The geographic distribution of the aquatic milfoil weevil (Euhrychiopsis lecontei)

and factors influencing its density in Wisconsin lakes

BY

LAURAL. JESTER

Wisconsin Cooperative Fishery Research Unit

A Thesis

Submitted in partial fulfillment of the

Requirements for the degree of

MASTER OF SCIENCE

In

Natural Resources

College of Natural Resources

University of Wisconsin

Stevens Point, Wisconsin

October 1998 APPROVED BY THE GRADUATE COMMITTEE OF:

. of Fisheries Committee Chair

Dr. Robert Freckmann Professor of Biology

-~ vJ, ft.df!izc,ytko Professor of Water Resources

Mr. Daniel Helsel Wisconsin Department of Natural Resources ABSTRACT

The exotic aquatic macrophyte, Eurasian ,._,atermilfoil (Myriophyllum spicatum L.) has invaded waterbodies throughout the United States and Canada and become a nuisance plant. Euhrychiopsis lecontei (Dietz), a native aquatic weevil has been associated with natural declines ofM spicatum and has the potential to be used as a biological control agent. The objectives of this study were I) to determine the geographic distribution of E. lecontei in Wisconsin lakes with M spicatum, and 2) to assess geographical and limnological characteristics associated with E. lecontei presence or absence and abundance.

M spicatum beds in 46 lakes were surveyed to determine the presence or absence of E. lecontei.

E. lecontei distribution was found to be widespread across the state with 45 new records identified in lakes with M spicatum.

Relations between lake and macroph)te characteristics and E. lecontei abundance were assessed in 31 lakes. E. lecontei densities were determined by collecting a total of 120 milfoil stems per lake and counting E. lecontei on each stem. E. lecontei densities varied from 0.0 to 2.5 weevils per stem across the lakes and varied among individual M spicatum beds within the lakes. Most geographical and limnological variables were collected at the lake level while other habitat variables were collected within the M spicatum beds (i.e., at the bed level). Only some bed level variables correlated with E. lecontei density including the distance from shore and depth of the milfoil beds, the number of apical tips per milfoil plant, the percentage of broken apical tips per milfoil stem and the percentage of natural shoreline adjacent to the milfoil bed.

The widespread distribution of E. lecontei and its herbivory on M spicatum make it an excellent organism to evaluate as a biological control agent. Factors influencing its density and predator-prey dynamics could provide insight into their potential as biological control agents for M spicatum.

iii ACKNOWLEDGEMENTS

I would like to thank the Wisconsin Department of Natural Resources (WDNR) and the following lake groups and associations for their funding of this study: Beaver Dam Lake, Big Sand Lake, Eagle

Lake, Gilbert Lake, Kangaroo Lake, Kusel Lake, LongTrade Lake, Lorraine Lake, Lower Spring Lake,

Nancy Lake, Pearl Lake, Waukesha County Parks System, and Whitewater Lake. Beyond their monetary support, the WDNR and the lake groups donated time, effort, and general support to the study. Their continuous involvement and generosity are greatly appreciated! Scholarships and research grants were also received from the 12 Apostles Musky Club and the Midwest Aquatic Plant Management Society and are also appreciated.

I would like to acknowledge the great support of the Wisconsin Cooperative Fishery Research

Unit and their donation of staff, resources and equipment. The boat may have some bullet holes and an old motor, but it gratefully got me across the lakes and lasted through two summers!

I heartily thank my graduate advisor, Dr. Michael Bozek, for his complete guidance, patience, praise, encouragement and good nature throughout this entire study and the graduate process! I cannot imagine working for a more thoughtful and devoted professional. I hope that I am lucky enough to work with such an ethical person again.

Another great thank you goes to Mr. Daniel Helsel for initiating this study and making it happen, suggesting that I apply for the position, and thoroughly encouraging me once I was accepted! Thank you for sitting on my committee even though you had to be forced. Thank you for remaining by my side, making me smile, helping me SCUBA dive, teaching me to fly fish and acting as my mentor. I look forward to many more years of your mentoring and friendship.

I also thank Drs. Robert Frcckmann, Stanley Szcz}tko, Ronald Crunkilton, and Michael Hansen in the College of Natural Resources, Dr. Sallie Sheldon at Middlebury College and Mr. Richard Lillie with the WDNR. Their guidance, help and patience on this study and their review of publications is much appreciated! It was a pleasure working with each of you.

iv Thank you also to the technicians who helped me over the past two summers including Todd

Johnson, Tracy Stephens, Peter Jester, Jennifer Zander and Michael Newbrey. I especially want to thank

Todd for his complete devotion to this project and all of his hard work. Thank you for all your long hours through hot days and long truck rides and thank you for keeping a smile on my face through it all!

Technicians and friends like you do not come along every day and I feel privileged to have worked with you. Thank you also for putting up with me for two summers in a row - how you did it, I'll never know!

I also want to thank the many graduate students in the College of Natural Resources whose comradery and sympathy during the long days and nights of graduate school were essential to my mental health. I can't imagine a better group of people anywhere on earth and ,...-ill miss everyone dearly.

And last, but certainly not least, I want to thank my husband, Steve and my family for their support, patience and understanding! I would not have survived the trials and tribulations of my study and studies had it not been for Steve's constant encouragement and praise. I thank him from the bottom of my heart for never doubting our decision to put our lives on hold so that I could complete this degree.

I promise this will be the last of our days and weeks apart!

V TABLE OF CONTENTS

TITLE PAGE ......

ABSTRACT...... Ill

ACKNOWLEDGEMENTS...... 1v

TABLE OF CONTENTS...... v1

LIST OF TABLES...... vm

LIST OF FIGURES...... 1x

LIST OF APPENDICES...... x

INTRODUCTION...... 1

General...... 1

M spicatum Invasion and Ecology ...... 2

Effect ofM spicatum on Aquatic Systems and Man ...... 4

M spicatum Control Methods ...... 6

E. lecontei Life History ...... 10

E. lecontei Distribution and Natural M spicatum Declines ...... 1 I

Study Objectives ...... I 6

METHODS...... 17

E. lecontei Distribution ...... 17

E. lecontei Densities and Associated Geographical and Limnological Characteristics ...... 18

E. lecontei densities ...... 18

Sample processing ...... 20

M spicatum density and biomass sampling ...... 20

vi Data Analysis ...... 23

E. lecontei distribution ...... 23

E. lecontei densities ...... 23

Correlations with E. lecontei densities ...... 23

RESULTS ...... 29

E. lecontei Distribution ...... 29

E. lecontei Densities and Correlated Variables ...... 33

DISCUSSION ...... 42

E. lecontei Distribution ...... 42

E. lecontei Life History Observations ...... 44

Factors Influencing E. lecontei Densities ...... 45

Bed level variables ...... 46

Conclusion ...... 51

Summary...... 52

LITERATURE CITED...... 53

vii LIST OF TABLES

Table 1. Macrophyte species used to quantify potential effects of E. lecontei on native species in previous studies (3 Sheldon and Creed 1995 and bSolarz and Newman 1996) ...... 9

Table 2. Lakes used in M spicatum density and biomass sampling ...... 21

Table 3. Dependent variables collected during E. lecontei density sampling ..... 25

Table 4. Independent variables tested for correlation at the lake level ...... 26

Table 5. Independent variables tested for correlation at the bed level ...... 27

Table 6. New records of E. lecontei in Wisconsin ...... 31

Table 7. E. lecontei densities in Wisconsin lakes ...... 35

Table 8. E. lecontei densities in different M. spicatum beds in 31 Wisconsin lakes ...... 36

Table 9. Spearman rank correlations between percent of E. lecontei lifestages and characteristics measured at the whole lake level ...... 37

Table 10. Spearman rank correlations between E. lecontei density at the bed level and characterisitics of macrophyte beds measured at the bed level ...... 39

Table 11. Spearman rank correlations between percent of E. lecontei lifestages and characteristics measured at the bed level ...... 40

Table 12. Spearman rank correlations between percentage of broken apical tips on M spicatum stems collected during E. lecontei density sampling and various bed level and lake level characteristics ...... 4 I

viii LIST OF FIGURES

Figure 1. Adult E. lecontei ...... 13

Figure 2. E. lecontei egg on apical tip ofM spicatum...... 13

Figure 3. First instar of E. lecontei larva in apical tip ofM spicatum ...... 14

Figure 4. Late instar of E. lecontei larva in M spicatum stem ...... 14

Figure 5. Stem of M spicatum with E. lecontei damage ...... 15

Figure 6. E. lecontei pupa in M spicatum stem ...... 15

Figure 7. Diagram depicting E. lecontei density sampling design on a hypothetical lake ...... 28

Figure 8. Known distribution of E. lecontei in Wisconsin (1997) ...... 30

Figure 9. Density of E. lecontei in Wisconsin lakes (1996 and 1997) ...... 34

ix LIST OF APPENDICES

Appendix A. Independent variables tested for correlation with E. lecontei densities at the lake level ...... • 59

Appendix B. Independent variables tested for correlation with E. lecontei densities at the bed level ...... 61

Appendix C. Raw data oflake level variables used in Spearman correlation analyses...... 63

Appendix D. Raw data of bed level variables used in Spearman correlation analyses ...... 66

Appendix E. Macrophytes other than M spicatum collected in study lakes during milfoil biomass sampling ...... 77

X INTRODUCTION

General

Recent interest in using the aquatic weevil, F.11hrychiopsis lecontei (Dietz) (= Euhrychiopsis lecontei) to control the exotic aquatic macroph~te Eurasian watermilfoil (Myriophyllum spicat11m L.) has increased the need to better understand the weevil's geographical distribution and abundance, life history requirements, and ecological niche. This herbivorous weevil, knmm to feed and reproduce on various species in the genus Myriophyllum, can significantly reduce the standing biomass of M spicatum by removing vascular tissue, causing a loss of stem buoyancy, and destroying apical growing tips (Creed et al.

1992, Creed and Sheldon 1993, 1995, Sheldon and Creed 1995, Ne,,man et al. 1996b ). F.. lecontei has been associated with episodic declines ofM spicatum in some Wisconsin, Illinois and Vermont lakes

(Creed and Sheldon 1991, Kirschner 1995, Lillie and Helsel 1997).

E. lecontei is native to North America and is thought to be widespread across the northern United

States and Canada, having been reported from Wisconsin, Illinois, Iowa, Ohio, Minnesota, Washington,

Michigan, Connecticut, Massachusetts, New York, Vern1ont, Alberta, British Columbia, Ontario and

Saskatchewan (Nemnan and Maher 1995, Sheldon and O'Bryan 1996, Creed 1998, Sheldon, Middlebury

College pers. comm.). However, specific distribution patterns in most states and provinces are unkno\\n.

In Wisconsin, it has previously been reported from only four lakes: Bierbrauer Pond in St. Croix County,

Wingra and Fish Lake in Dane County, and Devil's Lake in Sauk County (Lillie 1991, Ne\\man and Maher

1995, Lillie and Helsel 1997). Newman and Maher ( 1995) sampled additional Wisconsin lakes (Lake

Mallalieu, St. Croix Co., Lake Onalaska, LaCrosse Co., and Spring Lake, Buffalo Co.) for E. lecontei but failed to find the weevil.

1 This research was done to better document the geographic distribution of E. lecontei in the state, assess limnological and geographical characteristics associated with its abundance, and evaluate the effectiveness of stocking E. lecontei as a practical management tool for controlling nuisance M spicatum.

M spicatum Invasion and Ecologv

The submersed aquatic macrophytc Eurasian watermilfoil (Myriophyllum spicatum L.) was introduced into North American lakes in the early 20th century from Europe and Asia (Couch and Nelson

1985). Since that time, it has spread to lakes, ponds and rivers and has now been recorded in at least 40 states and three Canadian provinces (Sheldon and Creed 1995). M spicatum was first found in southeastern Wisconsin in the I 960's and has since been reported in lakes located in 39 of Wisconsin's 72 counties and it continues to spread. In Wisconsin it is also found in Lake Michigan, Lake Superior and the

Mississippi River (Bode et al. 1993).

Anthropogenic activities are the primary vector of M spicatum dispersal among lakes throughout

North America (Couch and Nelson 1985). Historically, aquarium traders, worm farmers, and fishermen were known to use and transplant M spicatum among lakes and ponds (Couch and Nelson 1985). Today, boaters and other recreationalists continue to inadvertently spread this exotic by transporting stem fragments on boat trailers, boat propellers, anchors, SCUBA gear and other recreational equipment (Recd

1977,Johnstoneetal.1985,Newroth 1985,Engcl 1994).

Physiological characteristics of M spicatum facilitate its rapid invasion and its ability to dominate macrophyte beds in littoral regions of many lakes. Vegetative reproduction of M spicatum through autofragmentation and stolon formation is thought to be the most significant means of dispersal throughout a lake (Nichols and Shaw 1986, Madsen ct al. 1988). After flowering in early summer and again in late summer, autofragmenting shoots develop adventitious roots. After abscising from the main stem, these fragments settle to the bottom and grow as independent plants (Kimbel 1982). Dispersal is augmented

2 through wind and wave action that carry fragmented stems great distances to colonize new areas.

Fragments can float in the water and stay viable for several weeks (Rawson 1985).

In addition to efficient and aggressive reproductive capabilities, M spicatum possesses other competitive adaptations which make it an effective invader. First, M spicatum is opportunistic in the uptake of essential nutrients, nitrogen and phosphorus. Although sediment is the primary source of nutrients (Smith and Barko 1990), it has the ability to utilize nutrients from both sediments and directly from water without being solely dependent on either (Nichols and Shmv 1986). Second, Af spicatum has the ability to use bicarbonate and recaptured carbon dioxide as carbon sources. The stems have a large lacuna} system, which acts as an internal gas reservoir. This allows it to retain respired carbon dioxide and permits diffusive gas exchange between the roots and shoots (Grace and Wetzel 1978). Third, M spicatum is a perennial and it overwinters under the ice as an entire plant, often with green shoots (Recd

1977, Kimbel I 982). As a result, in early spring, the root crm"n of M spicatum grows quickly before other species have had a chance to get started and the plants become well established by April (Aiken et al.

I 979). In addition to an early start, gro\\th is rapid and stands can be extremely dense. Recd ( 1977) reported that summer growth can reach a rate of 5-7 centimeters per day. M spicatum gro,vs at depths from 1-10 meters, often surfacing and fom1ing a dense canopy of entangled branches OR the water's surface. Stands can become so dense that the canopy can support the weight of frogs and wading birds and can shade out other plant species attempting to grow below (Aiken et al. 1979). Finally, M spicatum has a broad range of tolerances for a variety of water quality conditions and sediment types, although it prefers hard, nutrient-rich, alkaline waters and fine sediments with an organic content of I 0-25% (Recd 1977,

Aiken et al. 1979).

3 Effect ofM spicatum on Aquatic Svstems and Man

The aggressive and competitive nature of M spicatum alters aquatic communities. M spicatum can inhibit the growth of native plant species so that it dominates macrophyte communities often within two to three years after introduction and can even colonize previously unvegetated areas (Aiken et al. 1979).

According to Aiken et al. ( 1979), the only aquatic plant that is knO\m to successfully compete with M spicatum is Hydrilla verticilata (L. f.) Royle, another exotic. Although there is disagreement among researchers on whether disturbance is required for an Eurasian ,vatermilfoil invasion to be successful (see

Smith and Barko 1990), Madsen et al. ( 1991) documented the intrusion of M spicatum into a healthy and diverse native macroph~1e community in Lake George, New York. The study showed that within a dense macrophyte bed, percent coverage increased from 30% in 1987 to near 100% by 1989, while abundance of

M spicatum within the bed increased from 15% in 1987 to over 95% in 1989. In addition, macrophyte species richness in the bed decreased linearly over time, from 20 in 1987, to 14 in 1988 and 9 in 1989, documenting that M spicatum had significantly suppressed native macrophyte grov,rth in three years

(Madsen et al. I 991 ). Lillie and Mason (I 986) also reported a change in the macroph~1e community of

Devils Lake, Wisconsin with a large decline in Elodea canadensis Michx. concurrent with a significant increase in M spicatum between 1974 and 1984.

The effect of M spicatum on invertebrate communities has been equivocal. Hanson ( 1990) reported that plant beds composed of different macroph~1e species differ in the diversity and abundance of invertebrates inhabiting them. However, there is little evidence showing whether M spicatum increases or decreases the diversity and abundance of invertebrates compared with other submersed macroph~1es. Cyr and Downing ( 1988) found that the abundance of invertebrates was not related to the degree of leaf dissection and they reported no significant difference in various invertebrate abundances bet,vecn

Myriophyllum spp. and Potamogeton amplifolius Tuckerm.. Chilton (1990) also concluded that significant differences in invertebrate abundance did not exist among plant species (Vallisneria americana

4 Michx., Myriophyllum spicat11m. and Ceratophyllum demersum L.). In contrast, Keast (1984) reported that five major zoobenthos taxa were three to seven times as abundant in a Potamogeton-Vallisneria community compared to aM spicatum community. In addition, total numbers of insects emerging from the surface were t\\·ice as high over the Potamogeton-Vallisneria bed than the M spicah1m bed (Keast

1984). Pardue and Webb (1985) compared maeroinvertebrates between M spicatum beds and unvegetated portions of the littoral zone and found that although species richness of invertebrates was higher in milfoil areas, invertebrate biomass was greater in open water habitat. They also reported that one of the most important fish food invertebrates, Hexagenia hi!ineata, showed a preference for unvegetated littoral regions. However, overall differences between the M spicat11m community and the unvegetated littoral zone were not significant (Pardue and Webb 1985).

Dense, monotypic stands of M spicat11m may also alter fish communities by changing the structure and density of macrophytes which influence predator-prey interactions (Crowder and Cooper 1982, Savino and Stein 1982, Diehl 1988, Dionne and Folt 1991 ). Dense Eurasian watermilfoil beds can reduce the open-water areas bet\,,een plants which give larger fish access to prey within the macroph~te beds (Engel

1994). Engel (1994) suggested that with increasing M spicatum biomass, fish production could shift from a few gamefish species, such as northern pike (F.sox !11ci11s) and walleyes (Stizostedion vitreum), to sunfish

(Centrarchidae) and less "sporty" fish. Keast ( 1984) reported that during peak plant gro\\th, the percentages of total fish caught in Lake Op in icon, Ontario were lowest in M spicatum beds and that bluegill (Lepomis macrochints), pumpkinseed (Lepomis gihbosus), rockbass (Ambloplites n,pestris) and yellow perch (Percafluviafilis) tended not to use milfoil beds either inshore or at 2.0 to 3.5 meter depths.

