil abundance built up
slowly in the South Bed reaching a peak by the end of
August. In the West Bed, weevil abundance built up very
quickly, reaching a peak of over three individuals per stem
(both adults and larvae) by mid-July. This peak in weevil
abundance was followed by the . spicatum collapse in the
51
-------
West Bed. While these data are correlative, they support
the hypothesis that . lecontel is having a stronger effect
on M. spicatum abundance in Brownlngton Pond than Acentria.
Acentria obviously has some effect (see experimental
results) but these data suggest that at the densities
present in Brownington Pond the effect of Acentria may be
minor compared to that of the weevils.
At present, we do not know why there was a difference in
the rate at which weevil numbers increased in the two beds.
Water temperatures and dissolved oxygen are similar for the
two beds. Two possible factors that may contribute to the
difference are 1) the proximity of the two beds to the
overwintering site of the weevils and 2) the amount of
direct sunlight reaching the beds early in the season. The
weevils which infest !vj. spicatum in Brownington Pond may all
overwinter at a single site and this site may be closer to
the West Bed. Thus, adult weevils returning to the pond may
encounter the West Bed first. We have no data on weevil
overwintering behavior or dispersion in the terrestrial
environment. Alternatively, watermilfoil may begin growing
earlier in the West Bed due to a more direct exposure to
sunlight early in the growing season. The South Bed would
be shaded longer as it is closer to the south shore of the
pond. At present we are unable to explain this phenomenon.
The super sample data from 1990 also support the idea that
there is some environmental difference between these two
beds.
52
-------
The South Bed minisample data suggest that at least
three generations of weevils were present in Brownington
Pond during 1991. The West Bed data are not quite as clear
in this regard. However, the West Bed partially collapsed
shortly after the observed accumulation of weevils in July
which could have influenced population cycles. The South
Bed did not collapse until late August after weevils had
reached densities similar to those of the West Bed in July.
As weevils were already present and abundant throughout the
pond when we arrived on June 4 and there were still weevils
in the pond when we sample in mid-September, there may have
been as many as five generations of weevils during the
growing season. The stem transect data do not show the same
clear pattern of weevil generations that appears in the
minisample data. Based on the minisample data, there appear
to have been only two generations of Acentria .
The Effect of Herbivores on Watermilfoil
The larvae of Acentria and Eurhychiopsis had somewhat
different effects on watermilfoil. Acentria had a
pronounced effect on both measured components of
watermilfoil growth, i.e., change in length and weight. The
effect of Acentria appears to differ from that of the
Eurhvchio sis larvae due to the different feeding modes of
these animals. Acentria feed largely on the exterior of the
plant. While first instars of Acentria have been reported
53
-------
to burrow into the stem (Batra 1977), most of the larval
phase is spent on the outside of the plant feeding on leaf
material (note: we have observed late instar Acentria larvae
inside of watermilfoil stems collected from beneath the ice
in Brownington Pond in March). The effect on length and
weight is largely a result of larger Acentria larvae cutting
the stem to build their retreats. However, larvae may cut
the stem additional times. We noticed that two of the
plants with four larvae had been cut into more pieces than
would be expected for retreat construction. This also
occurred in the tank in the buoyancy experiment that had 18
Acentria . The plants were cut into numerous small pieces
that had settled to the bottom.
The effect of Eurhvchio sis larvae on watermilfoil
length and weight was not as strong as that of Acentria .
This is due, in part, to the fact that we used late instar
Eurhychiopsis larvae in this experiment. While first instar
weevil larvae feed on apical tissue and thus directly effect
apical growth, late instar larvae feed by burrowing through
the stem. As much of the plant weight is in stem tissue,
removal of substantial portions of stem tissue should have
an impact on weight. Late instar larvae do not feed on
meristematic tissue and thus should not be expected to have
a direct effect on length change. However, removal of stem
vascular tissue could indirectly Influence stem elongation
as a result of reduced or halted translocatiort of nutrients
from roots to actively growing portions of shoots. If
54
-------
adequate quantities of nutrients can be removed directly
from the water by the plant then this effect should be
neglible. While the sediments are an important source of
nutrients for rooted, aquatic inacrophytes (e.g., Barko and
Smart 1978, 1981, Carignan and Kaiff 1980), . spicatum can
absorb nutrients from the water column through stem and leaf
tissue (Best and Mantai 1978). The ability for j. spicatum
to take up water column nutrients appears to be a function
of nutrient concentration (Best and Mantal 1978, Carignan
and Kaiff 1980), with nutrients being absorbed from the
water when at higher concentrations. In their review of
Eurasian watermilf oil biology, Smith and Barko (1990) report
that most of the N and P taken up by this species comes from
the sediments. Thus, larval burrowing might have a more
pronounced effect on !f. spicatum growth in nutrient-poor
water bodies if the growing portion of the stems can not
obtain sediment nutrients.
While the above information suggests that feeding by
only late instar weevil larvae watermilfoil could produce a
variable response in length change, it does not account for
the consistent difference between the one and two larvae
treatments observed in our experiment for this response
variable. We are unable to explain this result. One
further comparison needs to be made between the effect of
Acentria and Eurhychiopsis larvae on watermilfoil. Larval
weevil burrowing does weaken the watermilfoll stem with the
result that burrowed stems are easily broken. While we have
55
-------
commonly encountered such broken stems in lakes and ponds,
this effect was not observed in our experiment as the stems
were protected from physical disturbance by the chambers.
Herbivores clearly had a strong effect on watermilfoil
buoyancy. This result is not surprising as both Acentria
and Eurhychiopsis (both adults and larvae) expose stem
vascular tissue while feeding. We have observed adult and
larval weevils feeding on several occasions and it is not
unusual to see a stream of bubbles emerging from the damaged
portion of the stem. That this small scale effect could
result in the collapse of the upper po tion of a
watermilfoil bed Is significant. The implication is that
this suite of herbivores does not have to remove
considerable amounts of stem and leaf tissue to have a
strong negative effect on watermifoll. The consequences of
leaf removal may be minor in comparison to the effect of
loss of buoyancy. If herbivore feeding can cause plants to
drop out of the photic zone (or at least well lit surface
waters) then plants may not be able to recover from this
damage. The effect becomes even more pronounced if damaged
plants can drag down undamaged ones. Indeed, loss of
buoyancy may prove to be one of the major mechanisms of
watermilfoil bed destruction by herbivores.
To date, three studies have found herbivorous insects
associated with declining watermilfoil populations (Painter
and McCabe 1988, MacRae et al. 1990, Creed and Sheldon
1991). In all three instances, the researchers conducted
56
-------
laboratory experiments that demonstrated that the associated
herbivores could have negative effects on watermilfoil.
Painter and McCabe (1988) found that watermilfoil could
tolerate and continue to grow with low densities of
Acentria . At higher densities, Acentria had a negative
effect on watermilfoil growth. Painter and McCabe did not
determine how Acentria feeding could generate these results,
however. MacRae et al. (1990) conducted more detailed
experiments which evaluated the effects of a midge larva
( Cricotopus myriophvlli ) on watermilf oil growth. They found
that this midge, which feeds on just the meristems, could
prevent the plants from increasing in length or weight.
Only one larva per apical tip was needed to produce this
effect. We have obtained similar results in an experiment
which examined the effects of adult weevils ( . lecontei ) on
watermilfoil (see Experiment 1, Creed and Sheldon 1991).
Adult weevils can influence elongation by destroying
meristems. In that experiment, some of the effect on change
in length was attributable to first instar larvae which also
feed on the meristem (see Creed and Sheldon, 1991, for
details). Adult weevils had an even stronger effect on
change in plant weight. At densities of four weevils per
plant, watermilfoil plants lost weight during the
experiment. This was due primarily to the removal of
several leaves. All of these studies have focused on the
impact of these herbivores on watermilf oil growth. However,
our field observations and the buoyancy experiment suggest
57
-------
that herbivores (at least weevils) may do more than just
suppress growth. They may have an additional negative
impact on watermilfoil plants if they can cause them to
sink.
The results discussed above, along with those presented
in this report, suggest that herbivores might be playing an
important role in either reducing the abundance of
watermilfoil. or maintaining populations at manageable
levels. While it remains to be demonstrated that any of
these herbivores can directly cause a decline, the data
collected from laboratory experiments are promising.
Continued effort should be focused on determining the
mechanisms by which these herbivores affect watermilfoil.
However, we are rapidly reaching the stage where we need to
conduct controlled herbivore introductions to fully assess
the ability of these insects to suppress or reduce nuisance
watermilfoil growth.
The Effect of Fish on the Abundance of Herbivores and
Potential Herbivore Predators
Fish do not appear to have a significant effect on the
abundance of Euhrvchio sis and Acentria In Brownington Pond.
These two herbivores have not shown up in yellow perch
stomachs In either 1990 or 1991. Neither species
demonstrated a response to the presence of fish in the fish
exclusion experiment. In fact, the weevils were more
abundant In the open parts of the watermilfoil bed than in
58
-------
the cages. These results support our previous contentions
that fish do not have a significant effect on watermilfoil
herbivores (see Creed and Sheldon 1991).
Brownington Pond is dominated by yellow perch which is
unusual for many temperate lakes and ponds. Centrarchids
(sunfish) are often the dominant family of freshwater fish
in these waterbodies. It is possible that in centrarchid
dominated systems a fish effect on these herbivores might be
observed, as centrarchids may be more effective at feeding
in vegetation than perch. The preliminary results of the
fish enclusion experiment in Lake Bomoseen suggest that
sunfish (bluegills) might have some effect on weevil
abundance. However, these herbivores are abundant in lakes
and ponds with abundant centrarchids. Thus, we still
believe that fish are not important in influencing the
distribution and abundance of these potentially important
watermilfoil herbivores.
The Effect of . lecontel on Myrlophyllum sibiricum
Weevils did feed on !! . sibiricum although they did not
have a significant effect on either change in length or
weight of this watermilfoil species. They did remove a
significant number of leaves, however. The pattern of leaf
removal was very similar to that observed for weevils
feeding on M. spicatum . The weevils used in this experiment
were collected from M. spicatum . Weevils collected from M.
59
-------
sibiricum might have different effects on this native
watermilfoil. We have yet to find a pond in Vermont that
has a population of . sibiricum with associated weevils but
that does not have Eurasian watermilfoil.
The oviposition experiment produced interesting results
with respect to selection of egg laying sites by female
weevils. Females demonstrated fairly consistent egg laying
patterns. Some females clearly preferred one species over
the other while others would lay on both watermilfoil
species. The preference for . sibiricum by one female
collected from . sibiricum was quite strong. She layed 15
eggs in six days and never laid one on . s icatum .
Apparently females can distinguish between these two
species. The mechanism by which they discriminate is not
yet clear. Also interesting was that in this small sample
of weevils we were able to obtain four females with three
different preferences. This suggests that a high degree of
variability in host preference is present in this species.
60
-------
RESEARCH AT LAI(E BOMOSEEN AND MIDDLEBURY
Research at Lake Bomoseen
Introduction
Lake Bomoseen is the largest lake (1128 hectares)
contained entirely within the boundaries of the state of
Vermont. Eurasian watermilfoil was first reported in Lake
Bomoseen in 1982. The lake currently has a serious
infestation of this species. Attempts at controlling the
watermilfoil have included harvesting and an overwinter
drawdown. Hydroraking and bottom barriers have also been
used by camp owners on an individual basis. One of the
primary objectives of this project is to determine if the
herbivores under study could be employed to control the .
spicatum infestation in Lake Bomoseen. The goals of the
1991 field season were to 1) determine the effect of fish on
the abundance of herbivores (primarily weevils) and 2)
collect data on the effects of watermilfoil harvesting on
herbivore abundance.
61
-------
The Effect of Fish on Herbivores
Introduction
Insectivorous fish can be an important structuring force
of littoral zone invertebrate assemblages. The success of a
weevil introduction in any given lake, therefore, may be
determined in part by the resident fish community.
Unfortunately, little Is known about interactions between
fish and weevils and It is difficult to predict how fish
will influence the impact of weevils on watermilfoil.
If fish do influence weevil populations, a relationship
might exist between fish density and weevil density. To
determine this relationship, experiments were conducted in
Lake Bomoseen during the summer of 1991. Laboratory trials
were also conducted to evaluate the potential f or direct
fish predation on adult weevils. The bluegill ( Le omis
macrochirus) , a common, Insectivorous species often
associated with aquatic macrophytes, was chosen to represent
the fish populations weevils are likely to encounter.
Materials and Methods
Field Experiment
Bluegill densities were manipulated using cages made of
1 cm mesh nylon netting attached to the top and sides of PVC
62
-------
frames measuring 2.5 m high by 1 m 2 . Cages were placed in
the unharvested section of the watermilfoil bed on the east
side of Neshobe Island. The cages were slid over the
watermilfoil plants, and secured with rocks placed around
the base. Twenty-four cages were arrayed in a line,
parallel to the shore of Neshobe Island in water 2.6 m to
3.4 m deep.
Bluegills measuring 65-90 mm were seined from Lake
Bomoseen, weighed, and tagged with Floy markers. A scale
sample was taken from each fish for aging. Bluegills were
stocked into the cages on June 24 at 3 densities: 0 fish per
cage (0-density), 2 fish per cage (low density), or 8 fish
per cage (high density). Eight replicates of each density
were randomly assigned to the cages. Three 0.18 m 2
watermilfoil samples were collected with a super sampler on
June 24 to quantify the initial invertebrate community
within the bed.
High fish mortality rates in both low (44%) and high
(61%) density cages over the next 3 weeks made it necessary
to re-evaluate stocking levels and procedures. Because a
maximum of 6 fish/cage survived in the high density cages,
stocking levels in these cages were reduced from B fish per
cage to 6 fish per cage. To reduce mortality associated
with handling stress, scale samples were not taken from
replacement fish. The presence of 1 bluegill in each of 2,
0-density cages, also made it necessary to randomly assign
these cages new stocking densities of either 2 or 6 fish per
63
-------
cage. An equal number of replicates of each stocking
density was maintained by re-assigning a density of 0 fish
per cage to 2 cages randomly chosen from among those in
which all fish had died. Fish were re-stocked on July 25,
and this is considered the starting date of the experiment.
To maintain the re-assigned bluegill densities, cages were
surveyed weekly for the remainder of the experiment (6
weeks) and the fish that died (11) were replaced. It was
assumed that the fish densities in each cage prior to July
25 did not affect results.
On September 6, the netting was removed from all cages
and a SCUBA diver used a super sampler to collect one
watermilfoil sample from each cage. Three watermilfoil
samples were also taken from the bed on the same date. All
samples were preserved in 70% ethanol.
In the laboratory, a series of
used to separate invertebrates from
were placed in the top, coarse mesh
tap water. Invertebrates and plant
middle (1 mm) and bottom (500 urn) sieves were transferred to
enamel pans where the invertebrates were separated from the
milfoil. and preserved in 70% ethanol. All
macroinvertebrates were later identif led to the lowest
feasible taxon. Plant stems were examined for endophytic
weevil larvae and pupae by holding them up to a bright
light. The biomass of milfoll in each sample was determined
3 graduated sieves was
the watermilfoil. Plants
sieve and rinsed with
debris caught on the
64
-------
by oven drying the plants at 85° C for approximately 24 hrs
and weighing them to the nearest 0.01 grams.
Laboratory Experiment I
Weevil density may be Influenced by fish, other
invertebrates, and/or interactions between fish and other
invertebrates. After preliminary trials established the
palatability of weevils to bluegills, laboratory aquaria
were used to evaluate the Influence of fish and a community
of selected non-weevil macroinvertebrates on adult weevil
survival. Two factors, fish and non-weevil invertebrates,
were tested at two levels, present and absent. A full
factorial design was used to test for an interactive effect
between fish and non-weevil Invertebrates on weevil
survival. A trial included four treatments in four separate
aquaria: (1) fish present, other invertebrates present; (2)
fish present, other invertebrates absent; (3) fish absent,
other invertebrates present (4) fish absent, other
invertebrates absent. The dependent variable was the number
of surviving weevils.
The non-weevil invertebrates included damselfly nymphs
( Enallaama sp.), dragonfly nymphs ( Tetragoneuria sp.),
amphipods ( Hyalella sp.), and snails ( Amnicola sp.).
Amphipods, dragonflies and damseif lies averaging
approximately 3 mm, 9 mm, and 6 mm, respectively, were
collected on Elodea plants In Shelburne Pond (Vt) between
65
-------
August 14 and August 22. Snails approximately 2 mm wide
were taken from watermilfoil plants in Shelburne Pond on
August 14. Weevils were collected from Glen Lake (Vt) on
July 30. BluegIlls measuring 57-78 mm were seined from Lake
Bomoseen between August 12 and August 22. The . spicatum
stems were collected from Lake Bomoseen. Fish, weevils, and
other invertebrates were held in separate aquaria between
trials, and each individual was used in only one trial.
