Northern Prairie Wildlife Research Center
Anderson and Low (1976) found that a highly significant regression of sago standing crop was a quadratic function of water depth in controlled plots in a Manitoba wetland, and the results agreed closely with field observations (Anderson 1978). Robel (1961b, 1962) also showed that peak sago biomass occurred over a narrow range of depths in managed Utah wetlands. Craner (1964) determined that biomass of sago vegetation and drupelet heads--but not turions--showed significant positive correlations with water depth. He found that best vegetative growth and greatest density of drupelet heads was at depths of 30-69 cm and 41-46 cm, respectively. Haag (1983) found that the greatest number of sago seedlings were at the shallowest sites in an Alberta lake, but seedling numbers were not significantly related to depth. He noted that sexual reproduction at such sites is often limited by insufficient light, short growing period, lack of nutrients, and absence of adequate protection from wave action.
Discontinuities in the depth distribution of sago within a wetland can be a result of wave action. In an African lake, sago did not occur at depths < 3.5 m except in sheltered areas, where plants grew at depths < 2.0 m (Bolts et al. 1969). Bimodal depth distribution of sago also can be caused by potential competitors. Denny (1972) found that peak (> =50%) sago frequencies were at 0.75-1.5 m and 3.5-3.75 m in a silty, clay-bottomed African lake, with the depth zone between (1.5-3.5 m) dominated by Chara sp. Two zones of sago were present in a sandy-bottomed African wetland studied by Taylor (1983). A shoreward zone in about 0.6 m water was separated from a lakeward zone in about 1.0 m water by a 40-m-wide zone of sparse Chara. The data of Gibbs (1973) suggest that the often seen cycles of phytoplankton-macrophyte dominance can be a function of water depth, the competitive advantage being with the phytoplankton in deeper wetlands. The depth distribution of sago can also change with season. Purohit (1981) recorded maximum sago density of 840 plants per square meter at a depth of 0.5-1.5 m where plants persisted for 9 months, whereas at 2.5-3.5 m, plants reached a density of only 490/m2 but persisted for 11 months. Voge (1987) found sago at depths up to 5 m in early summer, but by mid-summer no plants were found at > 3 m.
The U.S. Bureau of Reclamation (Garrison Diversion Unit Refuge Monitoring Annual Report 1989, unpublished) has 6 years of measurements on major environmental factors believed to affect sago production in a river in North Dakota and South Dakota. Their data suggest that annual water levels likely have little effect on turbidity, total dissolved solids (TDS), and nutrient loadings in the water column. Rather, poor sago production during years of high water levels seemed to be associated with a combination of depth and turbidity--which decreased photosynthesis--and lower water temperatures. These factors may have delayed sago growth during a critical late-May to mid-June period when plants must reach the surface and spread out horizontally to avoid competition for light from phytoplankton.
Ongoing experiments by Butler and Hanson (1985, 1986, 1988, unpublished) in a shallow freshwater Minnesota lake suggest that increases in depth combined with invasions of bottom-feeding rough fish can initiate a complex series of events that lead to increased turbidity and a permanent reduction of sago and other submersed macrophytes. After these events have occurred, they postulate that (1) greater water depth and increased bottom feeding combine to create an unfavorable light climate for submersed plants; (2) reduced plant biomass exposes more bottom sediments to wave energy and further reduces probability of winterkill; (3) reduced winterkill enlarges the fish community to include planktivorous species capable of overgrazing the zooplankton; (4) reduced zooplankton and increased nutrients caused by lessened macrophyte biomass increases phytoplankton; and (5) increased phytoplankton elevates summer pH and produces greater calcite precipitation which further increases turbidity.
To summarize, the depth distribution of sago is largely controlled by turbidity and wave action, but turbidity can be caused by a multitude of factors, many of which are poorly understood. Depth increases of only 40 cm can greatly reduce sago densities where light is limited by fine suspended silts or phytoplankton, and depth increases of only 10 cm can markedly reduce production where substrates are high in easily suspendible clays.
Turbidity caused by particulate matter and plankton combines with color from physical and biological (Juday and Birge 1933). Water itself and the agents that cause turbidity also selectively absorb wavelengths important for plant photosynthesis (Wetzel et al. 1982).
Light attenuation and wave action usually control the maximum depth of colonization and zonation of submersed macrophytes (Bolts et al. 1969; Spence and Chrystal 1970; Schiemer and Prosser 1976; Spence 1982), although chance occurrence sometimes explains the variety of species found at any given depth (Denny 1973). Light quantity (photosynthetically active radiation or PAR), along with temperature, are the principal factors governing photosynthesis and growth in some submersed macrophytes (Wetzel and Neckles 1986). It was accepted for years that rooted vascular plants would grow to depths where irradiance of bottoms was only 1-4% of that striking the surface (Sculthorpe 1967; Hutchinson 1975). However, recent work by Chambers and Kalff (1985) indicated that angiosperm colonization ceases at depths where, on the average, 21% of PAR incident on the water surface is received during the growing season. However, older established plants may survive under lesser PAR values.
Sago growth can be greatly reduced by light attenuation caused by both organic and largely inorganic materials. Sago often grows in environments where wind action often suspends inorganic materials in the water column for variable periods. Such particles increase turbidity and reduce light needed for photosynthesis, but greatest production, especially of turions, often occurs on substrates composed of easily suspendible materials (Jensen 1940; Low and Bellrose 1944; Craner 1964; Sincock 1965; Smeins 1967). Growth of all submerged plants at such sites is usually limited to relatively shallow (<1.7 m) waters (Chamberlain 1948; Sincock 1965). Decreased water levels can more than overcome the effects of turbidity. Thus Bailey and Titman (1984) recorded a nearly 200% increase in sago biomass following a 17-cm decrease in water levels, even though turbidity slightly increased. Nonetheless, increased turbidity is the likeliest main reason why sago has disappeared from many wetlands (Bellrose et al. 1979).
Agricultural, domestic, and industrial pollution are well known sources of turbidity. Blooms of phytoplankton also are a major factor in the depth distribution of submerged macrophytes (Denniston 1921). Phytoplankton blooms limited sago to depths of sources to decrease transparency of natural waters 0.35-< 1.5 m in the studies of Crum and Bachman (1973), Andersen (1976), and Jupp and Spence (1977a). An often overlooked result of such blooms is the restriction of submersed macrophytes to shallowest nearshore areas where significant reductions in biomass caused by waterfowl grazing and wave action can occur (Jupp and Spence 1977a,b; Peterka and Hanson 1978).
Reduced radiation through screening by epiphytes in eutrophic waters has been blamed for the reduction of sago (Schiemer and Prosser 1976), as has shading by early-sprouting species of higher plants (Gorman 1979). Martin and Uhler (1939) stated that filamentous algae can nearly exclude sunlight and greatly reduce sago production in calm water areas. Recent theory (Phillips et al. 1978) suggests that in nutrient-rich waters, epiphytes and filamentous algae first blanket and shade out the submersed macrophytes; then phytoplankton blooms following decomposition as nutrients are released.
Colorimetric turbidity is negatively associated with sago production (Robel 1961b). It has been claimed that sago will grow to depths where light is 2.5-5.0% of that striking the water surface (Bourn 1932; Jensen 1940; Howard-Williams and Liptrot 1980), but these estimates are probably low (Chambers and Kalff 1985). Bourn (1932) lowered solar energy in greenhouse tanks from 12% to 9.5% of that outside the greenhouse by increasing water depth 0.6 m; this resulted in a 45% decrease in sago biomass. Kulberg (1974) found that in streams sago is absent at turbidities > 153 Jackson Turbidity Units (JTU's). Otto and Enger (1960) showed that 100 ppm suspended sediments causes a 50% reduction in sago production in vessel tests. Sago was absent in parts of a New Zealand lake where suspended solids reached 100-300 mgL, but was present in sheltered areas in water < 1 m deep where 1-7% surface light occurred at 0.6 m (Gerbeaux and Ward 1986).
