Northern Prairie Wildlife Research Center
Early efforts to control sago in irrigation ditches through use of aromatic solvents eliminated top growth and reduced turion numbers and size but did not eliminate turions completely (Otto et al. 1964; Ogg et al. 1969). Commercial granular fertilizers have been recommended for their ability to encourage phytoplankton and thus decrease water transparency so as to prevent pondweed invasion (Walker 1959a). Control of established stands with 2,4-D granules, silvex, or sodium arsenite has also been advised (Davison et al. 1962).
CuSO4 algicide has been recommended at 0.05-0.11 ppm to control sago in still and flowing waters, but these rates are 10-20 times that needed to kill algae, and alkalinity of treated waters must be less than 150 ppm (Gangstad 1986). CuSO4 browns sago leaves, shortens internodes, causes a bushy appearance, necrosis, and slows growth, but plants can survive 100 mg/L Cu concentrations for at least 21 days (Ryan and Riemer 1975).
Yeo (1967) considered the threshold of sago toxicity in reservoirs 125 ppb for diquat and 250 ppb for paraquat. Diquat at 26.7 ppm in flowing water controlled sago during the study of Hesser et al. (1972). Brooker and Edwards (1973) reported that sago replaced Myriophyllum spicatum in deeper waters after being treated with paraquat.
Frank et al. (1961) tested the effects of exposure time of eight herbicides on sago; for short-term exposure (30 min), diquat at 100 ppm gave the best control. Frank et al. (1963) later tested 91 herbicides and experimental compounds on sago in the laboratory and found that only 2 (fenac and dichlobenil), at 2.24 g/m2, gave good control over four successive plantings. Fenac is easily absorbed by both above and below ground parts of sago (Frank and Hodgson 1964).
When compared with controls (Devlin and Yaklich 1971), uptake of the herbicide naptalum was greater in young (30 days old) sago plants that were deficient in N or P. Thus control can sometimes be achieved with less than normal amounts of herbicide in nutrient-poor waters.
Devlin and Karczmarczyk (1975) found that norflurazon reduced chlorophyll content of sago, but growth was sustained by turions for 17 days. Sago is very sensitive to low levels of fluridone (Marquis et al.1981). Sprouting plants can be controlled if sufficient light and contact time of 1-6 days can be attained (Anderson 1981). The Illinois Department of Conservation (1981) lists aquathol-K, aquizine, fenac, and ortho diquat in various forms and concentrations for sago control. Illinois sago was routinely treated with all these chemicals except fenac by 1984 (Tazik and Wiley 1984). Heavy sago infestations in Arizona municipal and industrial water canals have been controlled sucessfully with repeated applications of an endothall derivative at 0.2 ppm (Corbus 1982). Acrolein has given good control of sago and has been used for more than 10 years in Egyptian irrigation canals (Khattab and El-Charably 1986). Lawrence (1962) and Newroth (1974) listed herbicides used to control sago and other plants.
Sago was controlled with 2,4-D butoxethanol ester pellets (29% active ingredient) applied at a rate of 168 kg/ha (Dutta and Gupta 1976). Getsinger et al. (1982) used a helicopter to drop 112 kg/ha of 20% acid equivalent granular 2,4-D on a community of introduced Myriophyllum spicatum, sago, and other native submersed macrophytes. Once the M. spicatum disappeared, greater wave action stirred bottom sediments and increased turbidity, reducing Secchi disk transparency from > 1.5 m to < 0.3 m. Native submersed macrophytes began to disappear during the third post-treatment week, and 1 week later a phytoplankton bloom (which also contributed to turbidity) was observed. Sago was still absent from the area 4 years later, at which time M. spicatum had attained 10% of its former biomass. Westerdahl and Hall (1983) listed the threshold concentration of 2,4-D needed to control sago as 0.1-0.25 mg/L.
Correll and Wu (1982) found that photosynthesis of sago was inhibited by dissolved atrazine at 650 ug/L but stimulated by 75 ug/L. Hartman and Martin (1985) tested three common herbicides (glycophosphate, atrazine, and alachlor) and one insecticide (carbofuran) for their inhibiting effects on the growth and sprouting of sago turions. None of the chemicals affected sprouting; atrazine inhibited growth at all concentrations, glycophosphate stimulated growth at intermediate application levels, carbofuran had no effect, and alachlor was stimulatory at low levels and inhibitive at high levels. Sago was affected by a triazine herbicide at 125 ppb (Fowler 1977).