Conditions within a watermilfoil bed were simulated in
38 1 aquaria. Each aquarium was filled with 15 1 of tap
water treated with Hartz All-In-One Water Conditioner.
Watermilfoil stems were examined at 7x with a dissecting
microscope and all observed invertebrates were removed. A
stem density of 220 stems/rn 2 was achieved by passing the
base of 18 milfoil stems through uniformly distributed
openings in a 34.5x19.5 cm piece of plastic canvas. The
above canvas length of each stem was between 155 and 175 mm
and the canvas was weighted to the bottom of the aquarium
with several small rocks. Stems were left in the aerated
aquaria overnight and most had regained an upright position
by 8:00 am the following morning.
Treatments were randomly assigned to aquaria. Eight
weevils were haphazardly selected and placed In each
aquarium at 8:00 am. A combination of six dragonflies, 6
damself lies, 12 amphipods, and 6 snails were added to the
appropriate aquaria (Invertebrates Present Treatment) at
4:00 pm the same day. Two haphazardly selected bluegills
66
-------
were placed in the appropriate aquaria (one bluegill per
aquarium) at 8:00 pm the same day and allowed to feed for 12
hours. Throughout the trial, aquaria were covered with
netting to prevent weevils from escaping.
At the conclusion of a trial, bluegills were removed
and sacrificed (anesthetic overdose). Guts were examined
within one hour and contents recorded. Aquaria were
examined and the number and identity of surviving
invertebrates were noted.
Four trials were run. Space and equipment
limitations made it necessary to run each trial in a
separate time block. All trials were run between August 15
and August 23 with a maximum of 3 days between trials.
Laboratory Experiment II
The results of the first laboratory experiment
indicated that the other invertebrates chosen for the
experiment may have influenced bluegill predation on
weevils. It was not clear which of the four taxa was
creating this effect, but because 83% of the other
invertebrates consumed were either dragonflies or
damself lies, these two prey groups seemed more likely to be
influencing bluegills than either amphipods (8% of consumed
organisms) or snails (8% of consumed organisms). A
67
-------
second laboratory experiment was run in September to
evaluate the potential for dragonflies and damself lies to
influence bluegill consumption of adult weevils.
A full factorial design was used again and the four
treatments were: (1) damseif lies present, dragonflies
present (2) damself lies present, dragonflies absent (3)
damseif lies absent, dragonflies present (4) damseif lies
absent, dragonflies absent. The dependent variable was the
number of surviving adult weevils.
Dragonflies ( Tetraponeuria sp.) and damseif lies
( Enallaama sp.) were collected from Shelburne Pond on
September 19. On September 20, adult weevils were collected
from Glen Lake and bluegills measuring 60-79 mm were seined
from Lake Bomoseen.
Aquaria were set-up as described for the first lab
experiment. Two replicates of each treatment were run
concurrently. A trial, therefore, consisted of 8 aquaria
housing two replicates of each of the four treatments. Two
trials were run, resulting in a total of four replicates of
each treatment. The procedures for conducting and
concluding each trial were the same as those used in the
first experiment.
68
-------
Results
Field Experiment
The preliminary results of the fish enclosure experiment
suggest that bluegills may have some effect on weevil
abundance. Note that the values in Table 10 are just weevil
densities, however. They have not been standardized by
watermilfoil biomass. As weevil abundance may be positively
correlated with watermilfoil biomass, further interpretation
of these data is not possible until weevil abundances are
standardized.
Laboratory Experiment I
The percent of adult weevils, and the percent of all
taxa surviving under each treatment are summarized In Tables
11A&B, respectively. There was one hundred percent
survivorship of weevils In both treatments without fish,
supporting the assumption that all mortality In the presence
of fish was due to fish predation. A question of interest,
however, was whether predation on weevils was influenced by
the presence of other prey. To address this question, the
two treatments in which fish were present (fish present,
other invertebrates present, and fish present, other
invertebrates absent) were compared.
The data were non-normally distributed and treatment
69
-------
variances were not equal (Levenes test for equal variance,
p
-------
Discussion
The preliminary results from the field experiment
suggest that bluegills can have a negative effect on weevil
abundance. However, note that the values in Table 10 are
just weevil densities, i.e., they have not been standardized
for watermilfoil biomass. As waterinilfoil biomass did vary
between cages some of the differences in weevil abundance
between treatments may be attributable to this factor.
The results of Laboratory Experiment I suggest that the
presence of other invertebrates may stimulate feeding by
bluegills. As a result of this feeding activity some
weevils got ingested. This statement is based on the fact
that weevils had a much higher survivorship in the presence
of just bluegills. The results of both laboratory
experiments also suggest that predatory invertebrates (e.g.,
dragonflies and damseif lies) have little direct effect on
weevil survivorship. However, the presence of these other
invertebrates may indirectly effect weevil mortality as
their presence appears to stimulate bluegill feeding on
invertebrates in general. The results of all three of these
experiments lend further support to our idea that weevil
abundance is only weakly affected by predators, especially
insectivorous fish (see discussion in Brownington Pond
section).
71
-------
The Effect of Mechanical Harvesting on Herbivore Abundance
Introduction
Successful establishment of watermilfoil herbivore
populations can be affected by the use of mechanical
harvesters. Previous research (Creed and Sheldon 1991) has
demonstrated that both Euhrychiopsis and Acentria are most
likely to be found near the top of the plant and that
Euhrychiopsis adults and larvae feed on meristem, leaf and
stem tissue located at the top of the watermilfoil stem.
Thus, mechanical harvesters which remove the upper portion
of watermilf oil stems may reduce the densities of these
potentially important herbivores. Preliminary samples
collected in Lake Bomoseen during 1990 suggested that this
was the case (Creed and Sheldon 1991). The goal of the past
summers research was to better quantify this harvester
effect. We also expanded the study to include more sites
within the lake in order to incorporate any site to site
variation.
Study Sites
No watermilfoil harvesting was done in Lake Bomoseen at
selected sites at Neshobe Island and Eckley Bay in 1990. In
1991 these sites were again unharvested and two additional
no-harvest sites were added (Figure 30). The four sites
72
-------
were: the east side of Neshobe Island (approximately lOOm
long), the northwest side of Eckley Bay (approximately 40m
long), the north side of Indian Bay (approximately 15m
long), and the southwest edge of West Castleton Bay
(approximately 20m). An extended no-harvest bed was
maintained at Neshobe Island for the fish enclosure
experiment. The samples taken for the harvester impact
study were all taken south of the fish enclosures.
Materials and Methods
Watermilfoil Stem Transects
Watermilfoil stems were collected weekly from each of
the four sites. Transects were set-up parallel to the edge
of the harvested area. On each side of the line of harvest
the first transects were within 1-3 m of the line, and two
more transects were placed progressively farther from the
harvest line, resulting in six parallel transects, with
three transects located in both the harvested and
unharvested areas. Along a transect line snorkiers removed
the 0.3 m uppermost portion of a plant and placed it in a
ziplock bag. For each transect, five plants with intact
apical meristems and five with damaged apical meristems were
collected, resulting in the collection of 60 stem tops per
site per day. On returning to the lab, plants were examined
under a dissecting microscope. From the ten plants within a
73
-------
transect, every weevil was removed, preserved and the number
and stage recorded. The number of Acentria larvae was also
noted. Samples were collected from early July through mid-
September 1991. Harvesting first occurred at Eckley Bay in
mid-July, therefore the first samples were taken from the
harvested area on 29 July. Over the summer, watermilfoll
became increasingly hard to find at Indian Bay and after
August 18 no more samples were taken there. Differences
between dates, sites and harvest vs unharvested areas were
compared using an ANOVA.
Super Samples
Once a month from June through October six super samples
were collected at each site (for description of sampling
method see Brownington section). Three samples were taken
from the harvested bed and three from the unharvested area.
The sampler was placed haphazardly, although care was taken
that sample locations were spread over an area within lOm of
the line of harvest. Samples were preserved in 70% ethanol.
In the lab, all of the invertebrates are being removed from
the plants, identified and enumerated. The plants were
sorted to species, dried at 800 C and weighed. There were
no harvest super samples taken at Eckley Bay in July. No
super samples were taken in September and October at Indian
Bay. Differences between dates, sites and harvested and
unharvested areas were compared using an ANOVA.
74
-------
Results
Watermilfoil Stem Transects
From the transect data the differences in the number of
weevils between harvested and unharvested areas can be
compared. Significant overall effects of harvesting are
apparent from the data (Tables 13A&B, Figure 31). A
significant date effect is only seen when the last two dates
are retained In the data set. Weevils were leaving the lake
at this time, presumably to overwinter in the terrestrial
environment (Charles OBrien, pers. comm.). There was also
a significant site effect (Table 13B and Figure 32k-C).
Thus, the results from three of the four sites will be
discussed separately.
Neshobe Island
At this site the total number of weevils per date was
higher in the unharvested area (Figure 32A). The no-harvest
area had an average of 10.8 (± 1.49 S.E.) weevils per date
compared to 1.50 (± 0.40 S.E.) in the adjacent harvested
area. Weevil density over the summer did not vary
significantly until fall when no weevils were seen at any
site. We noticed that weevil densities appeared higher near
shore. However, samples were not taken in the shallows
because there would be too much distance between the harvest
75
-------
and no-harvest sites. Thus, our samples may have
underrepresented weevil densities.
West Castleton Bay
The highest densities of weevils were recorded from the
unharvested area at West Castleton Bay (Figure 32B). In
late July there was an average of one weevil per plant.
Again there were more weevils in the unharvested area.
Unlike the Neshobe site, weevil densities in West Castleton
Bay did vary markedly over the summer. The highest
densities, found in July, were composed of all life stages,
but were dominated by eggs. Sampling the West Castleton Bay
site was difficult. Large quantities of watermilfoil pieces
floated over the no-harvest site and it was difficult to
determine which plants were attached and could be sampled.
Because of this difficulty, this site will not be used in
1992.
Eckley Bay
Of the three sites, weevil densities were lowest at
Eckley Bay (Figure 32C). There were no harvest samples
taken until 29 July because the harvester had not yet gotten
to the demarcated area. There were more weevils In the
unharvested area (Figure 32C). Weevil abundance may be
underrepresented because like the Neshobe site, there
76
-------
appeared to be a dramatic increase in weevil abundance in
the shallows. This is not apparent from our data. We did
not sample these shallower areas as this would have resulted
in harvest and unharvest samples being taken from different
depths. By mid-September, . spicatum plants in the no
harvest area at Eckley Bay had collapsed; few upright plants
remained. Most of the plant material was in a consolidated
mat about a half meter thick which covered the bottom
(Creed, pers. obs.). The watermilf oil which had been cut by
the harvesters, and was only a few meters away from the
collapsed watermilfoil, was upright.
Data from the stem transects also allow us to plot the
frequency of . lecontei life stages by week. In Figures
33A&B the total numbers of each life stage for each week
were summed over all sites. Densities of each stage varied
over the summer. Egg and larval densities peaked three
times over the summer (Figure 33A). Adults were numerous in
the early August samples (Figure 33B), followed by an
increase in eggs, followed by an increase in larvae. By the
end of the sample period, no weevils were found.
Super Samples
There were significant differences among sites for all
sixteen major taxa (Table 14). Most of the site effects
were the result of Indian Bay being different from the other
sites, primarily Neshobe I. and Eckley Bay. Indian Bay had
77
-------
greater densities of Oligochaets, Chironomids, Caenis,
Oxvethira and Amnicola than the other three sites. Indian
Bay had lower densities of Orthotrichia than the other three
sites. Indian Bay and W. Castleton Bay had similar
densities of Ainphipods, Hydracarina, Anisoptera, Zygoptera,
Apray].ea , Planorbids and Physa ; densities of these same taxa
were lower at Neshobe I. and Eckley Bay. W. Castleton Bay
had higher densities of Isopoda and Acentria than the other
three sites. Acentria was only found at W. Castleton Bay.
The highest densites of Euhrvchio sis were also recorded at
W. Castleton Bay. In conclusion, Indian Bay was the most
different from the other sites, being different from two or
more sites for 13 different taxa. W. Castleton Bay had
similar densities of Invertebrates as Neshobe I. and Eckley
Bay for eight out of sixteen taxa. Neshobe I. and Eckley
Bay were the most similar with similar densities of
invertebrates for all taxa except Euhrvchiopsis and
Agraylea .
Significant differences between dates were observed for
twelve of the sixteen taxa (Table 14). Most taxa were more
abundant In July. Only Isopods and Acentria were more
abundant in June. Only four taxa (Isopoda, Euhrychiopsis ,
Planorbid snails and Physa ) showed a significant response to
harvesting. The abundances of the latter three taxa
decreased In the harvested areas. Euhrychiopsls were lox
less abundant on harvested watermilfoil. Isopoda, on the
other hand, Increased In abundance on harvested
78
-------
watermilfoil. Of the twelve taxa which did not display
significant responses to harvesting, ten had higher
abundances in the unharvested areas. Of these taxa, only
the Oligochaeta and Hydracarina were more abundant on
harvested watermilfoil.
Discussion
Our data show that there is a negative effect of
harvesting on weevil densities. This is not surprizing as
all life stages of . lecontel are spent on or in the upper
portions of the watermilfoil. Since harvesters remove the
tops of the plant, it is likely that they are removing
weevils also. Rapid recolonization of a harvested area will
probably be reduced since the apical meristems, as well as
several lateral meristems, will have been removed. Until
large numbers of meristems are present in a harvested area
weevil abundances will be low. If areas are harvested
repeatedly over a summer, as is the case with all sites in
Lake Bomoseen, weevils densities may never be high enough to
affect . spicatum growth,
Euhrychiopsis lecontel was most abundant at the W.
Castleton Bay site. The next highest densities were at
Neshobe I. The lowest densities were at Eckley Bay and
Indian Bay. The sampling site in W. Castleton Bay is
located near the inlet of the stream that drains nearby Glen
Lake. Glen Lake has a fairly abundant . lecontel
79
-------
population and a small decline was observed in areas of Glen
Lake this summer (S. Sheldon, pers. obs.). The abundance of
weevils at the W. Castleton Bay site suggests that .
lecontei may be colonizing L. Bomoseen from Glen Lake.
Whether the colonization occurs as a result of weevils
drifting downstream from Glen Lake on watermilfoil fragments
or by flying is unknown. Neshobe is the closest site to W.
Castleton Bay; Eckley Bay and Indian Bay are much further
away. The lower densities of . lecontei at these latter
sites supports the hypothesis that this weevil species
invaded L. Bomoseen from Glen Lake and is slowly expanding
its range within this large lake.
While harvesting reduced the abundance of many of the
invertebrates associated with watermilfoil, significant
reductions were only observed for two snail taxa (planorbids
and Physa ) in addition to Euhrychiopsis . We are not sure
why these snails decreased. It could be that they have a
lower rate of recolonization than a snail like Amnicola .
Alternatively, removal of the dense top of a watermilfoil
bed may make much of the bed accessible to molluscivorous
fish such as pumpkinseed sunfish. Phvsa and planorbids
(e.g., Helisoma ) are more vulnerable to pumpkinseeds than
Amnicola (Osenberg et al. 1992). Crushing and handling
times for pumpkinseeds feeding on these taxa are
approximately 3 times shorter than those for Amnicola
(Osenberg et al 1992). Fish are better able to forage in
areas with lower macrophyte densities (Crowder and Cooper
80
-------
1982), thus these two vulnerable snail species may have
suffered increased mortality after the tops of the plants
were harvested. We are unable to explain the significant
increase in isopods observed in the harvested areas.
A few shore owners from the Cedar Mountain area noticed
that the j. spicatum along the shore looked less healthy
than usual. We examined the west side of Rabbit Island and
the east side of Cedar Mountain late in the season after
weevils were no longer found on transects. Plants at both
sites showed evidence of weevil damage and an adult .
lecontei was seen at Rabbit Island. It is of note that
neither of these two sites had been harvested in 1991.
Samples in 1992 will be relocated to ensure
quantification of weevil densities in the shallows. In
addition, two other sites, one in North Bay and one near the
West Castleton Bay site (where floating plant material does
not accumulate) will be sampled.
Our life history data are consistent with OBriens
(pers. comm.) expectation that adult weevils overwinter in
the leaf litter. As temperatures increase in the spring,
one would expect to see adults, followed by eggs etc in the
lake. It is unclear from our data whether the adults in the
early July sample were overwintering adults or newly pupated
adults. In the summer of 1992, weevil transects will be
started in early May. There are three peaks in egg and
larval densities which might represent three generations
between 1 July and mid-September. The interval between the
81
-------
first and second peaks was approximately 23 days and the
second interval was 19 days. As water temperature was
increasing throughout this time it is not surprising that
the time from egg to adult would decrease. These
generation times are similar to those predicted from the
culturing studies.
