Northern Prairie Wildlife Research Center
Despite differences in the vertical and horizontal distribution of species in a wetland--likely related to competition for light, nutrients, or substrate-- an overlapping or mosaic distribution is the rule (Kornas et al. 1960). Submerged vegetation in some lakes may be considered in a state of permanent flux, initiated by chance and controlled by competition (Denny 1973). Maximum sago production only occurs where certain combinations of water chemistry, depth, light transmittance, substrate, and other habitat factors exist. Yet sago coexists with dozens of other species of submersed aquatic p'ants, both vascular and nonvascular. Sago beds are also frequently interspersed with stands of emergent or floating-leaved hydrophytes (Stewart and Kantrud 1971; Britton and Podlejski 1981; Liston et al. 1981). Some submersed macrophytes that grow in the same wetland with sago are listed in Table 8. Many of the references listed in this table contained information on possible competitive interactions discussed later in this section. Although the information in Table 8 does not reflect a random sample of sago habitat, genera showing high frequency of association with sago include Myriophyllum, Ceratophyllum, Ruppia, Ranunculus, Chara, and Najas, making them good candidates for the most important worldwide sago competitors. Common members of these genera, like sago, are distributed worldwide and have great environmental tolerance.
Of the 27 species of Potamogeton listed in Table 8, only P. pusillus, P. perfoliatus, and P. crispus probably are important individual competitors of sago in the Northern Hemisphere. In addition, P. richardsonii (considered by some to be a subspecies of P. perfoliatus) and P. zosteriformis likely are important in North America. Potamogeton richardsonii, P. friesii, P. vaginatus, and P. zosteriformis tend to associate with each other in central North America (Pip 1987) and thus may compete for the same resources.
The wide environmental tolerance of sago is shown by the extensive distributions of the 48 other taxa listed in Table 8. These plants occur in fresh to saline waters of coast and interior in a variety of climates. Waters represented on the right side of the table likely are mostly saline, alkaline, or turbid, whereas those on the left probably are fresh, neutral, or slightly acidic, and clear. Some other submersed macrophytes not shown in Table 8 are often associated with sago. These include the Charophytes Nitella and Tolypella, the primitive angiosperm Isoetes, and many large Chlorophytes.
Sago can alternate in dominance with other species on an annual basis, at least in some wetlands in the glaciated prairie region of North America. There, unstable hydrological conditions cause large fluctuations in salinity of surface waters. In some of the more saline wetlands, dilution of surface waters during periods of relatively high water levels results in dominance by sago (commonly accompanied by Chara), whereas evaporative losses can increase salinities so that only the more salt-tolerant Ruppia maritima is able to grow (H. A. Kantrud, personal observation). Similarly, some less-saline waters, when diluted, become dominated by less salt-tolerant macrophytes such as Utricularia vulgaris but return to dominance by sago when salinities again increase (Swanson et al. 1988). In addition, plant dominance can be partitioned by other means, including differential timing of the life cycle (Crowder et al. 1977). For example, in India, sago covered its greatest depth range and attained its peak biomass in summer, whereas its most abundant associate, Hydrilla verticillata, covered a smaller depth range and attained its peak biomass in the second half of the year (Purohit 1981).
Sago frequency showed a bimodal depth distribution in an African lake where Chara dominated the depths between the modal peaks (Denny 1972). Gibbs (1973) described a curious cycle that followed a drought that eliminated all surface water in a shallow (< 85 cm) New Zealand wetland. Emergent macrophytes germinated and a sparse sago stand survived in wet ooze. When 50-60 cm of water was replenished, Chara was dominant for more than a year but suddenly decayed with the onset of a phytoplankton bloom. That, in turn, diminished and allowed Chara to recover and then be replaced by a vigorous growth of sago and Potamogeton crispus at water depths nearly identical to those that had supported the Chara. Reduced water transparency caused by the phytoplankton bloom largely explained the first decline in Chara, but not the replacement of Chara by Potamogetons. Gibbs (1973) stated that the balance between phytoplankton, macroalgae, and macrophyte dominance in this wetland was delicate.