The data of Westlake (1967) was used by Kemp et al. (1981) to estimate the light compensation point for sago at 60 microE/m2/s; this value, compared with that of common associates such as Myriophyllum spicatum and Ceratophyllum demersum, suggested that sago photosynthesizes at relatively low rates. This is true even under light saturation, and as light intensity decreases, net production of sago ceases at relatively high light levels. Thus, sago could be at a competitive disadvantage in turbid waters. Experiments of Madsen (1986) with Wisconsin plants also showed that photosynthesis in sago is rapidly light-saturated and that its net photosynthetic rate rises slowly with increases in temperature across the range 10-35° C. He achieved maximum net photosynthesis of 1.39 mg C/g ash-free dry weight per hour at 28° C when plants were subjected to 2,000 microE/m²/s of PAR. Under controlled conditions, Hodgson and Otto (1963) observed that sago plants increased in weight but decreased in length at greater light intensities.
Symptoms of moderately reduced light intensity on sago include fewer and coarser leaves and stems and a lighter green color. Severe light reductions result in etiolation, lengthened internodes, stiffened leaves and stems, loss of branching, basal decay, and a tendency of leaves to protrude and wilt at the water surface (Bourn 1932). Anderson (1950) noticed that reduced light intensity increases the reproductive period of sago. Experiments of Van Wijk et al. (1988) showed that sago has adaptations to unfavorable light climates common to eutrophic or brackish waters. These adaptations likely differ among sago ecotypes and include increased relative turion production and increased shoot length. Plants can thus reach the water surface in an earlier stage and concentrate foliage in the surface layer.
Stuckey (1971) considered sago to have a wide tolerance to turbidity compared with other submersed macrophytes. This contention is supported by the surveys of Barker and Larson (1976), who found sago the only submersed hydrophyte in a muddy North Dakota river, but it was limited to calm sections where the sediment load was lessened. Reed (1979) found sago to be the only plant growing in murky water over chalky mud in a North Carolina estuary, and Bue (1956) observed sago to be the only submersed plant in a highly turbid South Dakota livestock pond overpopulated with fish. Reese and Lubinski (1983) found sago and Potamogeton nodosus the only remaining submersed macrophytes in the turbid lower reaches of the llinois River. Davis and Carey (1981) recorded a 157% increase in sago after storm damage and turbidity increases reduced total plant biomass 42% in Currituck Sound, North Carolina. A high tolerance of sago to shade is also indicated because the plant will grow under overhanging trees (Hynes 1970) as well as in the understory of emergent plants (Wilson 1958).
Haslam (1978) believed sago was intermediate in turbidity tolerance when compared with other submersed plants. This may be true under competition in grazed sites because sago grows new shoots from belowground rhizomes, whereas other species can grow new stems epically below clipped points (Cragg et al. 1980; Cohen et al. 1986).
In a polluted Egyptian lake, sago was absent where phytoplankton blooms reduced Secchi transparencies to 11-40 cm, but nearby, sago was abundant in water transparent to 60 cm (Aleem and Samaan 1969a). Similar results were reported by Gibbs (1973), who saw sago begin to recover in abundance in a New Zealand lake when Secchi disk readings exceeded 20 cm as a phytoplankton bloom diminished. Mason and Bryant (1975) found that sago was abundant in wetlands with Secchi transparencies as low as 66 cm, and the species was one of the major dominants in another water body where Secchi transparency was 60 cm (Engel 1984). Algal blooms that reduce light penetration to 10 cm can kill sago, but plants can grow well where Secchi transparency during the latter half of the growing season is 20-30 cm (Peterka and Hanson 1978). In an ongoing study, sago occurred in waters with Secchi disk readings as low as 33 cm (U.S. Bureau of Reclamation, Garrison Diversion Unit Refuge Monitoring Annual Report 1986, unpublished). Sago was absent where turbidity was > 11.8 Nephelometric Turbidity Units (NTU's). Secchi and NTU values were negatively and positively correlated, respectively, with chlorophyll a, suspended inorganic matter, and suspended organic matter. Sago and other submersed vascular plants were absent from other lakes where Secchi disk readings were < 20 cm because of algal blooms, but sago was abundant in a nearby lake where the Secchi transparency was 60 cm (Jenkin 1936). Rich (1966) found relatively low sago production and an almost total lack of turions in silty, carp-infested waters 45.7 cm deep where Secchi depths were < 30 cm.
Sago beds increase water transparency by reducing water movement (Kollman and Wali 1976; Schiemer and Prosser 1976) and by oxygenating the water column; thus, they can contribute greatly to maintaining water quality, especially in wetlands where much of the bottom substrate is anaerobic (Stewart and Davies 1986). Increased transparency also is frequently observed in shallow or protected sites where sago helps anchor the substrate (Jackson and Starrett 1959). Increases in water transparency are often associated with other drastic changes in the environment. Steffeck et al. (1985) saw sago and many other hydrophytes increase in coverage during a dry year, when sediment contributions to the Mississippi River were presumably reduced. Water level fluctuations during the growing season decreased from an average of 3.2 m during the 4 years previous to the study, to 0.3 m during the year of the study, and water clarity increased--from 21-40 cm Secchi during the 2 years previous to the year when sago increased to 80 cm during the year the increase occurred. In a large shallow lake with a nearly flat clay bottom, I observed dense beds of sago and associated filamentous algae to nearly eliminate water turbidity by late July, except in a central area of greatest fetch where water depths were a few centimeters shallower. I attributed the slight reduction in depth of the sago-free central area to the great swelling properties of the fine clays deposited there (Kantrud 1986c, unpublished).
Observations by Whitfield (1986) emphasized the difficulty of separating the effects of changes in water transparency on sago growth from simultaneous changes in water level fluctuations, salinity, algal growth, and supplies of Ca in the water column. The precipitation of calcite by both biogenic and physical means cause lake "whitings" that can greatly increase turbidity (Strong and Eadie 1978; Weidemann et al. 1985). Preliminary work of Butler and Hanson (1985, 1986, 1988, unpublished) indicated that shallow (< 1.5 m) freshwater wetlands having high proportions of Mg, Ca, and HCO3 in the water column, adequate nutrients, long fetches, and rough fish populations are prone to high growing-season turbidities. A similar effect might result from suspended calcite crystals and phytoplankton, which allow sago growth only at very shallow depths. They postulated that, at low conductivities, small, colloidal size calcite particles tend to repel each other and do not flocculate and settle. In turn, such small particles are easily suspended by rough fish and waves, an activity which adds to turbidity caused by phytoplankton. They further proposed that the dense phytoplankton populations are the result of reduced zooplankton grazers caused by an overabundance of planktivorous fish. In their studies, buckets of growing sago suspended 2 m deep in waters of 30-cm Secchi transparency received only 1.3% of surface irradiance and produced 4% the biomass of plants grown in buckets 0.5 m below the surface. Plants at the shallowest bucket depth received 25.7% of the incident light and were the only plants that produced drupelets.
In summary, the decreased transparency of waters inhabited by sago is caused by many common natural and man-made factors such as suspended organic and inorganic particles and phytoplankton. Sago biomass is also commonly reduced by shading or screening effects of filamentous algae or epiphytes. Sago photosynthetic rate is low and photosynthesis ceases at relatively high light levels, but the plant has important adaptations (some possibly genetically influenced) that allow it to succeed in highly turbid waters unfavorable to several common potential competitors. Nevertheless, Secchi transparencies less than 0.2 m usually indicate waters that will not support sago. According to the formula of Chambers and Kalff (1985), maximum depth for any angiosperm colonization under these conditions is 0.47 m. Sago beds can help increase water transparency by reducing water movement, providing a substrate for the growth of filamentous algae, and anchoring the bottom substrate. Recent research suggests that, at least in fresh waters, sago biomass can be limited by reduced water transparency caused by complex interactions among fish, invertebrate and phytoplankton populations, and water chemistry.
