Euhrychiopsis lecontei Distribution, Abundance, and Experimental Augmentations for Eurasian Watermilfoil Control in Wisconsin Lakes

Transcription

1 J. Aqut. Plnt Mnge. 38: Euhrychiopsis lecontei Distribution, Abundnce, nd Experimentl Augmenttions for Eursin Wtermilfoil Control in Wisconsin Lkes LAURA L. JESTER 1,5, MICHAEL A. BOZEK 1, DANIEL R. HELSEL 2, AND SALLIE P. SHELDON 3 ABSTRACT The specilist qutic herbivore Euhrychiopsis lecontei (Dietz) is currently being reserched s potentil biologicl control gent for Eursin wtermilfoil (Myriophyllum spictum L.). Our reserch in Wisconsin focused on 1) determining milfoil weevil distribution cross lkes, 2) ssessing limnologicl chrcteristics ssocited with their bundnce, nd 3) evluting milfoil weevil ugmenttion s prcticl mngement tool for controlling Eursin wtermilfoil. The geogrphic distribution of the milfoil weevil is widespred with 49 new records of the weevil mong Wisconsin lkes contining Eursin wtermilfoil. Among 31 of the Wisconsin lkes tht contined the milfoil weevil, their bundnce vried from non-detectble to 2.5 weevils per stem of Eursin wtermilfoil. No whole-lke chrcteristics nd only some milfoil bed chrcteristics such s the percentge of nturl shoreline, the depth nd distnce of the Eursin wtermilfoil bed from shore, the number of picl tips nd the percentge of broken picl tips per stem of Eursin wtermilfoil, were significntly correlted with milfoil weevil bundnce. Twelve Wisconsin lkes ugmented with one of three different tretment levels of weevils (1, 2 or 4 weevils per Eursin wtermilfoil stem) showed some significnt dmge to the Eursin wtermilfoil in smll study plots t the end of the 1 Wisconsin Coopertive Fishery Reserch Unit 4, College of Nturl Resources, UW Stevens Point, Stevens Point, WI Wisconsin Deprtment of Nturl Resources, 101 S. Webster St., Mdison, WI Middlebury College, Deprtment of Biology, Middlebury, VT U.S. Geologicl Survey, Wisconsin Deprtment of Nturl Resources nd University of Wisconsin Stevens Point. 5 Current ddress: Dkot County Soil nd Wter Conservtion District, th St. W. Suite 102, Frmington, MN Received for publiction October 22, 1998 nd revised form My 10, first tretment seson. Additionl smpling to ssess longterm effects of this ugmenttion is ongoing. Key words: Myriophyllum spictum, milfoil weevil, weevil ugmenttion, biologicl control. INTRODUCTION The exotic qutic plnt Eursin wtermilfoil (Myriophyllum spictum L.) invded Wisconsin lkes in the 1960s, nd continued to spred cross the stte (Bode et l. 1993). Most Eursin wtermilfoil infesttions spred quickly within lke nd often rech nuisnce levels; inhibiting recretion, displcing ntive plnt communities, nd ltering fish nd invertebrte communities. Although mechnicl hrvesting nd chemicl herbicides re often used s control methods for Eursin wtermilfoil, they provide only short-term relief (Aiken et l. 1979, Smith nd Brko 1990) nd hve potentilly negtive effects on non-trget plnts nd nimls (Engel 1990,b). Nturl declines of Eursin wtermilfoil in mny lkes round the United Sttes hve been concurrent with the presence nd incresed bundnce of the qutic milfoil weevil (Euhrychiopsis lecontei Dietz, Creed nd Sheldon 1995, Kirschner 1995, Lillie nd Helsel 1997, Creed 1998). The milfoil weevil is ntive to North Americ nd is known to exist in lkes cross the northern tier of the United Sttes nd southern Cnd (Newmn nd Mher 1995, Sheldon nd O Bryn 1996, Creed 1998), lthough little is known bout its specific distribution. Recent reserch hs demonstrted tht the milfoil weevil significntly reduces biomss nd density of Eursin wtermilfoil in numerous field nd lb experiments (Creed nd Sheldon 1993, 1995, Creed et l. 1992, Sheldon nd Creed 1995, Newmn et l. 1996). Previous studies hve detiled the life history of the milfoil weevil (see Creed et l. 1992, Creed nd Sheldon 1993, Solrz nd Newmn 1996, Newmn et l. 1996, 1997, Sheldon 88 J. Aqut. Plnt Mnge. 38: 2000.