Seasonal abundance of sago in two Wisconsin streams illustrated sago's ability to take advantage of the phenology of suspected competitors while simultaneously occupying stressed environments (Madsen and Adams 1985). In one stream, sago had highest relative cover in May before Elodea canadensis and other species began rapid growth, but in upper stream sections sago was not found to coexist with plants having lower nutrient requirements. In the other stream, sago increased in dominance in July after the senescence of P. crispus. In the lower sections of both streams. sago maintained dominance in areas of pollution, siltation, shifting sands, and increased turbidity.
Sago is rare or absent in some waters for unknown reasons. As discussed earlier, however, it is well documented that sago does poorly when growing in the same wetland with specialist taxa of acidic, nutrient-poor waters (Pip 1984) where free CO2 is the only carbon source (Hinneri 1976) or where calcium is mostly unavailable (Hutchinson 1975). The probable competitive advantage of sago generally occurs at high, but not extreme, values for common water chemistry parameters that include TDS, pH, total alkalinity, Cl, SO4, Kjeldahl N, and P (Pip 1987; Hammer and Heseltine 1988). Other physical factors often present at sites where sago seemingly lacks competitive ability include increased water levels (Harris and Marshall 1963; Boltt et al. 1969) and very soft (Ravanko 1972) or exposed, coarse grained bottom substrates (Purohit 1981). Other Potamogetons and members of the genera Ceratophyllum, Myriophyllum, Chara, and Ruppia were usually the main sago associates in these instances. Factors likely related to sago competition and succession are included in the following paragraphs, but declines or increases in plant abundance often are related to factors so completely interrelated that synergistic effects may be the best possible explanation for such events (Carter and Rybicki 1986).
The distribution of sago and other submersed macrophytes within some wetlands likely is determined by a combination of habitat factors and differences in the biology of the plants themselves. Thus, the inability of sago to grow well with Zannichellia palustris and Ranunculus baudotii in the shallower portions of a group of wetlands studied by Van Vierssen and Verhoeven (1983) could be attributed to sago's lesser abilty to produce viable diaspores in areas subject to desiccation, whereas the inabilty of the two other species to coexist with sago in the deeper portions of many ponds could be attributed to the higher levels of chlorinity found there.
Schiemer and Prosser (1976) studied an Austrian lake inhabited by sago and M. spicatum, which occupied a much greater area than sago. Exceptions were a narrow zone of greatly increased silt deposition--where sago likely was more competitive because it is better able to resist coatings of silt and epiphytic algae which cause loss of buoyancy--and again, in more windswept portions of the lake- where sago, unlike M. spicatum, was able to anchor itself to the firm substrates found there. Sago and Ceratophyllum demersum increased 2 times and 8 times in absolute frequency, respectively, in a Wisconsin lake as M. spicatum declined for unknown reasons (Carpenter 1979).
Haag and Gorham (1977) saw Elodea canadensis begin seasonal growth earlier in an area of thermal effluent discharge on a large Alberta lake, and then shade out or suppress the early growth of sago and other submersed macrophytes. By July though, storm damage and self-shading by Elodea weakened it sufficiently to produce thinning, which let the current year's growth of sago reach maximum standing crop by mid-August. An early-sprouting stand of Najas seemed to shade out sago in deeper waters of an Ohio impoundment following winter drawdown, but sago was more abundant at shallower sites than before the drawdown (Gorman 1979).
Sago can replace other macrophytes under certain conditions. McDonald (1951, 1955) reported that sago succeeded Utricularia after Typha mortality in a Michigan wetland. He postulated that if the Typha were to recover, sago would decline rapidly. Baldwin (1968) reported that an initial community dominated by Chara sp. would be replaced by a sago-Najas community in freshwater impoundments in the southeastern United States, but that succession to a Utricularia-Cabomba-Ceratophyllum community would occur if alkalinity were not maintained. Sago's rapid growth rate allows it to quickly occupy large water surfaces and smother potential competitors. This may be one reason sago often is not found with a diversity of species (Meriaux 1978).
Sago, M. spicatum, and Hippuris vulgaris replaced Najas, Chara, and other Potamogetons as an English lake high in Cl became eutrophic (Leah et al. 1978). Filbin and Barko (1985) suggested that competition for nutrients may be reduced when sago grows with poorly rooted species.
