J. Aquat. Plant Manage. 57: 33–37 The effects of predation on biological control of Eurasian watermilfoil DANIEL C. MILLER AND RONALD L. CRUNKILTON* ABSTRACT crowded canopy, it can inhibit water circulation and alter fish community composition by providing a refuge for prey Eurasian watermilfoil (EWM), Myriophyllum spicatum L., is fish (Diehl 1988). Eurasian watermilfoil can spread quickly an invasive aquatic macrophyte in North America. The because it has adventitious roots, produces seeds that can aquatic milfoil weevil, Euhrychiopsis lecontei, is a native establish new plants, and spreads through fragmentation. herbivore on milfoils that has been used as a biological Fragments that remain on boats and boat trailers can spread control agent for EWM. The objective of this study was to EWM to other lakes (Kimbel 1982). determine if predation by sunfishes (Lepomis spp.) can Euhrychiopsis lecontei (Coleoptera:Curculionidae) (hereaf- suppress milfoil weevil populations below the density ter milfoil weevil) is a specialist herbivore on milfoils necessary to control EWM. In Lake Joanis, Wisconsin, where (Myriophyllum spp.) (Solarz and Newman 1996). It is native to supplemental milfoil weevil stocking had not led to an North America but prefers EWM where it is present increase in weevil density, 944 L mesh exclusion cages (Newman et al. 1997). Adult milfoil weevils live on the stocked with milfoil weevils were used to manipulate upper portion of the plant and eat the stems and leaves. The densities of small bluegill (Lepomis machrochirus)to0,2,and females lay eggs on the tips of EWM. The larvae burrow into 4 per cage. Results indicated an inverse relationship the stem and eat vascular tissue (Newman et al. 1996). They between bluegill and milfoil weevil densities. Mean densities pupate inside the stem, further blocking nutrient transport of milfoil weevils and mean percent of EWM stems showing to the tips. The damage caused by immature milfoil weevils milfoil weevil damage were significantly different among often causes EWM to lose buoyancy (Creed and Sheldon treatments (ANOVA, P ¼ 0.005, P ¼ 0.0004). The average 1992). Although adults are fully aquatic, they overwinter on density of milfoil weevils in cages with no bluegill was 0.31 w shore under leaf litter (Newman et al. 2001) and can be À1 s (weevils per stem). Cages with two bluegill averaged 0.02 affected by shoreline disturbance. À1 À1 ws . Cages with four bluegill averaged 0.01 w s . Stem It is generally accepted that when milfoil weevils are damage ranged from 60.4% in control cages to 13.3% in present in sufficient densities, they will control the impact cages with four bluegill. These results indicate that sunfish, of EWM (Creed and Sheldon 1992, 1993, 1995; Newman et even at relatively low densities, substantially reduce milfoil al. 1996). The minimum density necessary for significant weevil densities and their ability to damage EWM. Protect- declines in EWM varies from 0.25 to 0.5 milfoil weevils per ing stocked milfoil weevils in cages could allow them to stem (w sÀ1) (Newman 2004). Complete collapses of EWM establish higher densities where they can serve as a control beds associated with natural milfoil weevil densities of 3 to 4 for milfoil in some lakes. Variable success of milfoil weevils wsÀ1 have been observed in Vermont (Creed and Sheldon in controlling EWM in different lakes reported in the 1995). Newman and Biesboer (2000) documented a decline literature could be attributable to variable densities of of EWM in Cenaiko Lake in Minnesota due to a natural sunfish populations. milfoil weevil density of about 1.6 w sÀ1. In both cases, native Key words: biocontrol, Euhrychiopsis lecontei, Lepomis macro- macrophytes regrew in open areas after the decline of EWM. chirus, Myriophyllum spicatum. Long-term studies suggest that milfoil weevil density and milfoil density can follow a predator-prey cycle as well as a INTRODUCTION seasonal cycle (Lillie 2000). Milfoil weevils are widely distributed across Wisconsin Eurasian watermilfoil (Myriophyllum spicatum, EWM), since (Jester et al. 2000), but only reach densities above 0.5 w sÀ1 in its introduction to the United States in the 1940s, is one of a few lakes. Jester et al. stocked as many as 10,000 milfoil the most problematic species of invasive aquatic plant weevils into study plots to try to achieve densities of 1, 2, (Couch and Nelson 1985). Eurasian watermilfoil can have and 4 milfoil w sÀ1. They found that 100% of plots stocked significant impacts on native plant species composition. It with 4 w sÀ1 and 60% with 1 w sÀ1 exhibited significant often outcompetes native species because it grows quickly declines in EWM. The density of