Control of submersed macrophytes with mechanical harvesters can sometimes result in increased density of sago the following growing season (Neel et al. 1973), likely because of reduced competition. Mechanical harvesting of sago and other macrophytes can also cause the number of morphologically shorter plants to increase (Engel 1984). Successful control of sago by underwater mowing can require three cuts per year in areas with long growing seasons (Weisser and Howard-Williams 1982). Mechanical harvesting in streams is often inefficient because of stream morphology (Madsen 1986). Madsen et al. (1988) found that sago in a stream needed to be mowed every 2 weeks to maintain a basal biomass level of 18% of preharvest levels and every 8 weeks to maintain 44%; harvesting delayed flowering and senescence.
Growth of sago having root zone in the upper 3 cm of substrate can be suppressed for about 6 months in irrigation canals in India by drying canal bottoms in the sun for 5 days (Malhotra 1976).
Many types of plants and animals have been used or suggested for use to control submersed hydrophytes. For sago control, only fish have been extensively studied. I found two suggestions for using vascular plants to control sago. Yeo (1976) indicated that plantings of the short emergent macrophytes Eleocharis acicularis and E. coloradoensis could prevent sago from establishing in shallow waters. Madsen (1986) indicated the possibility of controlling sago in streams by growing shade trees along the banks to reduce light.
The burrowing chrysomelid beetle Haemonia appendiculata (Donaciinae) has been shown to cause great, although temporary, damage to sago in fresh to oligosaline wetlands having permanently flooded water regimes and has been suggested as a promising biological control agent for the plant (Grilles 1988).
The grass carp (Ctenopharyngodon idella) was introduced to United States waters for aquatic plant control in 1963 and much controversy over the ecological consequences of this act has resulted (Leslie et al. 1983). Studies in Europe and Asia cited by Cross (1969), Opuszynski (1973), and Fowler and Robson (1978), and work in the United States (Duthu and Kilgen 1975; Swanson 1986a; Wiley et al.1986; Bowers et al. 1987) showed that sago is at least moderately, or more likely highly, preferred as food by the grass carp or hybirds of the grass carp and common carp or bighead carp (Hypophthalmichthys nobilis). Mehta et al. (1976) found that Indian sago became a highly preferred food of grass carp when the fish attained a weight of 200 g.
Grass carp (mean wt., 217 g) stocked at 116-688 kg/ha shifted from a diet of sago to that of nonpreferred species, but the sago recovered 1 year following fish removal (Fowler and Robson 1978). The feeding activities of grass carp reduced the July dry weight biomass of mixed stands of sago and P. nodosus from 1,400 g/m2 to 24 g/m2 in 4 years, but after about 3 years the grass carp biomass began to decline, and by the fourth year carp biomass decreased to levels attained during the second year after stocking (Mitz ner 1978). These fish ate large amounts of Potamogeton. Subsequent stockings of grass carp in many farm ponds in the central United States showed that this fish eliminated almost all submersed macrophytes and that the desired partial control of these plants is very difficult (L. Mitzner, personal communication). It is likely that grass carp singled out and consumed all the aboveground parts of a highly productive (aboveground biomass 829.6 g/m2) sago-Chara community in a Russian floodplain lake where the fish were common.
Latest research has focused on use of sterile diploid and triploid grass carp for control of sago and other submersed plants, because grass carp overpopulation has had adverse effects on gamefish habitat (Bowers et al. 1987). These fish retain more than 90% of the phosphorus consumed when artificially fed sago and may be an effective means to remove phosphorus from aquatic systems (Chapman et al. 1987). Modeling efforts by Wiley et al. (1984) indicated that triploid grass carp were far more cost effective than chemicals for either suppressing or eradicating a community dominated by sago and naiads (Najas spp.). In some areas of Egypt, harvesting of grass carp used for sago control in irrigation waterways resulted in a net profit (Khattab and El-Charably 1986).
Schwartz et al. (1986) found that sago dry weight was reduced from 4.9 g/m2 to 0.2 g/m2 and 2.1 g/m2 by the addition of (respectively) 500 and 2,500 blue tilapia (Tilapia aurea) per hectare in experimental ponds. However, sago was not the dominant plant in the ponds studied, and these authors postulated that control of vegetation in these ponds was through uprooting, deleafing, and increases in turbidity.
Submersed macrophytes are undesirable in artificial fishponds because they make fishing difficult and tie up nutrients. Fertilizers are often added to these ponds to stimulate phytoplankton, which then compete with and decrease vascular plants, but I could find no reference as to whether this technique is effective against sago.