M spicatum beds can also adversely effect fish spa\\ning areas by reducing hydrodynamic processes that aid in aeration of eggs and causing gravel spavming areas to become buried with organic material and muck (Keast 1984, Newroth 1985, Engel 1994).

Eurasian watermilfoil can also affect physical and chemical characteristics of aquatic ecosystems.

Water movement, surface mixing and reaeration, and light penetration can be hampered by dense M

5 spicatum beds (Carpenter and Lodge I 986). Dense plant gro\\th can cause a layer of heated water on the surface and a cooler layer near the lake bottom (Engel 1994). Engel (1990a) reported that on a hot summer day, water temperatures within dense plant beds can stratify 20 °F from surface to bottom, thus reducing metabolic rates and growth of bottom-dwelling prey. DcnseM spicatum beds can also cause high diel dissolved oxygen fluctuations by saturating the water with dissolved oxygen during the day and using large amounts of dissolved oxygen for respiration at night. This respiration also alters pH and free carbon dioxide (Engel 1994). Phosphorus released from living M spicatum tissues is found to be insignificant, although contributions from tissue sloughing during the growing season and year-end senescence represent a significant internal loading source (Newroth 1985, Carpenter and Lodge 1986). Newroth ( 1985) also reported that M spicatum has the potential to seasonally release more phosphorus to Cultus Lake (British

Columbia) than individual sources such as storm sewers and industries could contribute to the svstcm.

Eurasian watermilfoil can negatively affect and even preclude many recreational activities.

Swimming, boating, sailing, water skiing and fishing often become impossible in dense, surfacing mats of

M spicatum. Boaters have been kno\\n to become stranded in thick A1. spicatum beds (Lillie pcrs. communication) and in many instances, riparian o,,11crs arc unable to get boats from their piers to open water during an entire summer. Extensive, surfacing M spicatum beds arc aesthetically unpleasing and can increase the safety risk for swimmers (Ncwroth 1985). Such situations can lead to decreased property values and resort profits (Newroth 1985, Bates et al. 1985). Additionally, M spicat11m has also been known to clog industrial and potable water supply systems, interfere with commercial fishing operations, and increase mosquito populations (Aiken 1979, Bates ct al. 1985, Ncwroth 1985).

M spicatum Control Methods

Numerous methods arc currently used in an attempt to control Eurasian watermilfoil from spreading and creating nuisance conditions, or being transferred to additional waterbodics. Chemical

6 herbicides (e.g., 2,4-dichlorophcnoxy acetic acid), large mechanical harvesters, bottom barriers, rototillers, suction dredges, drav.-downs, ultrasound, and biological controls have been researched, tested, and in many cases, implemented to help control the grov.th of M spicatum in lakes, rivers and reservoirs throughout

North America (Bates et al. I 985, Maxnuk 1985, Rawson I 985, Soar 1985, Bode et al. I 993). Lake organizations and local governments continue to spend millions of dollars on harvesting machines, consulting fees, labor costs, and chemicals in an attempt to control M spicatum (Bode ct al. I 993). For example, between 196 I and 1985, over $6 million were spent on herbicides for M spicatum maintenance in

Tennessee Valley Authority reservoirs alone (Bates ct al. 1985).

In addition to being costly, most control methods provide only short-tenn reductions in biomass

(e.g., usually one season or just part of one season) (Aiken ct al. I 979, Smith and Barko 1990, Bode et al.

1993) and often have drawbacks associated with their use. For instance, chemical treatment can kill both target and non-target species, promote oxygen loss from rapid plant decay, and suppress less resistant (i.e., tolerant) native species (Engel 1990a). Mechanical harvesting can impact ecosystems by disturbing sediments, creating drifting plant fragments, removing and dislodging macroinvertebrates and fish, and altering fish feeding behavior (Engcl I 990a, Engel 1990b). Macrophytes, especially long-season species such as M spicatum, often recover from harvesting within a few weeks and can resurge to even greater densities following harvest (Engcl 1990b ). Moreover, controls such as rototilling and placing bottom barriers cannot practically be used for large areas.

Biological control of M spicatum using microbiological approaches and herbivorous macroinvertebrates has gained attention in recent years (Nc\\man and Ragsdale 1995). Unfortunately, microbiological control has been unsuccessful to date. Shearer (1994) attempted to control M spicatum using the fungus Mycoleptodiscus terristris (Gcrd.) Ostazcski (as a mycohcrbicide, Aqua-F)te™) and found it ineffective in reducing M spicatum biomass under natural field conditions. Natural declines ofM spicatum in various lakes, however, have corresponded with the presence of three herbivores: the naturalized moth Acentria ephemerella (Denis and Schiffcnnullcr) (=Acentria nivea Oliver), the native

7 midge Cricotopus myriophylli Oliver, and the native weevil Euhrychiopsis lecontei Dietz (=F:ubrychiopsis lecontei) (Sheldon 1994, Ne,\man and Ragsdale 1995). Although all three species feed on Eurasian watermilfoil, studies indicate that E. lecontei appeared to have the best potential for biological control

(Creed and Sheldon 1995, Sheldon and Creed 1995). Studies quantifying the effects of E. lecontei onM spicatum have been performed in New England and Minnesota in recent years. The work of Creed and

Sheldon (1993, 1995, Creed et al. 1992, Sheldon and Creed 1995) showed that E. lecontei suppressed M spicatum production, adversely affected its buoyancy, and lowered fragment viability both in the laboratory and in natural field conditions. Newman et al. ( 1996b) also found E. lecontei had a significant negative impact on M spicatum in experimental tanks. Feeding by the weevil reduced the percentage and absolute stocks of carbohydrates in both stems and roots (Ne,,man ct al. I 996b, Perry and Penner 1995). Newman et al. (1996b) speculated that plant injury by weevils may accumulate over several years by reducing root stocks and thus invoke longer term M spicatum declines with more lasting effects than harvesting.

While E. lecontei negatively affects M spicatum, its effects on native plants appear to be minimal.

In experiments by Shledon and Creed (1995) there was no evidence that weevils fed or reproduced on any native macrophytes except northern watermilfoil (Myriophyllum sibiricum Komarov) and in experiments by Solarz and Newman (1996), only 3 of 207 females laid eggs on Megalodonta bee/di (Table 1).

Although weevils fed and laid eggs on the northern watern1ilfoil, the quantitative effects on the M sibiricum stems (i.e., stem length, number of leaflets, added leaf whorls, biomass) were not significantly different from control treatments (those without weevils) (Sheldon and Creed 1995, Creed and Sheldon

1993). This suggests that M sihiricum and E. lecontei may have cocvolved with M sibiricum or other

Myriophyllum species as the original host plants. It is possible that E. lecontei is undergoing a host expansion or shift and it may serve as a good biocontrol agent due to its high specificity for M spicatum

(Solarz and Newman 1996).

8 Table I. Macrophyte species used to quantify potential effects of E. lecontei on native species in previous studies (3Sheldon and Creed 1995 and hSolarz and Newman 1996). The studies showed that £. lecontei had no effect on native plants.

Macrophvte species Common name Effect of E. lecontei 1 Myriophyllum sihiricum Komarov a.b Northern watermilfoil n.s. Ceratophyllum demersum L. a.b Coontail n.s. Elodea canadensis G. a.b Elodea n.s. Heteranthera dubia Jacq a Water stargrass n.s. Megalodonta beckii (Torre.) a.b Water marigold n.s. Najas jlexilis (Willd.) Rostk. and Schmidt a Naiad n.s. Potamogeton amplifolius Tuckerm. a Large-leaf pondweed n.s. Utricularia vulgaris L. a Bladderwort n.s. Chara sp. a Stonewort n.s. Vallisneria americana Michx. a Wild celcrv n.s. Potamogeton zosteriformis Fernald. h Flat-stemmed pondweed n.s. Potamogeton pectinatus L. b Sago pondweed n.s. Ranunculus aquatilis b Buttercup n.s. 1 n.s. indicates no significant effect

9 Solarz and Newman (1996) found that weevils raised on M spicatum had a high specificity for M spicatum, while weevils raised on M sibiricum had no preference between M spicatum and the native milfoil. Recent studies also indicate that Eurasian watcnnilfoil produces larger E. lecontei adults and promotes a faster development time from egg to adult, thus making M spicatum a superior host plant

(Newman et al. 1997a, Newman ct al. 1997b ). Solarz and Newman ( 1996) reported that the weevil is unlikely to shift to non-watennilfoil hosts and will have minimal damage on native plant species. However, ifM spicatum becomes rare, the weevil may likely use M sibiricum until other M spicatum becomes available.

E. Lecontei Life Historv

In order to understand the mechanism by which H. lecontei may control M spicatum, the life history of E. lecontei was studied in detail (Creed ct al. I 992, Creed and Sheldon 1993, Solarz and

Newman 1996, Newman et al. l 996a,b, Sheldon and O'Bryan 1996a). Adult weevils are 2 - 3 mm in length with black and yellow stripes along the back and a light ventral side (Figure 1). The adult female lays one, two or sometimes more tiny, yellow individual eggs (Figure 2) in the leafy apical mcristcm of a plant before moving onto another tip on the same plant or an adjacent plant to continue laying eggs. E. lecontei eggs are elliptical in shape and approximately 0.52 mm x 0.39 mm in size (Sheldon and O'Bryan

1996a). The larvae hatch after about four days and begin to cat the delicate meristematic tissues of the tip where they hatched (Figure 3). Later larval instars continue to burrow further down into the stem, consuming the vascular tissues and occasionally making exit and entry holes along the way (Figure 4). The larval lifestagc lasts approximately IO - 13 d and duration is temperature dependent (Sheldon and O'Bryan

1996a, Ne\\'111an et al. l 997a). The majority of E. lecontei damage comes from the destruction of the apical growing tip which suppresses the production of new plant biomass (Creed and Sheldon 1995), and the

10 ho II owing of the stem, ,vhich disrupts internal transport of carbohydrates and nutrients, suppresses root production, and reduces plant buoyancy (Figure 5) (Creed and Sheldon 1995, Sheldon and Creed 1995).

Following the larval stage, the weevil pupates (Figure 6) ,vithin the stem further down the plant stem(> 0.5 m) where the stem is thicker (1. 7 - 4.0 mm in diameter) (Sheldon and O'Bryan 1996a,

Newman et al. 1997a). Pupation lasts approximately IO - I 3 d and again, duration is temperature dependent (Sheldon and O'Bryan 1996a, Newman et al. 1997a). In the northeastern United States (i.e.,

Vermont) there are approximately three generations of weevils per summer and adult weevils live the entire season with females laying up to I .9 eggs per day and a hatching rate of 87% (Sheldon and O'Bryan

1996a). As such, there is the potential for rapid population gro,,th under optimal conditions. Ne\\man et al. (1997b) reported that the entire life cycle is temperature dependent ranging from 60d at I 5 °C to 17d at

27-31°C with 75% successful development above 15 °C.

E. lecontei adults move to shoreline overwintering habitat (the top 2.5 cm of soil/litter interface) between September and November (Ne,,man et al. I 996b) and overwinter survival can be as high as 40%

(Newman and Ragsdale 1995). Overwintering females are mated and hibernate with fat reserves but without developed eggs. These same adults emerge in the spring with depleted fat reserves and appear to develop eggs only after entering the water and beginning to feed (Nev,-man and Ragsdale 1995). Flight muscles in adults are not developed during the summer but appear to develop after reproduction ceases in late summer or early fall (Newman and Ragsdale 1995).

E. lecontei Distribution and Natural M spicatum Declines

Localized declines of M spicatum beds have been reported in various lakes throughout the U.S. and Canada, and although researchers have speculated on possible causes of these declines (e.g., turbidity, sediment nutrient depletion, pathogens and herbivores), definitive cause and effect relations have not been established (Smith and Barko I 990, Lillie and Helsel 1997). For instance, Carpenter (1980) reported that

11 the peak abundance ofM spicatum in many lakes is maintained for approximately IO years, followed by a decline in distribution and abundance of the plant. Carpenter (1980) investigated potential reasons for a decline ofM spicatum in Lake Wingra, Dane Co., Wisconsin including toxin accumulation, climate, nutrients, past control methods, epiph~tes, competition, parasites and pathogens and concluded that the decline was not caused by any single factor that he researched, and thus was probably an interaction of several factors. E. /econtei is suspected to be a contributing factor in some other natural M spicatum declines including Fish Lake (Dane Co., WI), McCullom Lake (McHenry Co., IL) and Brownington Pond

(Orleans Co., VT) (Lillie and Helsel 1997, Creed and Sheldon 1995, Kirschner 1995). In fact, a recent study by Creed (1998) examined the distribution of all reported, unexplained M spicatum declines with respect to the known distribution of E. lecontei and found that significantly more declines occurred within the range of E. lecontei than would be expected by chance (p < 0.001). This evidence supports the hypothesis that the weevil may be an important agent in the cause of these natural declines.

Circumstantial evidence of M spicatum declines concurrent with higher E. lecontei abundance warrants further research as a biological control. Bode ct al. ( 1993) suggested that research efforts on controlling M spicatum must be cost-effective and ecologically-sound. Hanson et al. ( 1995) reported that a major limitation in the current weevil research is a lack of quantitative data correlating weevil densities with M spicatum damage. Ne\\man and Ragsdale ( 1995) and Ne,,man et al. ( 1997b) added that it is essential to discover what limits weevil populations, what effects current weevil densities have on M spicatum, and what types of M spicatum populations are susceptible to E. lecontei control. Additional information is necessary to determine the current geographic distribution of E. lecontei in Wisconsin and to assess what factors may be controlling weevil populations in Wisconsin lakes. Understanding which lakes may be susceptible to a natural, weevil-induced M spicatum decline, may help determine whether E. lecontei could be a viable management tool for the control of M spicatum in Wisconsin.

12 Figure 1. Adult E. lecontei.

Figure 2. E. lecontei egg on·apical tip of M spicatum.

13 Figure 3. First instar of E. lecontei larva in apical tip of M spicatum.

Figure 4. Late instar of E. lecontei larva in M spicatum stem.

14 Figure 5. Stem of M spicatum with E. lecontei damage.

Figure 6. E. lecontei pupa in M spicatum stem.

15 Studv Objectives

The two objectives of this study were I) to determine the geographic distribution of E. lecontei among Wisconsin lakes that contain M spicatum, and 2) to assess geographical and limnological characteristics associated with the presence or absence and abundance of E. /econtei. The results of this study provided baseline information for a third project objective which will evaluate the effectiveness of stocking E. lecontei as a practical management tool for the selective control of M spicatum in two years.

The third objective is not part of this thesis.

16 METHODS

E. lecontei Distribution

The distribution (i.e., presence or absence) of F:. lecontei was assessed in 46 Wisconsin lakes containing M spicatum during the summers (June - September) of 1996 and l 997 using three methods.

First, 23 lakes were surveyed to assess presence of E. lecontei by visually searching for E. lecontei adults on the tips ofM spicatum plants. Surveys were most often conducted by snorkeling. During these surveys, diagnostic evidence of E. lecontei herbivory aided in the search for actual organisms by helping to identify areas to search more intensively. At each lake, a maximum of four man-hours were spent intensively searching before a lake was considered to be absent of weevils. On some lakes, snorkeling was not done, rather surveys were conducted from the boat if the milfoil was surfacing and easy to search over the side of the boat. Second, 6 additional lakes were surveyed during macroph)te sampling used to assess density of E. lecontei (described below). Third, 17 lakes were surveyed by WDNR personnel following a weevil sampling workshop and training session conducted on Fish Lake in Dane County during the summer of 1996. Specimens collected by WDNR personnel were collected either while snorkeling or from boats.

All adult E. lecontei specimens were preserved and maintained as voucher specimens for each lake except

Little Falls Lake, Mason Lake, and Parker Lake where adult specimens were not collected; only larvae from these lakes were found and kept as voucher specimens. All voucher specimens were preserved in 80% isopropyl alcohol in vials with an identification label {species, lake, county, collection date and collector) and are housed in the Museum of Natural History at the University of Wisconsin - Stevens Point.

The lakes sampled were selected due to the kno"n presence of M spicatum beds. Lakes surveyed

by the WDNR were those visited by Department personnel for other sampling procedures (e.g., water

chemistry or macrophyte surveys) which contained M spicatum.

17 E. lecontei Densities and Associated Geographical and Limnological Characteristics

E. lecontei Densities

Thirty-one lakes were used to assess factors influencing E. lecontei density in Wisconsin. Lakes

that were selected included: I) a subset of eight WDNR Ambient Lakes (i.e., Long Term Trend Lakes), 2)

seven non-Ambient lakes, 3) t\velve lakes participating in the ,veevil stocking portion of the study (objective

three), and 4) four "cooperating" lakes (objective three).

Study lakes were selected based on 1) the presence of M spicatum, 2) the lake's location within the

state, and 3) the amount of water quality data collected from the lake in the past. A widespread distribution

oflakes throughout the state was sought in order to test the geographic variation in weevil density.

Ambient Lakes were chosen, when possible, in each geographic region due to the availability background

data collected from these lakes in the past ten years (mostly chemical and some biological data). However,

because many of the Ambient Lakes in central and northern Wisconsin do not harbor M spicatum and

could not be surveyed, non-Ambient Lakes with M .\piccrtum were substituted.

Densities of E. lecontei were sampled once from each lake in July or August of 1996 or I 997.

Because E. lecontei densities may vary seasonally, (Ne\\man and Ragsdale 1995) samples were collected

during approximately the same time in the growing season to minimize seasonal sampling variation across

lakes. Accordingly, southern lakes were sampled first, followed by central lakes and finally northern lakes.

Within each study lake, only beds not subjected to control treatments such as mechanical harvesting or

chemical herbicides during the sampling season were chosen for density sampling. E. lecontei reside in the

upper portions of the milfoil stem, which is removed through mechanical harvesting, and significantly fewer

weevils are found in harvested rather than unharvested beds (Sheldon and O 'Bryan 1996b).