2 nd O Bryn 1996) nd its effects on Eursin wtermilfoil. The milfoil weevil overwinters in lef litter long the shoreline s dults nd returns to the milfoil bed in the spring to ly eggs in the picl meristems of the milfoil plnts. The destruction of the milfoil is cused by the lrve which consume the tissue of the picl tip nd burrow into the stem just below the tip. Lter lrvl instrs continue burrowing through the stem, consuming vsculr tissue nd eventully pupting within the stem. The hollow stem loses the bility to trnsport nutrients nd crbohydrtes nd loses buoyncy in the wter column. Reserch lso shows tht ntive qutic plnts re rrely used by the milfoil weevil for rering nd feeding nd thus the weevil does not impct their growth (Sheldon nd Creed 1995, Solrz nd Newmn 1996). Additionl reserch on milfoil weevil distribution, life history requirements nd potentil use s control gent is wrrnted. The objectives of this study were to better document the geogrphic distribution of the milfoil weevil in the Wisconsin, ssess limnologicl nd geogrphicl chrcteristics ssocited with its bundnce, nd evlute the effectiveness of ugmenting milfoil weevil popultions s prcticl mngement tool for Eursin wtermilfoil control. In this pper we report new records of the milfoil weevil in Wisconsin, report milfoil weevil bundnce nd ssocited vribles from subset of lkes, nd report preliminry results on the effectiveness of ugmenting milfoil weevil popultions s mngement tool. Geogrphic Distribution METHODS The geogrphic distribution of the milfoil weevil in Wisconsin ws studied by ssessing the presence of the weevil using two methods in 50 Wisconsin lkes contining Eursin wtermilfoil during the summers of 1996 nd First, we surveyed 44 lkes to ssess presence of the milfoil weevil by snorkeling or by bot in surfcing beds nd visully serched for milfoil weevil eggs, lrve, pupe or dults. During these surveys, dignostic evidence of milfoil weevil herbivory ided in the serch for ctul specimens by helping to identify res to serch more intensively. In ech lke, mximum of four mn-hours ws spent intensively serching for the milfoil weevil to identify their presence. Second, 6 dditionl lkes were surveyed during mcrophyte smpling used to ssess bundnce of the milfoil weevil (described lter). Adult milfoil weevil specimens were preserved nd mintined s voucher specimens for ech lke except Little Flls Lke, Mson Lke, nd Prker Lke where dult specimens were not collected; only lrve from these lkes were collected nd kept s voucher specimens. All specimens re housed in the Museum of Nturl History t the University of Wisconsin Stevens Point. Milfoil Weevil Abundnce The bundnce of the milfoil weevil ws evluted in 31 Wisconsin lkes once between mid-july nd mid-august 1996 or In ech lke, four Eursin wtermilfoil beds tht hd not been hrvested or treted with herbicides were selected for smpling. The distnce between the four mcrophyte beds ws mximized to provide n estimte of weevil bundnce for the entire lke. A totl of 120 picl stems of Eursin wtermilfoil ws collected from ech lke (4 beds 3 trnsects/bed 5 smpling points/trnsect 2 stems/smpling point). Within ech Eursin wtermilfoil bed, three equidistnt trnsects (reltive to the mcrophyte bed width) were smpled. Trnsects were perpendiculr to shore. If Eursin wtermilfoil ws distributed throughout the entire lke, then trnsects rn from the center of the lke to shore. Along ech trnsect, two stems from five equidistnt points in the Eursin wtermilfoil bed were smpled. At ech smpling point, the top 50 cm of the first two stems of Eursin wtermilfoil touching the snorkeler s hnd beneth the surfce were collected. Becuse only the top 50 cm of stem ws collected, these re referred to s picl stems herein. These picl stems often included severl lterl brnches. To minimize sesonl influences on bundnce, lkes in southern Wisconsin were smpled first followed by centrl Wisconsin lkes nd finlly northern Wisconsin lkes. In the lbortory, the stems were inspected for the presence of milfoil weevil eggs, lrve, pupe, nd dults nd for dmge by the weevils. All picl tips (i.e., picl meristems) nd portions of stems with dmge were inspected with dissecting microscope between 10 nd 20 power. Stems dmged by herbivory were sliced open length-wise with rzor blde nd lrve nd pupe were extrcted (Creed nd Sheldon 1995). The number of weevils in ech life stge (egg, lrv, pup, dult) on ech stem ws recorded nd ll were preserved. Abundnce ws recorded s the number of milfoil weevils (ll lifestges combined) per picl stem of Eursin wtermilfoil nd confidence intervls were clculted. Abundnce ws lso recorded s the number of milfoil weevils (ll lifestges combined) per picl tip of Eursin wtermilfoil by dividing the number of milfoil weevils collected by the number of tips (both dmged nd undmged) collected. Independent vribles tht might explin vrition in weevil bundnce were collected nd tested for correltions with weevil bundnce. Some vribles were whole-lke chrcteristics (lke-level vribles) nd other vribles were mesured in, or djcent to the bed of Eursin wtermilfoil being smpled (bed-level vribles). Lke-level vribles collected during weevil bundnce smpling included wter temperture, dissolved oxygen nd Secchi depth ech tken t the lke s deepest point. Additionl lke-level vribles were cquired from the Wisconsin Deprtment of Nturl Resources nd the Environmentl Protection Agency STORET dtbse. These vribles included lke loction, mximum nd men depth, re, nd type (seepge, dringe or spring-fed), time of the Eursin wtermilfoil invsion, nd wter chemistry vribles collected within few yers of weevil bundnce smpling such s totl phosphorus, totl lklinity, totl Kjeldhl nitrogen, nitrte plus nitrite, chlorophyll, ph, nd conductivity. Bed-level vribles collected during weevil bundnce smpling included the distnce to shore from the deep nd shllow edges of the bed nd the depth of ech trnsect smpling point. Riprin conditions were visully estimted using percentges of ech shoreline type present t the lnd-wter interfce (nturl shore, mowed lwn to wter s edge, snd, or sewll/rip-rp) prllel to the smpled bed or t the end of J. Aqut. Plnt Mnge. 38:

3 trnsect if the entire littorl zone ws smpled. Specific Eursin wtermilfoil chrcteristics such s biomss, density, stem length nd number of picl tips were lso mesured in subset of lkes s prt of the finl study objective (see below). Independent vribles were not normlly distributed nd thus were nlyzed for significnt ssocitions with milfoil weevil bundnce using Spermn rnk correltion nlysis (p 0.05). Associtions with lke-level vribles were tested using the men number of milfoil weevils per picl stem for the entire lke (i.e., cross ll four Eursin wtermilfoil beds) while bed-level vribles were tested using the men milfoil weevil per picl stem for tht bed. Milfoil Weevil Popultion Augmenttion In 12 Wisconsin lkes, existing milfoil weevil popultions were ugmented with dditionl weevils in designted plots to determine the effectiveness of stocking weevils to control Eursin wtermilfoil. In ech lke, four plots, 2 m 6 m, were estblished nd in July 1997 smll bundles of Eursin wtermilfoil contining estimted numbers of milfoil weevil eggs nd lrve were tied onto existing Eursin wtermilfoil plnts in three of the four plots. The fourth plot ws reference nd ws not ugmented. In ech lke existing weevil bundnce ws ugmented to obtin finl tretment level of 1, 2 or 4 weevils per stem of milfoil. Pst reserch suggested tht pproximtely two weevils per stem resulted in significnt Eursin wtermilfoil declines (Newmn et l. 1996). The ctul number of weevils relesed in ech plot rnged from 115 to lmost 10,000. Tretment levels were rndomly ssigned to the lkes unless the ugmenttion sites were found to hve greter thn the rndomly ssigned ugmenttion level. In these cses, higher ugmenttion levels were rndomly ssigned when possible. Plots within lke were replictes nd plots mong lkes were tretments. Milfoil weevil eggs nd lrve were cultured t Middlebury College in Vermont from dults collected in Wisconsin. Adults were shipped to Vermont in coolers, on ice vi overnight express; eggs nd lrve were returned in the sme mnner. The rectngulr plots were set prllel to shore with pproximtely 9 m between the plots. While we tried to keep depth constnt mong lkes, depth of the plots rnged from 1.0 m to 3.0 m depending on the Eursin wtermilfoil depth in ech lke, but plot depth within ech lke remined constnt. Although bot trffic ws encourged to sty wy from the plot res, neither enclosures nor exclosures were estblished. Prior to ugmenttion, milfoil weevil bundnce ws estimted in June 1997 to determine the number of weevils needed to bring popultions up to the desired tretment level. Approximtely five weeks post-ugmenttion (i.e., lte August 1997), milfoil weevil bundnce ws mesured gin mong the plots to ssess popultion levels. Milfoil weevil bundnce ws mesured mong the plots gin in June nd August Mcrophytes were smpled in the plot res using 0.15 m 2 qudrt smpler nd scub. In 1996, the yer before ugmenttion, 16 smples were collected long the ugmenttion plot contour. In 1997, pproximtely five weeks postugmenttion, three rndomly selected smples were collected from ech of the four plots for totl of 12 smples. Mcrophytes were collected during pek plnt biomss in August of both yers nd mesured for biomss, stem density, stem length, number of picl tips nd the percentge of broken tips s well s the biomss of ntive mcrophytes. In the lb, smples were processed by floting plnts lengthwise in n quculture rcewy. The number of plnts in ech smple ws counted to determine plnt density nd converted to number of plnts per squre meter. A sub-smple of 30 plnts per smple ws used to get verge plnt length (cm), number of tips per plnt, nd the number of broken tips per plnt. Plnts covered with clcium crbonte deposits were soked in 5% hydrochloric cid bth nd rinsed with tp wter to remove the deposits. Mcrophytes other thn Eursin wtermilfoil were seprted nd identified to species. All plnts of ech species were then ir dried for 30 minutes nd mesured for wet weight. To determine dry weight biomss, plnts from ech species nd ech smpling site were plced in pre-dried, tred pper bg, dried t 106C to constnt weight (pproximtely 56 hours), nd weighed. Wilcoxon Signed-Rnk Tests were used to determine significnt differences in Eursin wtermilfoil vribles between pre-ugmenttion (1996) to post-ugmenttion (1997) nd between reference nd tretment plots. Dt for only one vrible, the percentge of broken tips, ws normlly distributed, thus prmetric t-test ws performed. Alph ws set t p 0.05 for ll tests. Additionl reserch will determine chnges in Eursin wtermilfoil within the ugmenttion plots in future yers. RESULTS Milfoil Weevil Distribution The milfoil weevil ws widely distributed cross Wisconsin in lkes contining Eursin wtermilfoil (Figure 1, Tble 1). The milfoil weevil ws found in 49 lkes where it ws not previously recorded nd ws found in nerly every lke smpled except Silver Lke in Wupc County. After serching for four mn-hours in Silver Lke, no milfoil weevils t ny lifestge were found nd there ws no evidence of milfoil weevil herbivory. In most other lkes, milfoil weevil dult ws collected within ten minutes of serching. In three lkes where dults were not found, Mson nd Prker Lkes in Adms County nd Little Flls Lke in St. Croix County, only lrvl weevils were found during weevil bundnce smpling. Lkes with new records of the milfoil weevil were vrible in type, size nd loction (Tble 1). Both seepge nd dringe lkes were found to hrbor weevils, lke size rnged from 1.2 h (Lower Kelly Lke) to lmost 3,000 h (Big Green Lke) nd mximum lke depth rnged from 0.6 m (Goose Pond) to 71.9 m (Big Green Lke). The lkes were lso widely distributed geogrphiclly round the stte (Figure 1). Generl chrcteristics of Eursin wtermilfoil beds in lkes with the milfoil weevil lso vried. Mny lkes hd very dense nd brod beds of milfoil while others hd nrrow bnds of Eursin wtermilfoil in the littorl zone, or spordic distributions. Men depth of Eursin wtermilfoil beds vried s well s the height of the plnts in the wter column. Milfoil Weevil Abundnce nd Correlted Vribles Milfoil weevil bundnce cross 31 lkes vried from 0.0 (non-detectble) to 2.5 weevils per picl stem with men 90 J. Aqut. Plnt Mnge. 38: 2000.

4 lso negtively correlted with depth of the Eursin wtermilfoil bed (r = -0.30) nd the percentge of sndy shoreline (r = -0.29). The percentge of broken picl tips of Eursin wtermilfoil ws lso correlted with some lke-level nd bed-level vribles including lke surfce re (r = 0.49), mximum depth (r = 0.50), distnce from shore to the middle of the Eursin wtermilfoil bed (0.27), nd depth of the bed (r = 0.81). Figure 1. Known distribution of the milfoil weevil in Wisconsin. Previous loctions referenced in Lillie (1991), Newmn nd Mher (1995), nd Lillie nd Helsel (1997). of 0.65 weevils per picl stem (Figure 2, Tble 2). Milfoil weevil lrve nd eggs were found to be more bundnt thn dults or pupe in 30 of the 31 lkes. While dults never mde up more thn 36.4% nd pupe never mde up more thn 25% of weevil lifestges collected in ny one lke, lrvl bundnce reched s high s 100% nd eggs s high s 84.6% of weevils collected in some lkes. In 19 of 31 lkes, ll life stges of the milfoil weevil were collected during bundnce smpling. Within lkes, milfoil weevil bundnce lso vried mong Eursin wtermilfoil beds with men difference of 0.81 weevils per picl stem between the highest nd lowest