Student Projects
A series of student projects were carried out in the
summer of 1991. Project topics included 1) invertebrate
predation on weevils, 2) the effect of water temperature on
weevil egg laying, 3) weevil oviposition preference tests
using M. spicatum and . sibiricum . Unfortunately, none of
the experiments had adequate controls and replication. The
results will be stated briefly. In the predation trials, no
consumption of weevils was actually observed. However, some
weevils did disappear in the presence of one crayfish. The
temperature study found that weevils laid the most eggs at a
temperature of 160 C but more data are needed to determine
if this effect is real or not. Finally, the oviposition
experiment found little discrimination by the weevils for
host plant.
82
-------
Research at Middlebury
Establishment and Maintenance of Weevil
and Watermilfoil Cultures
Introduction
If insect herbivores (particularly the weevil .
lecontei ) are to be used in a biological control program, it
would be preferrable to produce large numbers of these
insects for sustained releases from batch cultures then to
have to collect them from the field. Last summer we
initiated some cultures at Middlebury. Our goals were 1) to
see if we could successfully establish weevil cultures in
the lab and 2) maintain cultures of weevils through the
winter. Obviously, maintaining a weevil culture requires
maintaining a watermilfoil culture so this was initiated as
well.
Materials and Methods
Weevil ( Euhrychiopsis lecontei ) and watermilfoil
( Myriophyllum spicatum ) cultures were kept in aquaria
(either 38 1 or 75 1) under a 16:8 hr photoperiod through
the summer, fall and winter. Aquaria were continuously
aerated to add dissolved oxygen and circulate water. Over
83
-------
the culturing period, water temperature ranged between 130 C
and 27° C.
Macrophytes necessary to support weevil populations were
cultured using two different methods: floating cultures and
rooted cultures. In floating cultures, . spicatum plants
were placed in clean aquaria. For rooted cultures, plant
rhizomes were planted in autoclaved sediment with marble
gravel to anchor the roots. A few tanks of . s icatum were
set up without weevil populations, in order to continue a
culture of weevil-free watermilfoil. Undamaged plants were
added as necessary to tanks containing plants with weevils.
Algal growth within aquaria was controlled using a variety
of methods: 1) snails (Physa), 2) hand removal, and 3) in-
tank filters.
E. lecontei were collected from M. spicatum plants from
June through August from several lakes in Vermont (Lake
Bomoseen, Brownington Pond and Glen Lake) although most came
from Glen Lake. Weevils were also collected on . sibiricum
in Iroquois Lake in August. In the laboratory, weevils were
maintained on macrophyte cultures of the plant species on
which they were found. Weevil populations were estimated
weekly by viewing plants from outside the tank. During the
months of September, October, and November, tanks with
weevil populations were covered with plastic or fine mesh
and sealed around the edges to prevent weevil escape.
Disturbance of tanks was kept to a minimum with the
exception of weekly maintenance.
84
-------
Results
M. spicatum cultures rooted in sediment appeared to be
relatively healthy. Floating cultures of . spicatum died.
High densities of weevils in these cultures may have
contributed to their decline. The rooted . sibiricum ,
under the same conditions as rooted . spicatum , did not
survive. Aquaria with sediment usually had algal blooms.
Algal abundance was controlled by the introduction of snails
and hand removal. Additional fresh plants were added to
declining cultures.
During September, approximately 50 weevil adults were
observed in several of the tanks of floating watermilfoil,
but few eggs and larvae were seen. In early October,
approximately 30 weevil adults were seen climbing up the
sides of each tank and several weevil adults were found in
culture room, outside the tanks. This emigration coincided
with a decline in health of the floating plants. Through
November only 2-4 adult weevils, and no eggs or larvae, were
observed in each of the tanks of unhealthy, floating
watermilfoil. Weevils collected from and cultured on .
sibiricum also disappeared in early November.
An adult weevil and two pupae were found in a tank with
healthy, rooted watermilf oil in November that previously had
no observed weevil populations. Adult weevils were then
added to rooted M. spicatum cultures. Since the
introduction, all life stages (8 eggs, 1 larva, 2 pupae and
85
-------
6 adults) have been observed in these cultures. The number
of larvae and pupae may be underestimated as these counts
were made by simply inspecting the plants, i.e., the plants
were not dissected. The hatching of eggs and pupation, as
well as mating pairs, have also been observed.
Discuss ion
The most successful watermilfoil cultures seem to be
those where the plants are rooted in sediment. Nuisance
algal growth can be a problem in these cultures with
sediment, however. All life stages of . lecontel have been
continually observed In at least one tank, and so it appears
that we have a successful weevil culture.
One of the major factors influencing weevil culturing
appears to be the emmigration of the weevils in October.
This exodus was predicted by Dr. Charles OBrien (pers.
comm.). However, all the weevils in culture at the time had
been raised under constant temperature and light conditions.
Thus, it Is unclear whether this emmigration was cued by
declining plant health, increased weevil population density,
or internal cues.
86
-------
Weevil Life History Studies
Introduction
Weevil life history information has a variety of uses.
It allows us to predict the rate at which weevil
populations, both in nature and In cultures, will increase.
This is useful, when combined with knowledge of how weevils
damage watermilfoil, in determining when a decline might
occur and how many weevils might be needed to produce the
decline. It also aids us In determining the number of
generations that are present in a lake or pond in a given
year. Our goals in these studies were to determine 1) the
duration of the different phases of . leconteis life
history (on both . spicatum and M. sibiricum ) and 2) the
fecundity of female weevils.
Materials and Methods
Adult E. lecontel were collected from Glen Lake.
Information on the life cycle of . lecontei was collected
by placing single watermilfoil stems with 1-3 weevil eggs on
the meristems into individual capsules. The eggs were laid
on the meristems by the weevils. The capsules were made of
a 5 cm cylinder of 500 urn Nytex mesh with removable plastic
caps at each end. All capsules were kept In a sIngle 3.8 1
aquarium. Each meristem was removed daily, examined under a
87
-------
dissecting microscope and the weevil life stage present was
recorded. Two life cycle trials were run during the course
of the summer. For the first trial, ten capsules containing
. spicatum were used. In this experiment, there was a
total of 21 eggs initially, which resulted in a mean of 2.63
eggs/meristem. For the second trial, eight capsules with .
spicatum and eight capsules with bj. sibiricum were used.
There was a total of 22 eggs on the . s icatum meristems
(mean number of eggs/meristem was 2.20). There were 12 eggs
on the M. sibiricum meristems (mean number of eggs/meristem
was 1.50).
Weevil fecundity was measured by counting the number of
eggs produced by virgin (isolated as pupae) female weevils.
Eight virgin females were added to individual containers and
paired with another weevil. If eggs were produced, then the
pair was assumed to be male/female and the number of eggs
produced per meristem was counted daily. Meristems on which
eggs were found were removed each day, and fresh meristems
without eggs were added to prevent any inhibiting effect the
presence of eggs might have had on female egg laying.
Results
The duration of each life stage was fairly similar for
weevils growing on . spicatum in the two trials (Figure
34). Similar development times were observed for eggs and
larvae on . sibiricum , but the pupal stage (n=2) did appear
88
-------
shorter. (Figure 34). See Figure 35A for the numbers of
individuals at each stage for each trial.
Survivorship data were also obtained. Weevils raised on
M. sibiricum appeared to have lower survivorship in each
stage than those raised on M. spicatum (Figures 35 A&B).
Pupal mortality was high for both trials, with only 2 of 18
pupae surviving to adult stage. Only two weevils
successfully pupated on . sibiricum . Both pupae died
within five days of the initiation of pupation.
To eliminate differences in starting date, the weevil
fecundity data were plotted as number of eggs produced per
female per week for the first four weeks (Figure 36). The
mean (± 1 S.E.) number of eggs laid per week ranged from 6 -
9.
Discussion
Pupal mortality was high so the estimate of pupal stage
length may not be accurate for any of the populations,
particularly so for the . sibiricum population (n=2).
During the experiment, it was noted that several pupal
chambers broke open while being handled, killing the pupae
inside.
Our data give a general idea of the fecundity of female
weevils. Unfortunately, the observations were not carried
out throughout the life of all adult pairs, and so do not
provide a complete estimate of lifetime fecundity. Also,
89
-------
due to difficulty in determining the sex of weevils, only
eight pairs were used f or fecundity experiments. To reduce
variability in further experiments, a greater number of
pairs should be used.
90
-------
The Effect of E. lecontei collected from M. sibiricum
on M. spicatum and M. sibiricum .
Introduction
This experiment is analagous to Weevil Bag Experiment II
which was described in Creed and Sheldon (1991). The
difference is that the adult weevils were collected from .
sibiricum in Lake Iroquois. An additional difference is
that we compared the effects of male and female weevils on
both watermilfoil species.
Materials and Methods
Pairs of watermilfoil stems (one M. spicatum and one M.
sibiricum stem) were placed into 12, labeled plastic bags
filled with dechlorinated tap water. The . spicatum was
collected from Lake Dunmore; the M. sibiricum was collected
from Lake Iroquois. Weevil eggs and other invertebrates
were removed from the stems before they were used In the
experiment. The bags also contained rocks to keep the bags
in a vertical position. A single, female weevil was added
to four of the bags, a single male weevil was added to
another four bags and the remaining four bags served as a
control. The bags were then placed in 38 1 aquaria in the
culture room at Middlebury.
91
-------
The experiment ran for 7 days. Plants were then removed
from the bags and all loose leaflets and leaves remaining in
a bag were removed and counted. The number of stem bites
was also recorded. An herbivorous trichopteran larva
( Triaenodes sp.) was found in one of the control bags. The
stems from this bag were not included in the analysis so n=3
for the control treatment. Leaflet, leaf and stem bite data
within a watermilfoil species were analyzed using an ANOVA
with planned, orthogonal contrasts. The first contrast
compared the control to the two weevil treatments. The
second contrast compared the effect of males vs females.
The preference of weevils for the two watermilfoil species
was evaluated using a chi-square goodness of fit test. The
null hypothesis for the chi-square tests was that equal
numbers of leaflets and leaves were removed from each
species or that equal numbers of stem bites were found on
each species. Males and females were compared separately.
Results
Weevils removed a significant number of leaflets from
M. spicatum (Table 15). No other significant (p
-------
sibiricum . Both males and females preferred . spicatum
over . sibiricum . Individuals of both sexes removed
significantly more leaflets from . spicatum (Chi-square
(males)=180.12, 1 df, p
-------
the case as it suggests that the weevils are capable of
rapidly shifting hosts from the native watermilfoil to the
exotic Eurasian watermilfoil. This issue can not be
resolved until weevils collected from a part of their range
that has not yet been invaded by M. spicatum have been
tested.
94
-------
MULTI-LAKE SURVEY
Introduction
In 1989 weevils were found in four Vermont lakes and
caterpillars were found in ten lakes. In order to determine
the distribution of . spicatum herbivores throughout
Vermont a number of lakes were visited in 1990. Qualitative
surveys were made of watermilfoil and its herbivores in
these lakes.
In 1991, thirty lakes in Vermont with nuisance
populations of . spicatum were visited; twenty two of these
lakes were quantitatively sampled to assess the abundance of
herbivorous invertebrates associated with watermilfoil.
Qualitative samples were taken in the remainder of the
lakes. Some of the lakes surveyed had been visited
previously in 1990.
Materials and Methods
Lakes determined by the Department of Environmental
Conservation to have watermilfoil populations were visited.
Based on prior information plus swimming and boating
inspections, the extent of the watermilfoil beds In each
lake was sketched. All plant species encountered were
identified and recorded with rough estimates of their
abundances ranked as: present (plant was seen), common
95
-------
(plant was frequently seen) or dominant (finding other
plants species was difficult). Swimmers also recorded
sightings of herbivores and herbivore damage.
A transect line was placed perpendicular to shore
through a selected watermilfoil bed. Samples of
watermilfoil were taken at five points spaced evenly along
the transect line. Length of the transect line was
determined by the length of the watermilfoil bed.
Two sampling methods, top of plant samples and mini
samples, were used at each of the five sites. A top of
plant sample consisted of a collection of all the plant
stems within a 25 x 25 cm quadrat lowered 20 cm from the top
of the tallest plant. Plant cuttings and associated
invertebrates were placed in plastic bags for later
identification and quantification. A minisample consisted
of enclosing one entire watermilfoil plant and associated
invertebrates within a 28 cm diameter mobile invertebrate
sampler. Plants were cut at the sediment surface and the
volume sieved through a 500 micron mesh. Plants and
associated invertebrates were washed into plastic bags to be
quantified later. Samples were stored at l0 C for one to
three days; then either the invertebrates were removed and
preserved, or the entire sample was preserved, in 70%
ethanol.
For both types of samples, the plants were separated by
species, dried, and weighed. From top samples, all
caterpillars, weevils and caddisflies were removed from
96
-------
plant material, identif led to genus (and when possible
species), and counted. From minisamples, all invertebrates
were removed from the plant material, Identified to family
(or genus and species when possible), and counted. Adult
weevils collected were Identif led as . lecontei or
Phytobius leucogaster , a native weevil similar in appearance
but larger than . lecontel . We do not yet know how to
differentiate between the eggs, larvae, and pupae of these
weevils. Thus, weevil numbers In the tables include adult
. lecontei and younger stages of either . lecontei and/or
. leucopaster .
Results
Of the twenty nine Vermont lakes sampled, . lecontei
was seen in twelve lakes and A. nivea was found in ten lakes
(Table 16). In 1991, . lecontel was collected in three
lakes (Berlin Pond, Lower Pond, Winona Lake) in which it was
not found in 1990 (Table 17). Tables 16 and 17 summarize
the more detailed results of the minisamples and top samples
presented in Tables 18 and 19. . lecontel was also
collected from three lakes not sampled In 1990 (McCuen
Slang, Lake Iroquois, North Montpelier Pond).
Unidentifiable weevil larvae were found in two additional
lakes, Lake St. Catherine and Sunrise Lake. . lecontel was
present in Arrowhead Mountain Lake, Echo Lake and Sunset
97
-------
Lake in 1990. In 1991, M. spicatum was so rare in all three
lakes that only visual surveys were performed; no samples
were taken and no E. lecontei were observed.
In 1991, . nivea was found in two lakes previously not
known to contain this species. . nivea was not found in
two lakes in 1991 from which it had been collected in 1990
(Table 17). Parapoynx badiusalis was only found in two
lakes in 1991 (Berlin Pond and Winona Pond) (Table 16).
This was the first time it was collected in Berlin Pond. In
1990, P. badiusalis was collected in Brownington Pond,
Shelburne Bay of Lake Champlain, Sunrise Lake and Winona
Pond.
The dried weights of the plant samples, both top
samples and mini samples were not significantly correlated
with either . lecontel and A nivea densities (top sample,
r=0.045, mini sample, r=0.077) or each other (r=0.427).
Discussion
Determining the extent of the distribution of macro-
invertebrates can be difficult. The regimented sampling
protocol used in 1991 may provide a more quantitative
measure of the presence or absence of invertebrates in the
surveyed lakes than the qualitative approach used in 1990.
However, due to within lake patchiness in both the
watermilfoil and herbivore populations, between year
variations in collection of invertebrates may not be
98
-------
representative of annual variation. Presence of herbivores
in lakes indicates that there is a population within the
lake but this does not mean that they are abundant.
Conversely, absence of invertebrates in lakes may simply be
the result of insufficient sampling. For example, no .
lecontel were collected in the quantitative sampling of Lake
Bomoseen at Crystal Beach. . lecontei have been collected
in Lake Bomoseen at four other locations (Eckley Bay, Indian
Bay, of f Neshobe Island, and near the State Park beach).
99
-------
RESEARCH AT CASTLETON STATE COLLEGE
Introduction
Researchers Anne Hampton, Michael Alfierl and Brian
Marsh conducted a series of studies examining the preference
of E. lecontei for Eurasian watermilfoil and a number of
native plant species (Hampton et al. 1992). In addition,
they established a second culture of weevils and examined
changes In the location of adults and oviposition sites as
fresh meristems aged in culture flasks. Finally, they
extensively recorded weevils on videotape, including adult
and larval feeding and copulation by adults.
Synopsis of Research Results
Chemical Orientation
Y-maze trials were conducted to determine if adult .
lecontel choose their location on the basis of water
chemistry. A plexiglass maze (Figure 37) was designed so
that two streams of water could be run through the maze and
remain unmixed through the chamber. An adult weevil was
placed In the chamber and its position with respect to the
streams of water noted. In preliminary trials it appeared
that weevils attached to the first surface with which they
came into contact and did not move (M. Aifierl, pers.
100
-------
comm.). A series of modifications were made resulting in
either a string or a net strung from left to right in the
weevil chamber, allowing a weevil to remain attached to a
substrate as it chose between sources of water. There were
no significant differences in weevil positions, even when
one of the waters contained bleach or deoxygenated water.