In each lake, four M spicatum beds were chosen for density sampling and were designated as beds

A-D (Figure 7). The distance between the sampled beds was maximized in order to have the whole lake

represented in the sample. In cases where M .\picatum did not grow in distinct beds but rather "ringed" the

18 entire lake in the littoral zone, the lake was divided into four quadrants, each serving as a "bed" for the purposes of transect placement. If a lake had less than four beds, larger beds were divided in half until there was a total of four beds.

A total of 120 apical stems of M spicatum were collected from each lake (4 beds x 3 transects x 5 sampling locations x 2 stems). Within each M spicatum bed, three roughly equidistant transects were placed perpendicular to shore from the deep edge of the bed (or from the center of the lake if M spicatum filled the entire lake) to the shoreline, marsh-water boundary or the shallow edge of the bed (Figure 7).

Because E. lecontei move from near shore to deeper water as the summer progresses (Newman and

Ragsdale 1995), transects were placed perpendicular to shore to systematically sample all depths in the bed and minimize the variation in distribution caused by weevil movement. Each transect was snorkeled and at five, roughly equidistant points along each transect the top~ 50 cm of the first two stems of M spicatum touched with the snorkeler's hand beneath the surface at that point were collected (Figure 7). In lakes where the milfoil did not reach the surface, the snorkeler dove beneath the surface and looked only at the base of the plants (or lower part of the stem) where weevil damage could not be assessed and two stems were randomly picked. In preliminary investigations, these collection methods minimized sampling bias associated with search image of damaged plants. Because only the top 50 cm of stem were retained, these were termed "apical stems." Hov,ever, it is important to note that each apical stem usually included several apical tips or small branches along its 50 cm length. The two apical stems picked at each location were carefully sealed in a pre-labeled, plastic bag.

During sampling, limnological characteristics of each lake and bed were recorded (Appendix A and

B). Water temperature and dissolved oxygen for the lake were measured just below the water's surface at the lake's deepest point ,·vith a calibrated YSI 5739 temperature and dissolved oxygen meter. The meter was unavailable for some lakes and therefore temperature and dissolved oxygen data from the

Environmental Protection Agency's STORET system from the same summer (and usually within one week of our sampling) was used for Fox Lake, Rock Lake and Silver Lake. Sccchi disc depth was measured at

19 the lake's deepest point with a standard Secchi disc. For each bed, the distance to shore was measured in one of three ways: I) \\ith a tape measure for distances < 40 m, 2) with a digital range finder, or 3) by visually estimating the distance when the other two options were unavailable. The depth of each sampling point and the depth of the deep and shallow edges of the bed were measured using a fishing reel fitted \\-ith a measured and marked nylon cord and a weight. The riparian conditions were visually estimated using the percentages of each shoreline type (natural, mo\\'ed la\\n, sand, or seawall/rip-rap) existing at the water - land interface adjacent to the sampled bed or at the end of a transect if the entire littoral zone was sampled.

Sample Processing

Plant samples were kept on ice until stems were processed. Any samples that could not be processed within seven days of collection were preserved in 95% ethanol. In the lab, the apical stems were placed in a glass Pyrex dish on top of a light table and carefully inspected for weevils and weevil damage.

Portions of stems with damage and all apical tips were inspected with a dissecting microscope for E. lecontei eggs, larvae and pupae. Eggs were identified as tiny ovoid yellow spheres sttached to the leaves of the apical tips. E. lecontei eggs arc not similar to any other invertebrate eggs and therefore were easily distinguished from other objects attached to the M. spicatum. Damaged sterns were slieed open length-wise with a razor blade so that larvae and pupae could be extracted from the stems. The number of weevils in each life stage (e.g., egg, larva, pupa, adult) on each stem ,rns recorded. The number of apical tips and the number of broken tips were also recorded for each stem. All weevils and eggs collected were preserved and labeled with species, collection date, collector, lake name and county.

M spicatum Density and Biomass Sampling

In a subset of 12 lakes (Table 2), plant density and biomass of M: spicatum were collected during peak plant biomass in late August of 1996. Density and biomass samples were collected along a single transect running parallel to shore extending either the length of the bed or between 50 and I 20 m long

20 Table 2. Lakes used in M spicatum density and biomass sampling.

Surface Area Maximum Alkalinity Lake Countv (hectares) Depth (m) (mg/L CaC03) Beaver Dam Lake Barron 444.8 32.3 77 Big Sand Lake Vilas 563.2 19.8 136 Eagle Lake Racine 206.0 3.7 153 Gilbert Lake Waushara 56.4 19.8 139 Kangaroo Lake Door 449.2 3.7 648 Kusel Lake Waushara 31.6 8.8 380 Lorraine Lake Walworth 53.2 2.4 517 Lower Spring Lake Jefferson 41.6 3.4 857 Mukwonago Park Pond Waukesha 6.8 0.9 753 Nancy Lake Washburn 308.8 11.9 117 Pearl Lake Waushara 36.8 15.2 470 Whitewater Lake Walworth 256.0 11.6 174

21 depending on the size of the bed. Depth along the transect was held constant in each lake; transect depths among lakes ranged from 1 to 3 m depending on the natural depth of the M spicatum bed.

Sh.teen quadrat samples along the transect were collected by a SCUBA diver using a 0. l 5m2 underwater quadrat sampler made from 1/4 inch pvc tubing. This sampler was constructed as a right triangle so that two sides of the triangle could be rotated (from vertical to horizontal) in the plot and the third piece could be slid into place to connect the sampler and enclose the plants. This ensured that plants from outside the sampling quadrat were not pulled into the sampler during placement. Plants were pulled from the sediment so that most root material was collected and the ,..-hole sample was brought to the surface and placed in an individual pre-labeled plastic garbage bag. Samples were kept on ice until being placed in an environmental chamber at 4 °C v,ithin 1-2 days of collection. During M spicatum sampling, the depth of each sampling point was measured in the bed while surface and bottom water temperature (°C) and dissolved oxygen (mg/L) and Secchi disc depth were collected at the lake's deepest point.

In the lab, samples were processed by "floating" the sample of plants lengthwise in an aquaculture raceway. The number of plants in each sample was counted for plant density which was converted to the number of plants per square meter. The roots were removed from the plants just above the root and rootlet portion of the stem. This was identified as the location where the stem gets distinctly lighter in color. A sub-sample of 30 plants per sample was used to get average plant length (cm), number of tips per plant, and the number of broken tips per plant. Plants covered with calcium carbonate deposits were soaked in a

5% hydrochloric acid bath and rinsed with tap water to remove the deposits. Macroph)tes other thanM spicatum were separated and identified to species and their stems were counted. All plants of each species were then air and drip-dried for 30 minutes and weighed for wet weight. To determine dry weight, plants from each species and each sampling site were placed in a pre-dried, tared paper bag, and dried at I 06 °C in a Blue M Electric Stabil-Them1 electric oven for 48 hours. After 48 hours, the bags were weighed, placed again in the drying oven, and re-weighed six to eight hours later to check that a constant weight was

22 achieved. Weights were measured on a Mettler PM4000 scale and recorded to the hundredths of grams

(0.00g).

Data Analysis

E. /econtei Distribution

The distribution of E. lecontei from previous collections and this study was plotted on a Wisconsin map and specific locations were tabulated.

E. lecontei Densities

E. lecontei densities were calculated as weevils per apical stem for each lake (whole lake) and plotted with the 95% confidence intervals. Density was also recorded as the number of weevils per apical tip (i.e., meristems) by dividing the number of E. lecontei collected in a lake by the number of apical tips collected in that lake (thus, confidence intervals could not be calculated).

Correlations with E. lecontei Densities

Spearman correlation analysis was conducted using SAS 6.10 to elucidate significant (p ~ 0.05) relationships between E. lecontei densities (and other dependent variables) and independent variables at the lake level and M spicatum bed level. Spearman correlation was used because most variables were not normally distributed and transformations often did not achieve normality.

Dependent variables were those collected regarding E. lecontei density, life stages and evidence of herbivory (Table 3). Independent variables included those previously described as collected during E. lecontei density sampling and M spicatum density and biomass sampling as well as several variables collected from additional sources. Two levels of variables were collected: those that applied to the lake as a

23 whole (lake level variables) (Table 4, Appendix A) and those that were measured in, or directly adjacent to the bed ofM spicatum being sampled (bed level variables) (Table 5, Appendix B).

24 Table 3. Dependent variables collected during E. lecontei density sampling.

Variable E. lecontei per apical stem of M spicatum E. lecontei per apical tip of M spicatum Percentage of each E. lecontei life stage from total number of E. lecontei Percentage ofbrokenM spicatum tips Percentage of stems with E. lecontei

25 Table 4. Independent variables tested for correlation at the lake level. Source, description and collection technique included in Appendix A.

Variable Northing coordinate (m) Easting coordinate (m) Year of E. lecontei density sampling Date of E. lecontei density sampling Julian date of E. lecontei density sampling Year ofM spicatum invasion Time of M spicatum invasion Maximum lake depth (m) Mean lake depth (m) Surface area of lake (hectares) Lake type Water temperature (°C) Dissolved oxygen (mg/L) Secchi disk depth (m) Total phosphorus (mg/L) Nitrogen (mg/L N02+N03) Total Kjeldahl nitrogen (mg/L) Chlorophyll a (ug/L) Total alkalinity (mg/L CaC03) pH Conductivitv (micromhos)

26 Table 5. Independent variables tested for correlation at the bed level. Source, description and collection technique included in Appendix B.

Variable Biomass ofM spicatum (dry g/m2) Biomass of native macroph)'tes (dry g/m2) Stem length ofM spicatum (cm) Stem density of M spicatum (stems/m2) Tips per M spicatum stem Percentage of broken M spicatum tips M spicatum tips per m2 Depth of M spicatum (m) Distance to shore from middle of M spicatum bed (m) Distance to shore from shallow edge of M spicatum bed (m) Distance to shore from deep edge of M spicatum bed (m) Percentage of natural, grass, sand or seawall/rip-rap shoreline

27 Bed B

Bed C

Figure 7. Diagram depicting E. lecontei density sampling design on a hypothetical lake. Lines indicate sampling transects which were roughly equidistant across the bed, dots indicate sampling points which were roughly equidistant from the deep edge of the bed to the shallow edge. Two apical stems were collected at each sampling point; 120 apical stems were

I-..) collected per lake. 00 RESULTS

F:. lecontei Distribution

E. lecontei was widely distributed across Wisconsin in lakes containing M spicatum (Figure 8,

Table 6). E. lecontei was found in 45 lakes where it was not previously recorded and was found in nearly every lake sampled except Silver Lake in Waupaca County. After searching for four man-hours in Silver

Lake, no E. lecontei of any lifcstagc was found and there was no evidence of damage to M spicatum that is typical of E. lecontei herbivory. In most other lakes, an E. lecontei adult was collected within a fc,v minutes of searching. In three lakes where adults were not found, Mason and Parker Lakes in Adams

County and Little Falls Lake in St. Croix County, only larval ,vccvils were found during weevil density sampling.

Lakes with new records of F:. lecontei were variable in type, size and location (Table 6). Both seepage and drainage lakes were found to harbor weevils, lake size ranged from 1.2 hectares (Lower Kelly

Lake) to almost 3,000 hectares (Big Green Lake) and depths were as shallow as 1.5 meters (Mukwonago

Park Pond) and as deep as 71.9 meters (Big Green Lake). The lakes were also widely distributed geographically around the state (Figure 8) from the southeastern counties of Kenosha and Racine, through central counties and into northern counties such as Washburn and Vilas.

M spicatum beds in lakes with E. lecontei were varied. Many lakes had very dense and broad beds of mil foil while others had narrow bands of M spicatum in the littoral zone, or very little mil foil at all.

Mean depth of M spicatum beds varied (sec Appendix for descriptions of 31 of the 45 lakes) as well as the distance occurring between the top of the plants and the surface. Although this characteristic was not measured, it was observed that many beds reached the surface and branched while others were well below the surface.

29 Figure 8. Kno\\n distribution of E. lecontei in Wisconsin. Previous locations referenced in Lillie (1991),

Ne\\man and Maher ( 1995), Lillie and Helsel ( 1997).

o Locations apparently absent of E. lecontei

o Previous locations of E. lecontei

• New records of E. lecontei

• •

0 • • ••• • • • ••

0

30 Table 6. New records of Euhr_ychiopsis lecontei in Wisconsin.

SURFACE AREA MAX. DATE OF LAKENA.\1E COCNTY TYPE' LOCATIO:'\ (hectares) DEPTH COLLECTION m Alpine Lake \\'aushara DG T19:'\RIIESee.4 22.4 5.5 I 7 July 1997

Bark Lake Washington SE T9:'\ R 19E Sec. 26 24.8 10.4 9 June 1997

Bass Bay Waukesha DG T5N R20E Sec. 15 40.0 7.0 12 July 1997

Beaver Dam Lake Barron SE T35N R 13\\' Secs. 5. 6. 7, 8 444.8 32.3 12 June 1996 T35N Rl4W Sec. I

Beulah lake Walworth DG T4N Rl8E Sec. 4 333.6 17.7 21 July 1997

Big Cedar Lake Washington SE TI0N Rl9E Sec. 5: 372.8 32.0 25 June 1996 Tl IN R 19E Secs. 20.20.30.31.32

Big Green Lake Green Lake DG Tl5.16N Rl2.13E 2.938.4 71.9 08 August 1997

Big Sand Lake Vila~ DG T41N R12E Secs. 2.3.4.9: 563.2 19.8 25 June 1996 T42 Rl2E Secs. 34.35

Camp J,ake Kenosha DG TIN R20E Secs. 20.21.28.29 184.4 5.2 02 August 1996

Crooked Lake Waukesha DG T7N R 17E Sec. 23 23.2 4.9 17 Sept. 1996

Crystal l~,ke Sheboygan SE T16N R21 E Secs. 31. 32 60.8 18.6 30 July 1997

Delavan Lake Walworth DG T2N R 16E Secs. 21. 22. 27, 28, 32, 33 828.8 17.1 28 July 1997

Eagle Lake Racine DG T3N R20E Secs. 21.22.27.28 208 3.6 06 June 1996

Elizaheth Lake Kenosha DG TIN R19E Sec-s. 28. 29. 32. 33 255.2 9.7 13 August 1997

Fox Lake Dodge DG Tl3'.\ Rl3E 1050.0 5.8 11 August 1997

Friendship Lake Adams DG Tl 7N R6E Sec. 5 46.0 4.9 27 June 1997

George Lake Kenosha DG TIN R21E Secs. 20.29 23.6 4.9 02 August 1996

Gilhert Lake Waushara SE T20:'.': R l IE S,x-s. JO.I 1.14.15 56.4 19.8 10 June 1996

Jordan l.ake Adams SE TISN R7E Sec. 34 85.2 24.1 27 June 1997

Kangaroo Lake Door DG T29'.\ R27E Sec. I 449.2 3.6 10 July 1997

Kusel l,ake Waushara SE T20N R 11 E Secs. 26.27.34.35 31.6 8.8 10 June 1996

Lac La Belle Waukesha DG T8N Rl7E 465.6 13.7 29 July 1997

Little Falls Lake St. Croix SE T29'.\ R 19\\' Secs. 4.8.9 68.8 5.5 14 August 1996

Long Trade J,ake Polk DG T36N R 18\\' Sec. 49 61.2 4.0 06 August 1996

31 Table 6. Continued

Sl1ff.-\CE AREA MAX. D.-\TEOF LAKE NAME COUNTY TYPE' LOCATION (hectares) DEPTH COLLECTION (m Lorraine Lake Walworth SE T3N Rl5E Sec. 29 53.2 2.4 06 June 1996

Lower Kelly Lake Waukesha SE T6N R20E Sec. 36 1.2 11.0 02 June 1997

Lower Spring Jefferson DG T5N Rl6E Secs. 22.23 41.6 3.3 06 June 1996 Lake Marie Lake Kenosha DG TIN Rl9E Sec. 21. 28 118.8 10.0 I 3 August 1997

Manson Lake Oneida DG T36N R7E Secs. 32.33 94.4 16.4 25 June 1996

Ma.son Lake Adams DG Tl3N R7E Secs. 25. 26. 35. 36 34.2 2.7 l 3 August 1996 Tl3N R8E Secs. 30. 3 I

Mctonga Lake Forest DG T35N RJ3E Sec. 8 862.8 24.1 14 Auguat I 997

Mukwonago Pond Waukesha SE T5N R l 8E Sec. 29 6.4 1.5 18 June 1996

Nancy Lake Wa.shhum SE T42N Rl3W Secs. 27.28.33 308.8 11.9 12 June 1996

North Lake Waukesha DG T8N R l 8E Secs. 16.17.20.21 174.8 23.8 14 August 1996

Paddock Lake Kenosha DG Tl N R20E Sec. 2 44.8 9.7 02 August I 996

Parker Lake Adams SE Tl 5N R7E Sec. l 4 23.6 9.1 27 June 1997

Pearl Lake Waushara SE Tl9N R 12E Sec. 30 36.8 15.2 JO June 1996

Pike I.A1ke Washington DG Tl ON R 18E Secs. 22.23.26.27 208.8 13.7 12 August 1996

Ripley Lake Jefferson SE T6N Rl3E Sec. 7.8 167.2 13.4 31 July 1997

Rock Lake Jefferson DG T7N R13E Secs. 2.10.11.14.15 548.4 17.1 I 3 August 1996

Sherwood Lake Adams DG T20N R6E S,-cs. 16. 17 98.4 8.2 12 August 1997

Whitewater Lake Walworth DG T3N R15E Sec. 3 256.0 11.6 06 June 1996 T4N Rl5E Secs. 25. 26. 27. 34. 35

Wind 1.ake Racine DG T4N R20E Secs. 3.4.8.9.10.16.17 374.4 15.2 29 July 1996

Wolf Lake Racine T2N R20E Secs. 15. 22 46.0 8 August 1996

Yellow Rirch Vila.~ DG T 40'.'-i R JOE Secs. 21.22 80.8 7.0 13 August 1997 I.ake a DG = Drainage lake. SE = Seepage lake

32 E. lecontei Densities and Correlated Variables

E. lecontei densities across 31 lakes varied from 0.0 (non-detectable) t-o 2.5 weevils per apical stem with a mean of 0.65 ,veevils per apical stem (Figure 9, Table 7). Using the index of weevils per apical tip

(i.e., meristem), densities ranged from 0.0 (non-detectable) to 0.5 weevils per tip with a mean of 0.15 weevils per apical tip. E. /econtei larvae and eggs were found to be more abundant than adults or pupae in

30 of the 31 lakes. Adults never made up more than 36.4% and pupae never made up more than 25% of weevils collected in any one lake, but larvae abundance reached as high as I 00% and eggs as high as

84.6% of weevils collected in some lakes. In 19 of 31 lakes, all life stages of E. lecontei were collected during density sampling. E. lecontei density also varied among M spicatum beds within lakes with a mean difference of 0.81 weevils per apical stem between the highest of lov.-est density found among beds within one lake (Table 8). The greatest difference between beds in one lake was 2.3 weevils per apical stem in

Lake Wingra.

