ABSTRACT
ABSTRACT
Havel JE, Knight SE, Miazga JR. 2017. Abundance of milfoil weevil in Wisconsin lakes: potential effects from herbicide control of Eurasian watermilfoil. Lake Reserve Manage. 00:1–10.
Although extensive research has revealed the milfoil weevil (Euhrychiopsis lecontei) to have potential for biological control of Eurasian watermilfoil (EWM, Myriophyllum spicatum), herbicides are widely used for controlling this invasive aquatic plant. We hypothesized that density of the milfoil weevil would be lower in lakes with a history of treating EWM with herbicides than in untreated lakes. We surveyed 3 groups of lakes in the Northern Highland Lake District of Wisconsin to see if densities of the milfoil weevil and extent of weevil damage to milfoil depended on history of herbicide control in the lakes or on species of milfoil (EWM or northern watermilfoil, NWM, Myriophyllum sibiricum). We detected milfoil weevils in 28 of 36 lakes surveyed. Despite healthy EWM beds, mean weevil density in lakes that had herbicide treatment of EWM within the past 10 yr was only one-fifth of that in comparable lakes without a recent history of herbicide control. Densities were also significantly higher in untreated EWM lakes than in lakes with only NWM and no EWM. The frequency of milfoil showing evidence of weevil damage depended strongly on the density of the milfoil weevil. The pattern of low weevil density in treated lakes may be explained either by slow recolonization of weevils after destruction of host EWM or by historically low densities of weevils releasing EWM to grow in abundance, triggering aggressive treatment with herbicides.
Eurasian watermilfoil (EWM, Myriophyllum spicatum) is widespread in North America (Pfingsten et al. Citation2016) and has been reported from 635 lakes in Wisconsin (SWIMS Citation2014). In about 15% of 413 surveyed lakes, EWM achieves nuisance levels (Fig. 1). Surface mats shade out native species and interfere with human recreation (Aiken et al. Citation1979, Madsen et al. Citation1991, Zhang and Boyle Citation2010). EWM is typically controlled by application of chemical herbicides or various physical management techniques, methods that are expensive and must be regularly repeated (Madsen Citation2000). Furthermore, harvesting may extend EWM populations through fragmentation (Aiken et al. Citation1979) and herbicides may be toxic to nontarget plants and fish (Nault et al. Citation2014, DeQuattro and Karasov Citation2016).
Published online:
05 July 2017Figure 1. Extent of Eurasian water-milfoil (EWM) cover in Wisconsin lakes. Littoral frequency of occurrence (FOO) was based on sampling 3–1017 points (mean 195) that fell within the depth zone of plant colonization per lake, with total point number scaled relative to lake size and shoreline complexity (Mikulyuk et al. Citation2010). A total of 413 lakes surveyed between 2005 and 2015 are included in the histogram. “Nuisance levels” indicate conditions where lake users notice problems with navigation, recreation, and aesthetics; in general, most users notice problems at about 35% FOO and problems grow more severe with increasing FOO. Source of histogram and other data: Wisconsin Department of Natural Resources Science Services.
Figure 1. Extent of Eurasian water-milfoil (EWM) cover in Wisconsin lakes. Littoral frequency of occurrence (FOO) was based on sampling 3–1017 points (mean 195) that fell within the depth zone of plant colonization per lake, with total point number scaled relative to lake size and shoreline complexity (Mikulyuk et al. Citation2010). A total of 413 lakes surveyed between 2005 and 2015 are included in the histogram. “Nuisance levels” indicate conditions where lake users notice problems with navigation, recreation, and aesthetics; in general, most users notice problems at about 35% FOO and problems grow more severe with increasing FOO. Source of histogram and other data: Wisconsin Department of Natural Resources Science Services.
As an alternative to physical and chemical control, the milfoil weevil (Euhrychiopsis lecontei) has received attention as a potential biocontrol agent (Newman Citation2004). This weevil is native to much of the northern US and Canada (Creed Citation1998), including many lakes in Wisconsin (Jester et al. Citation2000). All life stages are fully aquatic in summer, with adults later moving to shore to spend winter in leaf litter (Fig. 2). Adults and larvae feed on numerous native milfoil species, but show faster development and higher fecundity when fed EWM (Creed and Sheldon Citation1993, Newman Citation2004, Roley and Newman Citation2006). Weevil populations sometimes achieve sufficient densities to control EWM (Creed and Sheldon Citation1995, Newman Citation2004), although their densities are highly variable among EWM beds (Havel et al. Citation2017). Numerous environmental factors affect weevil populations (Newman et al. Citation1998) and some forms of lake management may impact the effectiveness of milfoil weevils for biocontrol. For example, mechanical harvesting both reduces EWM abundance and removes the apical meristem that milfoil weevils use to burrow into the stem (Sheldon and O'Bryan Citation1996, Newman and Inglis Citation2009). Chemical herbicides, by killing host plants, remove habitat and food and thus should indirectly reduce populations of weevils and other biocontrol arthropods (Newman et al. Citation1998, Cuda et al. Citation2008). Indeed, surveys in Florida of water hyacinth (Eichhornia crassipes) and 2 introduced weevils (Neochetina eichhorniae, N. bruchi) revealed lower populations of the weevils in lakes treated with herbicides than in untreated lakes, although differences in plant quality and weevil fecundity suggested a complex interaction (Center et al. Citation1999). Does a similar pattern occur in northern lakes with milfoil weevils, in which the adults must undergo a yearly diapause (Fig. 2)?
