Myriophyllum spicatum (spiked watermilfoil)

There is increasing evidence to suggest that taking action against invasive growth of M. spicatum may actually prolong the period and scale of nuisance caused by the plant, especially if disturbance-based methods of control (which the plant is superbly adapted to resist) are used. The arguments behind this are summarized by Murphy (1988) and more recent studies tend to support the "do-nothing" approach (where it can be tolerated) as effective. M. spicatum invasions seem to follow a standard pattern of rapid increase followed by stabilization, then decline to minor-nuisance status, or even disappearance, in many waterbodies which the plant has invaded across its introduced range (Nicholas and Lathrop, 1994; Smith and Barko, 1996). This appears, in part, at least, to be a result of the gradually increasing impact over time of action of natural enemies (Painter and McCabe, 1988). If nuisance growths of the plant cannot be tolerated then it is probably best to use augmented or introduced biological, or appropriate chemical control methods to suppress the perceived problem.
M. spicatum is an aquatic weed noted for its ability (under favourable growth conditions) to cause nuisance problems within its native range, either alone, with other native weeds, or (less commonly) alongside alien invasive species. France offers numerous recent examples of this phenomenon (e.g Peltre et al., 2002), as does the UK (Newman, 1999).

Highly susceptible to a range of standard submerged-use herbicides, including triazines, e.g. terbutryne, simazine (Murphy, 1982); diquat and dichlobenil (MAFF, 1986); and 2,4-D (MEBC, 1980). Using controlled-release formulations, Hall et al. (1984) found the minimum sustained concentration of fluridone was required to give >50% control of M. spicatum. Fluridone has been occasionally used (despite high costs) against weed populations of M. spicatum, either as whole-lake, or selective low-dose treatments, for example in Michigan and Minnesota lakes (e.g. Heilman et al., 2003; Pedlow et al., 2006; Valley et al., 2006). Getsinger et al. (1994) found that bensulfuron-methyl could give excellent control of M. spicatum, but a 12-week exposure period was required to produce >95% weed suppression. Using triclopyr, good control (>85% biomass reduction) was achieved at varying concentrations over exposure periods of 18-72 hours (Netherland and Getsinger, 1992; Madsen and Getsinger, 2005). For endothall, similar exposure periods were 12-48 hours for over 85% reduction in weed biomass (Netherland et al., 1991), whilst low application rates of endothall combined with 2,4-D have also been shown effective (Skogerboe and Getsinger, 2005). In the past 2,4-D was a widely preferred herbicide in many parts of its introduced range (e.g. Canada) but such programmes have largely been abandoned on grounds of perceived environmental hazard and cost (Dearden, 1985), leaving physical and biological control as the main control options. Nevertheless, 2,4-D-based control programmes (e.g. using spot applications) are still used in the USA (e.g. Bugbee and White, 2005). Getsinger (2002) provides a recent review of the effectiveness of triclopyr, fluridone and endothall as selective controls against M. spicatum in northern US lakes. Glomski et al. (2006) and Gray et al. (2007) provide evidence that carfentrazone-ethyl, alone or in combination with 2,4-D may be an effective herbicidal treatment against M. spicatum.

M. spicatum exhibits a classical example of a disturbance-tolerance strategy, elements of which are found in many submerged weeds (Murphy, 1995). It possesses a combination of physiological, morphological and reproductive attributes which make it highly resistant to control measures based on cutting or other physical disturbance, with regrowth being rapid, reaching pre-treatment abundance within 30 days to 4 months of mechanical clearance in spring or summer (Collett et al., 1981; Mikol, 1985; Filizadeh, 1999). Despite limited long-term success (e.g. Painter, 1988) a wide range of mechanical control measures continues to be employed against the weed, including specialized and expensive (US $100,000 or more) aquatic weed-harvesting systems, rototilling, shallow water cultivation, weed-cutting boats, diver-operated dredging and benthic shade barriers (e.g. McNabb, 1998; Boylen et al., 1996). A recent report of the impact of weed harvesting on M. spicatum (as part of a native-range weed problem, in Lake Geneva, Switzerland) is provided by Demierre and Perfetta (2002). Annual partial drawdown has proved especially effective in reservoirs, exposing the weed in the shallower areas to freezing or drying conditions during the cooler season (Murphy and Pieterse, 1993).

Species of Bagous have been investigated from Pakistan (Habib et al., 1969) but proved disappointing. Among pathogens, most success has been achieved with Mycoleptodiscus terrestris (Verma and Charudattan, 1993), especially in integrated treatment with 2,4-D (Nelson and Shearer, 2005), although Colletotrichum gloeosporioides has also achieved promising results. Integration of mycoherbicide and chemical (endothal) control has shown promise in small-scale trials (Sorsa et al., 1988). Grass carp (Ctenopharyngodon idella) have had mixed results against M. spicatum (e.g. Adamec and Husak, 2002). The plant tends to be low among the feeding preferences of triploid grass carp (Pine and Anderson, 1991), although normal grass carp appear to be less fussy. Julien (1992) records that attempts were made to transfer the stem-boring weevil Phytobius [Litodactylus] leucogaster from California to Florida for control of M. spicatum, but establishment was not confirmed. Other insects, such as Acentria ephemerella and Euhrychiopsis lecontei are among those more recently investigated as possible biocontrols for M. spicatum (Johnson and Blossey, 2002). The natural host of the latter is the US-native northern watermilfoil (Myriophyllum sibiricum), but the weevil has expanded its range to include M. spicatum (Roley & Newman, 2006), as well as the hybrid M. spicatum x sibiricum. Regulatory control In the northwestern USA and western Canada, which have suffered severely from M. spicatum infestation (Anderson, 1993), attempts have been made to quarantine areas against invasion by the weed, using public education programmes, warning notices at boat-launching sites, and checks on boats and trailers to try to minimize the risk of spreading propagules into uninfested lake and river systems. Success has at best been limited. Quarantine controls are also utilised against M. spicatum in New Zealand, in conjunction with a weed risk assessment protocol (Champion and Clayton, 2001). Attempts have been made to regulate sales of this and other invasive species by plant nurseries, again with limited success (e.g. in Florida – Caton, 2005).