E. lecontei density was not significantly correlated (P < 0.05) with any variables collected at the whole-lake level such as latitude of the lake, time since the invasion of M spicatum, lake depth (maximum or mean), lake type. or lake size. E. lecontei density also did not significantly correlate ,vith water quality

variables such as summer water temperature, dissolved oxygen, Secchi disc depth, total phosphorus, NO2 -

NO3 nitrogen, total Kjcldahl nitrogen, chlorophyll a, alkalinity, pH, nor conductivity. However, the percent of various E. lecontei lifcstages among all weevils collected per lake was significantly (P < 0.05) correlated with a few variables (Table 9). Specifically, the percentage of E. lecontei adults was negatively correlated

,vith the date of collection (R = -0.42), the percentage oflarvae was negatively correlated ·with total phosphorus (R = -0.44) and the percentage of eggs was positively correlated with summer water temperature (R = 0.51) and the year of collection (R = 0.46).

33 Figure 9. Density of E. lecontei in Wisconsin lakes; all weevil lifestages combined. Values indicate the mean density of E. lecontei per apical stem ± 95% confidence intervals. * Confidence intervals could not be calculated for Lower Spring and Eagle Lakes due to a different sampling method. No E. /econtei were collected in Silver Lake or Kangaroo Lake.

3.5

3.0

2.5 -, ------·------·----

E ,. .S! UI 2.0 - - -· - ni 0 ,. 'ii • ...IQ & 1.5 ------~--- ·-- ·-~ ·-1--- -- .. -- . C: 0 0 ~ 1.0 1.1.i

0.5 ·------r ! ! I I ! ! ! ■ I ! I I ! I 11-,-•-.-•- Table 7. E. lecontei densities in Wisconsin lakes. WEEVILS PER WEEVILS PER PERCENT PERCENT PERCENT PERCENT LAKE COUNTY APICAL STEM" APICAL TIPC ADULTS LARVAE PUPAE EGGS Alpine Waushara 0.2 ±0.13 0.1 6.9 34.5 0.0 58.6 Bea,·er Dam Barron 1.8 ± 0.44 0.5 6.4 42.7 10.1 40.8 Beulah Walworth 0.1 ± 0.08 0.1 6.3 75.0 12.5 6.3 Big Green Green Lake 0.4±0.15 0.1 2.1 34.0 6.4 57.4 Big Sand Vilas 0.3 ±0.15 (l.l 10.0 40.0 17.5 32.5 Camp Kenosha 0.7 ± 0.35 0.1 15.1 16.3 1.2 67.4 Crystal Sheboygan 0. I ± 0.06 0.0 12.5 50.0 0.0 37.5 Delavan Walworth 1.2 ±ll.31 0.2 16.9 16.2 11.5 55.4 Eagle Racine ().lb 0.0 0.0 66.7 0.0 33.3 Fox Dodge 0.8 ± 0.30 0.1 0.0 56.0 4.4 39.6 Gilbert Waushara 0.1 ± 0.09 0.0 0.0 8 l.8 0.0 18.2 Jordan Adams 0.3±0.13 0.1 6.5 58.l 12.9 22.6 Kangaroo Door 0.0 0.0 0.0 0.0 0.0 0.0 Kusel Waushara 0. I ± 0.07 0.0 33.3 25.0 25.0 16.7 Lower Spring JclTerson I .Ob 0.3 7.4 73.6 19.0 0.0 Lac La Belle Waukesha 0.8 ± 0.22 (l.l 2.1 50.5 5.3 42.1 Little Falls St. Croix 0.2 ±0.17 0.1 0.0 15.4 0.0 84.6 Lorraine Walworth 1.9±0.41 0.4 9.2 41.9 21.0 27.9 Manson Oneida 0.3 ± 0.19 0.1 6.1 57.6 9.1 27.3 Mason Adams < 0.1 ± 0.02 0.0 0.0 100.0 0.0 0.0 Metonga Forest 0.5 ± 0.21 0.1 3.4 59.3 3.4 33.9 Mukwonago Waukesha 0.3±0.12 0.0 36.4 54.5 9.1 0.0 Nancy Washburn 1.2 ± 0.31 0.4 3.5 52.5 6.4 37.6 Parker Adams 0.2 ±0.12 0.0 3.4 48.3 10.3 37.9 Pearl Waushara 0.3±0.17 0.1 0.0 70.3 2.7 27.0 Ripley JclTerson 0.9 ± 0.23 0.2 1.9 21.0 8.6 68.6 Rock JelTerson 0.1 ± 0.09 0.1 11.8 64.7 11.8 11.8 Silver Waupaca 0.0 0.0 0.0 0.0 0.0 0.0 Whitewater Walworth 1.4 ± 0.40 ().3 8.8 41.5 7.0 42.7 Wingra Dane 2.2 ± 0.69 0.5 4.1 18.7 3.4 73.9 Yellow Birch Vilas 2.5 ± 0.50 0.5 3.1 49. l 2.4 45.4 mean= 0.65 mean= 0.15 mean= 7.01 mean= 45.65 mean= 7.13 mean= 33.77 w • All lifostages combined per apical portion ofM. spicat11111 stem (approx. 50 cm) V, b JA1kes with different sampling method where confidence interval could not be calculated 'All lifostages combined per apical tip (apical mcristcm) ofM. spicafllm Table 8. E. lecontei densities in different M spicatum beds in 31 Wisconsin lakes. Mean values are the mean of four beds and the 95% confidence interval. WEEVILS PER APICAL STEM LAKE BED l BED2 BED 3• BED 4• MEAN DIFFERENCEb Alpine 0.0 0.0 0.7 0.3 0.25 ± 0.325 0.7 Bea,·erDam 2.6 2.6 l.0 1.0 1.80 ± 0.905 1.6 Beulah 0.0 0.0 0.2 0.3 0.12±0.147 0.3 Big Green 0.0 0.2 0.8 0.6 0.40 ± 0.358 0.8 Big Sand 0.1 0.7 0.0 0.6 0.35 ± 0.3-t4 0.7 Camp 0.7 0.2 0.7 1.2 0.70 ± 0.400 l.0 Crystal 0.2 0.1 0.0 0.0 0.07 ± 0.094 0.2 Delavan 0.2 2.1 1.0 l.6 1.22 ± 0.802 1.9 Eagle 0.2 <0.0l 0.0 <0.01 0.05 ± 0.098 0.2 Fox 0.3 l.3 0. l 1.4 0.77 ± 0.(,57 l.3 Gilbert <0.0 l 0.2 0.1 0.0 0.07 ± 0.094 0.2 Jordan 0.3 0.2 0.4 0.2 0.27 ± 0.094 0,2 Kangaroo 0.0 0.0 0.0 0.0 0.00 0.0 Kusel 0.1 0.1 0.0 0.2 0.10 ± 0.080 0.2 Lower Spring 0.2 2.0 0.7 I.I l.00 ± 0.746 l.8 Lac La Belle 0.5 0.4 I.I 1.2 0.80 ±0.400 0.8 Little Falls 0.1 0.3 na na 0.20 0.2 Lorraine l.2 2.0 1.8 2.6 1.90 ± 0.566 l.4 Manson 0.0 0.5 0.6 <0.01 0.27 ± 0.314 0.6 Mason <0.01 0.0 0.0 0.0 <0.01 <0.01 Metonga 0.1 O. l 0.2 l.6 0.50 ± 0.720 1.5 Mukwonago 0.3 0.2 <0.01 0.6 0.27 ± 0.245 0.6 Nancy 2.2 0.5 0.9 l.l l.17 ± 0.713 1.7 Parker 0.1 0.1 0.4 0.4 0.25 ± 0.170 0.3 Pearl 0.4 0.7 0.1 0.1 0.32 ± 0.281 0.6 Ripley 0.7 0.7 1.0 l.l 0.87 ± 0.202 0.4 Rock 0.2 0.3 0.0 0.1 0.15 ± 0.126 0.3 SilYcr 0.0 0.0 0.0 0.0 0.00 0 Whitewater 1.6 0.8 2.2 1.0 1.40 ± 0.620 1.4 Wingra 3.2 2.2 0.9 2.6 2.22 ± 0.955 2.3 Yellow Birch 1.9 1.6 2.7 3.6 2.45 ± 0.878 2.0 w mean= 0.81 °' • Only two M. spicatum beds were sampled on Little Falls Lake; b Difference between beds with the highest and lowest densities Table 9. Spcannan rank correlations between percent of E. fecontei lifcstagcs and characteristics measured at the whole lake level.

NUMBER OF CORRELATION p VARIABLES CORRELATED OBSERVATIONS COEFFICIENT Percent adults vs. Date of collection 31 -0.41998 0.0187

Percent larvae vs. Total phosphorus 28 - 0.44245 0.0187

Percent eggs vs. Year of collection 31 0.51343 0.0031

Percent eggs vs. Summer water temp. 23 0.45912 0.0275

37 E. lecontei density was significantly (P < 0.05) correlated with some variables associated \\ith individual macrophyte beds (Table 10). Density was positively correlated with the percentage of broken apical tips (R = 0.54), distance from shore to the middle (R = 0.25) and deep (R = 0.28) edge of the M spicatum bed, percentage of natural shoreline (R = 0.21), and number of apical tips per M spicatum plant

(R = 0.60). Weevil density was also negatively correlated with depth of the M spicatum bed (R = -0.30) and the percentage of sandy shoreline (R = -0.29). Other variables measured at the bed level were not significantly correlated with E. lecontei density.

Correlations were also found bet,veen macrophyte bed characteristics and the percentage of various

E. lecontei lifcstages among all weevils collected per lake (Table 11 ). For instance, the percentage of E. lecontei adults, pupae and eggs was positively correlated with the percentage of broken apical tips (R =

0.34, 0.53, 0.36, respectively). Additionally, the percentage of E. lecontei adults was positively correlated with the number of apical tips per M spicatum plant (R = 0.86), and negatively correlated ,,.·ith depth of the M spicatum bed (R = -0.20) and the percentage of sandy shoreline (R = -0.26). The percentage of E. lecontei eggs was positively correlated with the middle (R = 0.28) and deep (R = 0.30) edge of the M spicatum bed.

The percentage of broken apical tips of M spicatum collected during density sampling were also correlated with some lake level and bed level variables (Table 12). First, the percentage of broken tips measured on the 120 stems collected during R. lecontei density sampling was positively correlated with the percentage of broken tips collected during M spicatum biomass and density sampling (R = 0. 76). Second, the percentage of broken tips was positively correlated with the surface area (R = 0.49) and maximum depth (R = 0.50) of the lake. Third, among bed level variables, the percentage of broken tips was positively correlated with the distance from shore to the middle (R = 0.27) and deep (R = 0.28) edges of the

M spicatum bed and the number of apical tips per M spicatum plant (R = 0.81). Fina11y, the percentage of broken tips was negatively correlated with M spicatum depth (R = -0.20).

38 Table IO. Spearman rank correlations between E. lecontei density at the bed level and characteristics of macrophyte beds measured at the bed level. Alpha was set at P ~ 0.05.

NUMBER OF CORRELATION P" VARIABLE OBSERVATIONS COEFFICIENT Percent broken tips 104 0.34002 0.0001

Depth of M. spicatum 122 -0.30400 0.0007

Distance from shallow bed edge to shore 112 ns

Distance from middle of bed to shore 108 0.24957 0.0092

Distance from deep bed edge to shore 108 0.28417 0.0029

Percent natural shore 99 0.21407 0.0334

Percent mowed grass shore 99 ns

Percent sand shore 99 -0.28563 0.0042

Percent wal1 or rip-rap shore 99 ns

Biomass of M. spicatum 12 ns

Biomass of native macroph)1es 12 ns

Stem length of Af spicatum 12 ns

Stem density of Af spicatum 12 ns

No. of apical tips per M spicatum plant 12 0.59828 0.0399

No. of apical tips m·2 12 ns

• ns denotes a non-significant correlation

39 Table 11. Spearman rank correlations between percent of E. lecontei lifcstages and characteristics measured at the bed level. Alpha v-:as set at PS 0.05.

NUMBER OF CORRELATION p VARIABLES CORRELATED OBSERVATIONS COEFFICIENT Percent adults vs. Percent broken tips 104 0.34406 0.0003

Percent pupae vs. Percent broken tips 104 0.53486 0.000)

Percent eggs vs. Percent broken tips 104 0.36144 0.0002

Percent adults vs. Number of apical tips 12 0.85563 0.0004

Percent adults vs. Depth of milfoil bed 122 -0.20112 0.0263

Percent adults vs. Percent sand shore 99 - 0.26480 0.0081

Percent eggs vs. Middle of bed to shore 108 0.28315 0.0030

Percent eggs vs. Deep bed edge to shore 108 0.30259 0.0015

40 Table 12. Speannan rank correlations bet\vecn percentage of broken apical tips onM spicatum stems collected during E. lecontei density sampling and various bed level and lake level characteristics. Alpha was set at P 5 0.05.

NUMBER OF CORRELATION P VARIABLE OBS. COEFFICIENT Percent broken apical tips collected during M 8 0.76190 0.0280 spicatum biomass and density sampling

Lake surface area 31 0.48773 0.0054

Maximum lake depth 31 0.49550 0.0046

Distance from middle of bed to shore 94 0.27160 0.0081

Distance from deep edge of bed to shore 94 0.28399 0.0055

Number of apical tips per M spicatum plant 8 0.80952 0.0149

Depth of M spicatum 104 - 0.19775 0.0442

41 DISCUSSION

E. lecontei Distribution

Although E. lecontei is native to Wisconsin and appears to be widespread across the temperate regions of North America {Sheldon and O'Bryan 1996a, Creed 1998), its detailed distribution was not well known. Prior to this study, its kno\\n distribution in Wisconsin was limited to four lakes: Bierbauer Pond in St. Croix County (Lillie 1991 ), Lake Wingra in Dane County (Lillie and Helsel 1997), Devils Lake in

Sauk County and Fish Lake in Dane County {Nc\\n1an and Maher 1995). All four lakes found to harbor E. lecontei contained M spicatum at the time of E. lecontei surveys except for Bierbauer Pond, which contained northern watermilfoil (A1yriophyllum sibiricum). Ne,,man and Maher {1995) searched for E. lecontei in three additional Wisconsin lakes containing M spicatum: Lake Mallalieu in St. Croix County,

Lake Onalaska in La Crosse County and Spring Lake in Buffalo County, but failed to find the ,,·eevil in • these lakes. However, these lakes may harbor the weevil at low levels which were undetectable in their surveys. Nev.man and Maher (1995) also searched Bierbauer Pond (St. Croix County) and were unable to find the ,veevil despite an extensive search and previous collection of E. lecontei in this pond by Lillie

(1991).

This study shO\:vs that E. lecontei is geographically widespread throughout Wisconsin as 45 new records of E. lecontei from southern, central and northern parts of the state were found. More importantly, because nearly every lake surveyed in this study actually contained the weevil, it is likely that E. lecontei is widespread across most Wisconsin lakes containing M spicatum. In the only lake where weevils were not found, Silver Lake, the weevils may exist at a very low abundance, making it difficult to detect in our sampling. Our surveys also appear to show that the southeastern and central parts of the state had a higher number of lakes kno,,11 to harbor E. lecontei (Figure 8), but this may have been an artifact of sampling.

42 Higher numbers of lakes with M spicatum occur in these areas and there was additional help by the

WDNR in finding lakes with E. /econtei in these areas.

The widespread distribution of E. lecontei was not expected. Nev.man and Maher (1995) found E. lecontei in only IO of 25 lakes sampled in Minnesota and Wisconsin. In Washington, only 21 of 51 lakes surveyed were found to harbor E. lecontei (Mariana Tamayo, University of Washington, pers. comm.).

Generally, other states have not searched extensively for E. lecontei but, at least 12 Vermont, 7

Massachusetts, I New York, 2 Connecticut, 2 Michigan, 3 Ohio, 3 Illinois, and 5 British Columbia lakes are knO\vn to harbor E. /econtei (Creed 1998, Sallie Sheldon, Middlebury College, pers. comm.).

Although only lakes ,vith M spicatum were sampled in this study, previous studies have found E. lecontei associated ,vith northern watermilfoil, M sihiricum. Newman and Maher (1995) found E. lecontei in one lake having only M sihiricum during their surveys in Minnesota, and Creed and Sheldon

(1994) found E. /econtei associated with M sihiricum in 10 of 13 lakes sampled in Alberta, Canada.

These Canadian lakes contained no M spicatum and were sampled for the sole purpose of detcm1ining E. lecontei's original host plant. Preliminary research in Washington also found E. lecontei to be associated with M sihiricum (Mariana Tamayo, University of Washington, pers. comm.).

The decline of M spicatum in some lakes is being attributed to E. lecontei populations through association with natural declines, specific damage to the M spicatum typical of E. lecontei herbivory, and relatively high E. lecontei populations (Lillie and Helsel 1997, Creed and Sheldon 1995, Kirschner 1995).

Significantly more declines of M spicatum have occurred within the known geographic range of E. lecontei than would be expected by chance (G = 20.08 p < 0.00 I) (Creed 1998). Moreover, at least ten Wisconsin lakes found to harbor E. lecontei in this study or previous surveys (Ne,,man and Maher 1995, Lillie 1991 ), have experiencedM spicatum declines: Big Green Lake (Green Lake Co.), Devil's Lake (Sauk Co.), Fish

Lake and Lake Wingra (Dane Co.), Long-Trade Lake (Polk Co.), Kusel Lake (Waushara Co.), Whitewater

Lake (Walworth Co.), Yellow Birch Lake (Vilas Co.), Mukwonago Pond and Wind Lake (Waukesha Co.)

