The controlled tank experiment indicates the high reproductive and damage potential of the milfoil weevil, however, at many of our field sites we have not seen weevil densities approaching 200 per square meter, and at some sites densities have remained below 10 per square meter. Thus milfoil weevil populations are limited. Limiting factors could include predators and parasites, water temperatures (which affects development), habitat, and plant quality and suitability. Limiting factors could affect the population during the open water period when weevils are in the water, or on shore when weevils are overwintering. Research to identify limiting factors is ongoing.
Weevils depend on relatively dry sites close to shore to overwinter. Estimates of overwinter survival suggest that overwinter mortality is similar to other insect populations and even with the harsh winters of Minnesota may not be a significant limiting factor. We have no evidence of parasitism or significant onshore predation. Quantity of overwinter habitat may be important, and is worthy of further investigation. Two unresolved issues are how weevils get to overwinter habitat and back to the water in spring. Weevils can fly, at least during the spring and fall, however, we have captured relatively few in sticky traps and above ground beetle traps. Comparisons of shoreline soil and in lake densities in spring and fall show no consistent relationship, but further analysis is required.
Factors within the lake are clearly important. For example, at one site, spring in-lake densities were high relative to other sites during two springs, but weevil populations declined over the summer rather than increased. Water temperatures at our sites, in combination with temperature-development time data (Mazzei et al. 1999), suggest that 4-5 generations should be produced each summer, with successful development from mid-May until late September. Given the different population responses among our lake sites, differences in plant quality (Sheldon 1997) or host plant resistance (Newman et al. 1998) could be important and is worthy of further investigation; preliminary results by Johnson et al. (1998) suggest that milfoil quality as influenced by sediment may be important (but see Watson and Newman).
Predators and parasites are two potential limiting factors. We have no evidence of parasitism, either in-lake or on-shore. Fish predation could be a significant factor, particularly in low density populations (Sutter and Newman 1997) and perhaps during the spring when weevils are first entering the water with relatively little macrophyte cover. Experimental augmentations in small open plots at three different lake sites failed to result in increases in weevil density. Sutter and Newman (1997) showed that the milfoil weevil is a small component of sunfish diet; pupae, which reside in stems were not consumed, larvae were rarely consumed and adults were most vulnerable. Estimates of consumption rates of milfoil weevils by sunfish, in combination with estimates of weevil and sunfish densities, suggest that sunfish would have little effect on moderate to high weevil density populations, but could be a significant limiting factor at sites with low weevil densities. Fish exclosure experiments to test these hypotheses are under way and graduate student Darren Ward will address this issue.
Lastly, plant density and availability of good meristems for oviposition and stems for pupation are probably important. Sheldon and O'Bryan (1996b) demonstrated that harvesting, which removes the upper stems and meristems of Eurasian watermilfoil, can significantly reduce weevil abundance. Results from our tank experiment (Newman et al. 1996) indicate that high densities over time will eliminate meristems and pupation sites and thus likely limit populations. Densities of weevils after 4 weeks were the same for four stocking densities because the higher stocking rates resulted in elimination of oviposition and pupation sites.