At this point the maze trials were discontinued.
Choice Tests
A series of choice trials was run in aquaria. The first
set of experiments examined visual orientation. Three 5.7L
tanks were grouped with long sides parallel and touching,
and were filled with declorinated water. For each trial,
weevils (29-33) were placed in the middle aquarium. .
spicatum was placed in one of the outside tanks, and in the
other Ceratophyllum demersum , or a group of native plant
species, or no plants. Weevil location (side of tank) was
recorded. The results suggest that weevils might
distinguish between plants and lack of plants, but they were
not attracted to any particular plant species. For the
three treatments described, the percents of weevils found on
the watermilfoil side of the inner aquarium were 62%, 51%
and 76% respectively.
A more extensive series of trials was run to examine
weevil choice of plant species. A 5.6L tank was filled
with declorinated tap water and set-up in a light-proof box
101
-------
with a light over it. Aquatic plants were placed upright
in each of the four corners of the tanks. Basal portions of
the plants were held down by rocks. Five to seven weevils
were placed in each tank, and after 16-24 hours the position
of each weevil was recorded by plant species and if
appropriate, by whorl number. Plant combinations included
varying densities of . spicatum , !j. sibiricum, Elodea
canadensis, Cerato hvllum demersum and Potamopeton
robbinsil . When . spicatum was run against other plant
species in the absence of . sibiricum , there was a tendency
for weevils to be found on M. spicatum (78%, 69%, 80% of the
weevils). When placed with . sibiricum and . canadensis ,
83% of the weevils were on M. sibiricum . In the M. spicatum
- . sibiricum trials ratios of blomass were varied. When
the ratio was 1 M. spicatum:1 M. siblricum (1:1) weevils
were found twice as frequently on M. spicatum . For the 5:1
and 1:5 trials, weevils did not appear to show preference
and were found on plants in equal proportion to plant
abundance. Position of the weevils on the plants was
recorded by whorl or internode number. Over all trials that
included either M. spicatum or M. sibiricum there was a
tendency for weevils on . spicatum to be approximately one
internode closer to the apical meristem than those on .
sibiricum .
102
-------
Culture
A variety of conditions for culturing were tried. The
number of eggs, and position of the adults on the plant
(whorl number) were recorded daily. When plant meristems
were changed, weevils started at the apical portion of the
plant (position was usually on the first or second whorl)
and apparently moved down until the meristem was replaced
(typical pre-meristem replacement position was whorl 4 after
7 days and whorl 7 after 10 days). The frequency of egg
laying appeared to be greater on a new meristem and
decreased with increasing meristem age.
103
-------
COMMUNICATION OF RESEARCH RESULTS WITH
PUBLIC GROUPS, STATE AND FEDERAL AGENCIES,
AND AT SCIENTIFIC MEETINGS
Public Groups
S. Sheldon gave talks to two local groups on the
Eurasian watermilfoil biological control project. The
groups included Vermonts Southern Federation of Lakes
(September 1991) and the residents of Lake Bomoseen
(February 1992). R. Creed made three presentations, one to
the residents of Lake Bomoseen (June 1991), one to the L.
Wononscopomuc Lake Association in Connecticut (November
1991), and one to the Middlebury Rotary Club (February
1992). Creed and Sheldon put together a slide show on the
prospect of biological control of Eurasian watermilfoil.
The slide show has been given to the Vermont Department of
Environmental Conservation and will be made available to the
public through the VtDEC.
State and Federal Agencies
In November, S. Sheldon, R. Creed and H. Crosson
attended the annual Aquatic Plant Control Research Program
(APCRP) meetings sponsored by the Army Corps of Engineers in
Dallas, Texas. The results of some experiments conducted at
104
-------
Brownington Pond and the harvest/no harvest results from L.
Bomoseen were presented by S. Sheldon.
L. OBryan, S. Sheldon and R. Creed presented a poster
on the effects of mechanical harvesting on weevils at the
annual New England Association of Environmental Biologists
meetings in March 1992. L. OBryan presented the poster.
Scientific Meetings
R. Creed presented the results of the 1990 research at
Brownington Pond at two national, scientific meetings. S.
Sheldon was co-author on both papers. The first talk was
presented at the annual meetings of the North American
Benthological Society in Santa Fe, New Mexico (May 1991).
The second talk was presented at the annual meetings of the
Ecological Society of America in San Antonio, Texas (August
1991).
105
-------
SUMMARY DISCUSSION
The research undertaken during 1991 addressed five of
the six primary objectives proposed for this project (Table
20). Considerable progress was made at the Brownington Pond
(BP) field site where we are examining the watermilfoil
decline (Objective 1). We continued the plant, invertebrate
and fish surveys begun in 1990. We also began monitoring
water chemistry and temperature. Laboratory experiments
documented strong negative effects of herbivores ( .
lecontei and A. nivea ) on watermilfoil. Field experiments
were less successful with the exception of the fish
exclusion experiment which demonstrated that yellow perch
have little direct effect on the abundance of watermilf oil
herbivores. Once again, we did not work with Parapovnx as
this species was still rare in BP. While we have not yet
demonstrated the cause of the decline at Brownington Pond
the results of the 1991 field season lend further support to
the hypothesis that Euhrychiopsis , and possibly other
herbivores such as the Acentria , played some role in the
watermilfoil reduction. Correlations between weevil
abundance and the collapse of the West Bed point to .
lecontei as the primary agent responsible for the decline.
While weevils can influence photosynthesis by removing leaf
tissue and affect the movement of nutrients through the stem
by burrowing through it, their most detrimental effect on
106
-------
watermilfoil may be to reduce the ability of this macrophyte
to remain buoyant.
Research conducted at both Brownington Pond and at
Middlebury (M) examined the effect of herbivores on
watermilfoil and native macrophytes (Objective 2). Strong
effects of Acentria on M. spicatum were observed in the BP
experiments. Weevil larvae did not have as strong an effect
on M. spicatum as Acentria , however, burrowing by larval
weevils may contribute the most to the reduction in Eurasian
watermilfoll buoyancy. Adult . lecontei can influence the
growth of the native watermilfoil, . sibiricum and remove
significant amounts of leaf tissue. Weevils collected from
!j. sibiricum in Lake Iroquois displayed various preference
patterns for egg laying. These data do not provide
definitive evidence that 14. sibiricum is the native host,
however. That can only be determined by collecting .
lecontel from a part of Its range which has not yet been
Invaded by . spicatum .
Quantitative surveys found one or more of the herbivore
species present in several lakes in Vermont. These results
suggest that the herbivores should be amenable to
Introductions in lakes throughout this region (Objective 3).
However, our results from the Lake Boinoseen harvesting study
suggest that extensive harvesting In a lake will prevent the
successful establishment of E. lecontel In infested lakes.
We have also had success at establishing small cultures of
. lecontel at Middlebury. Successful culturing Is
107
-------
important if we are to undertake controlled introductions in
the future.
Our data from L. Bomoseen indicate that this lake
already supports a population of the weevil . lecontel
which suggests that a natural weevil introduction has
already begun (Objective 4). Furthermore, the weevils have
begun to reduce watermilfoil abundance In certain parts of
the lake (objective 5). As mentioned above, extensive
harvesting could prevent this weevil population from
expanding and affecting watermilfoil throughout the lake.
While fish such as bluegills may consume weevils they
probably have a minor impact on weevil abundance.
We have presented the results of our work to public
groups and at scientific meetings. We have also prepared a
slide show on watermilfoil control that is available to the
public through the Vermont DEC (Objective 6).
Finally, a list of equipment purchased on the grant from
June 1991 to April 1992 is presented in Table 21.
108
-------
Acknowledgments. We wish to thank Kristin Henshaw, Diana
Cheek and Gabe Gries for their invaluable help at
Brownington Pond in 1991. Kathy Newbrough, Bill Waddell and
Wendy Cox helped collect the 1990 super samples. Linda
OBryan, Creed Clayton, Molly Franz, McDuff Sheehy, Lori
Racha, Kim Kruse and Tammy Anthony helped with all of the
Middlebury research. We are also grateful to the relentless
bugpickers without whose dedication much of these data could
not be presented! This work was funded by the EPA Clean
Lakes Demonstration Program, the USACE (DAC W39-90-K-0028),
the Vermont Department of Environmental Conservation and
Middlebury College.
109
-------
LITERATURE CITED
Abrahamsson, S.A.A. 1966. Dynamics of an isolated
population of the crayfish Astacus astacus Linne.
Oikos 17:96107.
Aiken, S.G., P.R. Newroth and I. Wile. 1979. The biology of
Canadian weeds. 34. Myriophyllum spicatum L. Can. J.
Plant Sd. 59:201-215.
Anonymous. 1990. Eurasian watermilfoil in northern
latitudes: results of management workshop. The
Freshwater Foundation and Minnesota Department of
Natural Resources, Navarre, Minnesota.
Barko, J.W., and R.M. Smart. 1978. The growth and biomass
distribution of two emergent plants, Cvperus esculentus
and Scir us validus , on different sediments. Aquat.
Bot. 5:109117.
Barko, J.W., and R.M. Smart. 1981. Sediment-based
nutrition of submersed macrophytes. Aquat. Bot. 10:339-
352.
Batra, S.W.T. 1977. The bionomics of the aquatic moth
Acentro us niveus : a potential biological control agent
for Eurasian watermilfoil and Hydrilla . J. N.Y. Ent.
Soc. 85:143152.
Bayley, S., H. Rabin and C.H. Southwlck. 1968. Recent
decline in the distribution and abundance of Eurasian
milfoil in Chesapeake Bay. Ches. Sci. 9:173-181.
Best, M.D., and K.E. Mantai. 1978. Growth of Myriophyllum :
sediment or lake water as the source of nitrogen and
phosphorus. Ecology 59: 1075-1080.
Carignan, R., and J. Kaiff. 1980. Phosphorus sources for
aquatic weeds: water or sediment. Sd. 207:987-989.
Carpenter, S.R. 1980. The decline of Myriophyllum spicatum
in a eutrophic Wisconsin lake. Can. J. Hot. 58:527-
535.
Couch, R., and E. Nelson. 1986. Myrlophyllum spicatum in
North America. In The First International Symposium on
Watermilfoil ( Myriophyllum spicatum ) and Related
Haloragaceae Species. pp. 8-18.
110
-------
Creed, R.P., Jr., and S.P. Sheldon. 1991. The potential
for biological control of Eurasian watermilfoil
( Myriophyllum spicatum) : Results of the Research
Programs initiated in 1990. Prepared for Region 1,
U.S. EPA, Boston, Mass.
Crowder, L.B., and W.E. Cooper. 1982. Habitat structural
complexity and the interaction between bluegills and
their prey. Ecology 63:1802-1813.
Flint, R.W., and C.R. Goldman. 1975. The effects of a
benthic grazer on the primary productivity of the
littoral zone of Lake Tahoe. Limnol. Oceanogr. 20:935-
944.
Hampton, A., M. Alfieri and B. Marsh. 1992. BehavIoral
studies to determine the appropriateness of
Eurhychiopsis lecontei as a biological control for
Eurasian watermilfoil, Myriophyllum spicatum .
Unpublished report. Slpp.
Lodge, D.M., and J.G. Lorman. 1987. Reductions in
submerged macrophyte biomass and species richness by
the crayfish Orconectes rusticus . Can. J. Fish. Aquat.
Sd. 44:591597.
Lorman, J., and J. Magnuson. 1978. The role of crayfish in
aquatic ecosystems. Fish. 6:8-10
MacRae, I.V., N.N. Winchester and R.A. Ring. 1990. Feeding
activity and host preference of the milfoJ.l midge,
Cricotopus myriophylli Oliver (Diptera: Chironomidae).
J. Aquat. Plant Manage. 28:89-92.
Nichols, S.A., and G. Cottam. 1972. Harvesting as a
control f or aquatic plants. Water Res. Bull. 8:1205-
1210.
Nichols, S.A., and B.H. Shaw. 1986. Ecological life
histories of the three aquatic nuisance plants,
Myriophyllum spicatum, Potamopeton crispus and Elodea
canadensis . Hydrobiol. 131:3-21.
Osenberg, C.W., G.G. Mittelbach, and P.C. Wainwright. 1992.
Two-stage life histories in fish: the interaction
between juvenile competition and adult performance.
Ecology 73:255267.
Painter, D.S., and K.J. McCabe. 1988. Investigation into
the disappearance of Eurasian watermilfoil from the
Kawartha Lakes, Canada. J. Aquat. Plant Manage.
26: 312.
111
-------
Reed, C.F. 1977. History and distribution of Eurasian
watermilfoil in United States and Canada. Phyto].ogia
36:417436.
Smith, C.S., and J.W. Barko. 1990. Ecology of Eurasian
watermilfoil. J. Aquat. Plant Manage. 28:55-64.
Sokal, R.R., and F.J. Rohif. 1981. Biometry. 2nd Ed. W.H.
Freeman and Co., New York, N.Y.
Spencer, N.R., and M. Lekic. 1974. Prospects for
biological control of Eurasian watermilfoil. Weed Sd.
2:401404.
112
-------
TABLES
113
-------
Table 1. Results of the water chemistry survey conducted in
Brownington Pond on 25 June 1991. Samples were collected at
the surface and just above the bottom. Values in the table
are mean concentrations (+ 1 S.E.).
Site Orthophosphate
Nitrite
Nitrate
(mg/i)
(mg/l)
(mg/l)
East Side
Surface 0.003 0.01 0.01
(0.0006) (0.00) (0.00)
Bottom 0.002 0.01 0.01
(0.0003) (0.00) (0.00)
South Bed
Surface 0.002 0.01 0.01
(0.000) (0.00) (0.00)
Bottom 0.002 0.01 0.01
(0.000) (0.00) (0.00)
West Bed
Surface 0.002 0.01 0.01
(0.000) (0.00) (0.00)
Bottom 0.002 0.01 0.01
(0.000) (0.00) (0.00)
-------
Table 2A. Dominant macroinvertebrate taxa associated with
M. spicatum stems from the South Bed in Brownington Pond
(June 18 - July 23, 1991). Samples were collected using the
smaller MIS sampler (minisampler). Data in the table are
mean number of individuals per stem (± 1 S.E.). Five
samples were taken on each date.
Taxon Date
18/6 25/6 2/7 9/7 16/7 23/7
Annelida
Hirudinea 2.8 - 0.2
(1.6) (0.2)
Oligochaeta 3.6 7.2 1.0 14.2 9.4 33.0
(1.8) (2.9) (0.6) (4.1) (0.0) (8.3)
Arthropoda
Amphi poda
Hyallela 1.0 0.6 0.2 1.0 1.0
(0.6) (0.2) (0.2) (1.0) (1.0)
Cladocera 2.4 0.6 0.4 0.2 1.2 0.2
(1.0) (0.4) (1.8) (0.2) (0.7) (0.2)
Ostracoda 0.8
(0.5)
Hydracarina 0.4 0.6 0.2 0.6 1.4
(0.2) (0.2) (0.2) (0.4) (0.6)
Insecta
Coleoptera
Euhrychlopsis 0.2 0.4 1.0 0.6 0.2 2.4
(0.2) (1.8) (0.5) (0.4) (0.2) (0.5)
Diptera
Chironomidae 1.0 1.4 0.4 0.8 2.2 1.8
(0.6) (0.7) (1.8) (0.4) (0.7) (0.7)
Heliidae 0.4 0.4 0.2
(0.2) (1.8) (0.2)
Lepidoptera
Acentria 0.6 0.6 0.2
(0.4) (0.4) (0.2)
-------
Table 2 Continued .
Odonata
Anisoptera -
Zygoptera
Enallagma 0.2
(0.2)
Trichoptera
Oxyethira - 1.8
(0.9)
Oecetis -
Polv-
centropus -
Bryozoa * 20%
Coelenterata
Hydra 0.2
(0.2)
Platyhelminthes
Planarlidae 2.8 8.8
(0.7) (3.4)
Mol lusca
Gas tropoda
Amnicola 8.2
(2.2)
Gvraulus 4.0
(1.1)
Phvsa 2.2
(1.2)
* Bryozoa were
colonies.
0.2
0.4
0.6
(0.2)
(0.2)
(0.4)
0.2
0.4
(0.2)
(0.2)
0.4
1.0
2.0
1.2
(1.8)
(1.0)
(0.7)
(0.5)
0.2
0.2
0.4
(0.2)
(0.2)
(0.4)
60% 60% 60% 20% 40%
2.0 2.0 9.2 2.2
(0.8) (0.6) (2.9) (0.9)
3.2
(1.1)
5.0
(2.4)
6.6
(1.7)
9.8
(2.8)
15.6
(5.8)
1.2
(0.7)
3.2
(1.5)
0.4
(0.4)
4.8
(1.8)
6.4
(2.7)
1.2
(0.8)
1.4
(0.6)
1.4
(0.2)
1.6
(0.9)
0.8
(0.8)
quantified as percent of stems with Bryozoan
-------
Table 2B. Dominant macroinvertebrate taxa associated with
M. spicatum stems from the South Bed in Brownington Pond
(July 30 - August 27, 1991). Samples were collected using
the smaller MIS sampler. Data in the table are mean number
of individuals per stem (± 1 S.E.). Five samples were taken
on each date.