(Dan Helsel and Laura Herman, WDNR, pers. comm., personal observation). Interestingly, Carpenter

43 ( 1980) reported that the duration of peak abundance of M spicatum in many lakes is approximately 10 years before declines occur. However, the definitive causes of M spicatum declines have not been established (Lillie and Helsel 1997, Carpenter 1980) but warrant further investigation into predator-prey cycles between M spicatum and E. lecontei.

E. lecontei Life Historv Observations

The life history observations collected during density sampling agree with previous life history research on E. lecontei. Sheldon and O'Bryan (1996) reported that adult E. lecontei reside outside the stems, moving among plants as they cat and reproduce. They are usually found in the top meter of milfoil stems and lay eggs on the leaves of the apical tips of the plant. The hatching larvae consume and burrow into the apical meristems and subsequent instars burrow further dov,n inside the milfoil stems, consuming vascular tissues. Larvae sometimes burrow to the outside of a stem and crawl along it until burrowing back into the stem on the same plant, or they may crawl onto the stems of adjacent plants (Sheldon and

O'Bryan 1996). Pupation also occurs in the stem. usually at or greater than 50 cm from the top of the plant (Sheldon and O'Bryan 1996). Our observations corroborate these reports. Eggs were always observed in the apical tips of the plant, and larvae and pupae were found in the stems. Larvae were occasionally found heading into the meristem after hatching or into larger stems from the outside. Our surveys also indicate that eggs and larvae arc more abundant than adults and pupae. The observation of fc,ver adults is intuitive due to the population d~11amies of insects. The lower number of pupae recorded may be related to the collection of only 50 cm of the milfoil stem, as they may be found lower in the plant.

Furthermore, three generations per summer have been documented in Vermont (Sheldon and O'Bryan

1996) and because Wisconsin is similar in latitude, it is expected that three generations per summer occur here.

44 Factors Influencing E. lecontei Densities

E. lecontei density varied greatly across the 31 lakes sampled as wellas among different M spicatum beds ,vithin lakes. Results indicate that some variables measured ,vithin the M spicatum beds may influence E. lecontei densities more than large-scale geographical and limnological variables as £. lecontei density was not correlated with variables measured at the lake level. Time since M spicatum invaded a particular lake did not appear to be a factor influencing E. lecontei density although it was speculated that the longer M spicatum had resided in a lake (i.e., years), the more time the weevil had to increase its populations on the M spicatum. Latitude was also not correlated with E. lecontei density, perhaps for two reasons. First, although M spicatum invasions have progressively moved north in the state over the years, time since invasion docs not appear to influence E. lecontei density. Second, climatic differences from north to south (such as later ice-out and lower water temperatures) may not be significant enough to induce measurable differences in E. lecontei densities.

Variables related to water chemistry were not significantly correlated with E. lecontei densities.

Water temperature and dissolved oxygen might still be related to E. !econtei densities within the individual macroph)'1e bed, but because they were measured at the lake's deepest point rather than in the littoral area, these measurements may not reflect conditions in the bed. Furthermore, dicl fluctuations in temperature and oxygen were also not measured. It is possible that these variables may be correlated with E. lecontei densities at the bed level and may be limiting populations in some lakes. Ne\\lnan (1996b) reported that E. lecontei development was temperature dependent and ranged from 60 d at 15 °C to 17 d at 27 - 31 °C. It is interesting to note that the percentage of E. !econtei eggs within a given lake was positively correlated with water temperature which ranged from 20.3 °C to 27.5 °C across the lakes. E. lecontei egg production is higher and hatching rate is faster in ambient water temperatures (24 - 26 °C) than in higher temperatures

(28 - 34 °C) (Sheldon 1997). Therefore perhaps overall lake water temperature is a limiting factor in egg

45 production. Future work in this area should measure water temperature and dissolved oxygen within the habitat of the M spicatum bed.

The percentage of E. lecontei larvae was negatively correlated with total phosphorus. How€Ver, the opposite relationship might be expected due to the larval damage and subsequent dying of M spicatum stems. Decomposing plant material can be a significant source of phosphorus in littoral areas, although up to 75% ofleached phosphorus is in a soluble form ,\·hich can be rapidly assimilated by epiph~1on and periphyton (Carpenter and Lodge 1986). The total phosphorus measurement used in this analysis was collected from the deepest point in the lake, usually far removed from the littoral area. Furthermore, it was collected during the spring or fall turnover and ,vas not necessarily collected during the same year as weevil density sampling. Therefore, the phosphorus levels among dying plants in the M spicatum bed may not be reflected in this variable and cannot reasonably be attributed to the higher levels of E. lecontei larvae.

Although the percentage of adults and the percentage of eggs ,vere significantly correlated with the date of collection, sampling bias is not suspected to play a role in this relationship. Sampling methods remained consistent throughout each summer and between the two summers and the possible bias of collection time within the season was minimized by sampling southern lakes first, then central lakes, follO\vcd by northern lakes and completing the sampling on all lakes within a five-week period.

Furthermore, the positive correlation bet\vecn percentage of eggs and year of collection is related to higher lake water temperatures sampled during the second year. The year of sampling and water temperature was significantly positively correlated (R = 0.51 ).

Bed level Variables

Variables related to M spicatum stem density and biomass did not correlate with E. lecontei density, yet the herbivory of the weevil on M spicatum has been shO\vn to significantly reduce its standing biomass (Creed and Sheldon 1993, 1995, Creed ct al. 1992, Sheldon and Creed 1995, Newman et al.

1996b ). However, because M spicatum biomass and other data were only collected from 12 lakes, there

46 may not be enough observations among these variables to elucidate a true relationship. Furthermore, temporal trends in M spicatum density and biomass could not be detected as only one year of data was

_collected. Thus, the influence of the d}namiGs of predator-prey cycles between M spicatum and E. lecontei were not measurable in this study. It is possible that in lakes where we measured high E. lecontei densities, the M spicatum biomass during that year was less than in previous years and was actually being affected by the weevils. The two lakes with the highest measured E. lecontei densities (Lake Wingra and

Yellow Birch Lake) have experienced dramatic declines in M spicatum in the past. In fact, M spicatum was very difficult to find in Yellow Birch Lake (there was barely enough to sample) and it is possible that

E. lecontei is currently controlling the M spicatum in this lake.

E. lecontei densities were positively correlated with the percentage of broken apical tips. This relationship has biological relevance due to larvae damage apical tips and stems just below the tip. In samples, fragile tips were often observed in conjunction with past or present larval herbivory and these tips were often broken prior to sampling or broke off during handling. The percentage of broken tips was also positively correlated with the size and depth of the lake. This may indicate that greater wave action generated in larger, deeper lakes breaks more apical tips. However, it is also likely that the greater wave action is more effective at causing damaged and fragile tips to break off. Due to the significant correlation between the percentage of damaged tips and the density of E. lecontei and the biological relevance of the correlation, the amount of milfoil tip damage can be used as an indication of E. lecontei abundance.

E. lecontei densities were also positively correlated with the distance from shore to the middle and deep edges oftheM spicatum bed and negatively correlated ,vith the depth oftheM spicatum bed. These relationships indicate that weevils arc more abundant in large, shallow expanses of M spicatum rather than deep (and perhaps non-surfacing) M spicatum closer to shore. This may characterize the type of M spicatum bed where E. lecontei has the best reproductive success. Perhaps the weevils are able to disperse more effectively throughout large, shallow beds by easily moving among milfoil plants and laying eggs as they move through the bed. Plants in shallow water arc more likely to be at or near the water's surface

47 ,vhere adult weevils collect oxygen in a plastron (i.e., incompressible gill). Although many aquatic insects with plastrons never need to surface to replenish their air stores, insects in low oxygen conditions must hwe the ability to come to the surface for air (Eriksen et al. 1996). Shallow plants may allow more time for feeding and reproducing and less time traveling to the surface to acquire air. In addition, higher temperatures in shallow water and water near the surface may promote faster E. lecontei development and egg production, thus increasing their densities.

The positive correlation between E. lecontei densities and the distance from shore to the deep edge of the M spicatum bed was surprising. E. lecontei oveminter on shore and move back to the mil foil bed in the spring (Newman et al. 1997). Damage to milfoil beds from E. lecontei begins near shore in early summer, presumably as weevils begin to attack milfoil closest to their overwintering habitat; damage appears to move further from shore as the summer progresses (Newman and Ragsdale 1995, Richard

Lillie, WDNR, Sallie Sheldon, Middlebury College, pers. comm.). However, it is still unknown how the weevils move from shore to the milfoil beds. Ne,,man et al. (1997) reported that although they have developed wing muscles and are capable of flight, few E. lecontei were found flying in the spring and adults were readily found along the shoreline feeding on floating milfoil stems. However, it is possible that the weevil may not merely be moving onto milfoil closest to shore in the spring, but flyiHg to a spot within the milfoil bed. Thus, weevils may be dispersing randomly throughout the bed and starting to feed and reproduce where they land, which could be far from shore.

E. lecontei densities were also positively correlated with the number of apical tips per M spicatum plant. It is unkno\\-n which variable is the dependent variable in this relationship. Perhaps weevils can reproduce more effectively in milfoil beds with higher numbers of apical tips because there are more oviposition sites. In contrast, perhaps the M spicatum plants are responding to high weevil densities and greater amounts of tip damage by growing new apical tips. Johnson et al. (1998) reported that herbivore damage to the apical meristems of M spicatum negatively effects stem elongation and forces the plant to invest in side shoots and thus more apical tips.

48 Finally, at the bed level, E. lecontei densities were positively correlated with the percentage of

natural shoreline. E. lecontei overwinter in the leaf litter and mud along the shore within a few meters of

1he water (Newman et al. 1997). It is possible that E. lec-0ntei are more abundant along natural shorelines

because they are more successful at overwintering in these areas. Habitats other than a natural shore, such

as a seawall, rock rip-rap, sand, or mown grass, may not offer enough protection or burrowing capabilities

for hibernating adult weevils. Unfortunately, more specific characteristics of the natural shorelines were

not measured in this study and the majority of beds sampled were adjacent to natural shorelines.

Variation in E. lecontei density across lakes could be caused and compounded by many variables

not measured in this study. In fact, large fluctuations in population sizes are not unusual in many species

(Pielou 1974). Density-dependent factors may influence population sizes of E. lecontei, especially ,1vhen

small beds ofM spicatum prevent their dispersal and relief from overcrowding. Ovipostition sites may

also become a limiting factor in small M spicatum beds, thus slowing population gro,..,th after one or two generations. However, a more likely cause of fluctuating E. lecontei populations within a lake could be

explained by predator-prey cycles whereby an increase in M spicatum is followed by an increase in E.

lecontei causing a subsequent decrease inM spicatum followed by a decrease in E. lecontei. Fluctuating

relations between herbivores and plants is even more profound when the herbivore is species specific.

Thus, a decline in M spicatum would most certainly mean a subsequent decline in E. lecontei. Lillie

(WDNR, pers. comm.) reported large fluctuations in E. lecontei populations from an ongoing study on Fish

Lake in Dane Co., Wisconsin. Occupancy rates of E. lecontei in M spicatum stems there (i.e., the

percentage of stems with E. lecontei) fluctuated over the years from 18% in 1992, to 3% in 1995, and back

to 18% in 1997. Furthermore, Fish Lake M spicatum biomass seems to fluctuate in response to E.

lecontei herbivory. Lillie (WDNR, pcrs. comm.) and Sheldon (1997) have both reported a '"lag" bet\veen

E. lecontei abundance and M spicatum biomass which suggests the presence of predator-prey-induced

oscillations. The data collected in this study docs not allow us to decipher where E. lecontei density is

located on the predator-prey curve for a given lake because weevils were only sampled at one point in time

49 and only during one year. Hence, at the same level of abundance, their densities may be on the rise in one lake, at the apex in another and on the decline in yet another.

Predation on E. lecontei by. sunfish (Lepomis sp.) and other smaH fish may have influenced E. lecontei densities across the lakes. In Lake Auburn Minnesota, Sutter and Nc,,man ( 1997) found that the frequency of occurrence of E. lecontei adults and larvae in the stomachs of bluegills (Lepomis macrochints) and pumpkinseeds (Lepomis gibbosus) ranged from 10.3% in September to 28.6% in

August. They concluded that although there was no correlation between E. lecontei density and either mean number or percent occurrence in sunfish stomachs, sunfish predation could be an important factor regulating their abundance, especially in systems with low E. lecontei populations and high sunfish populations. Similarly, Newbrough (1993) found that total weevil density was reduced in enclosures with high L. macrochints densities. She concluded that high densities of littoral zone fish may reduce the survival of E. lecontei but low fish densities may not affect E. lecontei densities. It is probable that sunfish population size and size-structure vary widely across lakes sampled for E. lecontei and thus may exhibit varying degrees of pressure on the E. lecontei populations in these systems.

Finally, studies on the genetic variation among M spicatum populations indicate that more than one M spicatum genotype exists in the Midwest (Fumier and Mustaphi 1992, Fumier et al. 1995). In fact,

Fumier et al. ( 1995) suggests that genetic variations in M spicatum be considered when assessing the efficacy of control agents. It is possible that the various genotypes inhibit or promote the gro,,th of E. lecontei populations in different ways.

50 Conclusion

The widesprnad distribution- of E. lecontei in Wisconsin has implications for its potential use as a biological control agent for M spicatum as this organism may be effective across a wide range of lake types and locations. And, as a native species, the weevil could be widely stocked without introducing a new species to an aquatic system. As already documented in previous studies, E. /econtei would likely have little effect on native macroph}tcs because these species coevolved: there is a lack of suitable oviposition sites on most native macrophytes and E. lecontei has a high specificity for M. spicatum and thus is not likely to consume and destroy native foliage. Only one drwback has been raised regarding the use of E. /econtei to control M spicatum. Engel (l 995) argued that a significant reduction in M spicatum biomass might create excess open water habitat and negatively affect fish cover and forage in some lakes.

However, replacement ofM spicatum by native macroph)tes appears to occur quickly in Fish Lake (Lillie pers. comm.) and thus an excess of open water habitat is unlikely.

In addition to documenting the geographic distribution of E. lecontei, the main objective of this study was to determine if lake characteristics could help predict where the weevils are most abundant.

Although geographical and large-scale limnological factors do not appear to influence weevil abundance as measured in this study, specific characteristics within and adjacent to M spicatum beds are associated with varying weevil densities. The distance from shore and depth of the bed, the number of apical tips per plant and the percentage of natural shoreline adjacent to the bed appear to play a role in E. lecontei abundance in individual M spicatum beds across lakes and within lakes.

Natural declines in M spicatum are associated with the presence of E. lecontei in Wisconsin lakes and this may continue in the future. Future research needs to focus on determining factors that limit E. lecontei populations including ovenvintering habitat in riparian areas and factors affecting survival, predator-prey cycles between the milfoil and the weevil, and effects of fish predation on weevil abundance.

Although E. lecontei stocking may become a viable management tool for M spicatum control in some

51 lakes, there are bound to be lakes ,vhere this method of control will never work. Discovering the factors that limit natural E. lecontei populations will help lend insight into the types of lakes and M spicatum beds where large E. lecontei populations can be maintained and those where they can be established or managed through stocking.

Summarv

• E. lecontei are geographically widespread across Wisconsin lakes containing M spicatum and ,vere

found in 45 of 46 lakes surveyed for E. lecontei presence.

• E. /econtei densities vary among lakes but do not appear to be related to lake-wide characteristics such

as time since the M spicatum invasion, lake type, size and location, and water quality variables

including dissolved oxygen and water temperature.

• E. /econtei egg production may be temperature dependent due to a significant positive correlation

between percentage of eggs and lake water temperature.

• Factors influencing E. lecontei densities in specific M spicatum beds include depth of the bed, it's

distance to shore, amount of natural shoreline and the number of apical tips per M spicatum plant.

• The percentage of broken apical tips is significantly correlated with ,veevil densities and may be used

as an indicator of ,veevil abundance.

• Future research should focus on assessing additional factors regulating E. /econtei populations

including overwintering habitat, predator-prey relationships between M spicatum and E. /econtei and

fish predation.

52 LITERATURE CITED

Aiken, S. G., P.R. Newroth and I. Wile. 1979. The biology of Canadian weeds. 34. Myriophyllum spicatum L. Canadian Journal of Plant Science 59: 201-215.

Bates, A. L., E. R. Burns, and D. H. Webb. 1985. Eurasian watermilfoil (Myriophyllum spicatum L.) in the Tennessee Valley: an update on biology and control. Pages 104 - 115 in L. W. J. Anderson editor. Proceedings of the First International Symposium on the watermilfoil (Myriophyllum spicatum) and related Haloragaceae species. Aquatic Plant Management Society. Washington D.C.

Bode, J., S. Borman, S. Engel, D. Helsel, F. Koshere, and S. Nichols. 1993. Eurasian watermilfoil in Wisconsin: A report to the legislature. Wisconsin Department of Natural Resources.

Carpenter, S. R. 1980. The decline of Myriophyllum spicatum in a eutrophic Wisconsin lake. Canadian Journal of Botany 58: 527-535.

Carpenter, S. R. and D. M. Lodge. 1986. Effects of submersed macroph~1es on ecosystem processes. Aquatic Botany 26: 341-370.

Chilton II, E.W. 1990. Macroinvertebrate communities associated with three aquatic macroph~1es (Ceratophyllum demersum. Myriophyllum spicatum and Vallisneria americana) in Lake Onalaska, Wisconsin. Journal of Freshwater Ecology 5: 455-466.

Couch, R. and E. Nelson. 1985. Myriophyllum spicatum in North America. Pages 8 -18 in L. W. J. Anderson editor. Proceedings of the First International Symposium on the watermilfoil (Myriophyllum spicatum) and related Haloragaceae species. Aquatic Plant Management Society. Washington D.C. 8-18.

Creed, R. P., Jr. 1998. A biogeographic perspective on Eurasian watermilfoil declines: additional evidence of the role of herbivorous weevils in promoting declines? Journal of Aquatic Plant Management 36: 16-22.

Creed, R. P, Jr. and S. P. Sheldon. 1993. The effect of feeding by a North American weevil, Euhrychiopsis lecontei, on Eurasian watermilfoil (Myriophyllum spicatum). Aquatic Botany 45: 245-256.