Sago does not occur with Ruppia spp. in polysaline to hypersaline marine lagoons where water level fluctuations are severe, but in mixosaline (5,400-16,200 mg/L) estuaries with more stable water regimes, sago can easily displace Ruppia (Howard-Williams and Liptrot 1980; Verhoeven 1980a). This is attributed to sago's larger number of better adapted diaspores and the rapidity with which it fills both the substrate and water column. Verhoeven (1980a) stated that the autecology of sago and its competetive relations with Ruppia in fresh and slightly brackish waters need further study.
Dense floating beds of Chara that are independent of bottom substrate were observed to shade out, but not eliminate, sago in African wetlands (Denny 1973). In a Michigan wetland dominated by Chara, sago survived only at a stream inlet (Rich et al. 1971). Bolen (1964) thought that the physical preponderance of unbroken carpets of Chara and Ruppia maritima growing at near optimum salinities would allow little chance for sago or other aquatic plants to establish in Utah wetlands. In shallow sites, sago seedling mortality may be high from damage caused by semi-floating masses of plant material and burial by litter (Haag 1983).
Plants with large floating or semi-erect leaves, such as Nelumbo lutea, can kill sago by shading (Bellrose 1941), but sago can survive fairly dense emergent growth (Wilson 1958). Shading by large algae can also occur in marine waters (Ravanko 1972). Trees growing along streams can also greatly reduce sago biomass through shading (Madsen 1986).
Cohen et al. (1986) planted Myriophyllum exalbescens (M. spicatum) and sago together in aquaria and clipped stems to simulate grazing. Results of this experiment suggested that sago can be displaced by M. spicatum under competition for light because sago reacted to clipping by growing new shoots from roots, whereas M. spicatum grew new stems below the point of clipping. Similar abilities by other plants could be expected to adversely affect sago. Conversely, mechanical harvesting of a sago-Potamogeton pusillus community resulted in an increase in the much shorter Heteranthera dubia (Engel 1984). In this case, the taller Potamogeton community obviously could out-compete H. dubia for light.
A natural succession of submersed macrophytes is often seen in newly constructed reservoirs. In such a reservoir with stable water levels in Czechoslovakia, sago appeared as the wetland was filling, but populations showed no strong expansion until 5 years later, when populations of Elodea canadensis, Utricularia australis, and Lemna sp. had decreased considerably (Krahulec et al. 1987). Rapid replacement by sago and other Potamogetons of an initial growth of Utricularia was also noted by Henry (1939) in a newly flooded impoundment in North Dakota. Steenis (1939) saw the aggressive P. foliosus (which rapidly spreads when shoots break away and grow) thwart efforts to establish sago on some newly flooded wildlife refuges in the northern Great Plains, but sago became dominant as the impoundments aged.
Almost all hydrophyte communities deposit organic matter and trap mineral materials, so hydrarch succession normally proceeds towards less hydric habitat and submersed plants are succeeded by emergents (Walker 1959b; Jahn and Moyle 1964; Whitman 1976). Thus, the presence of emergent plants in and around sago beds can indicate such succession (Van Donselarr et al. 1961; Meriaux 1978; Verhoeven 1980a). Verhoeven (1980a) suggested that, in thalassic waters, such succession was restricted to waters where salinities are less than 9 g/L. Similarly, I have noted no large accumulations of organic matter in mesosaline wetlands of the North American prairies, where sago is often monodominant, but have seen hundreds of instances where sago and other submersed species were eliminated by succession to Typha monotypes in fresher wetlands where organic matter increased. Lack of fire and grazing and the spread of the Typha angustifolia or hybrid Typha likely was implicated in the buildup of organic matter (Kantrud 1986b,c, unpublished). Siltation from adjacent cropland can reduce water depth and encourage emergent vegetation in such wetlands. Vermaak et al. (1981) observed that sago rapidly replaced communities of emergent hydrophytes in the littoral zone of an oligosaline lake when its acidic waters (pH 4.5) were turned circumneutral. I assume, then, that succession back to emergent growth would occur if conditions were reversed.