St. John (1916) described sago habitat as brackish, alkaline, or sometimes fresh waters. Seddon (1972) considered sago intolerant of conductivity below 200 microS and suspected higher nutrient requirements limited the plant's distribution. Recent work by Pip (1987)--who sampled for 17 Potamogeton species at 430 sites scattered across a large area of central North America--showed that waters inhabited by sago and P. vaginatus were significantly higher in TDS than for 14 other species. Mean TDS concentrations of the waters inhabited by sago, P. vaginatus, and P. filiformis were highest among the Potamogetons. These three are the only members of the linear-leaved subgenus Coleogeton (at least in North America), which suggests that the linear-leaved growth form is an adaptation to increased salinity.
Optimum salinity for sago in thalassic waters is 5-14 g/L (Sincock 1965; Orth et al. 1979; Verhoeven 1980a), and the species is generally replaced by algal or Ruppia-dominated communities at 13-20 g/L in coastal areas (Verhoeven and Van Vierssen 1978a,b; Spence et al. 1979b). Optimum salinity in athalassic Cl- and SO4-dominated waters is 3-6 g/L and 2-15 g/L, respectively (Jensen 1940; Stewart and Kantrud 1972; Millar 1976). In North Dakota wetlands, sago is usually replaced by Ruppia in SO4-dominated waters at salinities more than 26 g/L and in HCO3-dominated waters by many types of other submersed macrophytes at salinities less than 0.7 g/L (Stewart and Kantrud 1972). However, Hammer and Heseltine (1988) found sago and Ruppia coexisting in waters of 53 g/L salinity in SO4-dominated lakes in the Canadian prairies. They suggested that sago does not flower in waters where salinity exceeds 45 g/L.
Measurements of salinity or electrical conductivity are often used to place approximate upper limits on the tolerance of sago for highly mineralized waters. Worldwide, sago grows--or at least its propagules survive--in natural waters with salinities up to 104 g/L TDS (Table 5). At the other extreme, sago has been recorded in natural waters with as little as 35 mg/L TDS (Pip 1987). In addition, sago is readily cultured in distilled water (Bourn and Jenkins 1928; Huebert and Gorham 1983). Upper limits for this species seem to vary greatly in soils or waters dominated by different anions, but this relation remains unproven. Hammer and Heseltine (1988) found no ionic effects on sago in prairie lakes of Canada dominated by various anions and cations. Nevertheless, sago seemingly survives 18-21 g/L TDS in Cl-dominated thalassic areas (Olsen 1950; Mathiesen and Nielsen 1956; Spence et al. 1979b; den Hartog 1981; Van Vierssen and Verhoeven 1983); 50 g/L TDS in inland areas rich in CO3 or HCO3 (McCarraher 1977); and 104 g/L in inland waters extremely high in SO4 (Ungar 1970).
Pip (1984) showed that sago maintains its affinity for waters high in TDS between regions of different geologic origin. Table 1 shows 19 optimum growth depths for sago in mixosaline waters, but only 3 for fresh waters. Downing (1975) opined that sago could not compete with submersed macrophytes in waters with only 280-490 mg/L TDS except in exposed sites. In the more arid regions of interior North America, sago grows in a regime of increasing water salinity during the growing season (Craner 1964; H. A. Kantrud, unpublished data).
Teeter (1963,1965) conducted extensive culture experiments to determine the effects of NaC1 salinity on the growth and reproduction of sago. He found that concentrations of 3 g/L stimulate but more than about 6 g/L curtail turion growth, and that vegetative production decreases and drupelet germination time increases at salinities greater than 3 g/L. Outdoor tank experiments of Lumsden et al. (1963) with thalassic waters indicated that drupelets are not produced at salinities greater than 3.8 g/L and that turion production increases uniformly as salinities increase from 0.6 to 5.4 g/L.
Van Wijk et al. (1988) cultured sago indoors, using plants taken from waters that varied from fresh to brackish (oligohaline to mesohaline) and grew the cultures in water varying from fresh to 9 ppt Cl-(16 g/L). Plants from freshwater habitats produced many double turions while plants from brackish habitats only rarely produced double turions. Plants from freshwater populations produced turions in an earlier stage of growth, compared with plants from brackish habitats. Plants from brackish-water populations usually had more shoots and biomass production when grown at 3 ppt Cl- (5.4 g/L), whereas plants from freshwater populations grew best in fresh water. Shoot length and biomass decreased considerably at 9 ppt Cl- in all populations.
In general, optimum shoot numbers and biomasses occurred at salinities coinciding with those found in the natural habitats for each population. The way in which sago adapts to salinity, and the actual salt tolerating mechanism involved, remains unclear. The experiments of Van Wijk et al. (1988) also indicated genetic differentiation and the existence of ecotypes.
Field studies in NaCl-dominated waters revealed negative correlations between salinities > 1.9 g/L and both vegetative growth and production of drupelet heads (Craner 1964). Sago grown from turions in distilled water does not produce inflorescences (Huebert and Gorham 1983). Results of several of the earlier culture experiments mentioned in this section were summarized by Christiansen and Low (1970).
Sago seems little affected by relatively slow salinity fluctuations within the species' range of tolerance. Thus sago persisted in an estuarine system where annual fluctuations of up to 14 g/L occurred (Howard-Williams 1978), and it survived in an inland, SO4-dominated wetland where salinity increased 81 g/L between growing seasons (Ungar 1970).
A regular cycle of sago growth and decline occurs in Lake Ichkeul, Tunisia, where salinities of 40 g/L occur by the end of summer as the sea flows in. Winter rains then freshen the lake for spring growth of sago and other hydrophytes (Skinner and Smart 1984). There is also evidence that sago can increase in estuaries after major storms lower salinities (Kerwin et al. 1976). Sago often shows great variation in annual abundance in many wetlands in the climatically unstable prairie region of North America where salinities are raised by evaporation and lowered by precipitation (H. A. Kantrud, personal observation).
Iversen (1929) considered sago characteristic of persistently alkaline waters (pH 7.0-9.0). Wiegleb (1978) recorded sago in seven of nine wetlands with maximum pH> 9.0. Mean pH of 116 sites inhabited by sago throughout a large area of central North America was 8.5; significantly higher pH was measured where the species was present than where it was absent (Pip 1987). She also showed that, with respect to pH, sago had a greater number of significant differences with other Potamogetons than all but the acidophilic P. epihydrsus, and that the mean pH of sago-inhabited waters was higher than for waters inhabited by all the other 16 Potamogetons studied except P. friesii.
Although sago will flourish in bottom substrates with pH as low as 4.6 (Denny 1972), the species has an aversion to acidic waters (Jeglum 1971; Merry et al. 1981; Kadono 1982; Pip 1984). Sago was lost from the flora of a reservoir when pH fell below neutrality (Hinneri 1976). Although sago has not been recorded in waters with pH< 6.3, it will photosynthesize at pH> 10.5 and has been recorded in waters up to pH 10.7 (Table 6). Penuelas and Sabater (1987) showed that sago abundance was negatively correlated with pH along the course of a river.
Luxuriant growth of submersed macrophytes can be accompanied by increases in water column pH as plants take up CO2 and HCO3. However, when pH is raised by algal blooms, precipitation of fine particles of calcite in the water column may add to high turbidity and cause poor sago production (Butler and Hanson 1985, 1986, 1988, unpublished).