Creed, R. P, Jr. and S. P. Sheldon. 1995. Weevils and watermilfoil: did a North American herbivore cause the decline of an exotic plant? Ecological Applications. 5: 1113-1121.

Creed, R. P., Jr., S. P. Sheldon and D. M. Check. 1992. The effect of herbivore feeding on the buoyancy of Eurasian watermilfoil. Journal of Aquatic Plant Management. 30: 75-76.

Crowder, L. B. and W. E. Cooper. 1982. Habitat structural complexity and the interaction between bluegills and their prey. Ecology 63: 1802-1813.

Cyr, H. and J. A. Do\\ning. 1988. The abundance of ph~1ophilous invertebrates on different species of submerged macroph~es. Freshwater Biology 20: 365-374.

53 Diehl, S. 1988. Foraging efficiency of three freshwater fishes: effects of structural complexity and light. Oikos 53: 207-214.

Dionne, M. and C. L. Folt. 1991. An experimental analysis of macrophyte growth forms as fish foraging habitat. Canadian Journal of Aquatic Science. 48: 123-131.

Engel, S. 1990a. Ecosystem responses to gro"th and control of submerged macroph)'tes: a literature review. Technical Bulletin Number 170. Wisconsin Department of Natural Resources 20 pp.

Engel, S. 1990b. Ecological impacts of harvesting macrophytcs in Halverson Lake, Wisconsin. Journal of Aquatic Plant Management. 28: 41-45.

Engel, S. 1994. Fish in troubled ,vaters: Eurasian watermilfoil as a threat to Wisconsin sport fish. Wisconsin Department of Natural Resources Fish Research Highlights January 1994.

Engel, S. 1995. Eurasian watermilfoil as a fishery management tool. Fisheries 20(3): 20-27.

Eriksen, C.H., G. A. Lamberti, V. H. Resh. 1996. Aquatic insect respiration. Page 29 in R.W. Merritt and K. W. Cummins, editors. An introduction tothe aquatic insects of North America. Dubuque, IA: Kendall Hunt.

Grace J. B. and R. G. Wetzel. 1978. The production of Eurasian watcrmilfoil (Myriophyllum spicatum L.): A review. Journal of Aquatic Plant Management. 16: 1-11.

Hanson, J.M. 1990. Macroinvcrtcbratc size-distributions of two contrasting fresh,vater macrophyte communities. Freshwater Biology 24: 481-49 l.

Hanson, T., C. Eliopoulos and A. Walker. 1995. Field collection, laboratory rearing and in-lake introductions of the herbivorous aquatic weevil, Euhrychiopsis lecontei, in Vermont: Year Two. Project report to the U.S. Environmental Protection Agency.

Johnson, R. L., E. M. Gross, N. G. Hairston, Jr. 1998. Decline of the invasive submersed macroph)'te Myriophyllum spicatum (Haloragaceacc) associated with hcrbivory by larvae of Acentria ephemerella (Lepidoptera). Aquatic Ecology. In press.

Johnstone, D. M., R. T. Coffey, and C. Howard-Williams. 1985. The role of recreational boat traffic in interlake dispersal ofmacrophy-tes: a New Zealand case study. Journal of Environmental Management. 20: 263-279.

Keast, A. 1984. The introduced aquatic macrophyte, Myriophyllum spicatum, as habitat for fish and their invertebrate prey. Canadian Journal of Zoology. 62: 1289-1303.

Kimbel, J. C. 1982. Factors influencing potential intralake colonization by M_vriophyllum spicatum L. Aquatic Botany 14: 295-307.

Kirschner, R. 1995. McCullom lake mil foil saga continues. Lake Waves 11: I - 5.

Lillie, R. A. 1991. The adult aquatic and scmiaquatic Coleoptera of nine northwestern Wisconsin wetlands. The Coleopterists Bulletin 45: 10 I -111.

54 Lillie, R. A. and D. Helsel. 1997. A native weevil attacks Eurasian watermilfoil. Wisconsin Department of Natural Resources Bureau of Research Management Findings No. 40. 4 pp.

Lillie, R. A. and J. W. Mason. 1986. Historical changes in water quality and biota of Devils Lake, Sauk County, 1866-1985. Wisconsin Academy of Sciences, Arts and Letters 74: 81-104.

Madsen, J. D., L. W. Eichler, and C. W. Boylen. 1988. Vegetative spread of Eurasian watermilfoil in Lake George, New York. Journal of Aquatic Plant Management. 26: 47-50.

Madsen, J. D., J. W. Sutherland, J. A. Bloomfield, L. W. Eichler and C. W. Boylen. 1991. The decline of native vegetation under dense Eurasian watermilfoil canopies. Journal of Aquatic Plant Management 29: 94-99.

Maxnuk, M. D. 1985. Bottom tillage treatments for Eurasian watermilfoil control. Pages 163- 172 in L. W. J. Anderson, editor. Proceedings of the First Jntemat_ional Symposium on the watermilfoil (Myriophyllum spicatum) and related Haloragaceae species. Aquatic Plant Management Society. Washington D.C.

Newbrough, K. L. 1993. The effect of bluegills (Lepomis macrochints) on the density and survival of an aquatic weevil. M.S. Thesis. University of Vermont, 55 pp.

Ne,vman, R. M. and D. W. Ragsdale. 1995. Evaluation of biological control agents for Eurasian watermilfoil. Report to Minnesota Department of Natural Resources, Ecological Services, St. Paul MN. 21 pp.

Newman, R. M. and L. M. Maher. 1995. New records and distribution of aquatic insect herbivores of watermilfoils (Haloragaceae: Myriophyl!um spp.) in Minnesota. Entomological News 106(1): 6- 12.

Newman, R. M, D. W. Ragsdale, and D. D. Biesbocr. 1996a. Can Eurasian watermilfoil be managed in Minnesota by biological control with native or naturalized insects? Second Progress Report submitted to Minnesota Department of Natural Resources, St. Paul, MN.

Newman, R. M., K. L. Holmberg, D. D. Bicsboer, and B. G. Penner. 1996b. Effects of a potential biocontrol agent, Euhrychiopsis lecontei, on Eurasian watermilfoil in experimental tanks. Aquatic Botany 53: 131-150.

Newman, R. M., M. E. Borman and S. W. Castro. 1997a. Developmental performance of the lveevil Euhrychiopsis lecontei on native and exotic watermilfoil host plants. Journal of the North American Benthological Society 16: 627-634.

Newman, R. M., D. W. Ragsdale, D. D. Bicsboer. 1997b. Can Eurasian watermilfoil be managed in Minnesota by biological control with native or naturalized insects? Final Report submitted to Minnesota Department of Natural Resources, St. Paul, MN.

Newroth, P.R. 1985. A review of Eurasian watermilfoil impacts and management in British Columbia. Pages 139 - 153 in L. W. J. Anderson, editor. Proceedings of the First International Symposium on the watcrmilfoil (Myriophyllum spicatum) and related Haloragaceae species. Aquatic Plant Management Society. Washington D.C.

55 Nichols, S. A. and B. H. Shaw. 1986. Ecological life histories of the three aquatic nuisance plants, Myriophyllum spicatum. Potamogeton crispus and Elodea canadensis. Hydrobiologia 131: 3- 21.

Pardue, W. and D. H. Webb. 1985. A comparison of aquatic macroinvertebrates occurring in association ,,-:-ith Eurasian watermilfoil (M_,vriophyllum spicatum L.) ,vith those found in the open littoral zone. Journal of Freshwater Ecology 3: 69-79.

Perry, J. A. and B. G. Penner. 1995. Evaluation of insects as biological control agents for Eurasian watermilfoil. Report to Minnesota Department of Natural Resources, Ecological Services, St. Paul,MN.

Piclou, E. C. 1974. Population and community ecology: principles and methods. New York: Gordon and Breach. p. 73 - 96.

Raw·son, R. M. 1985. History of the spread of Eurasian watermilfoil through the Okanogan and Columbia River systems (1978-1984). Pages 35 - 38 in L. W. J. Anderson, editor. Proceedings of the First International Symposium on the watermilfoil (Myriophyllum spicatum) and related Haloragaccae species. Aquatic Plant Management Society. Washington D.C.

Reed, C. F. 1977. History and distribution of Eurasian watermilfoil in United States and Canada. Ph:ytologia 36: 417-436.

Robinson, A. H., Editor. 1974. The atlas of Wisconsin: general maps and gazetteer. University of Wisconsin Press, Madison, WI p. 31 - I I 1.

Savino, J. F. and R. A. Stein. I 982. Predator-prey interaction between largemouth bass and bluegills as influenced by simulated, submersed vegetation. Transactions of the American Fisheries Society. 111: 255-266.

Shearer, J. F. 1994. Field test of first commercial formulation of Mycoleptodiscus terrestris (Gerd.) Ostazeski as a biocontrol for Eurasian watermilfoil. Miscellaneous Paper A-94-4, U.S. Army Engineer Waterways Experiment Station, Vicksburg, Mississippi.

Sheldon, S. P. 1997. Investigations on the potential use of an aquatic weevil to control Eurasian watermilfoil. Journal of Lake and Reservoir Management 13: 79-88.

Sheldon, S. P. 1994. Invasions and declines of submersed macroph)'tes in New England, with particular reference to Vermont lakes and herbivorous invertebrates in New England. Lake and Reservoir Management. I 0: 13-17.

Sheldon, S. P. and R. P. Creed, Jr. 1995. Use of a native insect as a biological control for an introduced weed. Ecological Applications 5: I 122-1132.

Sheldon, S. P. and L. M. O'Bryan I 996a. The life history of the weevil Euhrychiopsis lecontei, a potential biological control agent of Eurasian ,vatcrmilfoil. Entomological News 107: 16-22.

Sheldon, S. P. and L. M. O'Bryan 1996b. The effects of harvesting Eurasian watermilfoil on the aquatic weevil Euhrychiopsis lecontei. Journal of Aquatic Plant Management 34: 76-77.

56 Smith, C. S. and J. W. Barko. 1990. Ecology of Eurasian watermilfoil. Journal of Aquatic Plant Management. 28: 55-64.

Soar, R. J. 1985. Laboratory investigations on ultrasonic control of Eurasian watermilfoil. Pages 173 - 186 in L. W. J. Anderson, editor. Proceedings Qf the First International Symposium on the watermilfoil (Myriophyllum spicatum) and related Haloragaceae species. Aquatic Plant Management Society. Washington D.C.

Solarz, S. L. and R. M. Nev,man. 1996. Oviposition specificity and behavior of the watermilfoil specialist Euhrychiopsis lecontei. Oecologia 106: 337-344.

Sutter, T. J. and R. M. Newman. 1997. Is predation by sunfish (Lepomis spp.) an important source of mortality for the Eurasian watermilfoil biocontrol agent Euhrychiopsis lecontei? Journal of Freshwater Ecology. 12: 225-234.

Wisconsin Department ofNatural Resources. 1995. Wisconsin Lakes. Wisconsin Department of Natural Resources PUB-FM-800 95REV. I 82 pp.

57 APPENDICES

58 Appendix A. Independent variables tested for correlation with E. lecontei densities at the lake level.

Variable Units Description Method of Measurement 1/ Source

Northing Meters Universal Transverse Mercator northing coordinate Calculated in Geographic Calcula1tor 3.05 from latitude/longitude coordinates (Robinson 1974)

Easting Meters Universal Transverse Mercator easting coordinate Calculated in Geographic Calculaltor 3.05 from latitude/longitude coordinates (Robinson 197-l)

Year Two digit year Year data were collected on lake ( 1996 or 1997)

Date 3 digit within - Numerical representation of day and month (did not Recorded in field notes and transl!tted to 3 digit year Julian date consider year) E. lecontei density sample was collected numerical date

True Julian Date Integer Last three digits of actual Julian date when data were Retrieved from Julian calendar collected on lake: day, month and year

Invasion Year Last two digits Actual year ,U spicatum invaded (or was discovered in) Collected from WDNR stalfwhen known from of year lake historical records

Invasion Time Integer Approximate time period (within 0.5 decade) that M. Collected from WDNR stalfwhen known spicatum invaded lake l =l 965-1969; 2=1970-1974; 3=1975- 1979; 4= 1980-1984; 5= 1985-1989; 6= 1990-1994; 7 = 1995- 1999

Maximum Lake Depth Meters Maximum depth of lake WDNR(1995)

Mean Lake Depth Meters Mean depth of lake WDNR (1995)

Lake Type Integer Type of lake: drainage ( l ), seepage (2), spring fed (3) WDNR(1995)

Water Temperature Celcius Summer surface water temperature Collected at lake's deepest point, usually 011 day of weevil density collections with YSII probe

Dissolved Oxygen mg/L Summer dissolYed oxygen Collected on the surface at lake's deepest point, Vt \0 usually on day of weevil density collections with YSI robe . Appendix A. Continued

Variable Units Description Method of Measurement Y Source

Secchi Disk Depth Meters Water clarity based on the lowest depth the black and Collected at lake's deepest point, ,sually on day of white disk is visible weevil density collections with standard Secchi disk

The following water quality variables were primarily collected by WDNR for each lake between 1993 - 1997. If available, the spring sample was used in analyses. If more than one year of spring data existed, one year was randomly chosen. If a spring sample was not available, a fall sample was chosen. Data was gathered from EPA's STORET system.

Total Phosphorus mg/L Total concentration of all soluble phosphorus Collected as spot sample from surface at lake's deepest point

Nitrate plus Nitrite mg/L Concentration of nitrate - nitrite Collected as spot sample from surface at lake's deepest point

Total Kjeldahl Nitrogen Ill!.' I. Collected as spot sample from surface at lake's deepest point

Chlorophyll a ug/L Amount of chlorophyll a Collected as spot sample from sudhce at lake's deepest point

Secchi Disk Depth (DNR) Meters Water clarity based on the lowest depth the black and Collected at lake's deepest point, ,~•ith standard white disk is visible Secchi disk

Total Alkalinity mg/L CaCO3 Total concentration of alkalinity in water Collected as sample from surface at lake's deepest point

pH Standard Units Measurement of the amount of hydrogen ions Collected as sample from surface at lake's deepest point

Conductivity Micromhos at Concentration of conductive ions Collected as sample from surface Ht lake's deepest 25°C lOint

0\ 0 Appendix B. Independent variables tested for correlation with E. lecontei densities at the bed level.

Variable Units Description Method of Measurement / Source

Al. spicatum biomass g/m2 diy weight Mean biomass of M. spicatum in 16 samples Collected using SCUBA from XI. spicatu111 beds in a 0.15 m2 in size and extrapolated to 1.0 m2 subset of 12 lakes; dried 48 hrs in ov¢n

Native biomass g/m2 dty weight Mean biomass of native macroph)1es in 16 Collected using SCUBA from M. spiaatu111 beds in a samples 0.15 m2 in size and extrapolated to 1.0 subset of 12 lakes; dried 48 hrs in oven ll12

1\f. .,picatum density s,~111s/1112 Mean number of Al. jpicatum plants in 16 Collected using SCUBA from M. spiaatum beds in a samples 0.15 m2 in size and extrapolated to 1.0 subset of 12 lakes; plants counted while floating in nt2 long tank of water (aquaculture raceway)

A1f. spicatum stem length Centimeters Mean length of plant from 16 samples: 30 Plants floated aquaculture raceway; 30 randomly plants (stems) per sample measured for length selected plants from aboye biomass samples; measured for length with tape measmte

Tips per stem Tips/ Mean number of apical meristems in 16 Plants floated aquaculture raceway; 30 randomly Al.. ,picatum samples: tips (including those broken oil) selected plants from above biomass samples; counted stem counted on 30 plants (stems) per sample tips per plant

Tips per nl Al. spicatum Mean number of apical meristems per m2 in 16 Calculated by multiplying tips per stem by stems/m2 tips/m2 samples

Percentage broken tips Percent hrokcn Mean percentage of broken apical meristems in Calculated by dividing number of broJ

M. spicatum depth Meters Mean depth of M spicatum along weevil Measured at each sampling point using a fishing reel density transect fitted with measured and marked nylon cord and a weight

Al. spicatu111 center distance Meters Distance from the center of the 1\ I. .,picatum bed Measured with a tape measure. digital range finder or to shore visually estimated

°'.... M spicatum minimum distance Meters Distance from shallow edge of M. spicatum bed Measured with a tape measure, digital range finder or to shore visually estimated Appendix B. Continued

Variable Units Description Method of Measurement

AI. spicatum maximum Meters Distance from deep edge of M. spicatum bed to Measured with a tape measure, digital range finder or distance shore visually estimated

Natural shoreline Percent Percentage of shoreline left natural (e.g., cattail Visually estimated along the sampled bed or at the beds, bmsh, wooded areas) end of a single transect if entire littoral area was sampled for E. lecontei density

Grass shoreline Percent Percentage of shoreline with mowed grass to the Visually estimated along the sampled bed or at the water's edge end of a single transect if entire littoral area was sampled for E. lecontei density

Sand shoreline Percent Percentage of shoreline with sand at the water's Visually estimated along the sampled bed or at the edge end of a single transect if entire littoral area was sampled for E. lecontei density

Wall/riprap shoreline Percent Percentage of shoreline with rock riprap or sea Visually estimated along the sampled bed or at the wall end of a single transect if entire littoral area was sampled for E. lecontei density

0\ N Appendix C. Raw data oflake level variables used in Speam1an correlation analyses. "na" indicates that data was not available.