I conclude that sago is often found with a large variety of submersed and emergent angiosperms and macroalgae, but it has a marked tendency to occur in discrete beds in stressed environments or those high in ionic content and pH. Sago can alternate in dominance with other species seasonally and annually and show bimodal depth distribution in a single wetland. The competitive abilty of sago seemingly is enhanced by its phenoplasticity, leaf morphology, pollution tolerance, and ability to rapidly colonize unoccupied habitats. However, sago likely is at a disadvantage in acidic, Ca- or mineral-poor sites; sites subject to hypersalinity and water-level fluctuations; and those dominated by nonrooted species or species with large floating or semi-erect leaves. Replacement of sago by emergent plants is common in nearshore zones subject to deposition of organic and mineral matter, often as a result of man's use of uplands in the watershed.
I saw sago growth and drupelet production severely limited, likely due to shading by filamentous algae, along exposed shorelines in waters < 30 cm deep in West Stump Lake, North Dakota. Beyond that depth, sago prospered even though much filamentous algae was draped around the stems below the leafy branches, and benthic algae carpeted the bottom. Such algae may actually have benefitted sago through reductions in water turbidity.
Epiphytic algae and a host of small plants and animals use sago as a substrate in inland waters. Wisconsin sago was higher in algal aufwuchs than four other genera of aquatic macrophytes (Gough and Woelkerling 1976). Sago epiphytes in a Wisconsin impoundment weighed more than the sago itself (Filbin and Barko 1985). Epiphyte biomass on sago in a Manitoba wetland varied from 48.6 g C/m2 in high-density sago stands to 11.8 g C/m2 in low-density stands (Hooper and Robinson 1976). However, Hynes (1970) observed little epiphytic development on sago in streams, and Marvan and Komarek (1978) observed little development of epiphytes on living sago where filamentous algae were common.
Sago epiphytes also vary in abundance in thalassic waters. Kornas et al. (1960) found that epiphytic algae was rare on sago along the Baltic seacoast but adnate coelenterates were common. Epiphytes were seen by Lundegardh-Ericson (1972) only on sago growing in the deepest part of Baltic coves. Staver et al. (1981) estimated sago biomass to be 100-200 times that of the associated epiphytes in Chesapeake Bay in the United States. In contrast, Kiorboe (1980) in Finland and Kollar (1985) in Chesapeake Bay observed epiphytes so abundant as to suppress sago growth. A complex of epiphytes developed on sago leaves at an early stage of growth in a South African estuarine lake (Howard-Williams et al. 1978; Howard-Williams and Liptrot 1980); these algae may have helped sago absorb phosphorus (HowardWilliams and Allanson 1981).
Much remains to be learned about the effects of epiphytes on sago. Sago leaves 1.5 weeks old can be colonized by epiphytes, and cuticular damage can occur at 6 weeks (Howard-Williams et al. 1978). Cattaneo and Kalff (1980) and Moss (1981) suggested that plants like sago with undissected leaves are not conducive to epiphyte growth. Lesser epiphytic growth may allow survival in relatively eutrophic waters. With sago, it is usually difficult to separate true epiphyton from material that is simply entangled in stems and leaves or associated with calcareous coatings. Oborn (1964) attributed calcite deposition on sago to the bacterium Bacterium precipitatum and also found various desmids and diatoms in the calcite.
Some submersed angiosperms, including at least one Potamogeton, inhibit the growth of certain phytoplankton (Hasler and Jones 1949; Hegrash and Matvienko 1965). Fitzgerald (1969) showed that some submersed macrophytes and filamentous algae commonly found in sago-inhabited waters remained relatively free of epiphytes in nitrogen-limited culture, and that epiphytic growth on a common filamentous alga (Cladophora sp.) was related to conditions of surplus nitrogen compounds in the water column. He proposed that submersed macrophytes prevented the growth of epiphytes by competition when nitrogen is limited. However, bacteria-sized organisms that have selective toxicity to some algae were also found, so both nitrogen levels and toxins may have been involved in the epiphyte suppression. Hooper-Reid and Robinson (1978a,b) listed epiphytic algae and diatoms found at a sago-dominated site in a prairie wetland in Manitoba and found that this site had more physiological indicators of nutrient deficiency for the algae than did a nearby site dominated by emergents.
Phytoplankton are often abundant in sago inhabited waters and cause great increases in turbidity which can drastically lower production of sago, restrict it to shallow (< 35 cm) waters, or render its habitat completely unsuitable (Crum and Bachman 1973; Andersen 1976; Filbin and Barko 1985). Phytoplankton can be moved by wind and affect different areas of a wetland (Jenkin 1936). Martin and Uhler (1939) stated that decomposing blue-green algae weakened sago, but they gave no direct evidence. It may be normal for submersed angiosperms to alternate in dominance with phytoplankton in many lakes of intermediate depth, although the reasons for these changes are unknown (Mitchell 1971). Harrison (1962) opined that alternating sago and algal blooms were related to irregular scouring of a lake bottom by high water levels. Increased phytoplankton abundance has also been implicated in a shift in dominance from submersed to emergent hydrophytes (Niemeier and Hubert 1986).
Gorman (1979) proposed a long-term model of autotroph succession based on observations in a eutrophic, sago-inhabited impoundment in Ohio. A spring bloom of phytoplankton occurred when water levels were slightly deeper, followed in the next year by a summer increase in macroalgae after a partial winter drawdown. Both of these pathways led to partial or complete elimination of submersed angiosperms. Observations of Butler and Hanson (1985, 1986, 1988, unpublished) suggest that light-limiting blooms of Fragilaria in a Minnesota sago lake result from overgrazing of zooplankton by fish.
Sago can die-back early in the growing season, concurrent with extensive blooms of Gleotrichia, Anabaena, and Aphanizomenon in waters rich in P (Welch et al. 1979). Leah et al. (1978) recorded that high summer populations of Apanothece, Coelosphaerium, Lyngbya, and Anabaenopsis in P-rich waters where sago, now found in small amounts, was probably more abundant before eutrophication. Van Vierssen and Verhoeven (1983) also associated reduced coverage by macrophytes, primarily sago, to decreased light penetration caused by phytoplankton blooms in P-rich waters. Kaumeyer et al. (1981) estimated sago biomass to be only about twice that of phytoplankton in Chesapeake Bay in the United States. Nevertheless, phytoplankton production can be greatly reduced in portions of polluted lakes where sago is abundant (Aleem and Samaan 1969a). Chlorophyll a concentrations can be 80% lower in beds of sago and other submersed macrophytes than in nearby openwater areas (Godshalk et al. 1987). Sago growth reached maximum values when algal populations and chlorophyll concentrations fell to lower levels in a eutrophic South African wetland having a seasonal water regime (Coetzer 1987). The algal standing crop (excluding nano-algae and cyanophytes) seemed to be limited by turbidity associated with high water levels.
In summary, sago coexists with a wide array of periphytic and planktonic algae that change seasonally with sago biomass. Periphyton can lower sago biomass by shading, especially in shallow, sheltered locations, but some epiphytes may help sago assimilate P. Phytoplankton often seriously limit sago biomass through reductions in water transparency, and blooms likely can be caused by a variety of factors, including eutrophication, water-level fluctuations, storm damage, and imbalances in the trophic structure. The relations between algal populations and submersed angiosperms need further research.
Accounts of sago's value to invertebrates vary. Some have suggested that sago beds are poorly to moderately attractive to invertebrates (Krecker 1939; Andrews and Hasler 1943), especially in moving-water environments (Needham 1938, cited in Moyle 1961; Greze 1953, cited in Hynes 1970) or openwater areas free of emergent plants (Berg 1949; Andrikovics 1973). Others have remarked about extremely large invertebrate populations supported by sago beds (Bolts 1973, cited in Howard-Williams and Davies 1979; Howard-Williams and Liptrot 1980). Zooplankton biomass in sago stands can reach 0.5 g C/m2, with seasonal net annual production 3.5-6.0 g C/m2 (Buchlovska 1964, cited in Korinek et al. 1987).
Studies by Moyle (1961), Krull (1970), and Andrikovics (1973) indicated that fine or feathery-leaved forms of submersed macrophytes supported more abundant invertebrate populations than broad leaved forms, but a study by Korinkova (1971) seemed to disprove the idea that invertebrate densities were related to degree of leaf dissection. Moreover, Krecker (1939) observed lower invertebrate density and diversity on sago than on several broader-leaved Potamogetons.