Taxon Date
30/7 6/8 13/8 20/8 27/8
Anne 1 ida
Hirudinea 0.2 0.2
(0.2) (0.2)
O ligochaeta 15.2 19.8 14.8 13.6 9.4
(4.7) (3.3) (10.7) (4.0) (5.7)
Arthropoda
Ainphipoda
Hyallela 0.6 0.4 2.0 3.4 4.2
(0.4) (0.2) (0.8) (1.7) (2.0)
Cladocera 2.8 1.8 1.6 4.0 2.4
(1.0) (1.0) (0.4) (1.4) (0.5)
Ostracoda 0.6 1.2 0.4 0.6
(0.2) (1.2) (0.2) (0.3)
Hydracarina 0.2 0.8 0.2
(0.2) (0.4) (0.2)
Insecta
Coleoptera
Euhrychiopsis 1.2 0.2 1.0 2.6 2.4
(0.7) (0.2) (0.8) (1.7) (2.2)
Diptera
Chironomidae 1.4 0.4 0.4 0.2 2.4
(0.5) (0.2) (0.2) (0.2) (1.2)
Heliidae
Lepidoptera
Acentria 0.2 0.2 1.2 1.6 0.4
(0.2) (0.2) (0.6) (0.5) (0.2)
-------
Table 2.. Continued .
Odonata
Anisoptera 0.2 0.6
(0.2) (0.4)
Zygoptera
Enallagma 0.2 0.4
(0.2) (0.2)
Trichoptera
Oxyethira 0.2 1.0 2.6 1.0 0.8
(0.2) (0.6) (0.4) (0.6) (0.8)
Oecetis 0.2 0.8 0.4
(0.2) (0.6) (0.2)
Poly-
centropus 0.2 0.6 0.2 0.2
(0.2) (0.2) (0.2) (0.2)
Bryozoa * 0% 0% 0% 0% 0%
Coelenterata
Hydra
Platyhelminthes
Planariidae 5.8 0.4 6.8 10.4 11.0
(1.8) (0.2) (2.1) (3.8) (5.2)
Mol 1 us ca
Gastropoda
Ainnicola 14.4 9.4 20.4 18.0 13.0
(3.8) (2.9) (3.3) (2.9) (1.9)
Gyraulus 5.0 2.4 3.0 2.6 1.6
(1.0) (1.0) (0.6) (1.5) (0.8)
Phvsa 0.8 0.8 0.6 0.4 1.2
(0.4) (0.6) (0.2) (0.2) (0.8)
* Bryozoa were quantified as percent of stems with Bryozoan
colonies.
-------
Table 3A. Dominant macroinvertebrate taxa associated with
M. spicatum stems from the West Bed in Brownington Pond
(June 18 - July 23, 1991). Samples were collected using the
smaller MIS sampler. Data in the table are mean number of
individuals per stem (± 1 S.E.). Five samples were taken on
each date.
Taxon Date
18/6 25/6 2/7 9/7 16/7 23/7
Anne 1 Ida
Hlrudinea 0.6 0.2 0.2
(0.4) (0.2) (0.2)
Ol lgochaeta 6.4 11.6 3.0 15.0 4.0 2.6
(2.1) (6.0) (1.6) (6.8) (1.4) (1.3)
Arthropoda
Amphipoda
Hyallela 0.2 1.0 0.2
(0.2) (1.0) (0.2)
Cladocera 0.6 0.2 0.2 0.6
(0.2) (0.2) (0.2) (0.6)
Ostracoda 0.2 0.2 0.4 0.2
(0.2) (0.2) (0.2) (0.2)
Hydracarlna 0.6 0.4 0.4 1.2 0.2 0.4
(0.2) (0.2) (0.2) (0.7) (0.2) (0.4)
Insecta
Coleoptera
Euhrychiopsls 0.6 1.0 3.2 1.4 2.0 0.8
(0.2) (0.8) (1.5) (0.5) (1.1) (0.4)
Diptera
Chironomidae 0.4 1.0 1.0 0.4 0.4 0.6
(0.2) (0.8) (0.8) (0.2) (0.2) (0.4)
Heliidae 0.4 1.2
(0.4) (0.7)
Lepidoptera
Acentria 0.6 0.8 0.4 0.2 0.2 0.2
(0.2) (0.2) (0.2) (0.2) (0.2) (0.2)
-------
Table Continued .
Odonata
Anisoptera - 0.2 0.2 0.2
(0.2) (0.2)
Z ygoptera
Enal lagma
Trichoptera
Oxyethira 3.4 8.0 2.2 5.6 2.2
(0.7) (1.6) (1.5) (1.5) (1.0)
Oecetis 0.2 0.2 0.6
(0.2) (0.2) (0.6)
Poly-
centropus
Bryozoa * 0% 0% 20% 40% 20% 40%
Coelenterata
Hydra 0.4 0.6 0.2
(0.2) (0.4) (0.2)
Platyhelminthes
P lanariidae 9.8 6.0 8.4 12.6 7.4 4.6
(4.5) (4.6) (2.4) (4.5) (2.4) (2.3)
Mol lusca
Gas tropoda
Antnico la 36.0 29.8 24.4 35.0 25.0 25.0
(13.1) (7.9) (2.0) (4.4) (7.0) (6.3)
Gvraulus 4.0 1.6 1.2 3.2 0.6 0.6
(0.9) (1.0) (0.7) (1.4) (0.6) (0.2)
Phvsa 0.4 0.2 1.6 0.2 2.8 2.4
(0.2) (0.2) (0.5) (0.2) (1.0) (0.9)
* Bryozoa were quantified as percent of stems with Bryozoan
colonies.
-------
Table 3B. Dominant macroinvertebrate taxa associated with
. spicatum stems from the West Bed in Brownington Pond
(July 30 - August 27, 1991). Samples were collected using
the smaller MIS sampler. Data in the table are mean number
of individuals per stem (± 1 S.E.). Five samples were taken
on each date.
Taxon Date
30/7 6/8 13/8 20/8 27/8
Annelida
Hirudinea
Ollgochaeta 9.0 8.6 13.4 16.4 26.2
(2.3) (2.7) (2.7) (4.6) (13.3)
Arthropoda
Amphipoda
Hyallela 0.2 0.4 1.8 0.6
(0.2) (0.2) (0.6) (0.4)
Cladocera 0.2 0.2 0.6 0.4 0.4
(0.2) (0.2) (0.4) (0.2) (0.4)
Ostracoda 0.2
(0.2)
Hydracarina 0.4 0.2 1.6 0.6 0.6
(0.2) (0.2) (1.0) (0.4) (0.4)
Insecta
Coleoptera
Euhrychiopsis 0.4 1.0 0.6 1.8 1.2
(0.4) (0.8) (0.4) (1.0) (0.7)
Diptera
Chironomidae 1.2 0.2 0.4 1.6 1.6
(0.6) (0.2) (0.2) (0.2) (0.7)
HeliIdae 0.2 0.2
(0.2) (0.2)
Lepidoptera
Acentria 0.6 0.6 2.8 2.2 1.0
(0.4) (0.4) (1.0) (0.2) (0.6)
-------
Table Continued .
Odonata
Anisoptera 0.2 0.2 - 0.2
(0.2) (0.2) (0.2)
Zygoptera
Enal lagma
Trichoptera
Oxyethira 1.6 0.4 0.6 1.8 3.4
(0.5) (0.2) (0.2) (0.5) (1.2)
Oecetis 0.2 0.2 1.4
(0.2) (0.2) (0.2)
Poly-
centropus
Bryozoa * 0% 0% 0% 0% 0%
Coelenterata
Hydra 0.4
(0.4)
Platyhelminthes
Planariidae 3.4 5.0 7.0 16.8 18.6
(1.5) (2.1) (2.4) (4.0) (2.6)
Mo 1 lus ca
Gastropoda
Amnicola 17.8 17.2 19.0 25.2 25.4
(3.1) (5.9) (3.0) (2.2) (5.2)
Gyraulus 1.8 1.0 0.8 3.6 4.0
(0.9) (0.5) (0.4) (1.7) (1.3)
Physa 0.2 0.4 0.6 0.8 0.8
(0.2) (0.2) (0.4) (0.6) (0.4)
* Bryozoa were quantified as percent of stems with Bryozoan
colonies.
-------
Table 4. Dominant macroinvertebrate taxa associated with M.
spicatum in the South Bed in Brownington Pond in 1990.
Samples were collected using the Super Sampler (MIS). Data
in the table are mean number (± 1 S.E.) of individuals per
gram (dry weight) of . spicatum . Four samples were taken
on 27 June; three samples were taken on the other two dates.
Statistical comparisons were made using an ANOVA with
Tukeys test on log transformed data. Means with the same
letter are not significantly different.
Taxon Date
27/6 25/7 8/8
Annel ida
Oligochaeta 2.2 a 21.8 b 2.0 a
(1.1) (9.1) (1.3)
Arthropoda
Amphipoda
Hyallela 2.2 a 10.5 a 3.6 a
(1.4) (5.4) (1.1)
Ostracoda 0.0 a 1.5 a 1.2 a
(0.0) (0.5) (0.6)
Hydracarina 0.5 a 4.5 b 1.3 ab
(0.5) (1.9) (0.5)
Insecta
Coleoptera
Euhrychiopsis 0.0 a 0.6 ab 0.7 b
(0.0) (0.3) (0.2)
Diptera
Chironomidae 19.4 a 15.2 a 7.9 a
(6.8) (7.7) (2.3)
Ephemeroptera
Caenis 26.5 a 2.3 a 1.7 a
(16.9) (1.0) (0.8)
Lepidoptera
Acentria 0.0 a 0.3 a 0.6 a
(0.0) (0.1) (0.3)
-------
Table 4 Continued .
Odonata
Anlsoptera <0.1 a 2.4 a 0.5 a
(<0.1) (1.7) (0.3)
Zygoptera
Enallagma 0.0 a 1.1 a 2.1 a
(0.0) (0.6) (1.1)
Pr Ic ho pt e ra
Oxyethira <0.1 a 5.9 b 1.3 c
(<0.1) (1.3) (0.4)
Platyhelminthes
Planariidae 0.1 a 1.0 a 0.2 a
(0.1) (0.5) (0.1)
Mo lius ca
Gastropoda
Amnicola 15.3 a 28.9 a 11.5 a
(8.8) (10.1) (7.5)
Physa <0.1 a 0.5 b 0.3 ab
(<0.1) (0.2) (0.1)
Planorbidae 1.9 a 3.7 a 3.5 a
(0.7) (0.8) (2.2)
-------
Table 5. Dominant macroinvertebrate taxa associated with M.
spicatum in the West Bed in Brownington Pond in 1990.
Samples were collected using the Super Sampler (MIS). Data
in the table are mean number (± 1 S.E.) of individuals per
gram (dry weight) of ! . spicatum . Three samples were taken
on each date. Statistical comparisons were made using an
ANOVA with Tukeys test on log transformed data. Means with
the same letter are not significantly different.
Taxon Date
11/7 8/8 21/8
Anne 1 Ida
Oligochaeta 0.3 a 4.7 a 10.0 a
(0.2) (2.5) (7.5)
Arthropoda
Amphipoda
Hyallela 2.9 a 4.4 a 18.6 b
(1.5) (0.3) (3.3)
Ostracoda 0.0 a 1.3 ab 3.7 b
(0.0) (0.6) (1.1)
Hydracarina 0.6 a 1.6 ab 3.3 b
(0.2) (0.2) (0.8)
I nsecta
Coleoptera
Euhrychiopsis 0.0 a 0.6 a 0.4 a
(0.0) (0.1) (0.3)
Diptera
Chiroriomidae 5.4 a 14.2 a 12.4 a
(0.4) (4.4) (1.9)
Ephemeroptera
Caenis 0.6 a 0.7 a 1.1 a
(0.4) (0.5) (0.6)
Lepidoptera
Acentria 0.3 a 0.5 a 1.9 b
(0.2) (0.2) (0.4)
-------
Table 5 Continued .
Odonata
Anisoptera 0.1 a 0.8 a 2.2 b
(0.1) (0.2) (0.6)
Zygoptera
Enallagma 0.1 a 3.0 b 2.6 b
(<0.1) (0.8) (0.6)
Trichoptera
Oxyethira 2.6 a 1.7 a 4.6 a
(1.1) (0.4) (1.6)
Platyhelminthes
Planariidae 0.0 a 0.5 a 17.0 b
(0.0) (0.2) (8.2)
Mollusca
Gastropoda
Amnicola 16.1 a 60.8 b 130.8 C
(4.3) (7.4) (6.7)
Physa 0.2 a 1.9 b 4.7 b
(0.1) (0.4) (1.2)
Planorbidae 0.4 a 2.6 b 6.4 c
(0.1) (0.7) (0.9)
-------
Table 6A. Results of the ANOVA for the 1991 super sample
data from Brownington Pond. Samples were collected in mid-
June (12 and 13 June) and 5 August. Samples were collected
in the South and West watermilfoil beds. The ANOVA was
performed on log transformed data. The table presents the
level of significance (i.e., based on p values) for date and
site for the major invertebrate taxa found associated with
watermilfoil. N=5 for all sample collections except the
West Bed in June when N=3.
Taxon Effect
Date Site Date X Site
Oligochaeta ** *
Arthropoda
Crustacea
Ainphipoda
Hyallela
Ostracoda
Hydracarina
Insecta
Euhrychiopsis *
Chironomidae *** ** -
Caenis *** *** *
Acentria *
Parapovnx - - -
Anisoptera ** * -
Enallagma # - **
Oxyethira *** -
Platyhelminthes
Planaria
Mol lusca
Gastropoda
Planorbidae * *
Amnicola *** *** **
Physa *** - -
Significance levels: - not significant, # marginally
significant (p
-------
Table 6B. Abundances of . lecontei and . nivea in the
1991 super samples taken at Brownington Pond. Data are
presented for the two sample dates and the two beds. Values
in the table are mean (± 1 S.E.) number of individuals per
gram dry weight of watermilfoil.
Date
Euhrychiopsis
Acentria
South Bed
West Bed
South Bed
West Bed
12
June
0.88
(0.22)
2.30
(1.09)
0.08
(0.05)
0.65
(0.13)
5
August
1.73
(0.26)
1.95
(0.35)
0.80
(0.20)
0.63
(0.21)
-------
Table 7. Summary of characteristics of the yellow perch
collected in Brownington Pond for the fish diet survey in
1991.
Date Number of
perch collected
Mean
Total Length
(mm)
Mean
Weight
(g)
Mean
Age
(years)
26 June 29
232
157
4.6
3 August 28
220
124
3.7
27 September 13
239
162
4.8
All Dates 70
228
144
4.3
-------
Table 8. The response of the dominant macroinvertebrates
found on . spicatum in Brownington Pond to the exclusion of
fish. Values in the table are mean number of a taxon per
gram of watermilfoil (± 1 S.E.). The treatments were
compared using an ANOVA with planned, orthogonal contrasts.
The Fish contrast in the table is the comparison of fish vs
no fish, i.e., the cage treatment vs the cage control and
the control treatments. The Cage contrast tests for a cage
effect and compares the cage control vs the control. The
ANOVA was performed on log(X+1) transformed data.
Significance levels are as follows: # p<0.1 (marginally
significant), * p<0.05, ** p<0.01, pcO.001. Ct=Control,
Cc=Cage Control, Ca=Cage.
Taxon Treatment
Ct Cc Ca
Contrast
Fish
Cage
Insecta
Coleoptera
Euhrychiopsis 2.45 2.86 1.84
(0.38) (0.02) (0.16)
Diptera
Chironomidae 1.70 2.70 3.08
(0.47) (0.84) (1.23)
Ephemeroptera
Caenis 0.85 0.95 2.85
(0.55) (0.42) (1.71)
Lepidoptera
Acentria 0.58 0.69 0.67
(0.12) (0.09) (0.33)
Parapoynx 0.14 0.14 0.04
(0.05) (0.04) (0.02)
Odonata
Anisoptera 0.21 1.03 2.35 **
(0.09) (0.19) (0.60)
Zygoptera
Enallaqma 3.84 3.99 3.37
(0.55) (0.63) (0.34)
Tn choptera
Oecetis 0.69 0.91 0.58
(0.10) (0.29) (0.08)
-------
Table L Continued .