YearofM Ambient Yearof Date of Julian date of spicatum Lake name lakes1 County WBIC Northing Easting collection collcction2 collection invasion3 Alpine 0 Waushara 245650 4879604 330439.4 1997 198 646 na Beaver Dam 0 Barron 2081200 5056386 107119.5 1996 218 300 91 Beulah 0 Walworth 766600 4743235 386941.7 1997 202 650 na Big Green l Green Lake 146100 4851604 339112.7 1997 216 664 71 Big Sand 0 Vilas 1602600 5103367 34661 l.6 1996 222 304 85 Camp 0 Kenosha 7-t7I00 4707784 405533.5 1997 209 657 na Crystal 1 Sheboygan -t.:-200 4850162 418216.2 1997 211 659 na Delavan 0 Walworth 7')3(,00 4717638 368742.4 1997 209 657 na Eagle 0 Racine 759800 4728124 407175.2 1996 199 281 na Fox l Dodge 835800 4827412 343914.6 1997 223 671 74 Gilbert 0 Waushara 186400 4898221 326921.6 1996 211 293 93 Jordan 0 Adams 104000 4847520 286645.l 1997 219 667 92 Kangaroo 0 Door 98600 4986667 486871.7 1996 227 309 na Kusel 0 Waushara 189600 4892667 326775.2 1996 212 294 na L. Spring 0 Jefferson 820800 4749050 372061 1996 204 286 na Lac La Belle I Waukesha 848800 4776738 376645.7 1997 210 658 na Little Falls 0 St. Croix 2607400 5000629 50883.37 1996 227 309 na Lorraine 0 Walworth 777500 4732659 358107.7 1996 206 288 na Manson 0 Oneida 1517200 5049276 294510.4 1997 225 673 89 Mason 0 Adams 175750 4836285 290322.3 1996 226 308 92 Metonga 0 Forest 394400 5045807 351691 1997 226 674 94 Mukwonago 0 Waukesha 767400 4746937 387002.5 1996 192 274 na Nancy 0 Washburn 2691500 5117242 114843.2 1996 219 301 89 Parker 0 Adams 106500 4849328 288045.9 1997 219 667 95 Pearl 0 Waushara 195400 4883307 330534.5 1996 211 293 na Ripley l Jefferson 809600 4762723 338339.5 1997 212 660 na Rock l Jefferson 830700 4771883 342628.3 1996 226 308 na SilYer l Waupaca 198800 4925920 330307.4 1996 225 307 na Whitewater Walworth 81(,800 4736305 360911 1996 205 287 na Wingra 0 Dane 805000 4769202 303177.2 1997 213 661 na Y. Birch 0 Vilas 1599600 5089066 326861.6 1997 225 673 92 °'w 1Ambient lakes = I; non-ambient lakes = 0 2Julian date of collection within one year Appendix C. Continued

TimeofM. Maximum spicatum lake depth Mean Water Dissolved o;-..}'gen Secchi disk Lake name invasion1 (111) depth (m) Hectares Lake type2 temperature (0 C) (mg/L) depth (111)3 Alpine na 5.5 na 22.4 na 27.5 9.9 2.4 Beaver Dam 6 32.3 0.9 444.8 2 22.7 9.4 4 Beulah na 17.7 5.2 333.6 I 25.7 8.8 2.7 Big Green 2 71.9 31.7 2938.4 I na na na Big Sand 5 19.8 4.9 563.2 1 22.6 10.7 3 Camp na 5.2 1.5 184.4 I 27.2 8.6 l.4 Crystal 3 18.6 6.1 60.8 2 24.7 8.3 4.9 Delavan na 17.1 6.4 828.8 1 25.4 8.3 4 Eagle 5 3.7 l.8 206.0 1 24.9 8 0.9 Fox 2 5.8 2.1 1050.0 l 22.4 9.3 na Gilbert 6 19.8 na 56.4 2 24 10.7 5.2 Jordan 6 24.1 na 85.2 2 na na na Kangaroo na 3.7 l.8 449.2 1 23.4 9.9 1.7 Kusel 6 8.8 na 31.6 2 23.l 8.5 3.6 L. Spring 1 3.4 l.2 41.6 I 26.9 8.15 1.7 Lac La Belle na 13.7 3.4 465.6 l 25.5 7.6 2.3 Little Falls 6 5.5 2.4 68.8 2 na na na Lorraine na 2.4 na 53.2 2 23.4 8.8 1.4 Manson 5 16.5 na 94.4 l na na 2.7 Mason 6 2.7 2.1 342.0 l na na na Metonga 6 2-U 7.6 862.8 I na na na Mukwonago 5 ()_l) 0.6 6.8 3 20.3 7 na Nancy 5 l l.9 3.4 308.8 2 24.9 10.5 2.9 Parker 7 9.1 4.0 23.6 2 na na 3 Pearl na 15.2 7.6 36.8 2 23.9 9.2 5.5 Ripley 2 13.4 5.5 167.2 2 25.8 9.3 2.3 Rock 2 17. l 4.9 548.4 l 24.2 8.1 na Sih·er na 5.2 2.1 27.2 2 22.5 7.4 na Whitewater 3 I l.6 na 256.0 25 10.4 1.7 Wingra l 6A na 138.0 24.4 9.1 0.4 Y. Birch 6 7.0 3.4 80.8 na na na °'+" 1Time ofM spicat11m invasion l = late l 960's, 2 = early l 970's, 3 = late l 970's, 4 = early l 980's, 5 = late l 980's, 6 = early l 990's, 7 = late l 990's 2Lake type l = Drainage, 2 = Seepage, 3 = Spring 3Secchi disk depth measured during weevil density sampling Appendix C. Continued Season of Total water quality Total Nitrate plus Kjeldahl sample phosphorus nitrite nitrogen Chlorophyll a Secchi disk Total alkalinity Lake name collection (mg/L) (mg/L) (mg/L) (ug/L) depth (m) 1 (mg/L CaC03) pH Conductivity

Alpine fall 0.014 na na 6.44 na na na na BeaYerDam spring 0.01 0.007 0.3 5A9 3.6 77 8 165 Beulah spring 0.011 na na 2.6 na na na na Big Green spring 0.027 0.3 0.4 6.63 na 189 8.4 481 Big Sand fall o.1n2 na na na na 136 7.3 75 Camp spring 0,014 0.09 1 16.4 1.2 158 8.6 512 Crystal spring 0,(105 0.09 0.4 3.2 3.8 158 8.5 373 Delayan spring 0.049 na na na na na 8.4 796 Eagle spring 0.033 0.3 0.9 11 na 153 8.1 493 Fox spring 0.176 1.7 1.8 94.5 0.5 182 8.3 469 Gilbert spring 0.007 0.06 0.71 2.09 7.4 139 8.6 273 Jordan spring 0.013 na na na na na na na Kangaroo spring 0.013 na na 2.37 na 648 8.4 317 Kusel fall 0.029 na na na na 380 7.7 184 L. Spring fall 0.019 na na na na 857 8.5 474 Lac La Belle spring 0.012 0.6 0.9 7.93 2.8 207 8.3 540 Little Falls spring 0.084 1.8 0.6 47.4 l 154 8.7 338 Lorraine fall 0.049 na na na na 517 8.1 286 Manson spring 0.009 0.01 0.3 1.92 5.2 27 7.6 76 Mason spring 0.029 0.2 0.6 16.3 1.5 173 8.3 355 Metonga na na na na na na na na na Mukwonago spring 0.03 0.5 0.9 22 na 753 8.52 468 Nancy fall 0.027 na na na na 117 7.3 64 Parker na na na na na na na na na Pearl fall 0.011 na na na na 470 7.9 222 Ripley spring 0.021 0.4 1.1 12.5 2.3 207 8.3 504 Rock spring 0.01 0.08 0.9 5.25 188 8.4 428 SilYer spring 0.066 0.007 l 9.35 2.1 76 7.9 188 Whitewater spring 0.025 0.01 0.8 20.9 I 174 8.5 367 Wingra na na na na na na na na na °'V, Y. Birch fall 0.031 0.012 0.6 15.1 2.2 28 7.56 na Secchi disk depth measured by WDNR Appendix D. Raw data of bed level variables used in Spearman correlation analyses. "na" indicates that data was not available.

E. lecontei ~r agical stem E. lecontei ~r agical tiQ Lake Name Bed I Bed2 Bed 3 Bed4 Bed I Bed 2 Bed 3 Bed4 -- Alpine 0.0 0.0 0.7 0.3 0.0 0:-0 0.1 0.1 Beaver Dam 2.6 2.6 1.0 1.0 0.7 0.6 0.3 0.3 Beulah 0.0 0.0 0.2 0.3 0.0 0.0 0.1 0.1 Big Green 0.0 0.2 0.8 0.6 0.0 0.0 0.2 0.1 Big Sand 0.1 0.7 0.0 0.6 0.0 0.3 0.0 0.3 Camp 0.7 0.2 0.7 1.2 0.1 0.0 0.2 0.2 Crystal 0.2 0.1 0.0 0.0 0.0 0.0 0.0 0.0 Delavan 0.2 2.1 1.0 1.6 0.1 0.2 0.2 0.2 Eagle 0.2 0.0 0.0 0.0 na na na na Fox 0.3 1.3 0.1 1.4 0.0 0.2 0.0 0.2 Gilbert 0.0 0.2 0.1 0.0 0.0 0.1 0.1 0.0 Jordan 0.3 0.2 0.4 0.2 0.1 0.0 0.1 0.1 Kangaroo 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Kusel 0.1 0.1 0.0 0.2 0.1 0.0 0.0 0.1 L. Spring 0.2 2.0 0.7 1.1 0.0 0.5 0.3 0.3 Lac La Belle 0.5 0.4 1.1 1.2 0.1 0.1 0.1 0.2 Little Falls 0.1 0.3 na na 0.1 0.1 na na Lorraine 1.2 2.0 1.8 2.6 0.3 0.5 0.3 0.3 Manson 0.0 0.5 0.6 0.0 0.0 0.2 0.2 0.0 Mason 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Metonga 0.1 0.1 0.2 1.6 0.1 0.0 0.1 0.2 Mukwonago 0.3 0.2 0.0 0.6 na na na na Nancy 2.2 0.5 0.9 I.I 0.5 0.2 0.4 0.4 Parker 0.1 0.1 0.4 OA 0.0 0.0 0.1 0.1 Pearl 0.4 0.7 0.1 0.1 0.1 0.2 0.0 0.0 Ripley 0.7 0.7 1.0 I.I 0.3 0.2 0.2 0.3 Rock 0.2 0.3 0.0 0.1 0.1 0.1 0.0 0.0 Silver 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Whitewater 1.6 0.8 2.2 1.0 0.3 0.2 0.4 0.3 Wingra 3.2 2.2 0.9 2.6 0.5 0.4 0.3 0.5 Y. Birch 1.9 1.6 2.7 3.6 0.3 0.4 0.6 0.7

66 Appendix D. Continued

Percentage broken Ai. se_icatum tigs Percentage stems with r,·. lecontei Lake Name Bed 1 Bed2 Bcd3 Bed4 Bed I Bed2 Bed 3 Bed4 Alpine 8.5 9.6 25.2 11.7 0.0 0.0 37.9 20.0 Beaver Dam 32.7 37.6 31.5 32.2 70;0 83.3 46.7 56.7 Beulah 26.0 25.4 17.4 33.7 0.0 0.0 0.0 22.2 Big Green 20.7 16.2 39.9 24.6 0.0 16.7 50.0 36.7 Big Sand 16.0 14.3 1.7 14.3 10.0 46.7 0.0 22.6 Camp 11.5 15.4 13.3 17.6 33.3 13.3 33.3 26.7 Crystal 3.1 7.1 0.0 5.4 13.3 6.7 0.0 0.0 Delavan 10.3 25.5 29.4 30.5 13.3 76.7 46.7 63.3 Eagle na na na na na na na na Fox 12.3 14.7 27.5 27.0 26.7 40.0 3.3 46.7 Gilbert 4.1 2.3 3.8 1.5 3.3 10.0 6.7 0.0 Jordan 15.1 14.5 16.7 6.7 20.0 13.3 17.2 17.2 Kangaroo 10.2 8.3 16.7 17.7 0.0 0.0 0.0 0.0 Kusel 30.0 27.6 21.6 20.0 10.0 6.7 0.0 13.3 L. Spring na na na na na na na na Lac La Belle 33.8 30.8 32.0 30.7 30.0 23.3 56.7 53.3 Little Falls 18.2 1.0 na na 6.7 16.7 na na Lorraine na na 25.1 35.3 50.0 66.7 56.7 66.7 Manson 6.6 21.9 23.5 14.9 0.0 13.3 23.3 3.3 Mason 13.2 9.7 8.6 10.7 6.3 0.0 0.0 0.0 Metonga 11.3 17.6 9.0 24.9 6.7 6.7 16.7 80.0 Mukwonago na na na na 20.0 13.3 3.3 33.3 Nancy 21.0 23.0 28.2 35.6 83.3 33.3 40.0 50.0 Parker 12.5 15.3 22.0 10.2 6.7 3.3 26.7 26.7 Pearl 4.2 2.2 4.5 10.0 16.7 30.0 6.7 6.7 Ripley 25.0 31.4 37.6 34.5 30.0 50.0 40.0 56.7 Rock 21.1 35.9 5.8 25.9 16.7 20.0 0.0 3.3 Silver 9.3 4.2 4.9 1.8 0.0 0.0 0.0 0.0 Whitewater na na na na na 36.7 80.0 60.0 Wingra 15.1 5.4 14.1 15.9 76.7 26.7 46.7 53.3 Y. Birch 18.5 17.8 27.6 20.5 57.1 70.0 80.0 76.7

67 Appendix D. Continued

Percentage of adult E. /econtei Percentage oflarval E. lecontei Lake Name Bed I Bed 2 Bed3 Bcd4 Bed I Bcd2 Bed3 Bed4 Alpine 0.0 0.0 10. 0.0 0.0 0.0 20.0 66.7 IreaverDam 5. l 7.6 6:9 6.5 35.4 44.3 58.6 41.9 Beulah 0.0 0.0 14.3 0.0 0.0 0.0 71.4 77.8 Big Green 0.0 0.0 4.0 0.0 0.0 60.0 20.0 47.1 Big Sand 0.0 15.0 0.0 5.9 0.0 55.0 0.0 29.4 Camp 9.1 0.0 28.6 13.5 27.3 16.7 14.3 10.8 Crystal 0.0 50.0 0.0 0.0 50.0 50.0 0.0 0.0 Delavan 0.0 9.7 19.4 26.5 50.0 19.4 12.9 10.2 Eagle 0.0 0.0 0.0 0.0 57.1 100.0 0.0 100.0 Fox 0.0 0.0 0.0 0.0 88.9 84.2 0.0 26.2 Gilbert 0.0 0.0 0.0 0.0 100.0 100.0 50.0 0.0 Jordan 11.1 20.0 0.0 0.0 77.8 40.0 58.3 40.0 Kangaroo 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Kusel 0.0 0.0 0.0 57.1 33.3 50.0 0.0 14.3 L. Spring 16.7 11.7 4.8 0.0 83.3 63.3 81.0 85.3 Lac La Belle 0.0 0.0 0.0 5.7 37.5 45.5 48.5 60.0 Little Falls 0.0 0.0 na na 25.0 I 1.1 na na Lorraine 11.4 4.9 10.9 10.3 51.4 44.3 41.8 35.9 Manson 0.0 0.0 11.8 0.0 0.0 60.0 58.8 0.0 Mason 0.0 0.0 0.0 0.0 100.0 0.0 0.0 0.0 Metonga 0.0 0.0 0.0 4.2 75.0 50.0 100.0 54.2 Mukwonago 55.6 60.0 0.0 22.2 44.4 20.0 100.0 66.7 Nancy 6.0 6.3 0.0 0.0 49.3 56.3 53.8 56.3 Parker 0.0 25.0 0.0 0.0 50. 75.0 25.0 63.6 Pearl 0.0 0.0 0.0 0.0 58.3 85.0 66.7 0.0 Ripley 0.0 0.0 3.3 3.0 5.0 22.7 30.0 21.2 Rock 14.3 12.5 0.0 0.0 57.1 62.5 0.0 100.0 Silver 0.0 0.0 0. 0.0 0.0 0.0 0.0 0.0 Whitewater 0.0 12.0 10.6 16.1 59.2 40.0 22.7 54.8 Wingra 2.1 4.5 3.6 6.5 8.2 24.2 53.6 14.3 Y. Birch 3.8 4.2 2.4 2.8 44.2 62.5 53.7 42.2

68 Appendix D. Continued

Percentage of QUQal E. lecontei Percentage of E. lecontei eggs Lake Name Bed 1 Bed 2 Bed 3 Bed4 Bed I Bed 2 Bed 3 Bed4 Alpine 0.0 0.0 0.0 0.0 0.0 0.0 70.0 33.3 -Bcavei-9am 11.4 IO.I 3.4 12.9 48.1 38.0 31.0 38.7 Beulah 0.0 0.0 0.0 22.2 0.0 0.0 14.3 0.0 Big Green 0.0 0.0 4.0 11.8 0.0 40.0 72.0 41.2 Big Sand 33.3 10.0 0.0 23.5 66.7 20.0 0.0 41.2 Camp 0.0 0.0 0.0 2.7 63.6 83.3 57.1 73.0 Crystal 0.0 0.0 0.0 0.0 50.0 0.0 0.0 0.0 Delavan 0.0 22.6 3.2 4.1 50.0 48.4 64.5 59.2 Eagle 0.0 0.0 0.0 0.0 42.9 0.0 0.0 0.0 Fox 0.0 2.6 50.0 4.8 11.1 13.2 50.0 69.0 Gilbert 0.0 0.0 0.0 0.0 0.0 0.0 50.0 0.0 Jordan 0.0 0.0 25.0 20.0 11.1 40.0 16.7 40.0 Kangaroo 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Kusel 33.3 50.0 0.0 14.3 33.3 0.0 0.0 14.3 L.Spring 0.0 25.0 14.3 14.7 0.0 0.0 0.0 0.0 Lac La Belle 0.0 9.1 6.1 5.7 62.5 45.5 45.5 28.6 Little Falls 0.0 0.0 na na 75.0 88.9 na na Lorraine 17.1 18.0 20.0 25.6 20.0 32.8 27.3 28.2 Manson 0.0 0.0 11.8 I 00.0 0.0 40.0 17.6 0.0 Mason 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Mctonga 0.0 0.0 0.0 4.2 25.0 50.0 0.0 37.5 Mukwonago 0.0 20.0 0.0 11.1 na na na na Nancy 10.4 0.0 3.8 3.1 3-U 37.5 42.3 40.6 Parker 50.0 0.0 8.3 9. l 0.0 0.0 66.7 27.3 Pearl 0.0 0.0 0.0 50.0 41.7 15.0 33.3 50.0 Ripley 5.0 13.6 6.7 9.1 90.0 63.6 60.0 66.7 Rock 14.3 12.5 0.0 0.0 14.3 12.5 0.0 0.0 Silver 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Whitewater 12.2 0.0 7.6 ·'·-", 28.6 48.0 59.1 25.8 Wingra 6.2 1.5 7.1 0.0 83.5 69.7 35.7 79.2 Y. Birch 0.0 2.1 2.4 3.7 51.9 31.3 41.5 51.4