bundnce found mong beds within one lke (Tble 3). The gretest difference between beds in one lke ws 2.3 weevils per picl stem in Lke Wingr. Milfoil weevil bundnce ws not significntly correlted with ny lke-level vribles such s ltitude of the lke, time since the invsion of Eursin wtermilfoil, lke depth (mximum or men), lke type, or lke size. Milfoil weevil bundnce lso did not significntly correlte with wter qulity vribles such s summer wter temperture, dissolved oxygen, Secchi disc depth, totl phosphorus, NO 2 + NO 3 nitrogen, totl Kjeldhl nitrogen, chlorophyll, lklinity, ph, nor conductivity. Milfoil weevil bundnce ws significntly correlted with some bed-level vribles (Tble 4). Abundnce ws positively correlted with the percentge of broken picl tips (r = 0.54), distnce from shore to the middle (r = 0.25) nd deep (r = 0.28) edge of the Eursin wtermilfoil bed, percentge of nturl shoreline (r = 0.21), nd number of picl tips per Eursin wtermilfoil plnt (r = 0.60). Weevil bundnce ws First-yer Results of Milfoil Weevil Augmenttion Although ugmenttion ws designed to bring weevil bundnce up to the desired tretment level in the plots, the ctul observed weevil bundnce five weeks post-ugmenttion ws well below tretment levels in most lkes (Tble 5). Survivl of weevils fter shipping ws high (pproximtely 75%), nd would not ccount for declines. Anlysis of first yer post-ugmenttion dt indictes tht mny milfoil vribles collected in plots decresed significntly from pre-ugmenttion. Moreover, significnt declines were more common in lkes tht received the highest tretment level (4 weevils per stem) (Tble 5). For instnce, 60% of lkes given the lowest tretment level (1 weevil per stem) experienced significnt declines in Eursin wtermilfoil biomss while 67% of lkes given the medium tretment level (2 weevils per stem) nd 100% of lkes given the highest tretment level experienced significnt declines in Eursin wtermilfoil biomss. Almost the sme ws true for Eursin wtermilfoil stem density: 40% of lkes given the lowest tretment level, 67% of lkes given the medium tretment level nd 75% of lkes given the highest tretment level showed significnt declines in stem density. When compring post-ugmenttion Eursin wtermilfoil mesurements between ugmented plots nd reference plots, fr fewer significnt differences were detected (Tble 6). Only one lke showed significnt decrese in Eursin wtermilfoil biomss s compred to the reference plot nd no lkes showed significnt difference in Eursin wtermilfoil stem density between reference plots nd tretment plots. While significnt declines were pprent from the dt in both tretment nd reference plots, these observtions were not lwys pprent in the field. Although some lkes hd extensive lrvl dmge in nd round the plots severl weeks fter ugmenttion, there did not pper to be visully noticeble decline of Eursin wtermilfoil in most plot res. Three lkes did experience substntil declines in Eursin wtermilfoil tht were visully pprent (Mukwongo, Kusel, nd Lower Spring). However, ll three lkes were in the midst of nturl Eursin wtermilfoil decline t the onset of weevil ugmenttions. Milfoil Weevil Distribution DISCUSSION This study shows tht the milfoil weevil is geogrphiclly widespred throughout Wisconsin s 49 new records of the milfoil weevil from southern, centrl nd northern prts of the stte were found. More importntly, becuse nerly every lke surveyed in this study ctully contined the weevil, it is likely tht the milfoil weevil is widespred cross most Wis- J. Aqut. Plnt Mnge. 38:

5 TABLE 1. NEW RECORDS OF EUHRYCHIOPSIS LECONTEI IN WISCONSIN. Lke nme County Type Loction Surfce re (hectres) Mx. depth (m) Dte of collection Alpine Lke Wushr DG T19N R11E Sec July 1997 Brk Lke Wshington SE T9N R19E Sec June 1997 Bss By Wukesh DG T5N R20E Sec July 1997 Bss Lke St. Croix SE T30N R19E Secs. 23, June 1998 Bever Dm Lk. Brron SE T35N R13W Secs. 5, 6, 7, June 1996 T35N R14W Sec. 1 Beulh lke Wlworth DG T4N R18E Sec July 1997 Big Cedr Lke Wshington SE T10N R19E Sec. 5; June 1996 T11N R19E Secs. 20,20,30,31,32 Big Green Lke Green Lke DG T15,16N R12,13E August 1997 Big Snd Lke Vils DG T41N R12E Secs. 2,3,4,9; June 1996 T42 R12E Secs. 34,35 Cmp Lke Kenosh DG T1N R20E Secs. 20,21,28, August 1996 Crooked Lke Wukesh DG T7N R17E Sec Sept Crystl Lke Sheboygn SE T16N R21E Secs. 31, July 1997 Delvn Lke Wlworth DG T2N R16E Secs. 21, 22, 27, 28, 32, July 1997 Egle Lke Rcine DG T3N R20E Secs. 21,22,27, June 1996 Elizbeth Lke Kenosh DG T1N R19E Secs. 28, 29, 32, August 1997 Fox Lke Dodge DG T13N R13E August 1997 Friendship Lke Adms DG T17N R6E Sec June 1997 George Lke Kenosh DG T1N R21E Secs. 20, August 1996 Gilbert Lke Wushr SE T20N R11E Secs. 10,11,14, June 1996 Goose Pond St. Croix SE T28N R20W Sec June 1998 Johns Lke Wushr SE T19N R11E Sec June 1998 Jordn Lke Adms SE T15N R7E Sec June 1997 Kngroo Lke Door DG T29N R27E Sec July 1997 Kusel Lke Wushr SE T20N R11E Secs. 26,27,34, June 1996 Lc L Belle Wukesh DG T8N R17E July 1997 Little Flls Lke St. Croix SE T29N R19W Secs. 4,8, August 1996 Long Trde Lke Polk DG T36N R18W Sec August 1996 Lorrine Lke Wlworth SE T3N R15E Sec June 1996 Lower Kelly Lke Wukesh SE T6N R20E Sec June 1997 Lower Spring Lke Jefferson DG T5N R16E Secs. 22, June 1996 Mrie Lke Kenosh DG T1N R19E Sec. 21, August 1997 Mnson Lke Oneid DG T36N R7E Secs. 32, June 1996 Mson Lke Adms DG T13N R7E Secs. 25, 26, 35, August 1996 T13N R8E Secs. 30, 31 Metong Lke Forest DG T35N R13E Sec August 1997 Mukwongo Pond Wukesh SP T5N R18E Sec June 1996 Nncy Lke Wshburn SE T42N R13W Secs. 27,28, June 1996 North Lke Wukesh DG T8N R18E Secs. 16,17,20, August 1996 Pddock Lke Kenosh DG T1N R20E Sec August 1996 Prker Lke Adms SE T15N R7E Sec June 1997 Perl Lke Wushr SE T19N R12E Sec June 1996 Pike Lke Wshington DG T10N R18E Secs. 22,23,26, August 1996 Pine Lke Wukesh SP T8N R18E Secs. 27, 28, June 1998 Ripley Lke Jefferson SE T6N R13E Sec. 7, July 1997 Rock Lke Jefferson DG T7N R13E Secs. 2,10,11,14, August 1996 Sherwood Lke Adms DG T20N R6E Secs. 16, August 1997 Whitewter Lke Wlworth DG T3N R15E Sec June 1996 T4N R15E Secs. 25, 26, 27, 34, 35 Wind Lke Rcine DG T4N R20E Secs. 3,4,8,9,10,16, July 1996 Wolf Lke Rcine T2N R20E Secs. 15, August 1996 Yellow Birch Lk. Vils DG T40N R10E Secs. 21, August 1997 DG = Dringe lke, SE = Seepge lke, SP = Spring Lke. consin lkes contining Eursin wtermilfoil. In the only lke where weevils were not found, Silver Lke, the weevils my exist t very low bundnce, mking it difficult to detect in our smpling. The higher numbers of lkes with the milfoil weevil in the southest nd centrl prts of the stte merely reflect the greter distribution of Eursin wtermilfoil in those res nd extr effort by the WDNR in collecting specimens there. 92 J. Aqut. Plnt Mnge. 38: 2000.