Taxon Treatment
Ct Cc Ca
Contrast
Fish
Cage
Polycentropus 1.20 0.80 0.87
(0.22) (0.14) (0.11)
Crustacea
Amphipoda
Hyalella 6.47 14.76 14.31 *
(1.10) (2.61) (3.07)
Cladocera 3.95 3.66 2.89
(1.23) (0.98) (0.62)
Hydracarlna 2.14 1.03 1.57 #
(0.64) (0.15) (0.34)
Gastropoda
Amnicola 36.46 24.42 23.49 **
(2.62) (2.24) (4.83)
Physa 4.44 11.88 12.77 *
(0.58) (1.67) (1.01)
Immature
P lartorbidae 2 6.07 3.84 2.90 *
(0.94) (0.66) (0.46)
Platyhelminthes
Planaria 7.09 8.15 5.77
(1.96) (2.31) (1.36)
Oligochaeta 5.55 2.02 1.95
(1.45) (0.29) (0.50)
1 - E. lecontei adults and larvae combined.
2 - Immature Planorbidae consists of Gyraulus sp. and
Helisoma sp.
-------
Table 9. Distribution of eggs laid by ! . sibiricum female
weevils in the weevil oviposition experiment conducted at
Brownington Pond. As the j. spicatum weevils laid very few
eggs, only the results of the M. sibiricum weevils are
shown. N stands for Northern watermilifoil (M. sibiricum )
and E stands for Eurasian watermilfoil ( . spicatum) .
Weevil Date
Number
23/8/91 24/8/91 26/8/91 28/8/91
NWM EWM NWM EWM NWM EWM NWM EWM
1 2 0 5 0 4 0 4 0
2 2 1 1 2 3 1 0 1
3 0 4 0 1 0 4 1 1
4 1 4 0 2 0 5 0 5
5 0 0 0 2 1 0 0 0
6 0 0 0 0 0 0 0 0
Total 5 9 6 7 8 10 5 7
-------
Table 10. The effect of bluegill density on the abundance
of weevils in the Fish Enclosure experiment conducted in
Lake Bomoseen in 1991. Values in the table are the mean
number of individuals of each weevil life stage found in all
eight cages of a given treatment.
Bluegill
(no./c
Density
age)
Weevil Life
Stage
Mean
No./cage
Standard
Deviation
Larvae 4.00 3.63
Pupae 1.13 1.46
0
Adult 3.00 3.85
Total 8.13 7.81
Larvae
6.38
5.80
Pupae
4.13
4.26
2
Adult
3.88
3.78
Total
14.38
13.42
Larvae 1.38 1.41
Pupae 0.75 1.04
6
Adult 1.13 1.36
Total 3.25 2.43
-------
Table hA. The mean proportion (± 1 S.E.) of adult weevils
surviving in the various treatments in Laboratory Experiment
I. There were four replicates of each treatment and there
were eight weevils placed In each aquarium for each trial.
Treatment Proportion
Surviving
Other Invertebrates 0.66 (± 0.12)
Bluegihls Present
Present
Other Invertebrates 0.94 (± 0.06)
Absent
Other Invertebrates 1.00 (± 0.00)
Bluegills Present
Absent
Other Invertebrates 1.00 (± 0.00)
Absent
Table 11B. The mean proportion (± 1 S.E.) of all
invertebrates (adult weevils plus other invertebrates)
surviving in the Bluegill plus Other Invertebrates treatment
of Laboratory Experiment I. There were eight replicates of
this treatment. The numbers in parentheses are the total
number of individuals of each taxa placed in the aquarium at
the beginning of a trial.
Taxon Proportion Surviving
Adult Weevils (8) 0.66 (± 0.12)
Dragonflies (6) 0.42 (± 0.17)
Damseiflies (6) 0.75 (± 0.05)
Amphipods (12) 0.96 (± 0.04)
Snails (6) 0.92 (± 0.05)
-------
Table 12. The ANOVA table for Laboratory Experiment II.
The effect of dragonflies alone, damseif lies alone, and the
interaction of these two taxa together on adult weevil
survival in the presence of foraging bluegills is shown.
See text for additional information.
Source
df
SS
MS
F value
P value
Dragonflies
1
0.2197
0.2917
3.65
0.080
Damseiflies
1
0.0010
0.0010
0.02
0.901
Dragonfly
X Damseifly
1
0.0791
0.0791
1.31
0.274
Error
12
0.7227
0.0602
Total
15
1.0225
-------
Table 13 A. Results of the ANOVA on the stem transect data
collected from Eckley Bay, Neshobe Island and West Castleton
Bay. The response was analyzed for the number of 1) all
weevil life stages combined, 2) weevil eggs and 3) larvae,
pupae and adults combined. This table includes the last two
collection dates in September when few weevils were present.
Type III SS are shown.
1) All weevil life stages
Source
df
SS
MS
F value
P
value
Date
11
438.48
39.86
2.05
0.0401
Site
2
427.95
213.97
11.03
0.0001
Harvest
1
827.66
827.66
42.65
0.0001
Error
55
1067.27
19.41
Total
69
2749.14
2) Weevil
eggs
Source
df
SS
MS
F value
P
value
Date
11
121.65
11.06
1.55
0.1411
Site
2
56.46
28.23
3.95
0.0249
Harvest
1
99.57
99.57
13.94
0.0004
Error
55
392.88
7.14
Total
69
667.27
3) Weevil
larvae,
pupae and
adults
Source
df
SS
MS
F value
P
value
Date 11 169.44 15.40 2.42 0.0161
Site 2 192.25 96.13 15.09 0.0001
Harvest 1 360.65 360.65 56.60 0.0001
Error 53 337.70 6.37
Total 67 1025.94
-------
Table 13 B. Results of the ANOVA on the stem transect data
collected from Eckley Bay, Neshobe Island and West Castleton
Bay. The response was analyzed for the number of 1) all
weevil life stages combined, 2) weevil eggs and 3) larvae,
pupae and adults combined. This table does not include the
last two collection dates in September when few weevils were
present. Type III SS are shown.
1) All weevil life stages
Source
df
SS
MS
F value
P
value
Date
9
196.22
21.08
1.14
0.3544
Site
2
494.27
247.14
12.95
0.0001
Harvest
1
970.82
970.82
50.86
0.0001
Error
45
859.01
19.09
Total
57
2518.62
2) Weevil
eggs
Source
df
SS
MS
F value
P
value
Date
9
85.46
9.50
1.18
0.3280
Site
2
67.86
33.93
4.23
0.0207
Harvest
1
120.41
120.41
15.02
0.0003
Error
45
360.76
8.02
Total
57
632.16
3) Weevil
larvae,
pupae and
adults
Source
df
SS
MS
F value
P
value
Date 9 83.01 9.22 1.58 0.1507
Site 2 224.18 112.09 19.25 0.0001
Harvest 1 422.59 422.59 72.59 0.0001
Error 43 250.34 5.82
Total 55 930.00
-------
Table 14. Results of the ANOVA of the super sample data
from the harvested and unharvested areas at four sites in
Lake Bomoseen in 1991. Samples were collected on 27 June
and 29 July. The four sites were W. Castleton Bay, Neshobe
I., Eckley Bay and Indian Bay. Three samples were taken in
each of the areas at each of the sites on each date except
at Eckley Bay on 27 June. The harvesters had not yet been
through Eckley Bay on 27 June. The ANOVA was performed on
log transformed data. The table presents the level of
significance (i.e., based on p values) of watermilfoil
biomass and major taxa to the effects of date, site or
harvesting.
Taxon Effect
Date Site Harvesting
Milfoil Dry Weight
Invertebrates
Oligochaeta * ***
Arthropoda
Crustacea
Amphipoda * *** -
Isopoda - ** *
Hydracarina ** **
I nsecta
Euhrychiopsis * ** *
Chironomidae *
Caenis ** ***
Acentria - ***
Anisoptera ** **
Zygoptera ** **
Apraylea - *
Oxvethira **
Orthotrichia *
Mol lus ca
Gas tropoda
Planorbidae *** *
Amnicola *** -
Physa *** **
Significance levels: - not significant, # marginally
significant (p
-------
Table 15. Results of the experiment in which adult weevils
(E. lecoritei ) collected from M. sibiricum in Lake Iroquois
were given a choice of . spicatum and . sibiricum . The
data presented in the table under Treatment are the mean (±
1 S.E.) number of leaflets or leaves removed or the number
of bites on the stem. The data were analyzed using an ANOVA
with orthogonal contrasts. Leaflet data were log
transformed, the remaining data were not transformed.
Treatments: C=Control, F=Female weevils, M=Male weevils;
Contrasts: NW vs WN0 weevils vs weevils, F vs M=Females vs
Males. N=4 for F and M treatments; N=3 for the control.
Response
Variable
Treatment
Contrast
C
F
M
NW
vs
W
F
vs
M
Leaflets (Msp)
6.33
(2.33)
109.00
(32.10)
79.00
(16.74)
Leaflets (Msi)
4.00
(1.53)
14.75
(4.59)
8.33
(3.28)
Leaves (Msp)
1.67
(0.67)
2.00
(0.81)
1.67
(0.88)
Leaves (Msi)
0.67
(0.33)
1.75
(0.63)
0.33
(0.33)
#
Stem Bites
(Msp)
0.00
(0.00)
5.50
(1.19)
11.00
(5.56)
#
Stem Bites
(Msi)
0.00
(0.00)
0.50
(0.29)
1.67
(1.67)
Significance levels: - not significant, # marginally
significant (pc0.10), * p
-------
Table 16. Lakes visited in Vermont in 1991 with indications
of presence or absence of selected invertebrates. An X
indicates collection and identification of invertebrates.
An R Indicates a field siting but no collected individuals.
Weevils counts include the eggs, larvae, and pupae of
Euhrychiopsis lecontel and Phvtobius leucogaster and adult
. lecontei .
Lakes Weevils Acentria Parapoynx
Arrowhead Mtn. R
Berlin X X X
Black
Bomoseen X
Brownington X X
Burr
Carmi
Champlain
-McCuen Slang X X
-Shelburne Bay X X
Dunmore 1
Echo
Glen X
Hortonia
Iroquois X R
Little
Loves R
Lower X X
Memphramagog X X
Metcalf X
Mill (Kennedy)
North Montpelier X X
Norton Brook
Paran 2 x
Parsons Mill
Richv ii. le
Round X
St. Catherine X 3
Sunrise
Sunset 1
Winona X X X
No samples taken In 1991 due to little or no milfoil.
2 E lecontei known to be present (VT DEC) , but no
samples taken.
3 unidentified weevil larvae.
-------
Table 17. Comparison of lakes found to contain weevils
and/or caterpillars in 1990 or 1991 samples. An X indicates
collection and identification of invertebrates. An R
indicates a field siting but no collected individuals.
Weevil counts are totals of eggs, larvae, and pupae of
Euhrvchiopsis lecontel or Phytobius leucogaster and adult .
lecontei .
Lakes Weevils Acentria Parapoynx
90 91 90 91 90 91
Arrowhead Mtn. X X R
Berlin X X X X
Bomoseen X X X
Brownington X X X X X
Shelburne Bay X X X X x
Echo X
Glen X X
Lower X X
Memphramagog X X X X
Metcalf X X
Paran 2 x x
St. Catherine X 3
Sunrise X X 3 X X
Sunset X
Winona X X X
No samples taken in 1991 due to little or no milfoil.
2 E lecontei known to be present in 1991 (VT DEC) , but
no samples taken.
3 unidentified weevil larvae.
-------
Table 18. Dominant taxa collected in minisamples during the
multi-lake sampling in Vermont in 1991. Values in the table
are the mean number (± 1 S.E.). Not all of the Vermont
lakes visited were sampled using this method. Five
minisamples were taken In each lake.
Lake
Berlin Bomoseen Brownington
Taxon Mean SE Mean SE Mean SE
Anne 1 Ida
Oligochaeta 0 0 4.6 1.87 2.6 1.43
Arthropoda
Amphipoda 2.8 1.75 0.6 0.36 0.4 0.22
Hydracarlna 0.6 0.22 0 0 0.2 0.18
Insecta
Weevil
Eggs 0.4 0.36 0 0 0.4 0.36
-Larvae 0.4 0.36 0 0 0.4 0.36
-Pupae 0 0 0 0 0 0
-Adults 0.2 0.18 0 0 0.2 0.18
Chironomld 5.8 1.78 9.2 3.93 1.8 1.07
He leidae 0 0 0 0 0 0
Caenis 0.2 0.18 0.4 0.22 0 0
Acentria 0.2 0.18 0 0 0.2 0.18
Enallagma 1.4 0.78 0.2 0.18 0 0
Apraylea 0 0 0 0 0 0
Oxyethlra 0 0 0 0 0.2 0.18
Oecetis 0 0 0 0 0.2 1.08
Triaenodes 0 0 0 0 0 0
Platyhelmin-
thes
Planaria 0 0 0.2 0.18 3.4 1.04
Mollusca
Gastropoda
Planorbidae 0.6 0.36 0 0 10.6 3.21
Amnicola 0.2 0.18 0.2 0.18 37.0 9.00
Physa 0 0 0 0 1.2 0.44
-------
Table 18. Continued .
Lake
Glen Hortonla
Taxon Mean SE Mean SE
Annelida
Oligochaete 8.2 2.99 0.4 0.22
Arthropoda
Amphipoda 0 0 0 0
Hydracarlna 0.4 0.22 0 0
Insecta
Weevil
-Eggs 0 0 0 0
-Larvae 0 0 0 0
-Pupae 0.2 0.17 0 0
-Adults 0 0 0 0
Chironomids 6.8 2.41 9.6 5.10
Heleidae 0 0 0.2 0.18
Caenis 0 0 0 0
Acentria 0 0 0 0
Enallagma 0.2 0.18 0 0
Agraylea 0.4 0.22 0 0
Oxvethira 1.2 0.52 0 0
Oecetis 0.2 0.18 0 0
Triaenodes 0 0 0 0
Platyhelmin-
thes
Planaria 0 0 0 0
Mo 11 us ca
Gas tropoda
Planorbidae 1.6 0.83 0.8 0.33
Amnicola 4.0 0.85 1.0 0.69
Phvsa 0 0 0.4 0.36
-------
Table j . Continued .
Lake
Loves Lower Memphremagog
Taxon Mean SE Mean SE Mean SE
Annelida
Oligochaeta 0.2 0.18 0.8 0.72 11.6 5.25
Arthropoda
Amphipoda 0.8 0.33 1.2 0.52 0.2 0.18
Hydracarina 0.2 0.18 0.4 0.22 0.2 0.18
I nsecta
Weevil
Eggs 0 0 0.6 0.36 0 0
Larvae 0 0 0 0 0.6 0.36
Pupae 0 0 0 0 0.6 0.22
Adults 0 0 0.2 0.18 0.2 0.18
Chironomid 2.4 0.46 5.2 1.68 9.6 1.59
Heleidae 1.0 0.28 0 0 0 0
Caenis 0 0 0 0 0 0
Acentria 0 0 4.6 2.01 2.0 0.57
Enallagma 0.4 0.36 0.4 0.22 0 0
Aaravlea 0 0 2.4 0.92 0.6 0.36
Oxvethira 1.8 1.00 0.4 0.22 0.4 0.22
Oecetis 0.4 0.22 0.2 0.18 0 0
Triaenodes 0 0 1.0 0.49 0.2 0.18
Platyhelmin-
thes
Planaria 0.6 0.36 1.6 1.22 1.6 0.78
Mo lius ca
Gas tropoda
Planorbidae 0 0 1.8 1.18 7.4 3.45
Amnicola 0.6 0.36 1.8 0.66 17.4 5.58
Physa 0 0 0.4 0.36 0.4 0.22
-------
Table Continued .
Lake
Metcalf Mill N. Montepelier
Taxon Mean SE Mean SE Mean SE
Anne 1 Ida
Ollgochaeta 0 0 6.2 8.52 0 0
Arthropoda
Amphlpoda 0 0 0 0 1.6 0.61
Hydracarina 0 0 0 0 0.2 0.18
I nsecta
Weevil
Eggs 0 0 0 0 1.6 0.67
-Larvae 0 0 0 0 0 0
-Pupae 0 0 0 0 0 0
Adults 0 0 0 0 0 0
Chironomid 18.2 3.43 1.2 1.47 74.0 23.40
Heleidae 2.0 0.75 0.2 0.40 0 0
Caenis 0 0 0.2 0.40 0.8 0.33
Acentria 0.2 0.18 0 0 2.2 0.72
Enallagma 0 0 0 0 0.4 0.22
Agraylea 0.2 0.18 0 0 0 0
Oxyethira 0 0 0 0 0.2 0.18
Oecetls 0 0 0 0 0.2 0.18
Trlaenodes 0.4 0.22 0 0 0 0
Platyhelmin-
thes
Planarla 0 0 0 0 0.2 0.18
Mo 1 lu s ca
Gastropoda
Planorbid 0.2 0.18 0 0 0.2 0.18
Ainnicola 0.2 0.18 0 0 7.6 1.34
Physa 0 0 0 0 0 0
-------
Table j Continued .