Published online:
05 July 2017Figure 2. Life cycle of the milfoil weevil, Euhrychiopsis lecontei. Larvae spend most of their time feeding inside the stem, then develop into pupae that remain in a chamber until emerging from the stem as an adult. Dispersal of adults to and from shore involves some combination of flight, swimming, and rafting. Adult diapause is a period of dormancy during the winter.
Figure 2. Life cycle of the milfoil weevil, Euhrychiopsis lecontei. Larvae spend most of their time feeding inside the stem, then develop into pupae that remain in a chamber until emerging from the stem as an adult. Dispersal of adults to and from shore involves some combination of flight, swimming, and rafting. Adult diapause is a period of dormancy during the winter.
The objective of the current study was to compare density of the milfoil weevil in lakes with and without chemical herbicide control and between untreated lakes with EWM and lakes containing only northern watermilfoil (NWM, Myriophyllum sibiricum). Because of possible long-term effects from the chemicals, we hypothesized that weevil densities would be lower in chemically treated lakes than in those without recent (10 yr) herbicide control.
Materials and methods
Study site and lake descriptions
We assessed density of milfoil weevils in 36 lakes from the Northern Highland Lakes District of Wisconsin (Fig. 3). These lakes were generally moderate-sized kettle lakes, with low to moderate concentration of dissolved solids (as conductivity) and a wide range in transparency (0.6–6.1 m; Appendix 1). The 3 groups of lakes were well matched with respect to physical (area, maximum depth, transparency) and chemical (conductivity) attributes, as well as their distribution in the region (Fig. 3). A multivariate comparison of these 4 variables indicated no significant difference among the 3 groups of study lakes (MANOVA, Pillai's F = 1.29, P = 0.265) and comparisons for each separate variable also showed no significant difference among groups (all P values ≥0.25). Four of the lakes in this survey (Fig. 3) were sampled intensively as part of a larger experimental study (Havel et al. Citation2017).
Published online:
05 July 2017Figure 3. Study area in northern Wisconsin. Lake and county names are given in Appendix 1. Symbols: triangle = EWM chem, square = EWM no chem, and circle = NWM-no chem. Filled symbols = milfoil weevils present, open symbols = absent.
Figure 3. Study area in northern Wisconsin. Lake and county names are given in Appendix 1. Symbols: triangle = EWM chem, square = EWM no chem, and circle = NWM-no chem. Filled symbols = milfoil weevils present, open symbols = absent.
All 36 lakes had well-developed littoral zones and one or more species of milfoil. Here, we use the term milfoil to refer to either EWM or NWM. Both species are widespread in the region (S. Knight, pers. obs.). After the project began, we learned that genetic analysis of EWM plants from Little Bearskin Lake indicated this population consisted of NWM × EWM hybrid (M. Nault, Wisconsin Department of Natural Resources, 14 October 2016, pers. com.). Since the frequency of such hybrids in other Wisconsin lakes is unknown, we assigned hybrids to the EWM category.
Lakes were classified into 3 groups: lakes with EWM and a recent history of herbicide treatment (EWM-chem, n = 10 lakes); lakes with EWM and no recent chemical treatment (EWM-no chem, n = 13); and lakes containing northern watermilfoil and no records of EWM or chemical treatment (NWM-no chem, n = 13). In the first 2 categories of lakes, NWM may have been present at other sites in the lakes, but was not recorded nor sampled. We thus sampled only one species of milfoil from each lake.
We determined lake treatment histories by reviewing permit applications to the Wisconsin Department of Natural Resources (DNR; Wisconsin Administrative Code NR 107.04). Lakes in the EWM-chem group were treated with various formulations of the herbicide 2,4-D at least once in the previous 10 yr (). In large lakes (e.g., Lake Tomahawk), herbicide treatments were applied in different sections of the lake in different years (N. Greedy, Tomahawk Lake Association, 2 November 2016, pers. com.); not all EWM beds were treated each year. Details on extent of herbicide treatments are available online (Supplemental Appendixes S1, S2). Furthermore, we assume that lake associations did not apply herbicides in any lake without permission from DNR. Lakes in the EWM-no chem group had no record of chemical treatment in the 10 yr prior to and including our survey (2004–2015; ). We attempted to sample milfoil from about 12 lakes in each group, but we were hindered by low densities of EWM in some EWM-chem lakes and thus chose to omit these lakes from our survey.
Table 1. Abundance of milfoil weevils (total of all life stages per stem) and weevil damage to milfoil (%) in each lake. Lakes are grouped according to presence of Eurasian watermilfoil (EWM) or only northern watermilfoil (NWM) and whether or not the lake had received chemical herbicide treatment for EWM in the last 10 yr (chem). N = total number of milfoil stems sampled. Weevil abundance includes all life stages including eggs. “sc” = scattered stems not in discrete beds.
Sampling methods
In 2014 and 2015, we sampled most lakes during June or July, when weevils tend to be most abundant in the region (Havel et al. Citation2017). Exceptions included 4 lakes (Weber, Upper Gresham, Tomahawk, and Rest) that were sampled in late August 2014. Since these lakes followed similar trends in weevil abundance as others in their respective groups (below), we included these lakes in the survey.