Sago stands alone among the Potamogetons in tolerance to alkalinity (Hellquist 1975). Of 68 Wisconsin lakes surveyed by Steenis (1932), sago was abundant only in the lake with highest (32.5 mg/L) CaCO3 alkalinity. Moyle (1945) placed sago among a group of species that inhabited Minnesota waters with alkalinity always > 15 mg HCO3/L and that do best where minimum alkalinities are > 30 mg HCO3/L. Spence (1964) considered sago characteristic of lakes with CaCO3 alkalinities > 60 mg/L, and in later surveys Spence et al. (1979a) found no sago in Scottish lochs where total alkalinity was < 17.7 mg/L. Hellquist (1980) placed sago with only two other species (P. hillii and P. vaginatus) in a group of plants characteristic of New England waters with HCO3 alkalinity > 109.8 mg/L. Kadono (1982) and Pip (1987) found significantly higher alkalinity in waters supporting sago than those that did not. Among 17 species of Potamogeton, the mean total CaCO3 alkalinity for sago-inhabited sites (163 mg/L) was exceeded only by sites supporting P. vaginatus (179 mg/L; Pip 1987). Hellquist (1980) found the range of HCO3 alkalinity in waters supporting sago (36.6-282.5 mg/L) to be exceeded only by that of P. nodosus. Penuelas and Sabater (1987) found sago abundance positively correlated with alkalinity along the course of a river. Sago was lost from the flora of a reservoir when HCO3 reserves were replaced by free CO3 (Hinneri 1976). Luxuriant sago growth consumes large amounts of CO2 and can cause rapid decreases in total alkalinity of the water column (Aleem and Samaan 1969a).
Data of McCarraher (1977) show for the CO3/ HCO3-dominated lakes of Nebraska that sago at least survived in Reno Lake (CO3 alkalinity 25.4 g/L, total alkalinity 34.7 g/L) and Moffit Lake (HCO3 alkalinity 20.2 g/L), and plants were collected in other lakes with 9.0 g/L total alkalinity. Dominant cations in these lakes were Na and K, and salinities were high (33-37 g/L TDS). These lakes were low in SO4 and so do not support the hypothesis of Reynolds and Reynolds (1975) that sago and other euryhaline plant species can tolerate much higher alkalinities and salinities in wetlands where SO4 rather than CO3 or HCO3 are the dominant anions. Reynolds and Reynolds (1975) did not find sago in CO3/HCO3 lakes in British Columbia where Na or Mg were the dominant cations and salinities were > 24 g/L TDS.
Competition for the inorganic carbon associated with the alkalinity system may be one of the most important factors determining plant species composition in nutrient-rich fresh waters, especially if phytoplankton are involved (Maberly and Spence 1983). Kollman and Wali (1976) measured CO3 and HCO3 in a lake where sago was the monodominant submersed macrophyte and found that HCO3 levels may limit the productivity of sago in alkaline waters where the plants must use this ion as a carbon source (see Physiology). Pip (1984) showed that sago inhabited waters averaged higher in alkalinity in a region where levels were generally lower than in an area where levels averaged higher, postulating the reason to be more intense competition from the large number of specialist taxa that inhabit soft waters.
Culture experiments of Huebert and Gorham (1983) showed that sago cannot mobilize enough inorganic C from the bottom sediments and that plants need a minimum of 30.5 mg/L HCO3 in the water phase to allow survival and normal growth. Sago can be found in lakes where HCO3 cannot be detected during the growing season, but in these lakes CO3 is present at > =68 mg/L (McCarraher 1977). The U.S. Bureau of Reclamation (Garrison Diversion Unit Refuge Monitoring Annual Report, 1986, unpublished) found sago presence significantly related to greater and lesser amounts of CO3 and HCO3, respectively, which indicates uptake of HCO3. Uptake of HCO3 was also indicated in the studies of Purohit et al. (1986), who showed that harvest of sago plants increased concentrations of both free CO2 and HCO3 during months when peak sago biomass normally would occur.
Wiegleb (1978) differentiated the normal form of sago (var. vulgaris) from a very narrow-leaved form (var. scoparius) which grew in HCO3-poor waters.
The relation of turbidity to alkalinity is unclear. Butler and Hanson (1985, 1986, 1988, unpublished) suggested that the negative correlation they found between total alkalinity and turbidity in a freshwater lake dominated by HCO3 could indicate that turbidity is caused by colloidal particulates that act as binding sites for dissolved ions in the water column. They further postulated that, at the lake's low electrical conductivities, these particles repel one another and remain in suspension, rather than flocculating to a size that promotes settling. Even if extended calm weather allowed some of these particulates to settle, minimal sediment disturbance would quickly resuspend them, again raising turbidity. Perhaps the opposite situation occurs in the highly alkaline, mixosaline waters favored by sago in prairie wetlands. There, calcite particles aggregate and fall to the bottom as spring water temperatures increase, clearing the water and allowing rapid growth of sago and benthic algae, which in turn help protect the bottom from disturbances by wave action.
Although the distribution and production of sago likely is seldom limited by water column nutrients (Peltier and Welch 1969), growth of the plant has often been associated with polluted, oxygen-poor wetlands high in nutrient ions. Of 10 dense stands of sago investigated by Prejs (1986b), 5 were affected by municipal wastes, whereas none of 12 sparse stands were. Efforts have been made to control sago in eutrophic streams where the plant can cause unacceptably low nighttime O2 levels (Madsen et al. 1988).
Data in Appendix B suggest that sewage and agricultural effluents are more frequently associated with eutrophication of waters supporting sago than are domestic animal wastes and industrial effluents, but the two former sources of pollution probably are more common. Eutrophic, but not hypereutrophic, waters supported higher sago biomass in the lake surveys of Ozimek et al. (1986), who also found greater biomass and more rapid growth of the plant in polluted versus unpolluted sites in the same lake.
Growth of sago in nutrient-rich waters is frequently noted in the lower reaches of rivers and streams, where pollution loads are usually greatest. However, extremely high nutrient loadings can destroy or injure the plant and result in its replacement by algae (Pieczynska and Ozimek 1976; Howard-Williams 1981). In a group of polluted English wetlands that had suffered the loss of many macrophyte species, sago showed best survival in those with lowest input of agricultural fertilizer (Mason and Bryant 1975). As lake eutrophication proceeds, sago biomass decreases in shallower zones, and plants disappear from deeper zones where light intensities are lowest (Bumby 1977). In a turbid, highly eutrophic, carp-infested reservoir on the Minnesota-South Dakota border, small, unhealthy-appearing sago plants existed only at depths <8 cm (Kantrud 1984, unpublished).
Vigorous sago growth is often associated with decreased total N, NO2, NO3, or NH3 levels in the water or bottom sediments during the growing season (Jensen 1940; Aleem and Samaan 1969a; Ho 1979; Shubert 1982). This indicates plant uptake. However, sometimes no relation is evident (Wong and Clark 1976; Purohit and Singh 1981). For example, in a polluted Wisconsin stream, Madsen (1986) found that the luxuriant sago community held only 5.2% of the daily loading of N into the system. Purohit (1981) found that total N in the water column positively correlated with sago biomass, but that NO3-N negatively correlated; N in plant tissues seemed independent of total N in the water column.
Data in Table 6 show the wide ranges in various forms of N found in sago-inhabited waters. Concentrations of NO3-N in the water column of a sago dominated South African lake decreased 13% (from 1.69 mg/L to 1.47 mg/L), even though sago biomass increased 20 times during the 2 years (Vermaak et al. 1983). Concentrations of N in sago tissue can be 1.8 x 103-5.4 x 103 times that of the water column (Gopal and Kulshreshtha 1980). Even so, it is unlikely that sago growth is limited by inadequate N in the water column in most natural waters because of the plant's ability to mobilize this nutrient from bottom sediments (Peltier and Welch 1969; Huebert and Gorham 1983). The relation between N in sago and in the water column often is obscured by high N consumption by algae (Howard-Williams 1981) and inputs from plant decomposition and surface runoff (Hutchinson 1957; Paullin 1973).