6 Figure 2. Abundnce of the milfoil weevil in Wisconsin lkes prior to ugmenttion: ll weevil lifestges combined. Vlues indicte the men bundnce of the milfoil weevil per picl stem ±95% confidence intervls. *Confidence intervls were not clculted for Lower Spring nd Egle Lkes due to different smpling method. The widespred distribution of the milfoil weevil ws not expected. Newmn nd Mher (1995) found the milfoil weevil in only 10 of 25 lkes smpled in Minnesot nd Wisconsin. In Wshington, only 21 of 51 lkes surveyed were found to hrbor the milfoil weevil (Tmyo et l. 1999). Generlly, other sttes hve not serched extensively for the milfoil weevil but, t lest 12 Vermont, 7 Msschusetts, 1 New York, 2 Connecticut, 2 Michign, 3 Ohio, 3 Illinois, nd 5 British Columbi lkes re known to hrbor the milfoil weevil (Creed 1998, Sllie Sheldon pers. comm.). Declines of Eursin wtermilfoil in some lkes re being ttributed to milfoil weevil popultions due to specific evidence of dmge to the Eursin wtermilfoil typicl of milfoil weevil herbivory, concurrent with reltively high milfoil weevil popultions (Lillie nd Helsel 1997, Creed nd Sheldon 1995, Kirschner 1995). Creed (1998) lso showed significntly more declines of Eursin wtermilfoil hve occurred within the known geogrphic rnge of the milfoil weevil thn would be expected by chnce (Creed 1998). Moreover, we observed t lest ten Wisconsin lkes found to hrbor the milfoil weevil in this study or previous surveys (Newmn nd Mher 1995, Lillie 1991) tht hve experienced Eursin wtermilfoil declines: Big Green Lke (Green Lke Co.), Devil s Lke (Suk Co.), Fish Lke nd Lke Wingr (Dne Co.), Long-Trde Lke (Polk Co.), Kusel Lke (Wushr Co.), Whitewter Lke (Wlworth Co.), Yellow Birch Lke (Vils Co.), Mukwongo Pond nd Wind Lke (Wukesh Co.). Interestingly, Crpenter (1980) reported tht the durtion of pek bundnce of Eursin wtermilfoil in mny lkes is pproximtely 10 yers before declines re evident. However, the definitive cuses of Eursin wtermilfoil declines hve not been estblished (Lillie nd Helsel 1997, Crpenter 1980) but wrrnts further investigtion into predtor-prey cycles between Eursin wtermilfoil nd the milfoil weevil. Fctors Influencing Milfoil Weevil Abundnce Milfoil weevil bundnce vried gretly cross the 31 lkes smpled s well s mong different Eursin wtermilfoil beds within lkes. Our results indicte tht some vribles we mesured within the Eursin wtermilfoil beds my influence milfoil weevil bundnce more thn the lrge-scle geogrphicl nd limnologicl vribles. Milfoil weevil bundnce ws positively correlted with the percentge of broken picl tips. This reltionship hs biologicl relevnce becuse herbivory dmge to the picl tip nd stem just below the tip occurs s lrve burrow into the stem consuming picl nd vsculr tissues. Frgile tips were often observed in conjunction with pst or present lrvl herbivory nd these tips were often broken prior to smpling or during hndling. Due to the significnt positive correltion between the percentge of dmged tips nd the bundnce of the milfoil weevil nd the biologicl relevnce of the correltion, the mount of milfoil tip dmge ws clerly n indiction of milfoil weevil bundnce. Milfoil weevil bundnce ws lso positively correlted with the distnce from shore to the middle nd deep edges of the Eursin wtermilfoil bed nd negtively correlted with the depth of the Eursin wtermilfoil bed. These reltionships indicte tht weevils re more bundnt in lrge, shllow expnses of Eursin wtermilfoil rther thn deep (nd perhps non-surfcing) Eursin wtermilfoil closer to shore. This reltion my chrcterize the type of Eursin wtermilfoil bed where the milfoil weevil hs the best reproductive success. Milfoil weevil bundnce ws lso positively correlted with the number of picl tips per Eursin wtermilfoil plnt. However, it is unknown which response is the dependent vrible in this reltionship. On one hnd, weevils my reproduce more effectively in milfoil beds with high- J. Aqut. Plnt Mnge. 38:

7 TABLE 2. MEAN MILFOIL WEEVIL ABUNDANCE ±95% CONFIDENCE LIMITS IN WISCONSIN LAKES. WEEVILS PER APICAL STEM (TOP ~ 50 CM OF STEM) AND WEEVILS PER APICAL TIP (I.E., APICAL MERISTEM) INCLUDE ALL LIFESTAGES COMBINED. THE LAST FOUR COLUMNS REPORT THE PERCENTAGE OF WEEVILS IN EACH OF FOUR LIFESTAGES FOR EACH LAKE. WEEVILS PER APICAL STEM AND TIP REPRESENT THE MEAN OF ALL STEMS COLLECTED PER LAKE. Lke County Weevils per picl stem Weevils per picl tip Percent dults Percent lrve Percent pupe Percent eggs Alpine Wushr 0.2 ± Bever Dm Brron 1.8 ± Beulh Wlworth 0.1 ± Big Green Green Lke 0.4 ± Big Snd Vils 0.3 ± Cmp Kenosh 0.7 ± Crystl Sheboygn 0.1 ± Delvn Wlworth 1.2 ± Egle Rcine Fox Dodge 0.8 ± Gilbert Wushr 0.1 ± Jordn Adms 0.3 ± Kngroo Door Kusel Wushr 0.1 ± L. Spring Jefferson Lc L Belle Wukesh 0.8 ± Little Flls St. Croix 0.2 ± Lorrine Wlworth 1.9 ± Mnson Oneid 0.3 ± Mson Adms <0.1 ± Metong Forest 0.5 ± Mukwongo Wukesh 0.3 ± Nncy Wshburn 1.2 ± Prker Adms 0.2 ± Perl Wushr 0.3 ± Ripley Jefferson 0.9 ± Rock