5.1. THE EMERGENT VEGETATION
5.2. THE SUBMERSED VEGETATION
"Where shallow, previously aquatic plant-dominated lakes are to be restored, (...) the problems may be likened to those of restoring a tropical forest after clear-felling ..." (Moss & Leah, 1982)
Eutrophication, i.e. the increased inputs of PO4-P and NO3-N is the main cause of the deterioration of the aquatic habitat. Therefore, the lowering of the nutrient levels seems to be the panacea for the problem. For reed (Phragmites australis) a direct influence of high dissolved NO3-N concentration on the condition was established (see Section 4.1.). However, the growth of submersed aquatic macrophytes appeared to be possible at quite different levels of nutrient concentrations. While under moderately high eutrophicated conditions even "nuisance" growth may become a problem under certain management regimes (water transport, irrigation, angling, swimming purposes).
The need of aquatic vegetation for good water quality and as a habitat for fish populations raises two questions:
i. if water plants are present: how do we keep them and how do we control them to maintain optimal water quality?
ii. if the water plants have disappeared, how do we restore the aquatic habitat with aquatic vegetation?
The following Section will describe a number of control measures for submersed water plants within the framework of the conceptual model, as given in Section 4. For reed and other emergent aquatic vegetation; this model is not applicable. The most important for the conservation of reed are the control of nutrient regime (especially the nitrate concentration), development of fence systems and replanting.
Fencing
In the Havel lakes fences are constructed since 1971 to prevent losses in the reed stands alongside the shoreline (Markstein & Sukopp, 1983). These measures are very expensive. Depending on the construction (logs, palisades, wooden boxes) the prices range between 110 and 340 DM per metre. It proved not to be effective, since the rate of decline of the proportion of the shoreline covered with reed stands did not decrease (Fig. 8). The construction of special swimming sites did not prevent further losses either. Experiments are going on with the underwater construction of wattle fences for protection. Klotzli (1971) also reports the use of wattle fences composed of willow twigs or plastic laths and plastic nets to keep floating litter less than 15-30 kg per metre shore line.
Recently Grosch (per. comm.) reported the end of the decline in emergent vegetation alongside the Havel lakes. The total costs of a number of restoration measurements (including dephosphatization) amounted 4000 DM per ha within lo years.
Fig. 8. Reed stands along the Berlin Havel (1962-'82) o: Proportion of the shoreline covered in the Upper Havel lakes; .: Proportion of the shoreline covered in the Lower Havel lakes; D: Both; --: Extrapolation (Markstein & Sukopp. 1983. with permission of the authors).
Intensive agricultural practise results in more cattle per ha and an early start of the grazing. In these cases. fences are needed to keep the cattle out of the zone of emergent aquatic plants.
Removal of old reed stems by mowing, burning and "raking"
There are no general applicable management techniques to stimulate reed growth which are effective in all situations. The effect depends on several factors:
(i) the probability of ground frost in spring;
(ii) the degree of infestation by herbivorous insects;
(iii) the fluctuations in water level at the reed stands. The degree of exposure to wind and wave action is important too. Reed bordering a deep lake has to be managed differently compared to reed stands in a swampy lowland area with shallow lakes.
Burning during winter or early spring positively influences the bud, and shoot formation and growth of the shoots. The bud development : determines the shoot growth and also the density later in the season.
Burning in springtime will increase the shoot density, but decreases the shoot biomass because the damaged shoots are replaced by more shoots from other, but smaller buds (Haslam. 1971).
The removal of litter causes more rapid growth, but makes the stand more vulnerable to ground-frost and the invasion of other emergent plants especially at non-flooded sites (Mook & van der Toorn. 1982; van der Toorn & Mook, 1982). The larvae of stem- and rhizome-boring insects also can cause considerable damage. In case of the latter burning is an effective measure, provided that it is done every winter and not incidentally (van der Toorn et al. 1983).
In the Netherlands a machine, especially designed for reed management along the shores of canals, is successful. After mowing the reed, the other vegetation and sol between the remaining reed stems and rhizomes is scraped away by a 1.5 m wide mechanical rake. This rake is mounted on a movable machine, which can be tended from the landside of the reed bed. The effect is a lowering of the ground level and breaking the rhizomes in the reed beds. In the next season, in the "raked" reed beds the stem diameter and height of reed increased more than after mowing or burning. This treatment is repeated after 5, sometimes 7 or 8 years (Baart & Ross, 1984).
However, "raking" is risky at shores with weakened reed stands. Klotzli (1971, 1973) advised not to mow the 3-4 m wide zone, bordering the open water of deep lakes. Weakened stands and stands endangered by algal mats also should not be mowed (for the peri-alpine lakes see also Schroder. 1979).
Cultivation of reed swamps for the removal of nutrients
Reed can remove considerable amounts of phosphorus and nitrogen due .to the activity of bacteria in the rhizosphere of the reed bed. In the Dummer (shallow, hypertrophic lake; open water area 1240 ha, West Germany) a reed swamp area (200 ha) has been constructed ;n the mouth .of the river Hunte to "trap" the nutrients from this eutrophicated stream. About 90% of the annual sedimentation and 5 -10 tonnes phosphorus will be trapped in this swamp (Ripl, 1984). However, harvesting during winter time of the above ground parts of reed may reduce the annual p and N loading of the water in a Dutch hypertrophy lake by 1%, which is very small in comparison with the cost (Loenen & Koridon, 1978).
Replanting
In the Havel area the planting of Typha angustifolia rhizomes behind palisades, the fencing of remaining reed and protected plantings of Acorus calamus, Carex acuta and T. angustifolia was successful. The replanting of Phragmites and Scirpus lacustris was problematic (Markstein & Sukopp, 1983). In the Netherlands a combination of "hard" structures by stones and loam and plantations of reed sods are successfully applied alongside lakes and canals (Acht & Sessink, 1982). These plantings are used to protect the shore because reed and several other emergent species are able to subdue the wave action: 2- 3 m wide stands of Phragmites but also of A. calamus or angustifolia can absorb 60-75% of the wave energy (Bonham, 1980). However, reed can resist only a limited amount of mechanical stress. Klotzli (1973) indicated 20 m as the minimum distance for passing motorboats with a maximum speed of 5 km/h.
The effects of three common management techniques used to control aquatic weeds are discussed within the framework of the conceptual model (see Section 4.3. and Fig. 9.).
(i) Mechanical and chemical control. Mechanical control of waterweeds is an expensive method. It may also stimulate the predominance of Elodea sp. and Potamogeton pectinatus and thus impoverish the aquatic macrophytic community. Chemical control is 50% cheaper (van Zon, 1977), but the side-effects are worse. Effective removal of all aquatic macrophytes may lead to phytoplankton dominance (Fig. 9). The contribution to the rapid decline of aquatic macrophytes by large-scale application of chemicals is discussed by Nicholson (1981), Johnstone (1982) and Jones & Winchell (1984). The implications of the application of herbicides are extensively reviewed by Brooker & Edwards (1975) and Hellawell & Bryan (1982), of whom the latter also give recommendations for use. Within the conceptual model these control measurements are risky because one will probably end up in a phase III situation (Fig. 6, 7 and 9).
(ii) Biological control by grass carp (Ctenopharyngdon idella Val.). Stocking with grass carp is often recommended as an acceptable solution (Fig. 9).This method is cheap as it represents 50% of the cost for chemical control and 20-30% of the cost inherent to mechanical control (van Zon. 1977). The grass carp can have a beneficial effect upon the diversity of the aquatic vegetation if stocked at densities under 250 kg/ha. Small grass carps do not eat parts of emergent plants. An increase in Hottonia palustris has been reported after 2 years in an experiment with 250 kg/ha grass carp in fish ponds (Provoost et al., 1984). Rough-leaved plants like Stratiotes aloides, floating-leaved plants (Nuphar lutea and Nymphea spp.) and plants with a strong taste like Polygonum hydropiper and Ranunculus spp., are not consumed by fingerlings (van Zon, 1977). Stocking over a period of three years in a small fishpond diminished the coverage of the vegetation and increased the number of macrophytic species from 9 to 20 (Provoost et al. 1984). Ahling & Jernelov (1971) reported the return of chara and improved dissolved oxygen concentrations in a small Swedish lake, after stocking with grass carp. Better growth, production and survival of other fishes have been repeatedly reported (van Zon, 1977; Pierce, 1983; Provoost et al, 1984). However, the presence of grazing macro-invertebrates on the remaining aquatic macrophytes is extraordinarily important to maintain clear water. If the grass carp stocking (>250 kg/ha) results in complete removal of the aquatic vegetation the herbivorous macro-invertebrates will also disappear. Bottom dwelling detritivores like chironomids may compensate the feeding conditions for other fish species (van der Zweerde, 1982. 1983; Provoost et al., 1984), but then the pathway to phytoplankton dominated waters (Phase III) becomes inevitable (Fig. 9).