Pip's (1987) survey of 430 Potamogeton-inhabited sites showed that waters supporting sago and the closely related P. vaginatus were significantly higher in combined NO2 and NO3-N than waters where the two species were not found. Mean N concentrations for the two species were 1.55 and 1.48 mg/L, respectively-the highest among the 17 species of Potamogeton studied. Earlier work (Pip 1984) showed that sago-inhabited waters averaged higher in these ions in a geographic region where levels were generally lower than in a region where levels averaged higher.
Several other investigations of lakes, streams, and reservoirs (Kaul and Zutshi 1967; Kaul 1977; Janauer 1981; Peverly 1985; U.S. Bureau of Reclamation 1986, unpublished; Penuelas and Sabater 1987) have related the presence or abundance of sago to higher levels of various N ions. Other studies (Fetter et al. 1978; Wiegleb 1978; Kohler and Zeltner 1981, cited in Haslam 1987) indicated that certain levels of these ions can be toxic to sago, particularly the NH3 ion at levels > 0.4 mg/L. Purohit et al. (1986) suggested that forms of N other than NO3 can be used by sago, and Wiegleb (1978) considered sago indifferent to NO3 concentrations.
I conclude that, regardless of the forms of N most used or their sources in the environment, sago seems strongly associated with higher levels of this element. However, the plant has a great ability to absorb N and compete for it in aquatic ecosystems where concentrations of this important nutrient are low. Thus, the distribution and abundance of sago probably are seldom limited by the availability of N.
Phosphorus is widely regarded as the most important nutrient limiting primary productivity. Sago production is associated with high or elevated amounts of P in the water column (Zaky 1960; Jones and Cullimore 1973; Jupp and Spence 1977b; Anderson 1978; Pip 1978; Janauer 1981; Collins et al. 1987; Penuelas and Sabater 1987). Low or decreasing P concentrations during peak periods of sago growth have been recorded in waters of lakes and rivers (Jensen 1940; Aleem and Samaan 1969a; Paullin 1973; Kollman and Wali 1976; Ho 1979; Purohit 1981; Purohit et al. 1986). Changes in soluble reactive P (SRP) values in the water column can be difficult to detect or largely caused by nonbiological activity where values are low (< 5 microg/L), even in dense sago stands (Howard-Williams and Allanson 1981).
Sago has been found in lakes with up to 646 mg/L P (Deevey 1957, cited in Cole 1963) and in waters where SRP could not be detected (Pip 1979; Table 6). Low levels likely indicate uptake, as growing sago, with its epiphytic algae, is a net accumulator of P from the water column (Howard-Williams and Allanson 1981; Vermaak et al. 1983) and does not leak the element as some common submersed plants are thought to do (Madsen 1986).
Sago uses roots and shoots to obtain P (Welsh and Denny 1979). Sediment P is usually not as readily available to sago as sediment N (Peltier and Welch 1969; Jones and Cullimore 1973; Huebert and Gorham 1983). However, sediments are the major source of P for sago, under aerobic conditions and where water-column SRP is low and is rapidly taken up by other organisms, including periphyton on sago leaf surfaces (Howard-Williams and Allanson 1981). Thus, under such circumstances, direct transfer of phosphates from sediments to the water column by sago is unimportant, leading Vermaak et al. (1983) to consider the plant a P reservoir instead of a pump.
Sago shoots can contain as little as 5 times or as much as 7,000 times the P found in the water column (Gopal and Kulshreshtha 1980; Vermaak et al. 1983). However, P contained in leaf periphyton can easily confuse such measurements, and it is possible that this community may even assist in P absorption (Howard-Williams and Allanson 1981). Wong and Clark (1976) and Ho (1979) recorded good correlations between water column P and P concentrations in sago tissue. Pip (1987) found that sago differed from other species of Potamogeton with respect to P concentrations in the water column; she also found P concentrations in 116 sago-inhabited sites in North America (x = 3.83 mg/L, range 0-31.3 mg/L) significantly higher than at sites where sago was not present. Despite these results, water column P is not a significant factor in sago distribution over large geographic areas of North America (Hellquist 1975; Pip 1984).
The effects of P uptake by sago on the aquatic environment depend on the abundance and availability of the element. Purohit and Singh (1981) found no detectable differences in P between water sites where sago growth was dense or sparse. Madsen (1986) determined that, even at the peak of the growing season, a luxuriant sago-dominated community in a polluted Wisconsin stream held only 1.9% of the daily loading of P into the stream. Conversely, Purohit et al. (1986) estimated that, at the period of peak macrophyte biomass, a community of hydrophytes where sago was abundant contained 96% of the total P in the littoral system, excluding bottom substrate and other biota.
Jones and Cullimore (1973) and Getsinger et al. (1982) indicated that sago lacks the ability to compete with common associates such as Ceratophyllum demersum and Potamogeton richardsonii in aquatic systems low (perhaps < 1.0 mg/L) in water column P. The only exception to this trend was in a lake dominated by Myriophyllum spicatum, where P was high in both water column and sediments.
Where sago is absent and P is abundant in the water column, other factors, especially turbidity, probably limit sago growth (Fetter et al. 1978), although in some instances lush sago growth can control P availability and suppress light-limiting phytoplankton blooms (Vermaak et al. 1981). When P was added to a community of sago and filamentous algae, the algae quickly absorbed most of the P, and increases in P content of sago were noticed only at high levels of enrichment (Howard-Williams 1981). Vermaak et al. (1983) postulated that P can be absent from the water column of a sago dominated lake high in CaCO3 through precipitation of apatite (Ca5FP3O12).
Studies of Jupp and Spence (1977a,b) and Van Vierssen and Verhoeven (1983) showed that phytoplankton blooms, sufficient to lower sago productivity, have occurred in wetlands in temperate climates where P concentrations ranged from 0.05-1.5 mg/L. Sago and other submersed macrophytes can be restricted to shallow (< 1m) waters under such conditions (Jupp and Spence 1977a,b). However, phytoplankton productivity was very low in a South African lake perennially dominated by sago where water column P was < 0.2 mg/L (Vermaak et al. 1983). In this lake, much P was concentrated in the sediments, and benthic algae carpeted the bottom during the winter period of low sago biomass (Vermaak et al. 1983). Thus, the interactions of P, sago, and phytoplankton blooms differ with climate or perhaps other correlated factors.
I conclude that large amounts of water column P can be tolerated by sago and taken up by roots and shoots, but that the plant is poor at extracting the element from sediments. Sago is less competitive with other angiosperms in ecosystems low in P. The plant shows a greater affinity for waters high in P than other members of the genus but, in such waters, often suffers from turbidity caused by phytoplankton, especially at deeper sites. This could greatly restrict sago production in deeper waters of temperate climates where the plant is forced to regenerate from turions each spring. The effects of P may be associated with other aspects of water chemistry, and the element likely is easily lost from the water column.
Potassium is essential for sago growth (Devlin et al. 1972), but few studies of the effects of water column K on sago and its environment have been conducted. Sago has been recorded as most abundant in lakes and streams with higher amounts of this element in the water column or interstitial sediment water (Kaul and Zutshi 1967; Kaul 1977; Peverly 1985), but K does not seem to be limiting in natural waters. Madsen (1986) found that a luxuriant sago-dominated community held only 4.2% of the daily loading of K into a polluted Wisconsin stream, even during maximum plant biomass. Sago has been found in waters with an extreme range in K concentration (Table 6).