Jefferson 0.1 ± Silver Wupc Whitewter Wlworth 1.4 ± Wingr Dne 2.2 ± Y. Birch Vils 2.5 ± men = 0.65 men = 0.15 men = 7.01 men = men = 7.13 men = Lkes were smpled with different smpling method; confidence intervl could not be clculted. er numbers of picl tips becuse there re more oviposition sites. In contrst, perhps the Eursin wtermilfoil plnts respond to high weevil bundnce nd greter mounts of tip dmge by growing new picl tips. At the bed-level, milfoil weevil bundnce ws positively correlted with the percentge of nturl shoreline. Milfoil weevils overwinter in the lef litter nd mud long the shore within few meters of the wter (Newmn et l. 1997). It is possible tht the milfoil weevil is more bundnt long nturl shorelines becuse they re more successful t overwintering in these res. Hbitts other thn nturl shore, such s sewll, rock rip-rp, snd, or mown grss, my not offer enough protection or burrowing cpbilities for hibernting dult weevils. Unfortuntely, more specific chrcteristics of the nturl shorelines were not mesured in this study nd the mjority of beds smpled were djcent to nturl shorelines. More reserch needs to be conducted regrding the importnce of overwintering success in weevil distribution nd popultion dynmics. Vribles mesured t the lke-level were not correlted with milfoil weevil bundnce. Time since the Eursin wtermilfoil invsion ws not fctor in milfoil weevil bundnce lthough we speculted tht the longer Eursin wtermilfoil hd resided in lke (i.e., yers), the more time the weevil hd to increse its popultions in predtor response to incresed prey. However, fctors such s the rte of Eursin wtermilfoil expnsion following the initil invsion nd the mount of chemicl or mechnicl control used in the lke my ffect the bility of the milfoil weevil to effectively colonize nd increse in bundnce in even long-residing Eursin wtermilfoil beds. Vribles relted to wter chemistry were not significntly correlted with milfoil weevil bundnce. Wter temperture nd dissolved oxygen my be relted to (or limiting) milfoil weevil bundnce within the hbitt of individul mcrophyte beds, however these vribles were only collected from the lke s whole insted t the bed-level. Herbivory by the milfoil weevil on Eursin wtermilfoil hs been shown to significntly reduce its stnding biomss (Creed et l. 1992, Creed nd Sheldon 1993, 1995, Sheldon nd Creed 1995, Newmn et l. 1996). However, Eursin wtermilfoil biomss, density nd other vribles mesured in 94 J. Aqut. Plnt Mnge. 38: 2000.

8 TABLE 3. MILFOIL WEEVIL ABUNDANCE IN DIFFERENT EURASIAN WATERMILFOIL BEDS IN 31 WISCONSIN LAKES. MEAN VALUES ARE THE MEAN OF THE FOUR BEDS ±95% CONFIDENCE INTERVAL. Weevils per picl stem Lke Bed 1 Bed 2 Bed 3 Bed 4 Men Difference b Alpine ± Bever Dm ± Beulh ± Big Green ± Big Snd ± Cmp ± Crystl ± Delvn ± Egle 0.2 < < ± Fox ± Gilbert < ± Jordn ± Kngroo Kusel ± Lower Spring ± Lc L Belle ± Little Flls n n Lorrine ± Mnson < ± Mson < <0.01 <0.01 Metong ± Mukwongo < ± Nncy ± Prker ± Perl ± Ripley ± Rock ± Silver Whitewter ± Wingr ± Yellow Birch ± Only two Eursin wtermilfoil beds were smpled on Little Flls Lke. b Difference between beds with the highest nd lowest bundnce. men = lkes (nd used s pre-ugmenttion vribles) were not relted to nturl weevil bundnce in those lkes. Unfortuntely, trend in Eursin wtermilfoil density nd biomss could not be detected s only one yer of dt were collected, thus the dynmics of predtor-prey cycle between Eursin wtermilfoil nd nturl milfoil weevil bundnce were not mesurble in this study. It is possible tht lkes with higher milfoil weevil bundnce, lso hd lower Eursin wtermilfoil biomss compred to previous yers. In this study, the two lkes with the highest mesured milfoil weevil bundnce (Lke Wingr nd Yellow Birch Lke) hve experienced drmtic declines in Eursin wtermilfoil in the pst. In fct, Eursin wtermilfoil ws very difficult to find in Yellow Birch Lke (there ws brely enough to smple) nd it is possible tht the milfoil weevil is currently controlling the Eursin wtermilfoil in this lke. Vrition in milfoil weevil bundnce cross lkes could be cused nd compounded by mny vribles not mesured in this study. Milfoil weevil popultions, like most other orgnisms, my fluctute for vriety of resons. In fct, lrge fluctutions in popultions re not unusul in common species (Pielou 1974). Milfoil weevil popultions within lke could be explined by predtor-prey cycles with n increse in Eursin wtermilfoil followed by n increse in the milfoil weevil nd then subsequent decrese in Eursin wtermilfoil nd decrese in the milfoil weevil. This reltionship is even more probble due to the fct tht the milfoil weevil is primrily species-specific nd thus decline in Eursin wtermilfoil would most certinly men subsequent decline in the milfoil weevil. Lillie, WDNR (pers. comm.) reported lrge fluctutions in milfoil weevil popultions from n ongoing study on Fish Lke in Dne Co., Wisconsin. Occupncy rtes of the milfoil weevil in Eursin wtermilfoil stems fluctuted over the yers from 18% in 1992, to 3% in 1995, nd bck to 18% in Furthermore, Fish Lke Eursin wtermilfoil biomss seems to fluctute in response to milfoil weevil herbivory. Lillie reported lg between milfoil weevil bundnce nd Eursin wtermilfoil biomss which suggests the presence of predtor-prey-induced oscilltions. The dt collected in this study does not llow us to decipher J. Aqut. Plnt Mnge. 38:

9 TABLE 4. SPEARMAN RANK CORRELATIONS BETWEEN OBSERVED MILFOIL WEEVIL ABUNDANCE AND CHARACTERISTICS OF MACROPHYTE BEDS IN WHICH THE WEEVILS WERE COLLECTED. Vrible where milfoil weevil bundnce is locted on the predtorprey curve for given lke. Their bundnce my be on the rise in one lke, t the pex in nother nd on the decline in yet nother. Milfoil Weevil Augmenttion Correltion coefficient Percent broken tips Depth of Eursin wtermilfoil Distnce from shllow bed edge to shore ns Distnce from middle of bed to shore Distnce from deep bed edge to shore Percent nturl shore Percent mowed grss shore ns Percent snd shore Percent wll or rip-rp shore ns Biomss of Eursin wtermilfoil ns Biomss of ntive mcrophytes ns Stem length of Eursin wtermilfoil ns Stem density of Eursin wtermilfoil ns No. of picl tips per Eursin wtermilfoil plnt No. of picl tips m -2 ns ns denotes non-significnt correltion. P Preliminry (i.e., first yer) post-ugmenttion dt shows tht five-weeks fter weevil ugmenttion Eursin wtermilfoil biomss, stem density, stem length nd tips per stem in the stocked plots of mny lkes were significntly lower thn pre-ugmenttion (the yer before ugmenttion) levels. Although weevil bundnce did not rech desired tretment levels in most lkes, those lkes hving the highest tretment level (4 weevils per stem) showed greter number of significnt declines versus lkes with lower tretment levels. It is possible, given the weevil s life cycle, tht the stocked weevils could hve hd n effect on the Eursin wtermilfoil nd moved out of the plot res before post-ugmenttion smpling five weeks lter. However, despite the fct tht there were fr fewer significnt chnges in Eursin wtermilfoil between ugmented plots nd reference plots, lkes with higher levels of weevils overll showed more declines. It is very importnt to note tht the pprent discrepncy in results cn simply be explined by the fct tht ll plots were open nd thus weevils were ble to move mong the tretment nd reference plots. As result, reference plots in the proximity of tretment plots were n in-effective control. In the initil design of the project, the plots were intended to guge the rte t which weevils migrted out of tretment plots rther thn for comprison to tretment plots. Following study design workshops, prticipnts concluded tht no truly effective control plots could be estblished. Weevil bundnce ws lower in reference thn in treted plots in only hlf of the lkes (Tble 6). Also, Eursin wtermilfoil declines in lkes tht were stocked t lower levels did not exhibit significnt declines in the post-ugmenttion period nd thus would unlikely be different from reference plots. For these resons, pre- nd post-ugmenttion comprisons might offer better indiction of weevil effectiveness. Reserch on the effects of the weevil ugmenttions is preliminry t this point. Dt on Eursin wtermilfoil ws collected only five weeks fter the ugmenttion nd thus it is too erly to generlize these results. Dt collected over the next two-three yers should show how well the weevils overwintered, how mny returned to the ugmented res, nd the effect of ugmenttion on Eursin wtermilfoil. Dt gthered from this study my help gencies implement Eursin wtermilfoil mngement plns tht integrte the use of biocontrol gents such s the milfoil weevil. Additionl reserch is necessry to determine where nd in wht kind of lkes nd in wht types of Eursin wtermilfoil beds the milfoil weevil might estblish lrge, effective popultions. Al- TABLE 5. OBSERVED WEEVIL ABUNDANCE AMONG TREATED PLOTS PRE-AUGMENTATION AND 5 WEEKS POST-AUGMENTATION AND THE PERCENT CHANGE IN EURASIAN WATERMILFOIL MEASUREMENTS AFTER MILFOIL WEEVIL AUGMENTATION. A POSITIVE PERCENTAGE INDICATES A SIGNIFICANT INCREASE IN THAT VARIABLE FROM PRE- AUGMENTATION (1996) TO POST-AUGMENTATION (1997) WHILE A NEGATIVE PERCENTAGE INDICATES A SIGNIFICANT DECLINE. WILCOXON SIGNED-RANK TEST, P NO SIGNIFICANT DIFFERENCE IS INDICATED WITH N.S. Lke Observed weevils per stem pre-ugmenttion Tretment level (milfoil weevils per stem) Observed weevils per stem postugmenttion EWM biomss EWM stem density EWM stem length Apicl tips per EWM stem Ntive mcrophyte biomss Egle % -43% -21% -13% n.s. Kngroo % -27% -30% n.s. -100% Nncy n.s. n.s. n.s. +23% n.s. Perl n.s. n.s. n.s. n.s. -73% Whitewter % n.s. -20% n.s. n.s. Bever Dm n.s. n.s. n.s. n.s % Gilbert % -16% -37% -31% n.s. Lorrine % -43% n.s. -52% n.s. Big Snd % -39% -33% -51% n.s. Kusel % -49% -61% -40% n.s. L. Spring % -64% -60% -57% +3555% Mukwongo % +71% -52% n.s. n.s. In treted plots. 96 J. Aqut. Plnt Mnge. 38: 2000.