Milfoil weevils were sampled by collecting their host plants. Within each lake, we first toured shallow areas by boat to see how many milfoil beds had enough plants to sample. We attempted to sample 2–4 milfoil beds and collect 50 stems per bed, but this sampling intensity proved impractical. Sometimes the milfoil plants were rare and at other times they were so damaged that we could not sample them for counting weevils. The actual sampling intensity is shown in . In 2 lakes (Brandy, Seidel), only scattered plants were available and, in one (Seidel), the sample size (number of EWM stems) was very small. Lake Seidel is included for describing prevalence but not population estimation. For the remaining lakes, total sample sizes (N) ranged from 24 to 298 milfoil stems per lake. Detection limits on each sample date ranged from 0.007 (Manson Lake) to 0.041 (Anvil Lake) weevils per stem. Lakes were sampled on 1 or 2 dates.
To sample weevils, we randomly collected up to 50 milfoil stems per bed by rowing through the length of the bed and pulling plants with a short-handled rake. To include all weevil age classes, we counted only from stems that were at least 50 cm long and had not yet flowered. Stems that were missing their tips (apical meristem) were counted only for the purpose of damage estimates. The milfoil plants from each bed were bagged, kept in a cooler, and processed in the lab within 24 h. Although our method of collection could have caused weevil loss, observations in the laboratory indicated that eggs and adults were difficult to dislodge from milfoil stems. Larvae and pupae are typically endophytic.
While in the field, we took numerous precautions to avoid transmitting EWM and other invasive plants within and among lakes. To prevent spread by fragmentation, we avoided motoring through EWM beds and we collected any stray EWM fragments found floating in the water. We thoroughly cleaned the boat by hand at the launch site, removing all visible plant fragments and discarded them on land. Later at the field station, the boat was washed with steam and a power washer. Following processing of plants, all EWM were double bagged and disposed of in the trash.
Lab analysis
We first examined each milfoil stem for signs of weevil damage. These signs included hollowed stems, blackened or missing tips, and “blast holes” from pupae chewing through the stem. Stems were classified as damaged or not to calculate percentage of weevil damage per bed.
Using only intact stems (with apical meristem), we counted the abundance of weevils by summing counts from all 4 stage classes: eggs, larvae, pupae, and adults. Combining all 4 stages into weevil density estimates is widely reported in the literature (reviewed in Havel and Knight Citation2016). Using only intact stems allowed accurate estimates of eggs and adults, but tended to underestimate larvae and pupae to only 64% of their density relative to stems without apical meristem (Havel and Knight Citation2016).
Adults were counted in a tray of water over the light box or, if missed, later while examining the stem under the dissecting microscope. Other stages were counted under the microscope. We observed most weevil eggs attached to the apical meristem. Larvae and pupae were counted after slicing open the stem. Larvae were usually observed in regions of the stem that were clear, where the cortex had been eaten by the larva. Pupae were dissected from their pupal chamber, which appeared as an enlarged clear area in the stem.
Statistical analysis
We compared mean weevil densities among the 3 groups of lakes using one-way analysis of variance, followed by Tukey a posteriori comparisons, in Minitab (ver. 17). We also used these statistical methods to compare percentage of weevil damage among the 3 lake groups. Although checks on the homogeneous variance assumption (Levine's Test) indicated no significant departure for either response variable, residuals did not follow a normal distribution. Because ANOVA tends to be robust to violations of normality (Whitlock and Schluter Citation2009), we have reported results from ANOVA below.
Results
In most lakes with EWM, we found 2 or more discrete EWM beds and sufficient plants to estimate weevil abundance (). Since weevil abundance and damage were not substantially different in lakes with only one bed compared to other lakes, the single-bed lakes were included in summary statistics. Lakes with NWM also had distinct beds, though generally the plants did not form surface mats.
We found milfoil weevils were common in 28 of the 36 lakes surveyed (Fig. 3). The weevils were present in 40% of lakes with a history of chemical herbicide treatment and EWM that had since recovered. In contrast, 92% of the lakes with EWM and no recent history of herbicide treatment had detectable weevil populations (). Weevils were also detected in 92% of the lakes with only NWM.
Weevil abundance (including eggs) ranged from 0 to 1.54 weevils per stem () and mean abundance was significantly different among the 3 groups of lakes (ANOVA, F = 9.86, P < 0.001; Fig. 4). We found the lowest average weevil densities in the EWM-chem group (mean 0.17 weevils per stem). Most lakes in this group had very low weevil densities (Fig. 4). However, one lake in this group (Ellwood) had the highest density of weevils that we observed in the survey (1.54 weevils/stem), with high densities in both years surveyed (2.12 weevils/stem in 2014 and 1.27 in 2015). This outlier was not a consequence of weak herbicide treatment, as the proportion of total lake area treated (summed across all years of treatment) was among the highest of all the treated lakes (Supplement Appendix S1). EWM lakes without herbicide treatments (EWM-no chem) had the highest average weevil densities (mean 0.79 weevils/stem), nearly 5-fold higher than those in the EWM-chem group. Nevertheless weevil densities in the EWM-no chem group were highly variable (coefficient of variation = 57.7%). Weevil abundance was below detection in Anvil Lake in the EWM-no chem group. Weevils were also common in the NWM-no chem lakes (all but one lake), though their mean abundance (0.22 weevils/stem) was not significantly different from that of the EWM-chem group (Tukey pairwise comparison, Fig. 4).
Published online:
05 July 2017Figure 4. Box plots for weevil densities by lake group (defined in ), surveyed summer 2014–2015. Numbers of lakes surveyed are shown below each group. ANOVA F = 9.86, P < 0.001.
Figure 4. Box plots for weevil densities by lake group (defined in ), surveyed summer 2014–2015. Numbers of lakes surveyed are shown below each group. ANOVA F = 9.86, P < 0.001.
Weevil populations were dominated by early life stages in all the lakes surveyed. Based on 1995 total weevils counted, 93.5% were eggs and larvae. In a companion study of 4 main study lakes, eggs and larvae dominated the age structure of each population throughout the summer in every year studied (2013–2015; Havel et al. Citation2017).
Weevil damage was widespread in lakes with weevil populations. Across all lakes, weevil damage increased significantly with weevil abundance (Percent Damage = 3.28 + 25.5 Weevils/stem; R2 = 0.648), mostly as a consequence of larval stem mining. Damage was rarely observed in the EWM-chem group (mean = 5.2% of EWM stems), with Ellwood Lake again being an outlier (38%). Damage appeared to be more widespread in both the EWM-no chem group (20.4%) and NWM-no chem group (13.7%), ranging between 0 and 48% of the plants for both groups (). Nevertheless, differences in mean weevil damage were only marginally different among these groups (F = 3.17, P = 0.055).
Discussion
To be effective in the long term, biological control agents must persist in the natural community. By understanding the conditions that limit biocontrol agents in nature, biologists can employ conservation strategies to facilitate their survival and reproduction (Newman et al. Citation1998). For instance, maintaining certain habitat features can provide refuge from predators. For herbivorous biocontrol insects, maintaining a diversity of native plants may also provide alternate foods during periods when the invasive host plant is in low abundance. Such conservation strategies have been employed as part of integrated pest management in agroecosystems (cultural control; Batra Citation1982), in rangelands (Messersmith and Adkins Citation1995), and in aquatic ecosystems (Newman et al. Citation1998). In freshwater habitats, application of the herbicides 2,4-D and endothall at recommended dosages in early spring can destroy EWM before active growth of native macrophytes (Skogerboe and Getsinger Citation2006). This practice has been shown to have minimal impacts on nontarget macrophytes, invertebrates, and fish (Kovalenko et al. Citation2010; but see Nault et al. Citation2014). Nevertheless, we should expect negative indirect effects of herbicides on biocontrol arthropods that are highly host-specific, since their habitat and food source are destroyed (Cuda et al. Citation2008). Indeed, surveys in Florida revealed that populations of 2 weevils (Neochetina eichhorniae, N. bruchi) previously introduced to control water hyacinth tended to be lower in lakes treated with herbicides than in untreated lakes (Center et al. Citation1999). However, plant tissue quality (nutrients and water content) and weevil fecundity (eggs per female) were higher in these Florida treated lakes compared to untreated lakes. Center et al. (Citation1999) suggested that, although weevil populations declined in herbicide-treated lakes, fecundity improved as a function of plant tissue quality. They suggested that herbicide and biological management could be successful in integrated control of water hyacinth by maintaining a viable weevil population.
In the current study, the average density of milfoil weevils in lakes with EWM was substantially lower in lakes that had been treated with herbicides than in those lakes not chemically treated. Furthermore, the percentage of lakes with detectable weevil populations was substantially less in the treated lakes (40%) than in untreated EWM lakes (92%). This result is not a consequence of recent herbicide application as most lakes were treated at least 2 yr prior to the weevil survey and EWM plants had since grown to form distinct beds and some had surface mats. Thus, the conditions appeared to be suitable for weevils and yet weevils were sparse or undetected in most lakes that had been chemically treated. Two general reasons might explain this pattern. First, weevils may have been extirpated with the EWM and then slow to recolonize the lake. Following chemical destruction of EWM in spring, adult milfoil weevils migrating back from overwintering on shore (Fig. 2) would likely have no other food source and thus either starve or be forced to migrate to another lake. Dispersal rates and distances of adult milfoil weevils are poorly known and in need of further study. A second explanation for the pattern we observed (lower weevils in treated lakes) would be if weevil populations were low in these lakes for other reasons, such as poor overwintering habitat (Newman et al. Citation2001) or high densities of insectivorous fish (Sutter and Newman Citation1997, Parsons et al. Citation2011). In the absence of other milfoil herbivores, such as the milfoil midge (Cricotopus myrioplylli) and pyralid moth (Acentria ephemerella; Newman Citation2004), low weevil densities might have resulted in unhindered growth of EWM to high densities, resulting in the need for aggressive treatment with herbicides. In the current study, we did not determine weevil abundance prior to herbicide management. Future research is needed to test this second hypothesis.
While our results are intriguing, the current study has its limitations. The most important limitation is that, like other observational surveys, we can show patterns but not demonstrate causation (Whitlock and Schluter Citation2009). There are just too many possible confounding variables. Future research needs to investigate weevil densities before and after herbicide treatment to account for other factors regulating abundance, such as overwintering habitat and predation pressure. Such an approach would require long-term sampling of multiple weevil populations, paired with agency records of permits for herbicide application. Information on dispersal of adult weevils would also be useful. A second limitation of the current study was the sample size (number of lakes) having sufficient sampling intensity (number of milfoil stems) for detecting small weevil populations. The detection limit within each lake on each date was usually about 0.02 weevils/stem (if one weevil were counted on 50 stems), and was sometimes greater with fewer stems sampled. Future research in a similar study system should collect more stems for greater sampling precision.
The 5-fold lower weevil density that we observed in chemically treated lakes is important because damage to EWM (percentage of plants) increases with weevil density and others have shown that control of EWM is most likely at weevil densities ≥1.0 weevils/stem (reviewed in Newman Citation2004). Thus, chemical control may reduce the probability of weevil biocontrol. Nevertheless, we had one outlier (Ellwood Lake) in which weevils were abundant 2 yr after a 9-yr period of chemical control. The reason for this recovery is unknown, although the fact that we were able to survey a large number of plants from each of 6 EWM beds in this lake suggests that the EWM population recovered quickly from chemical application. Nothing in the permit application would indicate that dosage or area of treatment was low (Supplemental Appendixes S1, S2). More research is needed on Ellwood Lake to determine if the resident EWM is more resistant to herbicides than EWM populations in other lakes. Also, detailed surveys of this lake could determine if spatially segregated beds of native milfoils were common and thus acted as a possible refuge for weevils during periods when EWM was suppressed in other beds.
In the current study, weevil densities were lower in lakes with only NWM than in lakes invaded by EWM and not treated recently with chemicals. Our results are consistent with earlier experimental tests of weevil performance on different species of host plant (Solarz and Newman Citation2001, Roley and Newman Citation2006, Borrowman et al. Citation2015). Although NWM and other native milfoils are suitable food for both adults and larvae, the development time and fecundity of milfoil weevils are substantially better on EWM than on native milfoils (Solarz and Newman Citation2001). Also, milfoil weevils have recently been shown to have better performance on EWM × NWM hybrid milfoil than either EWM or NWM (Borrowman et al. Citation2015). Our survey results depart from that of a survey of 17 Washington lakes (Tamayo et al. Citation2004) that found no significant difference in densities of milfoil weevil between lakes with EWM and lakes with only NWM. However, their small sample size precluded detecting statistical significance between the 2 groups. Furthermore, the authors provided no information on chemical control of EWM in these lakes; if any of the EWM lakes received chemical control, based on the current study we would have expected lower densities of weevils, similar to those found in the NWM lakes. The weevil populations in our NWM-no chem lakes, though low in density, provided a source of colonists in the event EWM ever invades these lakes.
Despite their higher average densities, EWM-no chem lakes had high variation in weevil densities among lakes (Fig. 4), indicating that other factors influence weevil densities in these lakes. For instance, Cuda (Citation2008) suggested that aerial drifting of malathion, commonly applied in wetlands for mosquito control, impacted the biocontrol insect specific for hydrilla (Hydrellia pakistanae). A variety of other environmental conditions are important for growth and persistence of weevil populations (Newman et al. Citation1998). For example, small size classes of bluegill sunfish (Lepomis macrochirus) are well known to consume weevils (Sutter and Newman Citation1997, Maxson Citation2016). Lakes with large numbers of sunfish may have much lower densities of milfoil weevils than could be supported by a healthy EWM bed. Such an explanation might also shed light on the chemically treated outlier, Ellwood Lake, which experienced a depression in reproduction of some fish species during the 5 yr of chemical treatment (Greg Sass, Wisconsin DNR, 2 November 2016, pers. comm.). Rapid recovery in the weevil population following chemical treatment may have been associated with the absence of young bluegills in this lake.
The complex interplay between toxicity effects, herbivory, and insectivorous fish suggests that lake managers should use special care in selecting target lakes for weevil biocontrol. If conditions for weevil population persistence are adequate (good shoreline habitat, low fish abundance) and recolonization of weevils is slow following prior EWM removal, then weevils could be stocked into EWM beds before this plant grows to damaging levels.
| Number of milfoil beds | ||||||||
|---|---|---|---|---|---|---|---|---|
| Lake name by group | Sampled | With weevils | Total N | % stems damaged | Weevil density (no./stem) | Date weevils sampled | Years with chem treatment | |
| EWM-chem (n = 10) | ||||||||
| Arrowhead | 2 | 0 | 70 | 0 | 0.00 | Jul 2015 | 2009–2011 | |
| Big Sand | 3 | 2 | 74 | 4 | 0.06 | Jul 2014 | 2008–2010, 2012 | |
| Ellwood | 6 | 6 | 267 | 38 | 1.54 | Jul 2014 & Jun 2015 | 2004–2012 | |
| Kawagesaga | 1 | 0 | 36 | 0 | 0.00 | Jul 2015 | 2014 | |
| Kentuck | 2 | 0 | 54 | 0 | 0.00 | Jul 2015 | 2013 | |
| Little St. Germain | 3 | 0 | 70 | 3 | 0.00 | Jul 2015 | 2012–2015 | |
| Silver | 3 | 3 | 80 | 2 | 0.06 | Jul 2015 | 2007, 2009, 2012, 2014 | |
| South Twin | 2 | 0 | 51 | 0 | 0.00 | Jul 2014 | 2012 | |
| Tomahawk | 1 | 0 | 30 | 0 | 0.00 | Aug 2014 | 2007–2015 | |
| Upper Gresham | 5 | 2 | 142 | 5 | 0.04 | Aug 2014 & Jul 2015 | 2007–2010, 2012–2013 | |
| mean | 5.2 | 0.17 | ||||||
| EWM-no chem (n = 13) | ||||||||
| Anvil | 2 | 0 | 40 | 0 | 0.00 | Jul 2015 | none | |
| Boot | 4 | 4 | 206 | 12 | 0.96 | Jun 2014 & Jun 2015 | none | |
| Brandy | 1 | 1 | 34 | 26 | 1.10 | Jun 2015 | none | |
| Clearwater | 3 | 3 | 73 | 44 | 0.84 | Jul 2015 | none | |
| Kathan | 3 | 2 | 61 | 0 | 0.33 | Jul 2015 | none | |
| Little Bearskin | 4 | 4 | 249 | 48 | 1.52 | Jun 2014 & Jun 2015 | none | |
| Long | 4 | 4 | 276 | 25 | 0.90 | Jun 2014 & Jun 2015 | none | |
| Manson | 4 | 4 | 298 | 23 | 0.92 | Jun 2014 & Jun 2015 | none | |
| Pickerel | 3 | 3 | 54 | 42 | 1.40 | Jul 2014 | none | |
| Pine | 2 | 2 | 47 | 19 | 0.85 | Jul 2014 | none | |
| Seidel | sc | sc | 7 | 14 | 1.00 | Jul 2014 | none | |
| Upper Kaubashine | 4 | 3 | 112 | 3 | 0.08 | Jul 2015 | none | |
| Weber | 3 | 2 | 95 | 8 | 0.42 | Aug 2014 | none | |
| mean | 20.3 | 0.79 | ||||||
| NWM-no chem (n = 13) | ||||||||
| Big | 2 | 2 | 67 | 13 | 0.22 | Jul 2015 | none | |
| Black Oak | 1 | 1 | 30 | 0 | 0.13 | Jul 2014 | none | |
| Clear | 2 | 2 | 54 | 13 | 0.17 | Jul 2015 | none | |
| Forest | 2 | 0 | 51 | 0 | 0.00 | Jul 2015 | none | |
| High | 2 | 1 | 46 | 5 | 0.10 | Jul 2015 | none | |
| Johnson | 2 | 2 | 58 | 16 | 0.38 | Jul 2014 | none | |
| Palmer | 2 | 1 | 50 | 2 | 0.10 | Jul 2015 | none | |
| Plum | 2 | 2 | 49 | 20 | 0.20 | Jul 2015 | none | |
| Rest | 2 | 2 | 49 | 18 | 0.24 | Aug 2014 | none | |
| Round | 3 | 3 | 85 | 31 | 0.47 | Jul 2015 | none | |
| Van Vliet | 3 | 3 | 93 | 15 | 0.26 | Jul 2015 | none | |
| Wild Rice | 2 | 1 | 54 | 2 | 0.06 | Jul 2015 | none | |
| Wildcat | 2 | 2 | 50 | 48 | 0.54 | Jul 2015 | none | |
| mean | 14.1 | 0.22 | ||||||
| Location (GPS coordinates) | ||||||||
|---|---|---|---|---|---|---|---|---|
| Surface area | Max. depth | Secchi | Conduct. | |||||
| Lake name | WBIC | County | Lat (N) | Long (W) | (ha) | (m) | transpar. (m) | (µS) |
| EWM-chem (n = 10) | ||||||||
| Arrowhead | 1541500 | Vilas | 45.906 | −89.690 | 39 | 13.1 | 3.4 | 99 |
| Big Sand | 1602600 | Vilas | 46.063 | −88.982 | 577 | 17.1 | 2.1 | 77 |
| Ellwood | 650500 | Florence | 45.858 | −88.142 | 53 | 7.6 | 5.2 | 210 |
| Kawaguesaga | 1542300 | Oneida | 45.868 | −89.738 | 283 | 13.4 | 3.4 | 92 |
| Kentuck | 716800 | Vilas, Forest | 45.985 | −89.000 | 405 | 12.2 | 0.6 | 81 |
| Little Saint Germain | 1596300 | Vilas | 45.903 | −89.455 | 393 | 16.2 | 2.7 | 66 |
| Silver | 1599800 | Vilas | 45.921 | −89.238 | 23 | 5.8 | 1.8 | 31 |
| South Twin | 1623700 | Vilas | 46.031 | −89.171 | 254 | 13.1 | 3.0 | 89 |
| Tomahawk | 1542700 | Oneida | 45.831 | −89.648 | 1401 | 25.6 | 5.9 | 106 |
| Upper Gresham | 2330800 | Vilas | 46.068 | −89.737 | 146 | 9.8 | 2.4 | 87 |
| EWM-no chem (n = 13) | ||||||||
| Anvil | 968800 | Vilas | 45.943 | −89.064 | 153 | 9.8 | 2.7 | 36 |
| Boot | 1619100 | Vilas | 45.969 | −89.326 | 116 | 4.6 | 1.5 | 55 |
| Brandy | 1541300 | Vilas | 45.907 | −89.701 | 46 | 13.4 | 3.0 | 87 |
| Clearwater | 1616400 | Vilas, Oneida | 45.850 | −89.200 | 136 | 14.6 | 3.4 | 87 |
| Kathan | 1598300 | Oneida | 45.872 | −89.318 | 87 | 4.6 | 0.9 | 3 |
| Little Bearskin | 1523500 | Oneida | 45.711 | −89.699 | 74 | 8.2 | 1.8 | 109 |
| Long | 2303500 | Iron | 46.260 | −90.026 | 150 | 10.4 | 0.9 | 98 |
| Manson | 1517200 | Oneida | 45.562 | −89.633 | 96 | 16.5 | 4.9 | 66 |
| Pickerel | 388100 | Forest, Langlade | 45.396 | −88.909 | 515 | 5.8 | 3.4 | 157 |
| Pine | 406900 | Forest | 45.677 | −88.981 | 677 | 4.6 | 3.7 | 90 |
| Seidel | 672000 | Florence | 45.868 | −88.352 | 20 | 13.4 | 1.8 | 134 |
| Upper Kaubashine | 1535000 | Oneida | 45.789 | −89.741 | 73 | 17.1 | 3.7 | 104 |
| Weber | 2909000 | Iron | 46.411 | −90.401 | 26.3 | 11.9 | 4.0 | 50 |
| NWM-no chem (n = 13) | ||||||||
| Big | 2334700 | Vilas | 46.155 | −89.770 | 335 | 18.6 | 2.8 | 120 |
| Black Oak | 1630100 | Vilas | 46.163 | −89.316 | 228 | 25.9 | 5.2 | 85 |
| Clear | 2329000 | Vilas | 46.147 | −89.812 | 208 | 13.7 | 3.0 | 75 |
| Forest | 2762200 | Vilas | 46.149 | −89.376 | 189 | 18.3 | 4.6 | 69 |
| High | 2344000 | Vilas | 46.156 | −89.549 | 300 | 11.0 | 2.7 | 160 |
| Johnson | 1541100 | Vilas, Oneida | 45.900 | −89.721 | 34 | 12.8 | 1.8 | 52 |
| Palmer | 2962900 | Vilas | 46.200 | −89.500 | 261 | 4.0 | 1.8 | 115 |
| Plum | 1592400 | Vilas | 46.003 | −89.519 | 428 | 17.4 | 6.1 | 90 |
| Rest | 2327500 | Vilas | 46.141 | −89.875 | 265 | 16.2 | 3.0 | 83 |
| Round | 2334900 | Vilas | 46.172 | −89.710 | 70 | 7.6 | 4.6 | 112 |
| Van Vliet | 2956800 | Vilas | 46.193 | −89.758 | 93 | 6.1 | 1.5 | 130 |
| Wild Rice | 2329800 | Vilas | 46.065 | −89.797 | 155 | 7.9 | 3.0 | 87 |
| Wildcat | 2336800 | Vilas | 46.173 | −89.617 | 115 | 11.6 | 2.4 | 130 |
ulrm_a_1334250_sm4183.zip
Download Zip (20 KB)Related Research Data
Acknowledgments
We thank J. Bevington, E. Fruehling, C. Winter, and N. Winter for their help with field collections and laboratory analysis. C. Warden kindly helped with information on years that lakes received chemical treatments. We also thank the following people for help with the figures: A. Mikulyuk and M. Nault (Fig. 1), A. Wempe (Fig. 2), and K. Maxson (Fig. 3). Comments from anonymous reviewers and the editor helped us refine the ideas presented in this paper. Logistical support was provided by the UW-Madison Trout Lake Station and financial support by a grant from the Wisconsin Department of Natural Resources (grant ACE-122-12).
References
- Aiken SG, Newroth PR, Wile I. 1979. The biology of Canadian weeds. 34. Myriophyllum spicatum L. Can J Plant Sci. 59:201–215. [Google Scholar]
- Batra SWT. 1982. Biological control in agroecosystems. Science. 215:134–139. [Google Scholar]
- Borrowman KR, Sager EPS, Thum RA. 2015. Growth and developmental performance of the milfoil weevil on distinct lineages of Eurasian watermilfoil and a northern × Eurasian hybrid. J Aquat Plant Manage. 53:81–87. [Google Scholar]
- Center TD, Dray FA, Jubinsky GP, Grodowitz MJ. 1999. Biological control of water hyacinth under conditions of maintenance management: can herbicides and insects be integrated? Environ Manage. 23:241–256. [Google Scholar]
- Creed RP, Jr. 1998. A biogeographic perspective on Eurasian watermilfoil declines: additional evidence for the role of herbivorous weevils in promoting declines? J Aquat Plant Manage. 36:16–22. [Google Scholar]
- Creed RP, Jr., Sheldon SP. 1993. The effect of feeding by a North American weevil, Euhrychiopsis lecontei, on Eurasian watermilfoil (Myriophyllum spicatum). Aquat Bot. 45:245–256. [Google Scholar]
- Creed RP, Jr., Sheldon SP. 1995. Weevils and watermilfoil: did a North American herbivore cause the decline of an exotic plant? Ecol Appl. 5:1113–1121. [Google Scholar]
- Cuda JP, Charudattan R, Grodowitz MJ, Newman RM, Shearer JF, Tamayo ML, Villegas B. 2008. Recent advances in biological control of submersed aquatic weeds. J Aquat Plant Manage. 46:15–32. [Google Scholar]
- DeQuattro ZA, Karasov WH. 2016. Impacts of 2,4-dichlorophenoxyacetic acid aquatic herbicide formulations on reproduction and development of the fathead minnow (Pimephales promelas). Environ Toxicol Chem. 35:1478–1488. [Google Scholar]
- Havel JE, Knight SE. 2016. A field test on the effectiveness of milfoil weevil for controlling Eurasian watermilfoil in northern lakes. Final report to Wisconsin Department of Natural Resources, grant ACE-122-12, Madison, WI. [Google Scholar]
- Havel JE, Knight SE, Maxson KA. 2017. A field test on the effectiveness of milfoil weevil for controlling Eurasian watermilfoil in Wisconsin lakes. Hydrobiologia. DOI:10.1007/s10750-017-3142-2 [Crossref], [Google Scholar]
- Jester LL, Bozek MA, Helsel DR, Sheldon SP. 2000. Euhrychiopsis lecontei distribution, abundance, and experimental augmentations for Eurasian watermilfoil control in Wisconsin lakes. J Aquat Plant Manage. 38:88–97. [Google Scholar]
- Kovalenko KE, Dibble ED, Slade JG. 2010. Community effects of invasive macrophyte control: role of invasive plant abundance and habitat complexity. J Appl Ecol. 47:318–328. [Google Scholar]
- Madsen JD. 2000. Advantages and disadvantages of aquatic plant management techniques. Report Number ERDC/EL MP-00-1. Vicksburg (MS): US Army Corps of Engineers. 38 p. [Crossref], [Google Scholar]
- Madsen JD, Sutherland JW, Bloomfield JA, Eichler LW, Boylen CW. 1991. The decline of native vegetation under dense Eurasian watermilfoil canopies. J Aquat Plant Manage. 29:94–99. [Google Scholar]
- Maxson KA. 2016. Trophic interactions and the efficacy of milfoil weevils for biocontrol of Eurasian watermilfoil in Wisconsin lakes. Master's thesis. Missouri State University: Springfield, MO. [Google Scholar]
- Messersmith CG, Adkins SW. 1995. Integrating weed-feeding insects and herbicides for weed control. Weed Technol. 9:199–208. [Google Scholar]
- Mikulyuk A, Hauxwell J, Rasmussen P, Knight S, Wagner KI, Nault M, Ridgely D. 2010. Testing a methodology for assessing plant communities in temperate inland lakes. Lake Reserv Manage. 26:54–62. [Google Scholar]
- Nault ME, Netherland MD, Mikulyuk A, Skogerboe JG, Asplund T, Hauxwell J, Toshner P. 2014. Efficacy, selectivity, and herbicide concentrations following a whole-lake 2,4-D application targeting Eurasian watermilfoil in two adjacent northern Wisconsin lakes. Lake Reserv Manage. 30:1–10. [Google Scholar]
- Newman RM. 2004. Biological control of Eurasian watermilfoil by aquatic insects: basic insights from an applied problem. Archiv Hydrobiol. 159:145–184. [Google Scholar]
- Newman RM, Inglis WG. 2009. Distribution and abundance of the milfoil weevil, Euhrychiopsis lecontei, in Lake Minnetonka and relation to milfoil harvesting. J Aquat Plant Manage. 47:21–25. [Google Scholar]
- Newman RM, Ragsdale DW, Milles A, Oien C. 2001. Overwinter habitat and the relationship of overwinter to in-lake densities of the milfoil weevil, Euhrychiopsis lecontei, a Eurasian watermilfoil biological control agent. J Aquat Plant Manage. 39:63–67. [Google Scholar]
- Newman RM, Thompson DC, Richman DB. 1998. Conservation strategies for the biological control of weeds. P. 371–396. In: Barbosa P, (ed). Conservation biological control. Academic Press: New York (NY). [Google Scholar]
- Parsons JK, Marx GE, Divens M. 2011. A study of Eurasian watermilfoil, macroinvertebrates and fish in a Washington lake. J Aquat Plant Manage. 49:71–82. [Google Scholar]
- Pfingsten IA, Berent L, Jacono CC, Richerson MM. 2016. Myriophyllum spicatum. USGS Nonindigenous Aquatic Species Database. https://nas.er.usgs.gov/queries/FactSheet.aspx?speciesID=237. Accessed October 2016. [Google Scholar]
- Roley SS, Newman RM. 2006. Developmental performance of the milfoil weevil, Euhrychiopsis lecontei (Coleoptera: Curculionidae), on Northern watermilfoil, Eurasian watermilfoil, and hybrid (Northern × Eurasian) watermilfoil. Environ Ent. 35:121–126. [Google Scholar]
- Sheldon SP, O'Bryan LM. 1996. Life history of the weevil Euhrychiopsis lecontei, a potential biological control agent of Eurasian watermilfoil. Ent News. 107:16–22. [Google Scholar]
- Skogerboe JG, Getsinger KD. 2006. Selective control of Eurasian watermilfoil and curlyleaf pondweed using low doses of endothall combined with 2,4-D. Vicksburg (MS): US Army Corps of Engineers Research and Development Center; APCRP Technical Notes Collection ERDC/TN APCRP-CC-05. [Google Scholar]
- Solarz SL, Newman RM. 2001. Variation in hostplant preference and performance by the milfoil weevil, Euhrychiopsis lecontei Dietz, exposed to native and exotic watermilfoils. Oecologia. 126:66–75. [Google Scholar]
- Sutter TJ, Newman RM. 1997. Is predation by sunfish (Lepomis spp.) an important source of mortality for the Eurasian watermilfoil biocontrol agent Euhrychiopsis lecontei? J Freshwat Ecol. 12:225–234. [Google Scholar]
- SWIMS. 2014. Surface Water Integrated Monitoring System. Madison (WI): Wisconsin Department of Natural Resources. http://dnr.wi.gov/topic/surfacewater/swims/. Accessed 3 September 2014. [Google Scholar]
- Tamayo M, Grue CE, Hamel K. 2004. Densities of milfoil weevil (Euhrychiopsis lecontei) on native and exotic watermilfoils. J Freshwat Ecol. 19:203–211. [Google Scholar]
- Whitlock MC, Schluter D. 2009. The analysis of biological data. Roberts and Company: Greenwood Village (CO). [Google Scholar]
- Zhang C, Boyle KJ. 2010. The effect of an aquatic invasive species (Eurasian watermilfoil) on lakefront property values. Ecol Econ. 70:394–404. [Google Scholar]
Appendix 1
Table A1. Characteristics of lakes surveyed for milfoil weevil.