Culture experiments of Huebert and Gorham (1983) showed that when K is absent from the water column, Na replaces it, although plant vigor suffers. Kollman and Wali (1976) found increases in K in both the water column and sediments during the growing season, which they thought were caused by evaporation and equilibrium shifts of the element from water to sediment. At the same time, K concentrations in sago rose rapidly through the growth period, declined during the reproductive period, and then increased at senescence. K levels in the water column decreased only slightly as the amount in sago rose rather abruptly during the peak growth period (Ho 1979).
Sago can tolerate at least 10 g/L of either Cl or SO4, and also high levels of Na, Mg, and many other elements (Table 3). In some waters where Cl is the dominant cation, sago does best where Cl content is 2.6-5.0 g/L; below and above this level Myriophyllum spicatum and Ruppia spp., respectively, were the most abundant plants (Verhoeven 1975, 1980a). In other such waters, sago prospered where Cl was about 3-9 g/L and occurred as the only submersed macrophyte where summer concentrations were as high as 11.6 g/L (Van Vierssen and Verhoeven 1983). Here, Ruppia spp. were remarkably absent, even at higher Cl concentrations, and Zannichellia palustris and Ranunculus baudotii occurred where Cl levels were below 5 g/L.
Sago was the only one of 17 species of Potamogeton studied by Pip (1987) that showed significantly higher mean Cl levels (x = 70 mg/L, range 0-1,234 mg/L) and mean SO4 concentrations (x = 76 mg/L, range 0-3,403 mg/L) in waters where sago was present. In addition, it was the only species significantly different from the others with respect to Cl or SO4 concentrations. Kadono (1982) and the U.S. Bureau of Reclamation (Garrison Diversion Unit Refuge Monitoring Annual Report 1986, unpublished) also noted that Cl levels were higher in waters supporting sago. Penuelas and Sabater (1987) found that sago abundance was highly correlated with SO4 concentrations along the course of a river.
Sincock (1965) thought that sago growth and turion production were enhanced in estuarine waters of 7-14 g/L salinity in which Ca, Mg, and K were found in approximately equal amounts. Lohammar (1938) and Hutchinson (1975) stated that sago has a high (> =25 mg/L) requirement for water phase Ca. Kaul and Zutshi (1967), Kaul (1977), and Kadono (1982) also noted that larger amounts of Ca were associated with larger amounts of sago, although Purohit and Singh (1981) found no detectable differences in Ca concentration between sites in waters where sago growth was luxuriant or sparse. At least 2 ppm water phase Ca is required to partly or fully counteract the toxic effects of other ions present in the water phase (Huebert and Gorham 1983).
In rivers, sago can be restricted to lower reaches where waters are highest in Ca (Merry et al. 1981). In one instance, the inflow of Ca-poor river water into a lake where the plant occurred reduced Ca concentrations to limits lethal to sago (Allanson and Howard-Williams 1984; Whitfield 1986).
Sago, likely through a symbiont bacterium (Oborn 1964), precipitates calcite (CaCO3) on leaves, stems, and in lacunae; this mineral encrustation can compose nearly 12% of plant dry weight (Oborn 1964). Thus Ca concentrations in the water column decrease during periods of rapid sago growth (Ho 1979), or cause the presence of sago to correlate with lower Ca values in the water column (U.S. Bureau of Reclamation Garrison Diversion Unit Refuge Monitoring Annual Report 1986, unpublished). However, this phenomenon may not be observed in wetlands subject to large losses of water through evaporation (Kollman and Wali 1976). McCarraher (1977) listed sago for several lakes where Ca was not detected in the water. Gopal and Kulshreshtha (1980) found that sago tissue contained 438-1,860 times more Ca than the water column. Purohit et al. (1986) also found that a large amount (> 44%) of the Ca present in the littoral system (excluding bottom substrate and other biota) can be contained in a macrophyte community dominated by sago. Butler and Hanson (1985, 1986, 1988 unpublished) theorized that, in fresh waters, calcite formation caused by phytoplankton can increase turbidity and cause poor sago growth.
Marl is a complex mixture of clay particles, calcite, trace elements, materials of biological origin, and probably Mg and Fe (Kelly and Ehlmann 1980). It is commonly found on sago (Kollman and Wali 1976) and other Potamogetons (Wetzel 1960), often in lesser amounts in shallow than in deeper water. Oborn (1964) found much Na as well as barium, Co, Pe, K, Mg, Mn, and Sr in the mineral encrustations on sago.
On a large shallow lake in eastern North Dakota, sago in 3-30 cm of water was relatively free of calcareous encrustations; plants in 30-105 cm had thin gray coatings of a flaky texture (Kantrud 1986c, unpublished). However, at the edge of the centrally located area of greatest fetch, where waters were 95-100 cm deep, sago plants in thin stands were covered with a thick reddish coating of a rubbery texture, presumably a mixture of marl and algae. This coating weighted the plants down toward the euphotic zone, which was very shallow because of suspended or colloidal clay particles. It is likely, though, that sago mortality in this central zone was also related to plant age, as the first green sago leaves of the season were seen in this zone, and marl accumulations are directly proportional to plant age (Purohit et al. 1986).
Culture experiments of Huebert and Gorham (1983) showed that sago cannot mobilize enough Mg from the sediments to grow normally. They listed 10 mg/L in the water phase as necessary to meet the Mg requirement. Magnesium can be present in sago tissue at concentrations of 500 times that of the water column and 5,000 times that of the sediments (Gopal and Kulshreshtha 1980). Ho (1979) recorded that the Mg content in sago increased as levels of this element in the water column decreased. Nevertheless, sago occurred in natural waters where Mg was found in very small amounts or could not be detected (McCarraher 1977; Fraser and Morton 1983).
Dissolved O2 remained high year-round in a polluted lake dominated by sago, except in the most polluted portion where sago did not grow due to high levels of phytoplankton and increased turbidity (Aleem and Samaan 1969a). There, O2 was completely depleted during summer. Both O2 consumption and dissolved organic matter (DOM) were lowest year-round where sago was most abundant. Pip (1987) also showed that DOM in waters supporting sago (mean absorbance at 275 nm = 0.29) was significantly lower than in waters free of sago.
In summary, sago tolerates extremes of TDS in the water column. Optimum biomass occurs at 2-15 g/L TDS, but propagule production is greatest at the lower end of this range. The plant may have evolved ecotypes adapted to waters of differing salinity. Sago is characteristic of alkaline waters (pH 7.0-9.0), and has not been recorded in natural waters with pH < 6.3 or > 10.7. Sago has difficulty mobilizing inorganic C from bottom sediments, and thus favors waters high in CO3 or HCO3 ion, with a minimum of 18 mg/L. The plant is often found, and may be monodominant, in oxygen-poor, eutrophic waters, but low nutrient levels alone usually do not limit growth. Sago may have critical requirements for water-column Ca and Mg. The plant tolerates high concentrations of other elements (Table 6).
Temperature can be an important variable in growth and niche differentiation among hydrophytes (Auclair et al. 1973; Haller et al. 1974; Wetzel and Neckles 1986). In some years, warmer than normal summer temperatures increase seed germination among annual species, thus accounting for the annual differences in species composition in wetlands (Lundegardh-Ericson 1972). Moderate temperature increases are believed to enhance microbial activity, which leads to increased mobilization of nutrients (Niemi 1975, cited in Keskitalo and Ilus 1987).
Sago grew in deep lakes where summer water temperatures were < 12° C (Jimbo et al. 1955), yet produced extremely high biomass in shallow lakes where temperatures reached 31° C (Flowers 1934; Craner 1964; Aleem and Samaan 1969b; Olson 1979). Sago also inhabited a river whose temperature reached 32° C (Carter et al. 1985b). Stuckey (1971) noted that the warm waters tolerated by sago were poorly oxygenated.