10 TABLE 6. OBSERVED WEEVIL ABUNDANCE AMONG TREATED AND REFERENCE PLOTS 5 WEEKS POST-AUGMENTATION AND THE PERCENT CHANGES BETWEEN AUG- MENTED PLOTS AND REFERENCE PLOTS APPROXIMATELY 5 WEEKS AFTER MILFOIL WEEVIL AUGMENTATION. PERCENTAGES INDICATE A SIGNIFICANT DIFFERENCE IN THE AUGMENTED PLOTS AS COMPARED TO THE REFERENCE PLOT. WILCOXON SIGNED-RANK TEST, P 0.05 NO SIGNIFICANT DIFFERENCE IS INDICATED WITH N.S. Lke Tretment level (milfoil weevils per stem) Observed weevils per stem in treted nd (reference) plots Eursin wtermilfoil biomss Eursin wtermilfoil stem density Eursin wtermilfoil stem length Apicl tips per Eursin wtermilfoil stem Ntive mcrophyte biomss Egle (0.2) n.s. n.s. n.s. n.s. n.s. Kngroo 1 0 (0) n.s. n.s. n.s. n.s. n.s. Nncy (0) n.s. n.s. n.s. n.s. n.s. Perl (0.25) n.s. n.s. n.s. n.s. n.s. Whitewter (3.2) n.s. n.s. n.s. n.s. n.s. Bever Dm (0.25) +236% n.s. -26% n.s. n.s. Gilbert 2 0 (0) n.s. n.s. n.s. n.s. n.s. Lorrine (0.95) n.s. n.s. n.s. n.s. n.s. Big Snd (0.7) -46% n.s. n.s. n.s % Kusel (0.35) n.s. n.s. -9% n.s. n.s. L. Spring (2.25) n.s. n.s. n.s. n.s. -85% Mukwongo (0.15) n.s. n.s. n.s. n.s. n.s. though this study suggests tht ugmenting milfoil weevil popultions to 4 weevils per stem of Eursin wtermilfoil my cuse decline in the Eursin wtermilfoil in the immedite re of stocking, how fr these effects extend beyond the ugmenttion plots is unknown. If successful, it would be criticl to determine the number of weevils necessry for Eursin wtermilfoil control within specific size of mcrophyte bed so tht lke resource mngers nd property owners cn effectively implement this method or use it in conjunction with other control methods. ACKNOWLEDGEMENTS We thnk nonymous reviewers for their comments nd recommendtions. We lso thnk the Wisconsin Deprtment of Nturl Resources nd the following lke ssocitions nd groups for their support nd funding: Bever Dm Lke, Big Snd Lke, Egle Lke, Gilbert Lke, Kngroo Lke, Kusel Lke, Long Trde Lke, Lorrine Lke, Lower Spring Lke, Nncy Lke, Perl Lke, Wukesh County Prks System nd Whitewter Lke. LITERATURE CITED Aiken, S. G., P. R. Newroth, nd I. Wile The biology of Cndin weeds. 34. Myriophyllum spictum L. Cn. J. Plnt Sci. 59: Bode, J., S. Bormn, S. Engel, D. Helsel, F. Koshere nd S. Nichols Eursin wtermilfoil in Wisconsin: A report to the legislture. WI Dept. Ntur. Resourc. Mdison, WI. Crpenter, S. R The decline of Myriophyllum spictum in eutrophic Wisconsin lke. Cn. J. Bot. 58: Creed, R. P., Jr A biogeogrphic perspective on Eursin wtermilfoil declines: dditionl evidence of the role of herbivorous weevils in promoting declines? J. Aqut. Plnt Mnge. 36: Creed, R. P., Jr. nd S. P. Sheldon The effect of feeding by North Americn weevil, Euhrychiopsis lecontei, on Eursin wtermilfoil (Myriophyllum spictum). Aqut. Bot. 45: Creed, R. P., Jr. nd S. P. Sheldon Weevils nd wtermilfoil: did North Americn herbivore cuse the decline of n exotic plnt? Ecol. Appl. 5: Creed, R. P., Jr., S. P. Sheldon nd D. M. Cheek The effect of herbivore feeding on the buoyncy of Eursin wtermilfoil. J. Aqut. Plnt Mnge. 30: Engel, S Ecosystem responses to growth nd control of submerged mcrophytes: literture review. Technicl Bulletin Number 170. WI Dept. Ntur. Resourc. Mdison WI. 20 pp. Engel, S. 1990b. Ecologicl impcts of hrvesting mcrophytes in Hlverson Lke, Wisconsin. J. Aqut. Plnt Mnge. 28: Kirschner, R McCullom Lke milfoil sg continues. Lke Wves 11: 1-5. Lillie, R. A The dult qutic nd semiqutic Coleopter of nine northwestern Wisconsin wetlnds. The Coleopterists Bulletin 45: Lillie, R. A. nd D. Helsel A ntive weevil ttcks Eursin wtermilfoil. WI Dept. Ntur. Resourc. Bur. Res. Mnge. Findings No. 40. Mdison, WI. 4 pp. Newmn, R. M. nd L. M. Mher New records nd distribution of qutic insect herbivores of wtermilfoils (Hlorgcee: Myriophyllum spp.) in Minnesot. Entomol. News 106: Newmn, R. M., K. L. Holmberg, D. D. Biesboer nd B. G. Penner Effects of potentil biocontrol gent, Euhrychiopsis lecontei, on Eursin wtermilfoil in experimentl tnks. Aqut. Bot. 53: Newmn, R. M., M. E. Bormn nd S. W. Cstro Developmentl performnce of the weevil Euhrychiopsis lecontei on ntive nd exotic wtermilfoil host plnts. J. N. Amer. Benthol. Soc. 16: Pielou, E. C Popultion nd Community Ecology: Principles nd Methods. New York: Gordon nd Brech. 424 pp. Sheldon, S. P. nd R. P. Creed, Jr Use of ntive insect s biologicl control for n introduced weed. Ecol. Appl. 5: Sheldon, S. P. nd L. M. O Bryn The life history of the weevil Euhrychiopsis lecontei, potentil biologicl control gent of Eursin wtermilfoil. Entomol. News 107: Smith, C. S. nd J. W. Brko Ecology of Eursin wtermilfoil. J. Aqut. Plnt Mnge. 28: Solrz, S. L. nd R. M. Newmn Oviposition specificity nd behvior of the wtermilfoil specilist Euhrychiopsis lecontei. Oecologi 106: Tmyo, M., C. W. O Brien, R. P. Creed, C. E. Grue nd K. Hmel Distribution nd clssifiction of qutic weevils (Coleopter: Curculionide) in the genus Euhrychiopsis in Wshington Stte. Entomol. News J. Aqut. Plnt Mnge. 38: