Herbivorous fish have a limited ability to convert plant material into animal tissue and to remove excess nutrients from the system. But being relatively efficient grazers and inefficient assimilators, they may play a significant part in initiating the internal, biologically mobilized nutrient source, or accelerate their recirculation (Prejs, 1984). Thus, if with advancing eutrophication, the rate of macrophyte consumption increases due to the increase in number of herbivores, a new part of their biomass may finally be converted into additional algal biomass - the main troublemaker in lakes of the temperate zone.
Considerable information is available on the impact of grass carp grazing on water quality. Grass carp partially digests its food and the faeces are a large source of nutrients. Nutrients are readily utilised by phytoplankton, resulting in elevated algal biomass, and in some situations leading to algal blooms. Bettoli et al. (1993) monitored the changes in chlorophyll a concentration prior to and after the introduction of grass carp in Lake Conroe (Fig. 18).
Fig. 18. Chlorophyll a concentrations measured approximately biweekly at four stations in Lake Conroe, Texas, 1980–1986. ‘Stocking’ indicates the introduction of grass carp. (From Bettoli et al., 1993).
Not always the primary production changes under the impact of grass carp grazing on macrophytes. In Red Haw Lake, Iowa, where within three years the introduced grass carp reduced aquatic macrophytes from 2 438 g m-2 to 211 g m-2, mean nitrites, nitrates, biological oxygen demand, and turbidity showed significant decreases, while alkalinity increased significantly (Mitzner, 1978). Mean concentrations of organic and inorganic phosphates gradually increased, but were not statistically different. During the first two years the average primary production was nearly identical, followed by a decrease in the third year. Grass carp consumed all major plant groups, with preference for Najas and Potamogeton. Lembi et al. (1978) found that as much as 54% of the phosphorus and 42% of the nitrogen released by consumption of plants by grass carp in Indiana ponds were incorporated into new fish tissue. Most of the phosphorus not taken up by the fish was sequestered into components other than water or phytoplankton. The same authors also noted that potassium concentrations in water appeared to be an excellent indicator of vegetation consumption by grass carp. The authors suggested that the impact of the grass carp on water quality should be minimal in temperate zone small water bodies if the fish is stocked early in the season when vegetation biomass is low.
Leslie et al. (1983) detected a significant increase in nitrate-nitrite concentrations in two Florida lakes following macrophyte removal by grass carp. Increased alkalinity may have been related to decreased photosynthetic activity and the resulting release of carbonate ions following the elimination of macrophytes. Mitzner (1978) reported a significant increase in alkalinity following a 91% reduction of aquatic macrophytes by grass carp in a lake in Iowa. Elimination of macrophytes in a cooling reservoir in North Carolina using Tilapia zillii also resulted in an increase in nitrate/nitrite and alkalinity concentrations (Crutchfield et al., 1992). An intensive decomposition of macrophytes of their residues may result in an increase in ammonia, the concentration of which, if exceeding 1.0 mg L-1, would stop grass carp to feed (Kracko and Noble, 1993).
In experiments in aquaria the macrophyte Anacharis canadensis was eliminated by the grazing activity of rudd (Hansson et al., 1987). The grazing resulted in an increase in phosphorus concentration in water. In ponds this caused an increase in the biomass of epiphytic algae but no increase in phytoplankton.
Common carp reworking and digestion of bottom sediments may lead to a substantial increase in phosphorus in water. Lamarra (1975) suggested that a carp population of 200 kg ha-1 could internally load a lake with orthophosphate at 0.52 mg P m-2 day-1. This might result in an increase in phytoplankton where plant control using common carp is proposed.
In southwestern Finland, the small and shallow Lake Littoistenjarvi alternates between two alternative states: mesotrophic, characterised by well-developed stands of aquatic macrophytes, and eutrophic, when the macrophytes decline. The high abundance of submersed macrophytes and clear water in the early 1990s, was followed in 1992 by high nutrient and chlorophyll levels, high turbidity and reduced stands of macrophytes. The situation changed again during 1993–1996 when the lake water was clear, chlorophyll levels low relative to total phosphorus, and aquatic macrophytes moderately abundant. The roach stock was moderately strong (71 kg ha-1) after the high abundance of submersed macrophytes in the early 1990s, declining to about 28 kg ha-1 by 1996. Perch biomass was notable, but consisted mainly of large (>20 cm) individuals that were probably piscivorous (Sarvala et al., 1998). In 1994–1996 the total fish biomass did not decrease, but there was a clear shift in the size distribution towards larger fish; especially small roach were almost absent in 1995–1996. This size shift probably led to a lowered predation pressure on zooplankton, and the cladoceran biomass showed some increase in 1994–1996 (Fig. 19). In this Finnish lake cladoceran zooplankton was thus controlled by fish predation, but water quality was not dependent on zooplankton. Phytoplankton abundance was more tightly linked to long-term fluctuations of submersed plants. The abundance of large crustaceans in 1993 (and an almost absence of Daphnia) and in later years of the littoral cladocerans Sida and Simocephalus, probably originated from the submersed vegetation, which may also have offered a refuge from fish predation for other fish. Interestingly, during 1993–1996, phytoplankton biomass remained low in spite of very low crustacean zooplankton biomass during most of the summer. Similar role of submersed aquatic plants has been also reported for some other lake ecosystems (Blindow et al., 1993; Schriver et al., 1995).
Fig. 19. Average crustacean biomass (period 26 July–15 September, years 1989–1996, no samples taken in 1990 and 1991) in Lake Littoistenjarvi, Finland. Roach biomass estimates are shown for 1993–1996, vertical bars denote the 95% confidence limits for the estimate. Chlorophyll a concentrations shown above each year when measured. (From Sarvala et al., 1998).
Sand-Jensen and Borum (1984) and Sand-Jensen and Sondergaard (1981) belong to a number of workers who pointed out the importance of epiphytic (synonym: periphytic) algae coating submersed macrophytes as a factor which leads to the suppression of aquatic plants. Excessive growth of epiphytes is harmful to the macrophyte, because insufficient quantities of light and inorganic carbon (HCO3 and CO2) can reach the plant and consequently the macrophyte photosynthesis is hampered (Sand-Jensen and Sondergaard, 1981). Epiphytic algae are more efficient at obtaining nutrients and light due to their large surface area to volume ratio, high biotic potential and closer proximity to the water column. They may also have more efficient nutrient uptake mechanism (Maberly and Spence, 1983). Periphyton provides more nutritious substrates than sediments or vascular plant tissues, and consequently it is an important energy source for both detritus and grazing food chains (Gressens, 1995).
Epiphytic algae benefit from inorganic nutrients and dissolved organic compounds leaking from the host macrophytes, and these may be readily used and beneficial to algal and bacterial communities of the epiphyton, especially in oligotrophic waters (Bronmark and Vermaat, 1998). Difference in algal communities on different species of macrophytes under low nutrient availability has been suggested to be due to host-specific composition of the released organic and inorganic compounds (Eminson and Moss, 1980).
Periphyton is grazed by a diversity of both invertebrate and vertebrate aquatic organisms. The invertebrates include molluscs, nematodes, microcrustaceans (Cladocera, Copepoda, Ostracoda), amphipod and isopod crustaceans, crayfish, mysids, chironomid larvae, caddisfly and mayfly larvae. Invertebrate grazers may play an extremely important role in controlling periphyton. Grazing snails benefit macrophytes by removing the epiphytic cover, as shown in experiments by Bronmark (1985): the growth rate of Ceratophyllum demersum increased in the presence of snails by almost 30 % due to the removal of epiphytic cover. Similar results were obtained by Martin et al. (1992) in Bays Mountain Lake (Tennessee, USA) where the biomass of submersed macrophytes (Najas flexilis and Potamogeton diversifolius) increased 60x in cages where snail-feeding redear sunfish and small sunfish (Lepomis macrochirus and L. microlophus) were excluded. The results suggested a chain of strong interactions (i.e. from fish to snails to periphyton to macrophytes) that may be important in lake littoral system. This contrasts sharply with earlier predictions based on cascading trophic interactions that propose that fish predation on snails would enhance macrophyte biomass.
In an experimental study on a small pond in southern Sweden, with dense submersed vegetation dominated by Elodea canadensis, with much less abundant Potamogeton pectinatus, P. obtusifolius, P. crispus, Ranunculus trichophyllus, and Chara sp., the presence of tench indirectly increased the biomass of periphyton through a reduction of grazing pressure by snails, on which the fish fed (Bronmark, 1994) (Fig. 20). A similar study carried out on two North American lakes showed that the highly specialized moluscivorous fish Lepomis gibbosus dramatically reduced the density and biomass of snails, which in turn resulted in an increased biomass of epiphyton (Bronmark et al., 1992). These studies showed strong cascading interactions between fish, snails, algae and submersed macrophytes. Given the difference in shade tolerance between plant species, cascading trophic interactions initiated by a predatory fish may well influence distribution patterns and species composition of submersed macrophyte communities in lake littoral zones. With an increasing density of predators feeding on snails the most shade-intolerant macrophytes are predicted to disappear first, and at high levels of predation, possibly coupled with increased nutrient enrichment, the entire submersed macrophyte community may disappear.
Fig. 20. Direct and indirect effects of molluscivorous tench on snails, periphyton, and the submersed macrophyte Elodea canadensis. (From Bronmark and Vermaat, 1998).
Intermediate-sized herbivores - fish and crayfish - reduce macrophytes and periphyton similarly, with the literature suggesting that the periphyton consumed by these species is primarily filamentous macroalgae that may be intentionally ingested by the herbivores. But Scott and Crossman (1973) pointed out that many American fish are algivorous or use filamentous algae as a substantial food source, particularly in late summer. In the slow-flowing rivers of southeast England young-of-the-year roach (Rutilus rutilus) and chubb (Leuciscus cephalus) switch to feeding on periphyton when preferred zooplankton is not available, either due to low numbers or at night when visual feeding is precluded. This switch coincides with a decline in the growth rate of these fish (Jones et al., 1998). By contrast, the smallest herbivores - snails and insects - have little (insects) or no impact (snails) on macrophytes but do substantially reduce periphyton, especially microalgae (Lodge et al., 1998).
The seasonal timing of the presence of different grazers may differ substantially. Chironomid larvae often dominate in spring, and graze until their pupation and emergence. Snails are present all year around and may remain active even at low temperatures. In some water bodies oligochaetes appear to be the most important herbivores in terms of overall grazing impact (Bronmark and Vermaat, 1998). But seasonality may be weakly expressed, or completely lacking in tropical waters. Thus, the consequences of grazing may differ in subtropical and tropical lakes from those in temperate water bodies. Botts and Cowell (1993) found that abundance of epiphytic algae and invertebrate grazers were only weakly correlated in a subtropical lake, whereas in temperate lakes they usually show strong temporal correlation.
In Argentina, the three rivers, Uruguay, Paraguay and Parana, form Rio de la Plata. The characid fish Prochladius lineatus is the dominant species among the some 300 species that inhabit Rio de la Plata. It represents around 60% of the biomass and over 70% of fish catches. It has been suggested that one of the causes of the extraordinary success of this detritivorous fish is its capacity to selectively feed on nutritionally superior detritus, i.e. epiphytic detritus, rich in total amino acids. Also, it has been suggested that it might obtain periphyton associated with roots and shoots of aquatic macrophytes. As water hyacinth covers 80% of lakes and ponds of the middle Parana floodplain, Planas and Neiff (1998) investigated this plant, collected from the Parana River floodplain, for periphyton. They found that mats of water hyacinth let through insufficient light to allow full development of periphyton.
Weisner et al. (1997) investigated a combined impact of epiphytes and waterfowl grazing on submersed vegetation on three Swedish eutrophic shallow lakes in the clear-water state. They found that mechanisms hampering submersed vegetation were strongest at shallow and/or sheltered locations. The growth of Myriophyllum spicatum, planted in the same substrate and at the same water depth, was compared between sheltered and wave exposed sites in two lakes. After 6 weeks the plants were significantly smaller at the sheltered sites, where periphyton production was about 5 times higher than at the exposed sites. Enclosure experiments were conducted to evaluate the effects of waterfowl (mainly coot Fulica atra and mute swan Cygnus olor) grazing on macrophyte biomass. Potamogeton pectinatus growth was decreased by grazing, whereas M. spicatum was not affected. The effects were greater at a sheltered than at a wave-exposed site. The results suggest that competition from epiphytes and waterfowl grazing hamper the development of submersed vegetation at sheltered and/or shallow locations.
Aquatic macrophytes provide shelter or refuge to zooplankton, and the presence of especially large zooplankton controls small phytoplankton. If the larger zooplankton becomes scarce, the phytoplankton community will change to slow-growing bigger, mainly inedible blue-green species. These bigger, often colonial algae compete more successfully with the small algae for nutrients and establish high-density populations, sometimes leading to algal blooms. In turn, this restricts light penetration into deeper water and leads to the death of submersed macrophytes.
In the English lake Little Mere experiments were conducted to investigate the relative importance of top-down (zooplankton grazing) and bottom-up (nitrogen-limitation) control in limiting algal growth, and the role of aquatic macrophytes in this process (Stephen et al., 1998). It is generally recognized that submersed macrophytes provide a refuge for cladocerans and a habitat for plant-associated macroinvertebrates, both of which graze on phytoplankton and epiphytes. However, in experiments with a high stickleback (Gasterosteus aculeatus) density the fish controlled the zooplankton associated with the macrophytes. The same conclusion was reached by Jeppesen et al. (1998), who found that the refuge effect of submersed macrophytes almost disappeared - even for small-sized cladocerans - if fish density exceeded ca. 4 fish m-2. However, if plant density is high and fish are not forced into the vegetation, pelagic zooplankton may aggregate in the submersed macrophytes during the daytime. Floating-leaved plants and reed belts have a comparatively low refuge effect due to low stem density.
The efficiency of plant beds as refuge for zooplankton is often high in eutrophic lakes. In less eutrophic and mesotrophic lakes, the impact of predatory fish increases. In Europe, in such lakes perch predation becomes more important, and the foraging conditions for predatory fish improve, due to, for instance, increased transparency. Young planktivorous prey fish thus seek refuge in the vegetation, which does not provide effective refuge from the predator (Jeppesen et al., 1998). In mesotrophic-oligotrophic lakes the predatory pressure on planktivores becomes even stronger.
In some other English lakes (Cromes Broad, Hoveton Little Broad/Pound End, and Upton Broad) Daphnia spp exhibited an early summer peak of abundance but had declined rapidly by July, probably as a result of predation by fish. In dense macrophyte stands Daphnia spp persisted after their elimination in open water, indicating some refuge effect (Stansfield et al., 1997). At sites with macrophyte cover and/or low fish predation pressure, Daphnia was replaced by Ceriodaphnia, both with or without Simocephalus, thus maintaining large populations of grazing Cladocera, apparently exerting a high grazing pressure on phytoplankton. This occurred even under high predation pressure by fish. The authors have suggested that dense macrophytes offer the most suitable refuge for cladocerans through provision of predator-free space.
Cascading trophic interactions are known to play an important role in the transfer of fish predation effects via zooplankton to the phytoplankton community, and this may affect also the structure and function of the microbial community. Jurgens and Jeppesen (1998) followed the cascading effects on microbial food web structure in a dense macrophyte bed (mainly Potamogeton pectinatus) under experimental conditions. As a result of differing fish predation pressure (strong where no macrophytes present, low where macrophytes present), there was a high density of different cladocerans and cyclopoid copepods in the enclosure system with macrophytes, whereas ciliates, rotifers, and cyclopoid copepods dominated in the enclosures without macrophytes. In the macrophytes the abundance of phytoplankton, protozoans and bacteria was very low, whereas with no macrophytes a diverse assemblage of pico-, nano- and microplankton coexisted with the metazooplankton community. Ciliate density was approximately two orders of magnitude higher in the enclosures without macrophytes. The protozoans there probably exerted considerable grazing pressure on the bacterial and algal communities. Bacterial concentrations were four- to fivefold higher in enclosures without macrophytes than those with macrophytes. The experiments have shown that fish predation impact can cascade from fish to the bacterial level, and that this is probably a general feature of more eutrophic systems, such as shallow eutrophic lakes in the temperate climate of the northern hemisphere.
See also Section 6.
Aquatic macrophytes are an important habitat for aquatic macroinvertebrates. The plants provide protection from predators and current, and are a direct and indirect source of food of fish and aquatic birds. In plant beds, benthic invertebrates are generally more numerous and more diverse than in open water. Diehl and Kornijow (1998) showed the increase in epiphytic macroinvertebrates with increasing macrophyte densities, and how this is paralleled by an increase in fish biomass (Fig. 21). Species composition and biomass of macroinvertebrates usually differ from plant species to species, depending on the structural complexity of the plant (see also Section 2.1.1), changes in the morphology of macrophytes during the growing season, amount of the periphyton and detritus on plants, changes in water level, life cycle of the individual macroinvertebrate species, and some other factors. Differences in invertebrate richness, biomass and distribution will have a direct effect on the fish populations depending on aquatic macroinvertebrates as source of food. Indirectly, macrophytes are also important for the benthic macroinvertebrates, for which they provide protection against foraging by fish and aquatic birds, as well as providing the invertebrates with a supply of organic matter from the decomposing plants.
Kornijow and Gulati (1992) investigated aquatic macroinvertebrates associated with aquatic vegetation about 18 months after the biomanipulation of Zwemlust, a small Dutch lake. They found that the macrofauna did not differ notably from that in non-manipulated lakes. The species diversity differed on the different species of aquatic plants. The greatest diversity was on Chara sp., moderate on other two submersed plants Elodea nuttallii and Ceratophyllum demersum, and on Phragmites australis, and the lowest on filamentous green algae, primarily Mougeotia sp. Some taxa were restricted to certain plants, e.g. the mollusc Acroloxus lacustris found solely on reed, while the molluscs Armiger crista, Planorbis corneus, Valvata piscinalis and leeches, except Erpobdella sp., were recorded only on the submersed plants. Larvae of Tipulidae and Ephydridae occurred exclusively on filamentous algae. The density and biomass of the fauna per 100 g of plant wet weight on different plant species markedly differed, with the lowest values on reed and among Mougeotia. Much higher density and biomass occurred on submersed plants, especially on Chara sp. Generally, the densities of the phytomacrofauna were relatively low but the biomass values high, compared with many other lakes. Gastropoda, larvae of Chironomidae and Trichoptera were the main quantitative components of the phytomacrofauna. It was noted that the greater the fragmentation of macrophyte leaves the greater the faunal density, and usually higher the taxonomical variability of the associated animal populations. In this respect, Lake Zwemlust aquatic vegetation confirmed findings from some other lakes. There was a high degree of similarity between the fauna associated with Ceratophyllum and algal mats, with a high dominance of the chironomid Cricotopus larvae. The larvae feed primarily on the algae constituting mats and those covering the macrophyte (Kornijow and Gulati, 1992).
During strong stratification, aquatic macrophytes provide a diversity of microhabitats and refugia for aquatic macroinvertebrates which are temporarily or permanently excluded from the lake bottom by low dissolved oxygen concentrations. Differences in the habitats provided by the floating Pistia stratiotes and submersed Ceratophyllum demersum in Volta Lake (Ghana, West Africa) were reflected in species composition, dominance and standing crops of their associated aquatic macroinvertebrates. While grazers and filter-feeders, such as zygopteran nymphs and chironomid larvae, dominated Ceratophyllum, submersed Pistia roots were utilised by predatory anisopterans, the hemipteran Diplonychus, and the grazing molluscs Bulinus forskali and B. truncatus rohlfsi (Petr, 1968). The density of macroinvertebrates in Pistia roots reached a maximum of about 16,000 individuals m-2, which is close to the density of 18 620 individuals m-2 collected by Pelli and Barbosa (1998) from Salvinia molesta roots in a lake in Brazil. The dry mass of macroorganisms on Pistia at 12 g m-2 was higher than that reported in the literature for floating plant associated invertebrates from African, Asian or South American lakes (McLachlan, 1969; Lim and Furtado, 1975; Junk, 1973). The association of Ceratophyllum with shallows was found to represent a protective mechanism for the associated macroinvertebrate fauna: wind-blown Ceratophyllum, carried off-shore and found stranded in the crowns of drowned trees several hundred metres from the shore, supported considerably lower animal densities than those in the shallow littoral (Petr, 1968).
Fig. 21. Relationships between macrophyte density and (A) the wet biomass of benthic and epiphytic macroinvertebrates in Lake Paajarvi (Finland) (with 95% confidence intervals); (B) the percentage of total macroinvertebrate biomass composed of epiphytic macro-invertebrates in Lake Memphremagog; (C and D) relationships among the average wet biomasses of macrophytes, macroinvertebrates, and fish in the littoral zones of ten southern Quebec lakes (one lake marked by open triangles was excluded from the analyses because it had high biomasses (>12 g m-2) of two fish species that were absent from all other lakes). (From Diehl and Kornijow, 1998).
The free-floating submersed Ceratophyllum demersum is a common plant in tropical African waters. While it can be a nuisance plant as it provides a habitat for host snails of the parasite Schistosoma haematobium (see Section 14), it is considered to be beneficial for fish as it serves as a substrate for aquatic invertebrates and periphyton, on which many fish feed. Due to the vertical distribution of this plant, which may reach several metres of length, Petr (1968), rather than making an attempt of expressing the number and biomass per square meter, which would be irrelevant because a different length/depth of the individual plants, expressed the abundance and biomass of invertebrates associated with Ceratophyllum as per 1 000 g of dried plant. During three months of observations, the inshore Ceratophyllum contained 16 502, 39 020, and 92 156 organisms, respectively (corresponding to 34, 40 and 76 g wet weight), as compared to 12 095, 6 891, and 6 234 organisms (corresponding to 7.6, 5.9 and 5.2 g wet weight) in off-shore Ceratophyllum plants, stranded on drowned trees. The off-shore, usually isolated individual plants, are exposed to heavy fish predation, and their recolonisation appears to be minimal. If using conservatively the biomass per 1 000 g of dried plant as a biomass per m2 for the inshore habitat, the biomass exceeds twice the maximum biomass of invertebrates found on Pistia roots in the same lake, i.e. 35 g m-2. The rich aquatic invertebrates of aquatic plants in Volta Lake must play an important role in the high fish production in this reservoir.
High densities of macroinvertebrates associated with aquatic macrophytes have been recorded from the borders of macrophyte beds on floodplains of Orinoco (Venezuela) and Parana (Argentina), and there was considerable difference between the ‘isolation’ phase, i.e. when the floodplains were not connected with the river, flooding phase, and the inundation period (Blanco-Belmonte et al., 1998). In Orinoco floodplain lakes, where macrophytes develop only during the rainy season, there is only one density peak. The peak densities there were: for water hyacinth 38 000 org. m-2, 68 000 org. m-2 for Paspalum repens, and up to 600 000 org. m-2 for Nymphaea. Such high densities represent lucrative feeding grounds for numerous fish.
In a review on herbivory on aquatic macrophytes Lodge (1991) pointed out that while a majority of herbivorous invertebrates eat periphyton, many do eat macrophytes. Those who eat macrophytes may be functionally important in food webs and nutrient fluxes. As a trophic link they are significant in community and ecosystem functions.
Some caddisfly larvae are known to be among the few aquatic invertebrates directly feeding on aquatic macrophytes. High densities of herbivorous insects, such as a caddisfly, several moths, several weevils, and one or more chironomid species, are known to accompany the decline in Myriophyllum (Smith, 1991). Myriophyllum spicatum and Hydrilla verticillata have become serious aquatic weeds in many parts of continental USA, and a search is going on for effective insect biological control agents. Balciunas (1991) and Balciunas et al. (1989) reported on their search for such agents in China and Australia, which followed Balciunas previous search in Africa. In Lake Tanganyika, Polypedilum chironomid larvae damage apical meristem of Hydrilla, and together with grazing by fish, are the two factors mainly responsible for containing Hydrilla (Markham, 1986). Whether the emphemeropteran nymphs of Povilla adusta, which burrow in Polygonum stems in Volta Lake (Petr, 1970) and in rootlets of Pistia stratiotes in Lake Chad (Dejoux, 1969) cause a serious damage to their host plants, is not known. They do not feed on them and use them as shelter from which these larvae filter phytoplankton from the surrounding water, or graze on the periphyton. The nymphs are known to burrow into bark and wood of trees, which speeds up the destruction of trees inundated in some African tropical reservoirs (Photo 4).
PHOTO 4: Flooded tree in Volta Lake, a reservoir in Ghana, third year of flooding. Note the exposed part with burrows, resulting from the activity of the nymphs of Povilla adusta.
In Lake Victoria and Lake George (both in Uganda), where the nymphs burrow into papyrus plants, they also attack wooded canoes of fishermen, or even boats made of fibreglass. The nymphs require for their existence a relatively high dissolved oxygen concentration in the surrounding water, which limits their lateral distribution within the fringing papyrus swamp to only about 1.5 m distance from the edge (Fig. 22).
Fig. 22. Habitats occupied by Povilla in Lake George, Uganda (A), and in Lake Volta, Ghana (B). Horizontal distribution of Povilla in the papyrus of Lake George, in relation to the dissolved oxygen content in water. In Volta Lake, during water stratification the average maximum depth of Povilla distribution was 7 m, with an average minimum dissolved oxygen content 70% saturation. During water mixing, the maximum depth of Povilla occurrence was 15 m, with an average minimum dissolved oxygen content 40% saturation. (From Petr, 1973).
In laboratory conditions the trichopteran Anabolia nervosa larvae were found to show a preference for young leaves of Potamogeton perfoliatus, which were richer in nitrogen than leaves of terrestrial trees Alnus glutinosa and Fagus sylvatica (Jacobsen and Sand-Jensen, 1994). The larvae consumed 3 times more fresh Potamogeton than Alnus leaves, but the same amount in terms of dry mass. Larvae grown on Alnus accumulated more fat than larvae grown on Potamogeton, while the reverse was true for protein accumulation. The study demonstrated that fresh tissue of Potamogeton perfoliatus provided a valuable source to this caddisfly larva. In the field, i.e. a Danish stream, however, the plant constituted only 1–5% of the food intake by this trichopteran (Jacobsen and Sand-Jensen, 1994a). The highest plant consumption was from early May to mid-June, but the invertebrate herbivores harvested only 1.3 to 1.8% of the total annual plant production. Otto and Svensson (1981) discovered in experiments that the caddisfly larvae of Sericostoma personatum (Spence), found in Swedish waters, consumed significantly less leaves with high nitrogen content than leaves with low nitrogen content, and suggested that the low palatability of aquatic macrophytes may be due to chemical defense mechanisms. Larvae of the trichopteran Potamophylax cingulatus were also tested and confirmed that aquatic plants produce secondary plant substances, which reduce the attacks by aquatic herbivores.
A study of invertebrates associated with three macrophytes in the Macquarie River, Tasmania (Australia) showed major differences in the invertebrate assemblage in any one month, and also a large turnover of taxa associated with the species of macrophyte species from one month to the next (Humphries, 1996). The three macrophytes sampled were Myriophyllum simulans/variifolium, Triglochin procera and Eleocharis sphacelata, each one differing in the structural complexity and its location in the littoral zone. The greatest abundance of macroinvertebrates was found associated in all months (and at all water levels) in the structurally complex and shallowest Myriophyllum. However, this plant had a much lower number of taxa than the structurally simpler and deeper water Triglochin and Eleocharis. The invertebrate abundance and Myriophyllum biomass were found to initially increase together, but as the plant biomass became very high, i.e. the plant became dense, the invertebrate abundance declined. This may have an implication for fish feeding, which in such a case is limited by the plant density and the availability of the prey. A similar situation was also found on Myriophyllum-associated mosquito Anopheles, with the density of the mosquitoes decreasing with an increase in the stem density (Orr and Resh, 1992).
Cebrian and Duarte (1994) found that consumption of freshwater macrophytes, seegrasses and terrestrial grasses by invertebrates and vertebrates ranged fom 1 to 80% of plant production. In two studies of floating-leaved freshwater macrophytes the invertebrate consumption of plant production was less than 10% for Nymphoides peltata (Van der Velde et al., 1982), and 3–6% for Nuphar lutea (Setala and Makela, 1991). A number of authors as quoted in Jacobsen and Sand-Jensen (1994) have observed a heavy damage of submersed and floating-leaved macrophytes by invertebrate herbivores.
The presence of vegetation serves as a food resource and cover for many vulnerable macroinvertebrates in the mud, which otherwise would have been quickly eliminated in the absence of plants (Petridis, 1990). McLachlan (1969) found an increase in the number of pulmonates and bivalves in the mud of Lake Kariba, a reservoir in eastern tropical Africa, following an invasion of aquatic macrophytes. But Killgore et al. (1989) found that dense monospecific beds of Myriophyllum spicatum contained few benthos and could support less diverse fish community. In the USA, on the other hand, in ponds without vegetation, common carp and bluegill greatly reduced the abundance of oligochaetes and nematodes (Forester and Lawrence, 1978).
Availability of aquatic invertebrates as fish food depends on the effectiveness of the shelter provided to them by aquatic macrophytes, and their productivity. Their productivity will depend on the food availability. Epiphyton, the source of food of the majority of macrophyte dwelling macroinvertebrates, is at its optimal development when the surrounding water has intermediate nutrient concentrations (Carpenter et al., 1998).
The most important aspects of the periphyton-invertebrate grazer-fish interaction in lakes is that herbivorous invertebrates provide an alternative food source for fish and hence reduce the pressure on zooplankton (Jones et al., 1998). Without such invertebrates, the ontogenetic shifts seen in feeding adult fish would not occur and all fish would compete for the same food resource, a situation seen in plantless eutrophic lakes. Such feeding on invertebrates by fish will result to some extent in an increase in periphyton, but the overall cost to the plants may be outweighed by gains resulting from increased zooplankton abundance (Jones et al., 1998).
Insects, which are used for biological control of nuisance aquatic plants, may also become food for fish, thus negating their impact. To control Myriophyllum spicatum, the aquatic weevil Euhrychiopsis lecontei was released in Lake Auburn (Minnesota, USA). Fish stomach analysis of the sunfish (Lepomis macrochirus) and pumpkinseed (Lepomis gibbosus) has shown a frequency occurrence for the weevil larvae and adults of 10.3% to 28.6% (Sutter and Newman, 1997). The findings suggest that sunfish and pumpkinseed predation may decrease the effectiveness of the biological control of Myriophyllum using E. lecontei.
Additional information on herbivore and detritivory on freshwater macrophytes by invertebrates is given in a review by Newman (1991).
The proportion of aquatic macrophyte in crayfish food may represent up to 75% of their total food intake as documented by a number of workers since the 1980s. Crayfish can also substantially reduce or eliminate submersed vegetation, as documented, for example, from lakes and ponds in the northern USA and from Lake Naivasha, Kenya. In New Mexico the crayfish Orconectes causeyi successfully cleared several lakes of Potamogeton, Elodea, Ranunculus and Myriophyllum, but avoided Polygonum, Scripus and Typha (Schuytema, 1977, and references therein). In Arizona's Parker Lake, when the crayfish (Orconectes causeyi) population increased, their intensive grazing on submersed plants reduced the protective cover, which in turn increased the crayfish vulnerability to the predatory largemouth bass. As the predation by largemouth bass increased, the population of crayfish decreased, which led to the recovery of Myriophyllum and an increase of crayfish (Saiki and Tash, 1979). Investigations of prey capture rates of trout (Oncorhynchus mykiss) feeding on two species of amphipods in an environment also rich in crayfish (Orconectes propinquus) and smallmouth bass (Micropterus dolomieu) have shown that prey capture rates of trout fell drastically with increasing cover. Capture rates of crayfish by smallmouth bass also declined (Crowder and Cooper, 1979a).
In England, the introduced signal crayfish Pacifastacus leniusculus fed on the following five main diets, in order of their importance: vascular detritus, filamentous green alga Cladophora, crayfish fragments, chironomid larvae, and mayfly nymphs (Guan and Wiles, 1998). The larger the crayfish, the more vascular detritus was ingested. The frequency of occurrence of Cladophora was increased with crayfish size in the winter.
Reductions of macrophyte root or shoot biomass by grazers range up to 100%, with reductions exceeding 50% reported for crayfish, insect larvae, snails, fish, and waterfowl (Lodge, 1991). In laboratory experiments, non-consumptive destruction by Orconectes rusticus comprised 30–95% of cut macrophyte biomass, with the percentage differing among macrophyte species. O. rusticus is known to feed selectively, as tested on 14 different plants species by Lodge (1991). Evergreen aquatic plants are not preferred, while those preferred are usually perennial. Avoidance of evergreens by grazers is also a common pattern among terrestrial grazers (MacLean and Jensen, 1985). This appears to result from secondary chemical defenses.
Chambers et al. (1990, 1991) studied the impact of crayfish introduction on submersed plant communities and associated aquatic macroinvertebrates. In feeding preference trials with O. virilis, out of ten submersed aquatic plant species Chara sp. and Lemna trisulca were the preferred food, followed by Myriophyllum exalbescens or Utricularia vulgaris. Chemical analysis of the plants have suggested that the feeding preference of this crayfish is given by the ease of handling small, short, bottom-dwelling plants (Chara and Lemna), and not by the plant chemistry. The same species of crayfish had a significant impact on both the aquatic macrophytes and macroinvertebrate community, even at a low biomass of less than 20 g m-2. Tests carried out under semi-natural conditions in plastic pools containing Potamogeton richardsonii, Myriophyllum exalbescens, Nuphar variegatum and Sparganium eurycarpum have shown that there was a major difference between the impact of the females and the males on the vegetation. While the authors described the effect of female crayfish as generally stimulatory, leading to an increase in the biomass of Myriophyllum and Potamogeton, and to an increase in Myriophyllum length, the plant growth was negatively impacted by male crayfish. The experiments have also shown the feeding selectivity of the crayfish: of the four macrophytes only Myriophyllum was consumed, while the crayfish just clipped the shoots of the other three plants which were left to float on the water surface. Female crayfish O. virilis were found to virtually eliminate snails from the surfaces of the macrophytes. Thus, the observed increase in the biomass of Potamogeton and Myriophyllum seemed to be related, at least in part, to the reduction in herbivorous snail abundance. Crayfish thus seems to indirectly alter the macroinvertebrates by changing the abundance and species composition of the aquatic macrophyte community.
Cronin (1998) followed the feeding choices of the Louisiana crayfish Procambarus clarkii, a commercially valuable species now cultivated in a number of other countries. When plants were offered as fresh tissue, Procambarus consumed large amounts of Chara, Ceratophyllum and Spirogyra; intermediate amounts of Typha; and little (if any) Potamogeton, Batrachospermum, Nuphar, Carex, or Iris. The three preferred species are finely branched or filamentous suggesting that plant structural properties are important determinant of Procambarus feeding decision. Cronin (1998) also tested crude extracts from plants: those from Ceratophyllum, Nuphar and Iris significantly reduced consumption, while extract from Typha significantly stimulated feeding. But according to Bolser et al. (1998), while lipophylic crude extract from Typha significantly deterred feeding of this species of crayfish, when offered it in an agar-based diet lacking structural defenses, the crayfish showed great preference for it. According to Cronin (1998) among plants without chemical defense mechanisms Procambarus will target those with high nitrogen content. Cronin concluded that generalist herbivores such as Procambarus, base their feeding decisions on multiple plant traits such as morphology, structure, chemical defenses, and nutritional value. Procambarus will eat only plants that they can handle, shred, and digest.
In experiments (Nystrom and Perez, 1998), crayfish Pacifastacus leniusculus had a strong impact on snail mortality. Their study suggested that crayfish can structure the abundance and size distribution of thin-shelled snails (Lymnaea stagnalis) through size-selective predation and reduction of macrophytes. Hanson et al. (1990) observed a decrease in the density of amphipods and the total biomass of non-molluscan invertebrates at male crayfish biomasses of 5 and 10 g m-2 and suggested that this may be directly related to crayfish predation or indirectly to reduction in macrophyte biomass and species richness. In laboratory experiments with O. rusticus and O. virilis Olsen et al. (1991) observed that O. rusticus damaged more snails than O. virilis, and O. rusticus had a greater weight-specific ingestion rate than O. virilis on snails but not macrophytes. Feeding beef liver to O. rusticus significantly reduced consumption and non-consumptive destruction of macrophyte Vallisneria (Lorman, 1980). Lodge and Lorman (1987) suggested that the presence of snails probably reduces macrophyte consumption by O. rusticus in the littoral zone of northern Wisconsin lakes. In Australia in experiments carried out by Whisson (1997), Vallisneria sp. was found to have a protective function for juvenile and adult crayfish marron (Cherax tenuimanus), reducing the predatory impact on them of silver perch (Bidyanus bidyanus). The author suggested that there could be a useful role for this plant in commercial marron production.
Changes in macrophyte communities due to both the direct and indirect effects of crayfish grazing may, in turn, affect fish communities, through the loss of the opportunity of some fish to hide from predators, change in or shortage of suitable spawning grounds, changes in or loss of food organisms such as periphyton, macroinvertebrates, zooplankton. In Wisconsin lakes, the introduction of O. virilis has been linked to detrimental changes in the fish community (Lodge et al., 1985). Carpenter and Lodge (1986) summarized the impacts of omnivorous crayfish on a lake ecosystem (Fig. 23). Rickett (1974) warned of the possibility of excessive vegetation elimination in lakes or ponds by uncontrolled crayfish populations. In northern Wisconsin lakes, the probably inadvertent introduction of the crayfish Orconectes rusticus and the subsequent increase in its numbers went in parallel with a decline in submersed macrophytes (Lorman, 1975). In Lake Metonga, the disappearance of macrophytes stopped the recruitment of the fish walleye (Stizostedion vitreum) (Lodge et al., 1985), and in some other northern Wisconsin lakes the indigenous crayfish Orconectes virilis and O. propinquus were displaced by the intruder (Olsen et al., 1991). Where the population of crayfish crashed, as in Lake Naivasha, Kenya, due most probably to overfishing of the introduced Procambarus clarkii, this was followed by the recovery of rooted aquatics (Harper et al., 1990).
Crowder et al. (1998) constructed a simplified generalized food web for submersed macrophyte system (Fig. 24). Boxes contain the major types of interactors. Lines ending in arrows indicate positive effects to the upper trophic level of energy flow up the web. Lines ending in closed circles indicate negative effects on the lower trophic level due to the level above. These may include the potential for predators to “control” prey densities. Dashed lines indicate potential interaction such as leaking of nutrients from macrophytes to epiphytes or indirect effects of cover on predator-prey interactions. Dotted lines indicate possible routes of nutrient recycling from consumers. The authors point out that individual systems will diverge from this diagramme, and many species shift from one role to another through ontogeny. In many systems, the herbivores include both mesograzers (primarily oriented toward epiphytic algae) and macrograzers (primarily oriented toward macrophytes). In systems dominated by macrograzers, the hypothetical effects of reduction of large predators would be an increase in macrograzers, which could lead directly to losses of macrophytes along with their associated epiphytes and fauna.
Fig. 23. Hypothesized impact of omnivorous crayfish on a lake ecosystem. (From Carpenter and Lodge, 1986).
Fig. 24. Generalized food web for submersed macrophytes, including crayfish. Boxes contain the major types of interactors. (From Crowder et al., 1998).