milfoil weevils stocked and begins growing earlier in the spring than most native declined within 5 wk. Their study did not make use of plants (Aiken et al. 1979). Because it grows in a tangled, enclosures, so the milfoil weevils might have simply moved out of the plots. The authors also speculated that sunfish *First author and second author: Graduate Research Assistant and predation could have been a factor, reducing weevil Professor, College of Natural Resources, University of Wisconsin– Stevens Point, 800 Reserve Street, Stevens Point, WI 54481. Correspond- densities. ing author’s E-mail: [email protected]. Received for publication May Several attempts at milfoil weevil stocking to control 17, 2018 and in revised form August 14, 2018. EWM in North America have been made. Reeves et al. (2008) J. Aquat. Plant Manage. 57: 2019 33 looked at data from 30 lakes stocked with milfoil weevils in MATERIALS AND METHODS Michigan and Wisconsin. The results were highly variable and several lakes did not have detectable milfoil weevils at Lake Joanis is a human-made lake built in 1976 on the the end of the season. Thorstenson (2011), reared and then University of Wisconsin–Stevens Point campus. It has a maximum depth of 7.6 m (25 ft) and a surface area of 9.3 ha stocked 13,000 milfoil weevils in August 2008 and 9,000 in (23 ac). It is a seepage lake with inputs from groundwater August 2009 into Lake Joanis, a 9.3 ha lake on the University and precipitation (Portage County Lake Study 2005). The of Wisconsin–Stevens Point campus that harbored nuisance average secchi depth is 4.27 m (14 ft). It has a very large levels of EWM. This stocking was estimated to be sufficient population of small sunfish that is almost entirely bluegill to bring the natural density of milfoil weevils from 0.01 to À1 À1 (Lepomis macrochirus). A single pass with a 2.6-m (8-ft) seine 0.03 w s up to over 2 w s and represented the highest through one of the milfoil beds typically collect 60 to 80 stocking rate achieved per unit area of a whole-lake bluegill. There are also small numbers of pumpkinseed treatment reported in the literature. The post-stocking À1 (Lepomis gibbosus) and green sunfish (Lepomis cyanellus). The density in September, 2008 was 0.03 w s . Milfoil weevils lake is surrounded by a nature reserve, thus its shoreline is migrate to the shore in September and October (Newman et mostly natural with no development. The EWM beds on al. 2001), so it is possible that they had already left the lake. Lake Joanis occur in 0.3 to 4.6 m (1 to 15 ft) of water. The density of milfoil weevils the following June, however, Because of its shallow depth and clear water, EWM occurred À1 was 0.06 w s and continued to fall over the summer. in approximately 90% of the lake at the time of the study. Because Lake Joanis is part of a protected reserve, its In this study, conducted in 2014, the exclusion cages used shoreline is almost 100% natural (Thorstenson 2011) and consisted of perforated, galvanized angle iron bolted should provide excellent overwintering habitat for milfoil together to form a 0.91 m (3 ft) by 0.91 m by 1.22 m (4 ft) weevils (Newman et al. 2001). The density 23 d after stocking tall cube, with an open bottom. The approximate volume of in August 2009 was 0.03 w sÀ1. Subsequent surveys water in the cages was 944 L (250 gal). At the bottom of the conducted through August 2015 indicated that milfoil cage, additional pieces of angle iron were bolted upside weevils were present in the lake, but only at very low down to secure the cages in the sediment over the EWM. A 3 (, 0.03 w sÀ1) densities. mm (1/8 in) plastic aquaculture mesh was attached to the Previous experiments with exclusion cages have conclud- outside walls of the cage with zip ties. The mesh was also ed that fish predation can have a significant effect on milfoil sewn on all sides with 13.6 kg (30 lb) test-braided fishing line. weevil populations. Ward and Newman (2006) constructed The lids of the cages were created by folding mesh and zip- PVC cages with plastic mesh to exclude or include sunfish in tying it at the corners. They were fastened to the cages with two lakes in Minnesota. One had naturally high milfoil custom-made bungee cords. Twenty-four exclusion cages weevil density and low sunfish density, the other high were placed in Lake Joanis on 3 July. There were eight sunfish density and low weevil density. They stocked milfoil replications of three treatments. The number of bluegill in weevils and sunfish into the cages where necessary. There each treatment was 0, 2, or 4. Placement of the cages was were initially five sunfish in half of the cages and none in the dependent on water levels and accessibility.