Sago is one of the first submersed plants to sprout in spring at temperate latitudes (Moore 1915). Hodgson and Otto (1963) considered 5°C as the lower threshold of sago growth, and Van Wijk (1983) found that 52% of sago turions had sprouted when lake water temperatures reached 5.5° C but concluded, through indoor experiments, that 25° C was optimum for turion germination. Hodgson (1966) showed that at this temperature the carbohydrate reserves in turions were exhausted 16-23 days after germination. Hammer and Heseltine (1988) noted that sago drupelets sprouted and flowering began in Canadian sago when water temperatures reached 8° C and 15° C, respectively. Sago and other submersed plants died in an experimental pond when temperatures reached 38° C (Bourn 1932). Optimum temperature for growth of young sago plants in the laboratory was 23-30° C; growth occured at 10° C, but no leaves were produced; little or no growth oc cured at 37° C (Spencer 1986a).
Haag (1983) collected drupelets in sediment samples from a large Canadian lake and found that more sago seedlings emerged from sediment in the warmer waters near a power plant than from samples taken in the cooler or deeper waters. Restratification of the drupelets resulted in further emergence of seedlings from the warmwater samples, which suggested that thermal discharge interferes with stratification. The warmer temperatures seemed to have a greater effect on seedling density than did sediment type and depth at the collection sites, probably because of increases in length of the growing season and enhanced seed production. In the lake, sago reproduction was mostly from turions, and seedling survival was minimal. Sago increased in areas affected by thermal pollution near a nuclear power plant on the Finnish coast but, except at shallowest sites, was mostly replaced by Myriophyllum spicatum and Cladophora glomerata in areas of peak increases in temperature and current velocity (Keskitalo and Ilus 1987). Sago also decreased at another site where a combination of lack of winter ice and increased nutrients allowed several species of algae to prosper.
The seasonal phenology of sago followed mean weekly air and water temperatures more closely than mean sunlight energy, and drupelets matured when air and water temperatures were maximum (Yeo 1965). Spencer's (1986a) controlled experiments suggested that, as long as stored carbohydrates are available from the turions, young sago plants are less influenced by light. This supports the hypothesis that the effects of temperature and irradiance change with sago's growth stage. His study showed that temperature, more than irradiance, affected relative rates of shoot elongation, leaf production, and the production of important pigments. These results indicated that early growth of sago from turions was better suited to cooler waters where carotenoids better protected chlorophyll.
Results of ongoing surveys by the U.S. Bureau of Reclamation (Garrison Diversion Unit Refuge Monitoring Annual Report 1989, unpublished) along the James River, North Dakota and South Dakota, suggest that, in deeper waters, water temperatures during a critical spring period of turion germination and growth can influence peak summer biomass and allocation of resources to propagules. These studies indicate that low water temperatures associated with high water levels during late May to mid-June cause lowered sago biomass but result in the production of heavier turions. Conversely, good turion germination and growth, rapid horizontal deployment of leaves near the water surface, high peak biomass, and increased drupelet production resulted when spring water temperatures were high and water levels were low.
In polluted Wisconsin streams there usually were spring and fall biomass peaks of filamentous algae and a net carbon gain for sago only when water temperatures exceeded 15° C; longer day lengths also likely were required (Madsen 1986). Wong et al.(1978) also found that when water temperatures reached 18° C, an algae-dominated community began to give way to a sago-dominated community and that, at 23° C, there were few other angiosperms present.
Kiorboe (1980) observed suppressed sago growth in shallow thalassic waters where temperatures approached 25° C and epiphytes covered the plants. Sago generally replaced M. spicatum in a Canadian lake in the area with thermal effluent discharge (Dobson 1964, cited by Haag and Gorham 1977). Nevertheless, Madsen (1986) found that sago had maximum net photosynthesis at 25-28° C (Topt) in a Wisconsin stream where the plant was an earlyseason dominant. This temperature range was lower than for late-season dominants such as Ceratophyllum demersum and M. spicatum. He suggested that the role of environment or selection in the control of Topt in sago could warrant research because the plant has adapted to a wide range of temperature conditions worldwide.
Sago plants influence the temperature of their own environment. In summer, sago beds shaded bottom sediments and prevented water mixing, which made surface waters warmer in a North Dakota wetland < 1 m deep; later in the season, cool surface waters were temporarily stabilized by the sago beds until late fall when surface waters fell through the decomposing plants and complete mixing occurred (Kollman and Wali 1976). Sago beds likely have less effect on water temperature in deeper sites where plants are ineffective in stabilizing bottom soils (Paullin 1973), although it is probable that a thick surface canopy of sago leaves would affect a temperature profile at almost any depth conducive to sago growth (M. G. Anderson, personal communication).
In short, sago's cosmopolitan distribution indicates that the plant has wide temperature tolerance. Temperature has a great regulatory effect on sago growth, phenology, and resource allocation to propagules. Plants have adaptations that allow reproduction under a wide range of temperature extremes and fluctuations. Sago also influences the temperature of the water column, especially in shallow, protected sites.
Wave action greatly reduces the number of species of submersed macrophytes in lacustrine systems (Schiemer and Prosser 1976). Wave action on some large lakes restricts sago and other macrophytes to river inlets and protected bays (Sheffer and Robinson 1939; Jaworski et al.1979). It is likely that wave action reduces productivity of Potamogetons by direct damage and by the transport of fine nutrient rich sediments to more sheltered areas (Jupp and Spence 1977b).
Schiemer and Prosser (1976) postulated that a long-term increase in water level from 1.25 m to 1.65 m created sufficient additional wave action that nearly destroyed a formerly wide zone of sago. Nevertheless, sago increased in more sheltered sites by replacing species less prone to mortality from the increased silt accumulations that occurred. Kautsky (1987) showed that sago in sheltered mud bottoms produced about 4 times the standing crop of plants found on exposed sandy and gravelly bottoms. She found that turion production was lower and drupelet production was higher in the sheltered location and postulated that the exposed sites made pollination difficult and caused high mortality of adult plants. An alternative hypothesis is that plants in exposed sites survived by allocating more resources to asexual reproduction (R. G. Anderson, personal communication).
Although sago growth can be limited by severe wave action in shallow waters (Tryon 1954; Boltt et al. 1969; Anderson and Jones 1976; Anderson 1978; Jupp and Spence 1977b), the species often grows well in-or is sometimes even restricted to- deeper portions of wetlands that have a great amount of wave action (Kornas et al. 1960; Downing 1975; Verhoeven 1980a; Clayton and Bagyaraj 1984; Ozimek and Kowalczewski 1984). Wave action in shallow waters tends to erode sediments where plants are rooted, whereas in deeper waters, waves tend to whip and pull the vegetation. Seedlings have difficulty becoming established under such conditions. Turbulence reduces the number of plants found near the center of large bodies of open water but allows considerable growth of individual plants that become established. Spence (1982) stated that it is the depth of the wave-mixed zone that is of great importance to macrophyte distribution in wetlands. The wave-mixed zone is where erosion, sorting, and some particle deposition occur, and the depth and breadth of the zone depend on slope and aspect of the shore and on lake size, principally fetch.
In a brackish coastal lake in New Zealand 22 years after a storm, sago was just beginning to recover; the likeliest causes for failure of plants to regenerate were wave action that removed sediments from the then unprotected lake bottom and increased turbidity due to phytoplankton blooms (Gerbeaux and Ward 1986). However, sago can prosper after storm damage and increased turbidity have reduced the biomass of other species, but such response is difficult to predict in different wetlands (Ogelsby et al. 1976; Stevenson and Confer 1978; Davis and Carey 1981). Extensive stands of sago can sufficiently stagnate the water column as to reduce phytoplankton production (Aleem and Samaan 1969b).