69 Appendix D. Continued

Mean deQth of A/. se_icatum bed (m} Distance to shore from middle of bed (m} Lake Name Bed 1 Bed2 Bed3 Bed 4 Bed 1 Bed2 Bed 3 Bed4 Alpine 1.97 1.36 1.17 1.4 8.6 9.7 4.8 8.6 Beaver Dam L60 1.83 2.48 2.32 76.0 55.5 63.5 48.2 Beulah 1.68 1.31 1.13 1.15 64.5 7.6 6.1 6.1 Big Green 3.51 2.64 1.13 3.13 187.5 360.0 127.0 39.0 Big Sand 2.27 1.95 2.37 2.23 63.5 19.0 38.1 55.9 Camp 1.35 1.59 1.41 1.70 199.2 264.7 63.5 80.8 Crystal 3.22 3.25 4.33 3.59 39.0 32.5 32.5 38.5 Delavan 3.19 1.45 1.89 1.02 17.5 86.7 35.0 38.5 Eagle 1.45 1.25 1.22 1.26 100.0 30.0 30.0 100.0 Fox 0.47 0.448 0.45 0.47 193.7 283.3 110.0 128.5 Gilbert 2.55 2.10 2.85 3. 14 26.7 26.7 26.7 26.7 Jordan 3.03 2.60 2.08 3.26 46.5 116.3 27.2 63.0 Kangaroo 2.00 2.40 2.50 2.00 na na na na Kusel 3.13 3.15 2.93 2.52 36.0 30.3 21.4 19.0 L. Spring 1.19 1.44 1.19 1.27 86.9 30.5 57.3 30.5 Lac La Belle 2.65 2.64 1.26 1.13 25.5 II8.0 36.7 40.0 Little Falls 1.52 1.41 na na na na na na Lorraine 1.60 1.70 2.03 1.54 47.2 62.7 55.9 83.8 Manson 3.58 2.26 1.80 1.69 56.7 170.0 120.0 51.5 Mason 2.00 2.25 2.13 2.05 na na na na Metonga 2.67 3.18 2.74 1.8 I 120.0 120.0 75.0 165.0 Mukwonago 1.03 1.00 0.96 0.94 na na na na Nancy 2.54 2.50 1.86 1.90 102.0 44.0 91.0 61.0 Parker 0.96 1.60 2.03 1.83 7.5 20.0 20.0 20.0 Pearl 2.23 2.38 3.18 2.95 7.6 7.6 11.4 15.2 Ripley 1.38 2.27 2.07 2.63 99.0 37.5 37.5 124.0 Rock 2.90 3.50 2.98 2.93 97.6 97.2 152.4 64.8 Silver 1.39 1.25 1.43 1.55 9.4 5.6 17.6 12.7 Whitewater 2.04 1.85 1.45 2.61 44.2 9.5 76.2 22.9 Wingra 0.93 1.17 1.40 1.47 141.8 42.0 73.4 120.3 Y. Birch 1.03 1.02 1.27 1.58 40.5 40.5 13.5 9.8

70 Appendix D. Continued

Distance from shallow bed edge to shore (m) Distance from deeu bed edge to shore (m) Lake Name Bed I Bed 2 Bed 3 Bed 4 Bed I Bed2 Bed 3 Bed4 Alpine 0.0 0.0 0.0 0.0 17.3 19.3 9.7 17.3 -Beaver Dam (W -50.0 50.8 33.0 1.:11.0 61.0 76.2 63.5 Beulah 63.0 3.0 3.0 3.0 66.0 12.2 9.1 9.1 Big Green 175.0 340.0 147.0 69.0 200.0 380.0 167.0 99.0 Big Sand 40.6 7.6 30.5 38.1 86.4 30.5 45.7 73.7 Camp 0.0 0.0 0.0 0.0 398.3 529.3 127.0 161.7 Crystal 20.0 20.0 20.0 19.0 58.0 45.0 45.0 58.0 Delavan 0.0 0.0 0.0 0.0 35.0 173.3 70.0 77.0 Eagle 0.0 0.0 0.0 0.0 200.0 60.0 60.0 200.0 Fox 148.7 197.7 52.3 54.0 238.7 369.0 169.7 203.0 Gilbert 24.7 24.7 24.7 24.7 28.7 28.7 28.7 28.7 Jordan 29.0 97.7 0.0 0.0 64.0 135.0 54.3 126.0 Kangaroo na na na na na na na na Kusel 22.5 26.5 15.0 10.3 49.5 34.0 27.7 27.7 L. Spring 39.7 0.0 0.0 0.0 134.0 61.0 114.7 61.0 Lac La Belle 16.3 96.7 0.0 0.0 34.7 139.3 73.3 80.0 Little Falls na na na na na na na na Lorraine 0.0 0.0 0.0 0.0 94.5 125.5 111.8 167.6 Manson 28.3 140.0 100.0 20.0 85.0 200.0 140.0 83.0 Mason na na na na na na na na Metonga 90.0 90.0 50.0 140.0 150.0 150.0 100.0 190.0 Mukwonago 0.0 0.0 0.0 37.0 na na na na Nancy 96.0 38.0 79.0 55.0 108.0 50.0 103.0 67.0 Parker 0.0 0.0 0.0 0.0 15.0 40.0 40.0 40.0 Pearl 0.0 0.0 0.0 0.0 15.2 15.2 22.9 30.5 Ripley 37.3 20.0 20.0 118.0 160.7 55.0 55.0 130.0 Rock 93.5 87.7 147.3 53.3 101.6 106.7 157.5 76.2 Silver 0.0 0.0 12.2 0.0 18.8 11.2 22.9 25.4 Whitewater 0.0 0.0 0.0 0.0 88.4 19.1 152.4 45.7 Wingra 127.3 25.3 57.7 77.3 156.3 58.7 89.0 163.3 Y. Birch 36.0 36.0 12.0 7.5 45.0 45.0 15.0 12.0

71 Appendix D. Continued

Percentage of natural shoreline Percentage of mown grass shoreline Lake Name Bed I Bed 2 Bed 3 Bed4 Bed I Bed 2 Bed 3 Bed4 Alpine 50.0 60.0 5.0 10.0 45.0 35.0 80.0 60.0 . .Beaver Dam Ila 1-00.0 100.0 t00.-6 na 0.0 0.0 0.0 Beulah 60.0 60.0 100.0 100.0 40.0 0.0 0.0 0.0 Big Green 60.0 100.0 0.0 0.0 0.0 0.0 0.0 0.0 Big Sand 100.0 100.0 100.0 100.0 0.0 0.0 0.0 0.0 Camp 75.0 75.0 40.0 40.0 0.0 0.0 10.0 10.0 Crystal 100.0 30.0 0.0 0.0 0.0 0.0 0.0 0.0 Delavan 0.0 0.0 40.0 100.0 0.0 0.0 0.0 0.0 Eagle 100.0 100.0 100.0 100.0 0.0 0.0 0.0 0.0 Fox 100.0 100.0 100.0 100.0 0.0 0.0 0.0 0.0 Gilbert na 100.0 75.0 75.0 na 0.0 0.0 0.0 Jordan 100.0 63.3 40.0 0.0 0.0 23.3 43.3 50.0 Kangaroo 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Kusel 0.0 0.0 33.3 100.0 0.0 0.0 0.0 0.0 L. Spring 100.0 100.0 100.0 100.0 0.0 0.0 0.0 0.0 Lac La Bene 0.0 20.0 100.0 100.0 0.0 0.0 0.0 0.0 Little Falls na na na na na na na na Lorraine 66.7 na 100.0 100.0 33.3 na 0.0 0.0 Manson 70.0 100.0 100.0 100.0 0.0 0.0 0.0 0.0 Mason na na na na na na na na Metonga 0.0 0.0 70.0 100.0 0.0 0.0 0.0 0.0 Mukwonago 0.0 0.0 100.0 100.0 100.0 100.0 0.0 0.0 Nancy na 100.0 na 100.0 na 0.0 na 0.0 Parker 100.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Pearl na na na na na na na na Ripley 100.0 10.0 100.0 IOO.O 0.0 0.0 0.0 0.0 Rock na na na na na na na na Silver 100.0 100.0 100.0 100.0 0.0 0.0 0.0 0.0 Whitewater 100.0 100.0 100.0 100.0 0.0 0.0 0.0 0.0 Wingra 100.0 100.0 100.0 100.0 0.0 0.0 0.0 0.0 Y. Birch 100.0 100.0 0.0 0.0 0.0 0.0 0.0 0.0

72 Appendix D. Continued

Percentage of sand shoreline Percentage of seawall or riR-raR shoreline Lake Name Bed 1 Bed 2 Bed3 Bed4 Bed 1 Bed2 Bed 3 Bed4 Alpine 5.0 5.0 0.0 15.0 0.0 0.0 15.0 15.0 . _Beaye__r Dam na 0.0 0.0 0.0 na OJ> {);O 0;0 Beulah 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Big Green 40.0 0.0 0.0 100.0 0.0 0.0 100.0 0.0 Big Sand 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Camp 12.0 12.0 0.0 0.0 13.0 13.0 50.0 50.0 Crystal 0.0 30.0 0.0 0.0 0.0 40.0 100.0 100.0 Delavan 0.0 0.0 0.0 0.0 100.0 100.0 60.0 0.0 Eagle 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Fox 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Gilbert na 0.0 0.0 0.0 na 0.0 0.0 0.0 Jordan 0.0 0.0 0.0 50.0 0.0 13.3 16.7 0.0 Kangaroo na na na na na na na na Kusel 33.3 83.3 33.3 0.0 66.7 16.7 33.3 0.0 L. Spring 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Lac La Belle 0.0 0.0 0.0 0.0 100.0 80.0 0.0 0.0 Little Falls na na na na na na na na Lorraine 0.0 na 0.0 0.0 0.0 na 0.0 0.0 Manson 0.0 0.0 0.0 0.0 30.0 0.0 0.0 0.0 Mason na na na na na na na na Metonga 100.0 100.0 30.0 0.0 0.0 0.0 0.0 0.0 Mukwonago 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Nancy na 0.0 na 0.0 na 0.0 na 0.0 Parker 0.0 0.0 0.0 0.0 0.0 100.0 100.0 100.0 Pearl na na na na na na na na Ripley 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Rock na na na na na na na na Silver 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Whitewater 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0. 0 Wingra 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Y. Birch 0.0 0.0 50.0 0.0 0.0 0.0 50.0 100.0

73 Appendix D. Continued

M. s12fcatum biomass (drv gl'.m:} Native biomass (drv gl'.m:} Lake Name Bed 1 Bed2 Bed 3 Bed 4 Bed 1 Bed 2 Bed 3 Bed4 Beaver Dam na na 405,675 255.175 na na 0.150 1.917 -Rig Saru:i- 1-11.838 Ra rut na 9-:579 na 113 na Eagle 551.529 na na na 8.413 na na na Gilbert 498.583 na 541.583 682.033 15.075 na 16.133 3.567 Kangaroo 159.433 316.250 181.717 199.017 37.450 31.917 1.450 10.900 Kusel na 362.392 376.108 na na 18.817 3.042 na L. Spring 159.471 na na na 5.242 na na na Lorraine na na 200.767 na na na 26.258 na Mukwonago 241.733 319.883 190.583 158.867 1.500 0.250 0.817 0.000 Nancy 360.379 na na 1.004 na na na Pearl na 366.867 662.152 na na 31.450 24.717 na Whitewater na na na 430.421 na na na 37.654

74 Appendix D. Continued

A1. seicatum stem length (cm) JI. seicatum stem densitv (stcms/m2) Lake Name Bed I Bcd2 Bed 3 Bed 4 Bed I Bcd2 Bed 3 Bcd4 Beaver Dam na na 173.1 109.3 na na 373.3 337.5 ----Ei g Sand 141.9 na na na 262.5 na na na Eagle 120.8 na na na 442.9 na na na Gilbert 185.7 na 178.1 212.5 450.0 na 313.3 323.3 Kangaroo 92.4 108.2 97.0 112.9 220.0 356.7 265.0 283.3 Kusel na 208.6 202.9 na na 319.2 327.5 na L. Spring 125.3 na na na 190.4 na na na Lorraine na na 158.6 na na na 261.7 na Mukwonago 109.3 122.4 104.7 108.3 391.7 343.3 335.0 333.3 Nancy 181.0 na na na 379.2 na na na Pearl na 158.6 230.7 na na 292.5 382.5 na Whitewater na na na 157.1 na na na 270.0

75 Appendix D. Continued

Al. seJcatum tiQs Q!.r stem Percentage of broken tiQS Lake Name Bed 1 Bed2 Bed 3 Bcd4 Bed I Bed 2 Bed 3 Bed4 Beaver Dam na na 5.7 4.2 na na 48.4 47.0 ·--Sig Sand 4.e na fl8 na 43.4 · na na na Eagle 3.4 na na na 30.2 na na na Gilbert 3.4 na 3.3 3.7 24.5 na 26.0 39.3 Kangaroo 2.9 3.1 2.7 3.0 31.2 30.7 34.6 28.3 Kusel na 4.1 4.5 na na 45.6 53.4 na L. Spring 6.9 na na na 51.3 na na na Lorraine na na 6.0 na na na 57.5 na Mukwonago 6.3 8.7 7.0 5.2 44.7 39.1 29.5 46.2 Nancy 3.9 na na na 41.4 na na na Pearl na 3.8 4.1 na na 19.8 25.5 na Whitewater na na na 6.6 na na na 37.5

76 Appendix E. Macrophytes other than M spicatum collected in study lakes during milfoil biomass sampling.

Lake Name Macrophvte (common name) Date(s) collected Beaver Dam Elodea canadensis G. (elodea) 26 Aug 1926, 26 Aug 1997 Poiamogeton gramineus (variable pondweed) 26 Aug 1996, 26 Aug 1997 Ceratophyllum demersum L. (coontail) 26 Aug 1996, 26 Aug 1997 Potamogeton amplifolius Tuckerm. (large leaf pondweed) 26 Aug 1997 Btdens beckii G. (water marigold) 26 Aug 1997 Potamogeton robinsii (fcrnleaved pondweed) 26 Aug 1996, 26 Aug 1997 Potamo)!eton a/pinus G, B. 26 Aug 1997 Big Sand Potamogeton pusi/lus (slender pondweed) 21 Aug 1996 Ceratophyllum demersum 21 Aug 1996, 24 Aug 1997 Potamogeton zostertformis Fernald. (flatstemmed 21 Aug 1996, 24 Aug 1997 pondweed) Elodea canadensis 21 Aug 1996, 24 Aug 1997 Potamogeton pectinatus L. (sago pondweed) 24 Aug 1997 Potamogeton berchtoldii Ficher 24 Aug 1997 Potamo)!eton robinsii 24 Aug 1997 Eagle Ceratophyllum demersum 16 Aug 1996, 18 Aug 1997 Potamogeton pectinatus 18 Aug 1997 PotamoJ!eton vusillus 16 Au~ 1996 Gilbert Potamogeton alpinus 22 Aug 1997 Najasfl,exilis (Willd.) Rostk. & Schmidt. (bushy 22 Aug 1996, 22 Aug 1997 pondweed) Elodea canadensis 22 Aug 1996 Myriophyllum sibiricum Komarov (northern watem1ilfoil) 22 Aug 1996 Potamogeton pectinatus 22 Aug 1996 Potamogeton zoster~formis 22 Aug 1996 Chara sp. 22 Aug 1996 Potamogeton illinoensis Morong. (lllinois pondweed) 22 Aug 1996 Ceratophyllum demersum 22 Aug 1996 Nitella sp. 22 Au~ 1996 Kangaroo Potamogeton pectinatus 29 Aug 1996 Chara sp. 29 Aug 1996 Myriophyllum sihiricum 29 Aug 1996 Zosterella duhia (,,·ater stargrass) 29 Aug 1996 Na;as fl.exilis 29 Au_g 1996 Kusel Potamogeton crispus L. (curly leaf pondweed) 22 Aug 1996, 22 Aug 1997 Najas fl,exilis 22 Aug 1996, 22 Aug 1997 Potamogeton zosteriformis 22 Aug 1997 Ceratophyllum demersum 22 Aug 1996, 22 Aug 1997 Chara sp. 22 Aug 1997

77 Appendix E. Continued

Lake Name Macroph°\'te (common name) Date(s) collected L. Spring Potamogeton pectinatus 20 Aug 1997 Ceratophyllum demersum 24 Aug 1996, 20 Aug 1997 - - Elodea canaaensis 20 Aug 1997 Najas flexilis 20 Aug 1997 Utricularia vulgaris L. (bladderwort) 20 Aug 1997 Potamof!efon crispus 20 Aug; 1997 Lorraine Ceratophyllum demersum · 24 Aug 1996, 30 June 1997, 26 Aug 1997 Chara sp. 24 Aug I 996, 3 0 June 1997, 26 Aug 1997 Elodea canadensis 30 June 1997, 26 Aug 1997 Potamogeton crispus 30 June 1997 Mukwonago Potamogeton pectinatus 25 Aug 1996, 16 Aug 1997 Nancy Myriophyllum sihiricum 26 Aug 1996 Potamogeton amplifolius 26 Aug 1996 Elodea canadensis 26 Aug 1997 Potamogeton zosteriformes 26Aug 1997 Ceratophvllum demersum 26 Aug 1997 Pearl Ceratophyllum demersum 23 Aug 1996, 21 Aug 1997 Potamogeton zosteriformes 23 Aug I 996, 21 Aug 1997 Potamogeton filiosus Raf. (leafy pondweed) 21 Aug 1997 Potamogeton pectinatus 23 Aug 1996, 21 Aug 1997 Najas flexilis 23 Aug 1996, 21 Aug 1997 Elodea canadensis 23 Aug 1996 Myriophyllum sihiricum 23 Aug 1996 Potamogeton natans L. (floating-leaf pondweed) 23 Aug 1996 Potamotgeton pusillus 23 Aug 1996 Potamogeton gramineus 23 Aug 1996 Chara sp. 23 Aug 1996 Zosterella duhia 23 Aug 1996 Whitewater Ceratophyllum demersum 19 Aug 1996, 19 Aug 1997 Potamogeton crispus 19 Aug 1996, 19 Aug 1997

78