Whatever the case, sago leaves are undissected but narrow and numerous, and, often in combination with algae, sago beds are heavily used feeding sites for waterfowl broods (Hochbaum 1944; Monda and Ratti 1988). These beds are prime sources of protein for young birds because of abundant and easily obtainable populations of macroinvertebrates, including diptera, trichoptera, odonata, chironomidae, and crustacea (G. A. Swanson and H. A. Kantrud, personal observation). Sago communities also provide escape cover for macroinvertebrates, thus allowing them to thrive in the presence of small fish. Jarvis et al. (1985) studied sago in a Nevada lake heavily used by largemouth bass (Micropterus salmoides), trout (Salmo spp.), and waterfowl broods and found that sago beds were attractive to ephemeroptera, tricoptera, chironomidae, and gastropoda during various seasons.
Filter-feeding and grazing forms predominate on sago (Howard-Williams and Davies 1979). In a large Saskatchewan lake where sago was the only submersed angiosperm, Huntsman (1922) found amphipod, copepod, and cladoceran crustaceans and notonectid, chironomid, and odonate insects abundant. The amphipod Hyalella knickerbockeri was most abundant in the sago beds, and the tricopteran Phryganea interrupta was restricted to them. Chironomid larvae were most numerous at stations occupied by sago in a shallow Iowa lake (Tebo 1955), and emerging adults were more abundant in sago than in either emergent communities or openwater areas of a shallow Manitoba marsh (Wrubleski 1987; Wrubleski and Roback 1987). Rich (1966) found Tendipes, Oronectes, Anodonta, and Placobdella in a Michigan wetland supporting sago. Abundant decapoda were associated with sago by Pirnie (1935).
In a Wisconsin reservoir, benthic macroinvertebrates numbered 25,000/m2 in the upper 5 cm of sediment under sago beds, and some forms were recorded 10-15 cm deep (Miller et al. 1987). This density was only slightly less than in nearby beds of Ceratophyllum demersum, but > 10 times that found at nearby unvegetated sites of approximately the same depth. Oligochaete worms were most common, followed by gastropods and chironomids.
Information on abundance of invertebrates in European, English, and African wetlands where I sago was dominant or an important member of the submersed macrophyte community can be found in the studies of Hoffman (1958), Aleem and Samaan (1969a), Mason and Bryant (1975), Verhoeven (1980a), Howard-Williams and Allanson (1981), Davies (1982), Van Vierssen (1982a), Van Vierssen and Verhoeven (1983), Driscoll (1986), and Coetzer (1987). These sources indicate greatest occurrence in sago of crustacea, mollusca, diptera (mostly chironomid) larvae, and annelida. Other insect orders, especially coleoptera, hemiptera, heteroptera, and odonata, were often important.
The effects of underwater mowing on invertebrates associated with monospecific stands of sago were investigated by Stewart and Davies (1986) in an African estuary. The wetland was rather unusual because a tube-dwelling polychaete contributed most to the invertebrate biomass in the sago beds. With the exception of this organism, which colonized sago stems and leaf bases, mowing was deleterious to the other important forms, especially the long-lived, attached filter-feeders. The authors urged that some sago sites be left completely undisturbed in order that these organisms could fulfill their role as important food sources for birds and fish and possibly as reducers of algae-related water turbidity.
Direct burrowing, defoliating, feeding, and egg laying by invertebrates on living sago plants can be relatively unimportant, according to Berg (1949), who found that sago and several other narrow-leaved Potamogetons were among the least used of 17 species of Potamogeton in Michigan lakes and rivers. However, laboratory studies show that Minnesota sago mass significantly decreases by snail (Physa gyrina) grazing (Sheldon 1987). Overall, the extent and effects of herbivory on submersed macrophytes is poorly understood.
Prejs (1986a,b, 1987) listed nematodes that penetrated the undergound parts of sago. The most common species, Hirschmanniella gracilis, causes necrotic yellow patches on rhizomes that can make up 10-50% of the rhizome mass. These nematodes may be most abundant on sago stressed by filamentous algae in polluted waters.
The chrysomelid beetle Haemonia appendiculata (Donaciinae) has recently become common in some marshes in the French Camargue, where damage to sago on the scale of hectares has been recorded (Grilles 1988). Artificial increases in water permanency and reductions in salinity presumably lead to the increase in dominance by sago and allow the development of high populations of this burrowing insect.
In summary, sago beds can be an important source of organic matter for zoobenthic food chains and provide many of the needs of a wide variety of invertebrates. Crustaceans, insects, and molluscs seem to be the most common macroinvertebrates associated with sago. These are also important foods of young waterfowl. Direct consumption of sago by invertebrates is relatively unimportant, but recent research reveals that a few species can significantly decrease sago biomass.
Sheldon (1987) found that excluding fish from aquatic plant communities in a Minnesota lake produced high snail densities and dominance by Ceratophyllum demersum, a common potential sago competitor. However, it is possible the plant increased because of lower light conditions in the exclosures. Gamefish are compatible with sago, but rooting species such as carp (Cyprinus carpio) will seriously lower sago production. Much information is available on carp-sago interactions, not only because this fish is a notorious destroyer of plants valuable to waterfowl and gamefish, but also because carp are often raised commercially in shallow wetlands where beds of sago and other submersed macrophytes form critical habitat for the invertebrate foods of carp. In addition, sago itself is a carp food.
In order to maintain sago and invertebrate production, carp must not be more than 1 year old when stocked and must be harvested at least every 2 years (Putshog 1973). They must not be stocked too heavily (Robel 1961a). Adult carp stocked at 530 kg/ha almost completely destroyed a well-established community of narrow-leaved Potamogetons and other submersed macrophytes in 51 days, but carp removed early in spring before intensive feeding began allowed rapid reestablishment of vegetation (Black 1946). In sago-dominated French wetlands, Crivelli (1983) found that 675 kg/ha carp destroyed half the dry weight of vegetation in 71 days. A prolific sago stand in an Australian lake was reduced to a few small patches after carp were introduced (Fletcher et al. 1985).
Adult carp feed and spawn in shallow (20-50 cm) water, where vegetation is prone to heavy damage (Crivelli 1983). Struthers (1930) found worst sago damage in shallow (< 1.5 m) New York lake waters during the mid-May to mid-July carp breeding season; after breeding, the fish moved to luxuriant sago beds in 2.1-4.6 m of water where they caused little damage. Carp eggs can be abundant in sago beds (Harrison 1962) and deposited on the plants (Verhoeven 1980a). Struthers (1930) noted that the shallow-water stands of sago on which egg deposition occurred also provided young carp important escape cover from their fish predators. In some cases, carp can be restricted to deeper areas and have little effect on sago production (Sterling 1970).
Uprooting, not consumption, likely is the main direct effect of carp on sago; such uprooting is most critical when plants are immature (King and Hunt 1967). One-year-old carp are planktivorous (Matlak and Matlak 1976) and so should have little adverse effect on sago.
Carp feeding is evidenced by a dimpled appearance on bottom substrates where mouthfuls of sediment have been extracted and by bare trails among any remaining vegetation (Black 1946). King (1965) found depressions, up to 1.2 m wide and 15 cm deep, caused by carp activity in Michigan waters that supported sago, and Rich (1966), working in the same wetland, found such depressions could occupy up to 50% of the bottom area.
Common carp eat sago, but, unlike the herbivorous grass carp (Ctenopharyngodon idella) or its hybrids (see Control Methods) do not seem to consume large amounts except when stressed. Sago leaves composed 2.2% of the stomach content volume and were present in three of nine carp taken in Michigan sago beds (King 1965). In shallow mixosaline waters where invertebrate foods were plentiful, Sigler (1958) found that plants, including sago, composed 1-13% by volume of the diet of adult carp and up to 23% where normal foods were scarce. Nearly all the plant food was debris, although one fish contained 50 sago drupelets. Sigler (1958) also showed that young carp consume only small amounts of plant material, but that such material could be far more important in the diet of adult carp in cold infertile waters where invertebrate foods were scarce. Crivelli (1981, 1983) found no green vegetation in carp, but did find plant detritus and up to 1,000 seeds or drupelets of sago, Scirpus maritimus, and Ranunculus baudotii. These propagules were found in 88-93% of carp stomachs. Sago may be preferred as food by adult carp over Ceratophyllum demersum, P. richardsonii, and Elodea canadensis, in that order (Black 1946).
Sago is likely less affected by carp than several other submersed macrophytes because of the plant's prolific system of underground rhizomes and turions (Moyle and Kuehn 1964; King and Hunt 1967; MacCrimmon 1968). Nevertheless, carp exclusion or removal usually produces increased frequency or production of sago, probably because of improvements in water clarity (Anderson 1950; Tryon 1954; Jessen and Kuehn 1960; Moyle 1961; Robel 1961a; Baldwin 1968). Beule (1979) found that there was an inverse relation between carp populations and rake sample densities of submersed macrophytes, including sago.
Carp exclosures sometimes result in such greatly improved growing conditions for submersed macrophytes that sago abundance is reduced, probably by competition from other less turbidity tolerant species (Threinen and Helm 1954). In at least some cases, sago can regain abundance l year after carp removal (Titcomb 1923) or reduction with toxicants (Garlick 1956). Carp population reduction with rotenone at Malheur Lake, Oregon, increased the sago-dominated area by more than 6,000 ha the following year (Duebbert 1969).
According to Crivelli (1983)--who could not detect carp-caused turbidity in his study ponds but cited studies showing opposite results--high turbidity in carp ponds can depend mostly on bottom substrate and meterological conditions. In a turbid (Secchi depth 25-33 cm) Michigan wetland with carp present, sago occurred only at depths < 51 cm but was found at depths up to 3.66 m after carp were removed (Lutz 1960). Rich (1966) cited unpublished data gathered by G. S. Hunt in a sago-inhabited Michigan wetland that showed Secchi depths increased from 33 cm to 117 cm in only 5 days after carp were poisoned, and that winds 32-40 km/in did not produce turbidity after such poisoning. King (1965) stated that, in a Michigan wetland that supported sago, carp-caused turbidity was not important at depths < 45 cm and where Secchi depths were > 30 cm. When carp occur in eutrophic wetlands subject to phytoplankton blooms, sago can be restricted to extremely shallow (< 8 cm) sites (Kantrud 1984, unpublished). The activities of carp superimposed on that of domestic livestock nearly eliminated sago in one shallow wetland (Barrett 1964).
I conclude that sago beds are important to many species of fish. Young sago plants, especially in soft substrates in shallow water, can be harmed by the activities of bottom feeders like the common carp. Old plants can also be damaged, although reproduction from underground turions makes sago more resistant to carp damage than many other submersed angiosperms. Carp damage is greatest where fish are large and numerous. Only a few species of fish normally eat large amounts of sago.
Several studies attempt to document the effects feeding waterfowl have on stands of sago. Waterfowl can excavate holes up to 10 m wide and 0.3 m deep in search of sago turions (Wetmore 1921). Sincock (1962) estimated that 9% of the fall sago crop was consumed by waterfowl, but his estimate did not include forage wasted during the feeding process. Sterling (1970) believed that a 52% removal of fall turions would not significantly decrease second season productivity. Anderson and Low (1976) reported that waterfowl feeding activities reduced peak standing crop and turions 40% and 43%, respectively, and postulated that a single stand could take heavy use over several seasons before production decreased. However, there was some evidence that heavy feeding on turions could reduce second season standing crops. They also noted that reduction of that degree would be unlikely, because feeding waterfowl tended to move to more productive sites. Jupp and Spence (1977b) attributed 21% of the biomass removal in unenclosed stands of sago to waterfowl grazing. In their study area, underground turions were held tightly by clay bottoms, and waterfowl mostly clipped aboveground portions of the plants. Beds of sago were believed to have flourished for at least 20 years in a wetland that received very heavy use by sago-consuming waterfowl (Kantrud 1986a).
I found no references to significant decreases in sago production caused by activities of other groups of aquatic birds. Thus, given the extremely high reproductive potential of sago and its ability to shield at least some of its propagules from feeding birds, it is unlikely that birds are often a significant factor limiting sago production.