Lake
Norton Brook Parsons Mill Shelburne
Taxon Mean SE Mean SE Mean SE
Annelida
Oligochaet 0 0 23.8 14.71 17.6 9.43
Arthropoda
Amphipoda 0 0 4.4 3.12 10.2 3.98
Hydracarina 0.2 0.18 0.2 0.18 1.0 0.49
Insecta
Weevil
Eggs 0 0 0 0 0 0
-Larvae 0 0 0 0 0 0
-Pupae 0 0 0 0 0 0
Adults 0 0 0 0 0 0
Chironomid 0.4 0.36 20.4 4.35 1.2 0.66
Heleidae 0.2 0.18 5.0 2.62 0.4 0.36
Caen is 0 0 0.4 0.22 0 0
Acentria 0 0 0 0 0 0
Enallagma 0 0 9.8 2.86 0 0
Agraylea 0 0 0.4 0.36 19.6 4.41
Oxyethira 0 0 0 0 0.6 0.36
Oecetis 0 0 0 0 0 0
Triaenodes 0 0 0 0 0 0
Platyhelmin-
thes
Planaria 0 0 5.8 1.73 0.8 0.52
Mollusca
Gas tropoda
Planorbidae 0 0 3.8 0.77 2.0 0.63
Aninlcola 0 0 2.8 1.25 2.4 0.78
Physa 0 0 0.6 0.36 0 0
-------
Table j Continued .
Lake
St. Catherine Sunrise
Taxon Mean SE Mean SE
Annelida
Ollgochaeta 4.2 1.97 0 0
Arthropoda
Ainphipoda 0 0 0 0
Hydracarina 0.2 0.18 0.6 0.36
I nsecta
Weevil
Eggs 0 0 0 0
-Larvae 0 0 0 0
-Pupae 0 0 0 0
-Adults 0 0 0 0
Chironomid 6.6 0.83 0.2 0.18
Heleidae 0 0 0 0
Caenis 0 0 0 0
Acentria 0 0 0 0
Enallagma 0 0 0 0
Apraylea 0.8 0.18 0 0
Oxvethira 0 0 0.4 0.22
Oecetis 0 0 0 0
Trienodes 0 0 0 0
Platyhelmin-
thes
Planaria 0 0 0.2 0.18
Mo ilus ca
Gas tropoda
Planorbid 0.2 0.18 0.2 0.18
Amnicola 1.0 0.28 0.2 0.18
Phvsa 0 0 0 0
-------
Table 19. Selected taxa collected in top samples during the
multi-lake sampling in Vermont in 1991. Values in the table
are the mean number (± 1 S.E.). Not all of the Vermont
lakes visited were sampled using this method. Five top
samples were taken in each lake.
Lake
Berlin Bomoseen Brownington
Taxon
Mean SE Mean SE Mean SE
Arthropoda
Insecta
Weevil
Eggs 0 0 0 0 1.4 0.61
Larvae 0.8 0.44 0 0 2.8 0.66
-Pupae 0.4 0.22 0 0 0 0
-Adults 0 0 0 0 0.6 0.22
Lepidoptera
Acentria 0.4 0.22 0 0 2.4 0.36
Para ovnx 0.6 0.36 0 0 0 0
Tricopteran
Apraylea 0 0 0.2 0.18 0 0
Oxyethira 0 0 0.2 0.18 0 0
Oecetis 0 0 0 0 0 0
Triaenodes 0.2 0.18 0 0 0 0
Lake
Burr Glen Hortonia
Taxon
Mean SE Mean SE Mean SE
Arthropoda
Insecta
Weevil
Eggs 0 0 0 0 0 0
-Larvae 0 0 0 0 0 0
Pupae 0 0 0.8 0.33 0 0
Adults 0 0 0.2 0.18 0 0
Lepidoptera
Acentria 0 0 0 0 0 0
Parapoynx 0 0 0 0 0 0
Tricopteran
Agraylea 3.2 1.82 0 0 0 0
Oxyethira 1.0 0.69 0.6 0.36 0 0
Oecetis 0 0 0 0 0 0
Triaenodes 0 0 0 0 0 0
-------
Table Continued .
Lepidoptera
Acentria
Parapoynx
Tricopteran
Agraylea
Oxvethira
Oecetis
Triaenodes
0.8
0
0
0 0.2
0.52
0
0
0.18
Taxon
Lepidoptera
Acentria
Parapovnx
Love s
McCuen Slang
Memphremegog
Lake
Little
Lower
Taxon
Arthropoda
Insecta
Weevil
-Eggs
-Larvae
-Pupae
-Adults
Mean SE Mean SE
0
0
0.2
0
1.8
0
0
0
0
0
0
0
2.2
1.0
0.2
0
0
0.18
0
0.52
0
0
0
0
0
0
0
0.52
0.28
0.18
0
Lake
Arthropoda
Insecta
Weevil
-Eggs
-Larvae
-Pupae
-Adults
Mean
SE
Mean SE Mean SE
0
0
0
0
0
0
0
0
0
0
0
0.2
0
0
8.2
0
0
0
Tricopteran
Agraylea
Oxyethira
Oecetis
Triaenodes
0
0
0
0.18
2.20
0
0.83
0.18
0
0
0.4
1.6
1.6
1.2
4.6
0
0.4
0.2
0
0
0
13.4
0.4
0.2
0.22
0.67
1.43
0.52
1.12
0
0.22
0.18
0
0
0
4.46
0.36
0.18
1.6
0.2
0
0
-------
Table j . Continued .
Lake
Metcalf Mill N. Montepelier
Taxon
Mean SE Mean SE Mean SE
Arthropoda
Insecta
Weevil
Eggs 0 0 0 0 0.2 0.18
Larvae 0 0 0 0 0.6 0.36
-Pupae 0 0 0 0 0.4 0.36
Adults 0 0 0 0 1.4 0.46
Lepidoptera
Acentria 1.8 1.00 0 0 8.4 1.85
Parapoynx 0 0 0 0 0 0
Tricopteran
Apraylea 0.2 0.18 0 0 0 0
Oxvethira 0 0 0 0 0 0
Oecetis 0 0 0 0 0.6 0.36
Triaenodes 0 0 0 0 0 0
Lake
Norton Brook Parsons Round
Taxon
Mean SE Mean SE Mean SE
Arthropoda
Insecta
Weevil
Eggs 0 0 0 0 0 0
-Larvae 0 0 0 0 0 0
-Pupae 0 0 0 0 0 0
Adults 0 0 0 0 0.2 0.18
Lepidoptera
Acentria 0 0 0 0 0 0
Parapoynx 0 0 0 0 0 0
Tricopteran
Aaravlea 0 0 0.2 0.18 0 0
Oxyethira 0 0 0.2 0.18 0.2 0.18
Oecetis 0 0 0 0 0 0
Triaenodes 0 0 0 0 0.6 0.54
-------
Table j Continued .
Lake
Shelburne St. Catherine Sunrise
Taxon
Mean SE Mean SE Mean SE
Arthropoda
Insecta
Weevil
Eggs 0 0 0.2 0.18 0 0
Larvae 0.2 0.18 0.4 0.22 0.2 0.18
-Pupae 0 0 0 0 0 0
-Adults 0 0 0 0 0 0
Lepidoptera
Acentria 0.8 0.52 0 0 0 0
Parapoynx 0 0 0 0 0 0
Tricopteran
Agraylea 9.8 2.88 0.2 0.18 0 0
Oxyethira 0.4 0.22 0.2 0.18 0 0
Oecetis 0 0 0.2 0.18 0 0
Triaenodes 0 0 0 0 0 0
Lake
Winona
Taxon
Mean SE
Arthropoda
Insecta
Weevil
Eggs 1.8 0.72
-Larvae 0.8 0.33
Pupae 0.2 0.18
Adults 0.4 0.22
Lepidoptera
Acentria 0.8 0.72
Para ovnx 0.6 0.36
Tricopteran
Aaravlea 0.2 0.18
Oxyethira 1.4 1.25
Oecetis 0.2 0.18
Triaenodes 0.2 0.18
-------
Table 20. A list of the six primary objectives of this study
and the work conducted during the 1991 field season that
addresses the goals of these objectives. As the ideas in
the objectives overlap, some projects are listed under two
or more objectives.
Objective 1. Determine the probable cause(s) of the
Eurasian watermilfoil decline in Brownington Pond.
-all Brownington Pond (BP) research
Objective 2. Examine the grazing/boring effects of all
major herbivores on Eurasian watermilfoil and native aquatic
plant species.
-Wading Pool Experiments (BP)
-Buoyancy Experiment (BP)
-Herbivore Enclosure and Crayfish Experiments (BP)
-Weevil Bag Experiment (Middlebury (M))
Objective 3. Determine the feasibility of herbivore
introductions into other milfoil-infested lakes in Vermont.
-Buoyancy data (BP)
-Weevil culture data (M and Castleton (C))
-Weevil transect data (BP and Lake Bomoseen (LB))
-Harvest/No Harvest data (LB)
-Multi-lake, multi-state survey results
-Choice tests (C)
Objective 4. Determine if Lake Bomoseen is a suitable site
for herbivore introductions/collect pre-introduction base-
line data.
-Bomoseen harvest/no harvest data (LB)
-Fish enclusion experiment (LB)
Objective 5. If determined to be feasible and appropriate
based on previous research (a high likelihood of success and
relatively free from causing negative impacts to non-target
species), use herbivorous insects to control Eurasian
watermilfoil in Lake Bomoseen.
-no research conducted in 1991 which directly addresses
this objective (appears to be happening naturally)
Objective 6. Develop a public education program to keep
Vermonts citizens abreast of the results of the research.
-presentations given by Sheldon and Creed
-slide show on watermilfoil control prepared (M)
-------
Table 21. EquIpment purchased on the EPA grant from June
1991 to April 1992.
Item
Amount
Zenith Laptop Computer $2,128
-------
FIGURE LEGENDS
Figure 1. The distribution of watermilfoil in Brownington
Pond in 1990 and 1991.
Figure 2 A and B. Dissolved oxygen concentrations in
Brownington Pond for 1991. Measurements were taken 0.5 m
below the surface and 0.5 m above the bottom. Values in the
figures are the means (± 1 S.E.) for readings taken at four
sites around the pond. A. Dissolved oxygen values for all
dates. B. Dissolved oxygen values for all dates except
October 30, 1991.
Figure 3 A and B. Water temperatures in Brownington Pond
for 1991. Temperatures were recorded with maximum/minimum
thermometers suspended 0.5 m below the surface and 0.5 m
above the bottom. Values in figures are means (+ 1 S.E.)
f or three pairs of thermometers located around the pond. A.
Surface temperatures. B. Bottom temperatures.
Figure 4 A - C. Results of the plant transects for the West
Bed in 1991. Bars represent the mean (± 1 S.E.) biomass of
watermilfoil or other macrophyte species.
Figure 5 A - C. Results of the plant transects for the
South Bed in 1991. Bars represent the mean (± 1 S.E.)
biomass of watermilfoil or combined native macrophyte
species (=Other).
Figure 6. Maps of the percent cover of Eurasian
watermilf oil in the Middle Grid, West Bed for one date in
1990 and three dates in 1991.
Figure 7. Maps of the percent cover of Eurasian
watermilfoil in the North Grid, West Bed for one date in
1990 and three dates in 1991.
Figure 8. Maps of the percent cover of Eurasian
watermilfoil in the West Grid, South Bed for one date in
1990 and three dates in 1991.
Figure 9. Maps of the percent cover of Eurasian
watermilfoil in the East Grid, South Bed for one date in
1990 and three dates in 1991.
Figure 10 A and B. Results of the stem transects in the
South Bed in Brownington Pond in 1991. The data in the
figure are the mean (± 1 S.E.) number of eggs found
associated with A) watermilfoil stems with intact apical
meristems and B) watermilfoil stems without intact apical
meristems.
-------
Figure 11 A and B. Results of the stem transects in the
South Bed in Brownington Pond in 1991. The data in the
figure are the mean (± 1 S.E.) number of meristem larvae
found associated with A) watermilfoil stems with intact
apical meristems and B) watermilfoil stems without intact
apical meristems.
Figure 12 A and B. Results of the stem transects in the
South Bed in Brownington Pond in 1991. The data in the
figure are the mean (± 1 S.E.) number of stem larvae found
associated with A) watermilfoil stems with intact apical
meristems and B) watermilfoil stems without intact apical
meristems.
Figure 13 A and B. Results of the stem transects in the
West Bed in Brownington Pond in 1991. The data in the
figure are the mean (± 1 S.E.) number of eggs found
associated with A) watermilfoil stems with intact apical
meristems and B) watermilf oil stems without Intact apical
meristems.
Figure 14 A and B. Results of the stem transects in the
West Bed in Brownington Pond in 1991. The data in the
figure are the mean (± 1 S.E.) number of meristem larvae
found associated with A) watermilfoil stems with intact
apical meristems and B) watermilfoil stems without intact
apical meristems.
Figure 15 A and B. Results of the stem transects in the
West Bed in Brownington Pond in 1991. The data in the
figure are the mean (± 1 S.E.) number of stem larvae found
associated with A) watermilfoil stems with intact apical
meristems and B) watermilfoil stems without Intact apical
meristems.
Figure 16 A and B. The abundance of . lecontel (adults and
larvae) in minisamples collected in the South and West Beds
of Brownington Pond in 1991. Bars in histograms are means
(± 1 S.E.).
Figure 17 A and B. The abundance of . nivea larvae in
minisamples collected in the South and West Beds of
Brownington Pond in 1991. Bars in histograms are means (± 1
S.E.).
Figure 18. The diet of large yellow perch collected in the
gill net surveys conducted in Brownington Pond in 1991. The
figure shows the frequency of occurrence of major prey Items
(see text for details) for the June, August and September
collections and all dates combined.
Figure 19. The diet of large yellow perch collected in the
gill net surveys conducted in Brownington Pond in June of
1991.
-------
Figure 20. The diet of large yellow perch collected in the
gill net surveys conducted in Brownington Pond in August of
1991.
Figure 21. The diet of large yellow perch collected in the
gill net surveys conducted in Brownington Pond in September
of 1991.
Figure 22. The diet of large yellow perch collected in the
gill net surveys conducted in Brownington Pond in 1990 (this
was Figure 11 in Creed and Sheldon 1991). The figure shows
the frequency of occurrence of major prey items (FO>5%) for
the July and August collections and both dates combined.
Figure 23. The percent contribution of each yellow perch
age class to the total number of fish caught in the 1991
gill net surveys in Brownington Pond.
Figure 24. The mean total length of each of the age classes
of yellow perch caught in the the 1991 gill net surveys in
Brownington Pond. Numbers above each of the points
represent the number of fish of that age captured.
Figure 25 A - C. The effect of feeding by Acentria larvae
on watermilfoil plants. The bars in the histogram represent
the mean change in a response variable (± 1 S.E.) for each
treatment. The lines with significance values above the
histograms show the results of ANOVA comparisons with
orthogonal contrasts. In each figure, the upper line
represents the comparison of the control vs the Acentria
treatments; the lower line represents the comparison of the
two vs the four Acentria treatment. A. Change in plant
length (in millimeters). B. Change in the number of whorls
per plant. C. Change in plant weight (in grams).
Figure 26. A and B. The effect of feeding by Eurhychiopsis
larvae on the watermilfoil plants. The bars in the
histogram represent the mean change in a response variable
(± 1 S.E.) for each treatment. The lines with significance
values above the histograms show the results of ANOVA
comparisons with orthogonal contrasts. In each figure, the
upper line represents the comparison of the control vs the
weevil treatments; the lower line represents the comparison
of the one vs the two weevil treatment. A. Change in plant
length (in millimeters). B. Change in plant weight (in
grams).
-------
Figure 27. A - C. The effect of feeding by adult .
lecontel collected from . s icatum on northern watermilfoil
plants ( . sibiricuin) . The bars in the histogram represent
the mean change in a response variable (± 1 S.E.) for each
treatment. The lines with significance values above the
histograms show the results of ANOVA comparisons with
orthogonal contrasts. In each figure, the upper line
represents the comparison of the control vs the weevil
treatments; the lower line represents the comparison of the
two vs the four weevil treatment. A. Change in plant length
(in millimeters). B. Change in plant weight (in grams). C.
The number of leaves removed.
Figure 28 A - C. The distribution of leaves removed from .
sibiricum stems by adult weevils for each treatment. The
bars in the histograms represent the total number of leaves
lost from a given whorl for all six replicate plants in each
treatment (maximum number of leaves that could be removed is
between 24 and 30 as M. sibiricum occasionally has five
leaves in a whorl). Whorl position (X axis) denotes
location of leaf whorls on the stem with whorl 1 being the
whorl adjacent to the apical meristem. A. Control (0
weevils) treatment. B. Two weevil treatment. C. Four
weevil treatment.