Dense growths of sago can greatly reduce wave action from winds < =25 km/in (Putschog 1973; Howard-Williams and Liptrot 1980). But Zaky (1960) showed that winds > 29.6 km/in effectively uproot sago and determine its distribution and density in a large (> 500 ha) reservoir. In this wetland, sago densities in a protected area were more than 4 times greater than densities in an exposed area. The wind tolerance of sago cannot be stated from these isolated examples; however, Spence (1982) showed that peak wave height at a given wind speed seems to vary with the square root of fetch, and how fetch varies with the width of the water body and irregularities of shoreline as well as with length. He cited an example where plant cover could be as much as 13 times greater on a sheltered shore than on an exposed shore.
Sago has been observed in the plant understory at sheltered sites and in nearly pure stands in more exposed sites (Wilson 1958). Haag (1983) suggested that wave action likely was one of the main restrictions on sago seedling survival in deeper (> 2 m) waters in a large Canadian lake where the plant grew mostly as a perennial. In deep (3 m), clear New York waters subject to vigorous wave action, sago plants nearly 3 m long grew from turions borne on tough perennial rootstalks (Moore 1915).
Sago withstands moderate to fast currents but is intermediate in tolerance to turbulence (Butcher 1933; Haslam 1978). Although rated low in anchoring strength compared with other submersed macrophytes, sago is difficult to erode once established (Haslam et al. 1975; Haslam 1978). However, higher current velocities accompanied by increases in depth and turbidity, reduces sago abundance without greatly affecting potential competitors such as Vallisneria americana or Heteranthera dubia (McConville et al. 1986).
Current speeds > 1 m/s were required to limit growth of sago and other macrophytes (Howard-Williams and Liptrot 1980), but sago displaced Elodea and Myriophyllum at two power plant discharge sites where swift currents were created (Allen and Gorham 1973). Madsen (1986) found that sago added biomass in a Wisconsin stream even in early April when current velocities reached 2 m/s. Sago prospered and retarded waterflow in canals 10-15 m wide, having flows as high as 13,450 L/s (Corbus 1982).
The common occurrence of sago in flowing waters is attributed to several factors. Van der Voo and Westhoff (1961) found dense sago stands in only 9% of 46 stagnant water areas they investigated, but flourishing stands were present in 24% and 31%, respectively, of 66 river and 13 tidal areas where significant water movement occurred. Sago may survive in flowing water because the plant's narrow leaves enable it to resist silt accumulations (McCombie and Wile 1971). Because of sago's low photosynthetic rate, Westlake (1967) found it difficult to measure increases in photosynthesis of the plant caused by increased water velocity, but found that the rate increased 1.5 times when current velocities increased from 0.2 mm/s to > =0.4 mm/s. Krausch (1976) postulated that higher sago densities in rivers could be caused by increased contact with nutrients and where there is a lack of annual ice scour.
The degree of water level stability may account for as much as 30% of the annual change in sago abundance (Bellrose et al. 1979). Sago's tolerance to fluctuations seems partly dependent on water clarity and the rate at which the fluctuations occur. Sago withstands water level fluctuations of at least 0.5 to 1.75 m in thalassic waters (Doherty and La Roi 1973; Kulberg 1974; Getsinger et al. 1982; Carter et al.1985a), and plants can endure periodic exposure under tidal conditions (Kiorboe 1980). The species persisted under fluctuations of at least 1.2 m in clear athalassic waters (Harrison 1962). Sago likely prospered for at least 10 years in a North Dakota lake where water levels fluctuated about 35 cm above and below an average depth of about 70 cm (Kantrud 1986c, unpublished); the plant persisted for decades in a river which fluctuated about 1.0 m (Hunt 1963). Lutz (1960) found no turion production where water was turbid and fluctuated outside a diked area from which carp were removed. Inside a diked area where water levels were stable--and from which carp had also been removed--drupelet production was 280 times greater by weight and 240 times greater by number.
Robel (1962) noted that water-level increases in turbid waters caused sago mortality through light extinction. A water level increase of less than 10 cm in turbid mixosaline waters resulted in 35% less sago production than the previous year. Similar reductions occurred with 0.4-1.0 m increases in water levels in fresher, less turbid waters (Anderson and Jones 1976; Carpenter 1980). Higher water levels in nutrient-rich lakes favor phytoplankton blooms over sago (Gibbs 1973). Higher water levels sometimes allow ingress of rough fish that caused uprooting and turbidity which lowered sago production (Anderson and Jones 1976). Otherwise, sago production may be increased in sheltered coves due to higher water levels when filamentous algae become less abundant than in shallower water (Anderson and Jones 1976). Sago growth can also increase greatly when higher water levels drown out or otherwise reduce densities of emergent hydrophytes (McDonald 1951).
Purohit (1981) related temporal trends in sago density to changes in water depth in an Indian lake. Sudden increases in water level reduced sago density more in areas where substantial parts of the plants floated than in areas where plants were not tall enough to float. A water leveldncrease of 1.5 m from June to July destroyed most sago shoots. When water levels were low, sago density declined rapidly as the depth gradient increased. However, when water levels were high, a reverse trend was evident, as the site of maximum sago density shifted towards greater depths. This was caused by the time required for sago to colonize formerly dry shorelines. By the time water levels began to recede in October, sago density was roughly similar across the 0.5-6.5-m depth gradient.
Few other data are available on the effects of lowered water levels on sago. Harris and Marshall (1963) saw great increases in drupelet production in remnant pools of impoundments when water levels were lowered artificially and by evaporation. The greater increases that occurred during the first 2 years of reflooding were suspected to be caused by changes in soil chemistry and nutrient availability. Bailey and Titman (1984) saw sago biomass nearly double as water levels fell 17 cm between growing seasons, even though water clarity was slightly decreased.
Stable water levels allow maximum sago reproduction (Bellrose 1941; Jackson and Starrett 1959; Kuflikowski 1977; Steffeck et al. 1985). Although Low and Bellrose (1944) found greater foliage production of sago in fluctuating water, more drupelets were produced under stable water conditions.
Wind-driven ice (ice scour) can severely damage beds of submersed plants (Martin and Uhler 1939), but I doubt whether ice scour has significant longterm effects on sago because it is resistant to disturbance and has rapid propagule dissemination from established stands. Martin and Uhler (1939) also opined that ice lift would carry away large areas of sago in sluggish streams during spring breakup. In the shallow prairie wetlands of interior North America, it is common in spring to see large amounts of undecomposed sago and bottom sediment lifted to the surface with bottom ice, but I have noticed no long-term changes in sago distribution in these areas. However, Kautsky (1987) mentioned that ice erosion may reduce sago populations in thalassic waters and noticed smallest standing crops at exposed sites. Krausch (1976) felt that lack of winter ice was a factor favoring retention of sago biomass in a river.
In summary, sago is adapted to and highly tolerant of currents and water level fluctuations, and sometimes may be benefited by the increased nutrient supplies and lack of potential macrophyte competitors at sites where water movement is substantial. Presence and abundance, of course, depend on other, often closely associated, environmental factors such as water transparency and sedimentation rates. Sago can survive periodic exposure to air and water level fluctuations of nearly 2 m in clear tidal waters, yet depth increases of as little as 10 cm can greatly reduce production in highly turbid waters. These limits probably are invalid for wetlands with fetches > 500 m where an increase in damage from wave action may occur. Sago is moderately tolerant of turbulence caused by wave action, and sometimes prospers after storm damage has reduced populations of other macrophytes. Sago beds can dampen wave action sufficiently to reduce phytoplankton production and likely cause many other biological changes in the littoral.