Figure 29. The effect of herbivores on Eurasian
watermilfoil buoyancy. The bars in the histogram represent
the mean (± 1 S.E.) percent of total watermilfoil weight
found floating in each of the two treatments.
Figure 30. Sites in Lake Bomoseen which were not harvested
in 1991.
Figure 31. The total number of all . lecontel life stages
sampled in harvested and unharvested areas for all three
sites (Neshobe I., W. Castleton Bay and Eckley Bay)
combined.
Figure 32. A comparisom of the total number of all .
lecontei life stages sampled In harvested and unharvested
areas at the three sites. A. Neshobe I. B. W. Castleton
Bay. C. Eckley Bay.
Figure 33. The number of each life stage of . lecontei ,
summed for all sites, for each week. A. Eggs and larvae.
B. Pupae and adults.
Figure 34. Mean (± 1 S.E.) number of days spent in egg,
larval, and pupal stages for . lecontei cultured on i.
spicatum and M. slbirlcum .
-------
Figure 35 A and B. A. The number of . lecontel surviving
the egg, larval and pupal stages. The black bar is for .
lecontel raised on . spicatum (trial 1). The cross hatched
and grey bars are for . lecontel raised on . spicatum and
. sibiricum in Trial 2. 8. The percent of . lecontei
surviving to the next life stage for Trials 1 and 2.
Figure 36. The mean (± 1 S.E.) number of eggs produced by
each virgin weevil female during the first four weeks that
that pair was mated. The data from 7 pairs were used.
Figure 37. The design of the Y-maze which allowed for water
from two separate sources to flow through a chamber unmixed.
-------
FIGURES
-------
Figure 1.
BROWNINGTON POND VT
100%
50%
i o Milfolli
El
0
IN
lOOm
I 1991 miiioiij
100%
El
D
-------
Figure 2.
Brownington Pond 1991
Dissolved Oxygen
12
10
8
Date
B. Brownington Pond 1991
Dissolved Oxygen
1O.5
9.5.
8.5
75.
CO U) CJ 0) U) CJ 0) (0
csJ - - C%J C J e- 1- C%J
. N- N- - - - CO -
(0 N N N- CO CO CO
Date
D Surface
Bottom
U-- Surface
Bottom
18
16
A.
-J
C
4)
14
0
V
4)
0
U)
U)
CO U) OJ 0) U) 01 0) C D 0
01 - . - 01 01 -. i- - 01
-J
C
4)
0
5
4)
0
U)
U)
-------
Figure 3.
A.
30-
C-)
0
20-
a)
I-
a)
a
E
I-
10-
0
30
20
B.
a)
a)
a)
a 10-
E
a)
I.-
0-
Brownington Pond 1991
Surface Temperature
I .I .I.II.I IIIII I
* - I L ) OJ a) ) CIL 0) (0*0
- i- OJ OJ - OJ C )
F F - - (0 - -
co r F F-- (0 (0 (0 0
Date
Brownington Pond 1991
Bottom Temperature
l.u.I.I.I.I.u.I.,.uI I
.* ,- IL) C J 0) IL) OJ 0) (0*0
. ,- C d Cd - ,- Cd ,- C )
Date
g- Maximum
Mnimum
I Maximum
Minimum
-------
Figure 4.
June 20, 1991
WEST BED
Lii
0.5 1.0 1.5 2.0 2.5 3.0 3.5
July
30
I
31, 1991
Iii
/
0.5 1.0 1.5 2.0 2.5 3.0 3.5
A
30
20-
10 -
0
other
D £wtsDIcatum
U
I C
>1
S
0
-C
>1
S
0
IC
>1
S
0
20
B.
C.
10
0
other
M. s catum
other
M.spicatum
September 14, 1991
30
20
10
0
0.5 1.0 1.5 2.0 2.5 3.0 3.5
Depth (m)
-------
SOUTH BED
Figure 5.
June 26, 1991
0.5 1.0 1.5 2.0 2.5 3.0 3.5
September 14, 1991
ilit
I
other
D it spicatum
other
M.sη catum
other
2 M.spicatum
C
w
I-
0
A
B.
C.
August 7, 1991
30
20 -
10-
0-
30
20
10
0
30-
20-
10-
0-
>1
0
>1
0
0.5 1.0 1.5 2.0 2.5 3.0 3.5
0.5 1.0 1.5 2.0 2.5 3.0 3.5
Depth (m)
-------
Figure 6.
West Bed, Middle Grid
September 9, 1990
A
E
Al
E
July 25, 1991
A
June 21, 1991
3 4
2
<25% Cover
25-50% Cover
50-75% Cover
>75% Cover
August 26, 1991
2 3 4 Al 2 3 4
1
E
I,
,
E
U
-------
West Bed, North Grid
Figure 7.
September 9, 1990
2
3
4
July 24, i99i
Al 2 3 4
!!TT c.:::
I
Al
E
June 21,1991
2
3
::
<25 Cover
25-50 Cover
50-75X Cover
>75 Cover
A
August 26, 1991
1:!;
Al
4
- -
E
I.
c.
I 4 i J
:
..)
E
E
-------
1 :.:::TE*::j
:E:
South Bed, West Grid
Figure 8.
September 9, 1990
June 18, 1991
A
2 3 4
A
E
E
<25% Cover
25-50% Cover
50-75% Cover
>75% Cover
July 25, 1991
Al
2 3 4
August 26, 1991
41 2 3 4
E
E
-------
Figure 9.
South Bed, Eest Grid
September 9, 1990
June 18, 1991
A
<25% Cover
25-50% Cover
50-75% Cover
75% Cover
July 26, 1991
A
August 26, 1991
Al 2
E
E
A
E
E
-------
Figure 10.
South Bed
Date
South Bed
No apical meristem
ii i
/
A
apical meristem
4
3
2
0
T
E
C l)
a
U)
w
0
L.
.0
E
z
E
Cl)
I-
a
U)
a)
a)
w
0
.0
E
z
(0(0(0 N. N. N. N. (0(0(0(0 O 0
-S -S S s
-(0 in C J 0 ) C D Ψ CD C ) -
- c J c J I.- C J - CsJ ,-
C)
B
3.
2
1
0
4
(0 (0 (0 N. N. N. N. (0CO(0(00)0
. . 5. s s -S 5
-(0 U) C J 0) (0 j CD C )
,- c,4 CU - - CU i- CU i- -
C ,)
Date
-------
Figure 11.
A South Bed
2.0
apical meristem
1.8
1.6
1.4
4)
m...
tCl) 1.2
4)
1.0
0.8
0.6
.L hI L
Z 0.4
0.2
0.0
CO C D (0 1- F - F F . (0(0(0(0 O 0
. - - -S S
(0 ifl CsJ O CD CsJ CD C)
,- OJ C.J i- ,- C J CsJ - -
C v)
Date
B South Bed
1.0-
No apical meristem
0.9 -
0.8 -
E
0.7-
0.6-
4)
0. 0.5-
04)
0.4-
E. i 0.3-
z o.2 -T I
0.1.1 H __
0.0- .
CO C D (OF F - F- F- (0(0(0(0 d 0
- - S - S S. 5 S S
-COin OJ 0) CD j CO C)
c J c J ,- 1 C J ,- C.J
Cv)
Date
-------
Figure 12.
apical meristem
South Bed
Date
South Bed
No apical meristem
nnρ ρ
A
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3 -
0.1 -
0.0
I
EE
C l)
r4I-
it
E
II
B
2.50-
2.25 -
2.00 -
1.75-
1.50-
1.25-
1.00-
0.75
0.50-
0.25-
0.00-
o ooco a 0
- . - . - , - ... -
- c..lJ c J e- - C 1 i- C.J ,-
C)
Date
-------
Figure 13.
apical meristem
West Bed
Date
West Bed
No apical meristem
ii
ii
I
(0(0 (ON. N. N. N. CD
. - - - . ,
i-CO NO) (0 N
N N - - N
0
/
4
/
CO CD CO 010
, ,.- -
0) CD C)
N - -
C ,)
A
3.
E
w
Cl)
w
0.
0,
w
P4
0
.0
E
z
2
1_TI
(0(0 (ON. N.
- - - -
-CO N
-N N
N. N. CD CD CD CD 0)0
..- - -.- .- . -
0) C D N 0) CD C )
i-N i-N--
C)
B
E
Cl)
I-
a
a )
w
0
.0
E
z
1.4
1.2
1.0.
0.8
0.6
0.4
0.2
0.0
I
0
4
Date
-------
Figure 14.
West Bed
Date
West Bed
No apical meristem
liii
I IL
(0(0 CON N N N (0(0(0(0 C) 0
. , .- .- .- . . -... ,. .. . I-
u, C J C) C D CO C)
I- C%J C sJ i C i C J
C,)
Date
A
apical meristem
E
(0(0(0 N N- N. N- (0(0(0(0 a) 0
% . .-. - . - -
-(0 fl C.J 0 ) (0 ( j 0) (0 C)
- C%J C%J - - CsJ - CsJ ,- -
C !)
1.50
1.25
1.00
0.75
0.50
0.25
0.00
B
1.0
0.9
0.8
0.7
0.6
0.5
O.4
O.3
0.2
0.1
0.0
E
LI
z
/
4
-------
Figure 15.
A
West Bed
1.0
0.9
apical meristem
0.8
E
E
4)
11
II
III
61
I
(0(0(0 F N- N. N- (0(0
- . -. .- -.- -
CO u, CsJ 0 ) C D j
cg -
(0(00)
..- -
CO C )
,- CJ -
0
C v)
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
B
1.50-
1.25-
1.00 -
0.75
0.50
0.25
0.00
Date
West Bed
No apical meristem
EE
11% 4)
ji
/
/
0
4
/
A
4
/
I
CO 00 00 ) 0
- - - . -
CJ 0 ) (0O)1
,- c. J ,- -
Cv)
D ate
-------
Figure 16.
5-
4-
3-
2-
1
0-
South Bed
E
C l)
01)
0
a
E
z
C
a)
a)
B. West Bed
E
a)
C l)
a)
0.
01)
a)
a)
0
I-
a)
a
E
z
C
a)
a)
Date
A.
5
4
3
I
o r. r r . r Co
It) o (0 C ) 0 co 0 )0
,- C J w- C J C) C J C 1
Date
-------
Figure 17.
South Bed
Date
West Bed
(0(0 N- N- N- N- N- (0(0(0(0
..-. - . - ,... -. - , . .
(0 fl c O (0 C ) O p C) ON-
,- C J .- c 1 C) ,- CJ
A.
B.
C D CD N- N- N- N- N- (0 (0 (0 (0
- .
CO I L) C%J (0 C) 0 C) ON-
,- - c j C) .- csj csJ
E
0)
C l)
4)
0. 2.0
0)
C
4)
U
01.0
I-
0)
.0
E
z
C
0) 0.0
0)
E
4)
C l)
I-
0)
0.
0)
L.
C
0)
U
4
I .-
0
0)
.0
E
z
C
0)
a)
4-
3-
2-
T
1
0
I
Date
-------
Figure 18.
100
80
60
40
20
0
w
0
z
w
cr
0
0
0
IL
0
a
w
IL
0
F
0
PREY GROUP
-------
Figure 19.
JUNE
LU
0
z
LU
0
0
0
U-
0
0
LU
LL
100
80
60
40
20
0
PREY TYPE
PREY WITH FO < 20% EXCLUDED
-------
Figure 20.
AUGUST
100
w
C)
z
w
0
C)
0
IL
0
0
w
IL
80
60
40
20
0
PREY TYPE
PREY WITH FO < 20% EXCLUDED
-------
Figure 21.
SEPTEMBER
w
0
z
w
0
0
0
U-
0
0
w
IL
100
80
60
40
20
0
0
0
PREY TYPE
PREY WITH FO < 20% EXCLUDED
-------
Figure 22.
100
Ui 80
0
Z
U i
6O - _____
S
O DATE (SAMPLE SIZE)
0
o 1JULY26 (15)
40 - ::.:. * AUG 24 (58)
2 JULY&AUG(73)
L i.. 20 - - - :..: * -
, / / / , /
PREY TYPE
PREY WITH FO < 5% EXCLUDED
-------
Figure 23.
AGE4
AGE7
AGE6
PERCENT OF FISH IN EACH AGE CLASS (TOTAL N=70)
AGE5
-------
Figure 24.
320 ... . ... ...
N=2
300
E280 ___
N=3
I
260 N 26 I 95% Ci
240 N 22 MEAN
i-. 220
N=17
200 ._...*.....
180
3 4 5 6 7
AGE CLASS
-------
p
-------
n.s.
Figure 26.
Number of Weevil Larvae
n.s.
ns.
Number of Weevil Larvae
A.
p=O.05
0 1 2
a,
C
0
-j
C
0
a.
C
0
a,
C
0
C.)
a,
0
C
0
a.
C
0
a,
C
0
C.)
20
10
0
B.
0.2 -
0.1 -
0.0 -
0 1 2
-------
Figure 27 A&B.
n.s.
n.s.
Number of Weevils
Number of Weevils
n.s.
A
n.s.
I
I C
C
-J
C
a.
C
C
0
IC
C)
40
30
20
10
0
0.3 -
0 2
4
B
0.2-
IC
C
0
a.
C
0
C
0
IC
CI )
0.1
0 2
4
-------
0 C
70 p
-------
Figure 28.
30
1 A. 0 Weevils
25
1
15
-j
10 ILUL.LIhulI iiII
ocow)cor
z
Whorl Number
30-
1 B. 2 Weevils
j i: _ __ I,IttI1I,i,I,IIitIII,1,tiL , i_r .,.. t
z
Whorl Number
30- i
C. 4 Weevils
1 20
i 1: IIhIIlbiiiiii. 11 1 11 . 41 Iii
z
C )CV)CV)C )
Whorl Number
-------
Figure 29.
0
E
c
0
LI
C
Q
C.
100
80
60
40
20
0
Control Herbivores
Treatment
-------
Figure 30.
WEST
BAY
ECKLEY BAY
NESHOBE ISLAND
NO HARVEST SITES
LAKE BOMOSEEN
IN
INDIAN DAY
-------
Figure 31.
WEEVILS IN HARVESTED AND UNHARVESTED AREAS
50
HARVEST
40 NOHARVEST
U
U.
030
U
l:.j
c..J c J - - c J 1 C J
r . r- Co Co
-------
Figure 32.
NESHOBE ISLAND A
40
harves*
30
LL.
0
20
N N i--N - N
F.. - .-
r . - r- F.- 0
WEST CASTLETON BAY B
40
U)
- N N
P.- 0 - G -.
P.- F.- F.- 0 0 0
ECKLEY BAY C
40
U)
-a
30
IL l
I
0
20
L i i
2
1 .i. .1. .i. .I. II.
N N - - N N
F- 0 - . - - 0 0 -
F F.- F.- 0 0 0 0 0
-------
Figure 33.
-0--
LARVAE
fr- - PUPAE
ADULTS
A
B
C ) U) N 4)
a) 14) 0)
4) c J c J U)
30
20
10
0
10
8
6
4.
2
w
C,
lii
a.
U i
z
-J
I-
0
I-
U i
C,
I- ,
Ui
z
-l
I-
0
I-
0
C ) U) r in
CI)C1)
_) -) -) _)
-------
Figure 34.
4)
C l)
Trial I
20
M.spicatum
o TrIal 2
ψj 10 tspicatum
.0
E
z
C ______________________________
0
Eggs Larvae Pi ae
Stage
-------
Figure 35.
A.
B.
C
C l)
.0
E
z
C l) 60-
I C
U
Lu
C
Cl)
C
U
L.
C.
40-
20
0-
Stage
Trial 1
Mspicatum
TrIal 2
M.spicatum
O Ms incum
Trial I
M.spcatum
Trial 2
Pvtsp catum
E:: M sibhicum
30
20
eggs
larvae
I
adults
_ I
eggs->larvae larvae-> pupae pupae->adults
Stage
-------
Figure 36.
20 -
18 -
16 -
14 -
12 -
10 -
8
6-
4-
2
0
E
a)
V
a)
U
I D
0
Q.
a)
Lu
0
a)
.
E
z
C
a)
a)
I
r
1 2 3 4
Week Number
-------
Figure 37.
TUBES IN WHICH THE WATER
SAMPLES ENTER THE MAZE
ARMS OF THE MAZE
AREA WHERE THE ORGANISM
INTRODUCED INTO THE MAZE
LEG OF THE MAZE
MESH GATE TO STOP THE
ORGANISM FROM SWIMMING
OUT OF THE MAZE
SCREW CLAMP TO REGULATE
RATE OF WATER LEAVING THE MAZE
DIRECTION OF WATER
4.5
11cm
STRING
6.5
SIRING
DIRECTION OF WATER
------- |