Krzywosz et al., (1980) carefully documented negative effects of a long-lasting, large
stocking project in Dgal Wielki (95 ha, Mazurian Lakes, Poland). In 1966 about 6 kg/ha fingerlings were stocked in the lake. More and bigger fish (up to 6 kg per individual) were added between 1970 and 1977. The average grass carp biomass over 1966 to 1978 in the open water zone was 84 kg/ha. Between 1966 and 1974 the macrophytic production remained on the same level and no essential quantitative changes were observed until the warm summer of 1976. In that year the vegetation decreased 9 to 17 fold. Chara sp., Potamogeton natans, P. pectinatus, P. perfoliatus, Stratiotes aloides and Elodea canadensis disappeared. Only Fontinalis antipyretica, Nuphar lutea and small traces of Myriophyllum sp. remained. The avifauna was pushed out because of a reduction of the reed swamps by 66%. The native rudd (Scardinius erythrophthalmus) and tench (Tinca tinca) considerably decreased in number, while the condition of bream (Abramis brama) became poor. The weight increase of the grass carp also became lower than under natural conditions in the Amur. It is unclear whether other factors (like the decreased water volume in the lake during the warm summer of 1976) influenced the decrease of aquatic macrophytes, as was reported for other Mazurian lakes (Ozimek & Kowalczewski, 1984).
The direct effect of moderate stocking with grass carp and other non- native phytophageous fish on phytoplankton remains unclear (Provoost et al., 1984). Heavy stocking causes turbidity because of disturbance of the bottom substrate and faeces excretion, which will enhance the growth of bluegreen algae (Januszko, 1974). Heavy stocking (more than 250 kg/ha) of the common carp (Cyprinus carpio) also affects the vegetation (Crivelli, 1983; ten Winkel & Meulemans, 1984).
(iii) Shading. Dawson & Kern-Hansen (1979) give many helpful suggestions for weed management in streams. They plead for moderate shading of the stream water surface by careful management of the (emergent) bank vegetation. In the absence of such vegetation they recommend shading of small streams by an artificial canopy made of plastic mesh or blind cloth. However, shading by fast growing willow trees around a small lake may cause a strong decrease in submersed aquatic macrophytes (Best, 1982).
Summary: Control measures against "nuisance" growth are risky because an irreversible situation without any aquatic vegetation may be created. Beneficial effects of grass carp stocking can be expected with stocking densities below 250 kg/ha and strictly maintained regulations and management. Artificial shading also is a mean to control nuisance growth. If conservation interests are involved, these methods must be considered very carefully.
Replanting
Artificial recolonization of aquatic plants by replanting will often be interfered by waterfowl grazing. Therefore, young plantations should be protected by nettings (Moss & Leah, 1982); although in the normal situation healthy plant populations can sustain high grazing rates (Kiorboe, 1980). Crawford (1979, 1981) successfully inoculated Chara oogonia in artificial fish ponds. The water quality significantly improved in these ponds.
Furtherance of grazing on phytoplankton
Turbidity, caused by abundant phytoplankton growth, may be lessened by stimulating the grazing of zooplankton (see Section 2.5.4.). Timms J & Moss (1984) found during summer in one of the Norfolk Broads t rapidly growing small phytoplanktonic algae in sparse densities (<10 �g/l chlorophyll-a) and "normal" densities of fish. They showed that, the period of clear water coincided with the presence of a large stand of Nuphar lutea and plant-associated large-bodied cladocerans. The nutrient level of the water was very high (186 �g/l total-P and 6.5 mg/l total-N) and could support great phytoplankton growth and did so in spring and autumn. Thus, aquatic macrophytes may be important to maintain a good water quality in coexistence with fish at high nutrient concentrations. In the Norfolk Broads experiments are going on with bundles of alder tree (Alnus glutinosa) twigs. These submerged bundles harboured (in 1986) good populations of large cladocerans (Moss, pers. comm.).
The manipulation of young fish populations is an acknowledged management tool to decrease the predation on zooplanktonic grazers and thus reducing the phytoplankton. Ripl (1984) mentioned the stocking of 1-2 year old pike (Esox lucens) to reduce the abundant but badly growing population of zooplanktivorous roach (Rutilus rutilus) and bream (Abramis brama). This measure is part of the lake restoration project in the Dummer. The Dutch Organisation of Improvement of the Inland Fisheries (OVB) is trying to control small planktivorous fish by stocking them with artificially propagated pikeperch (Stizostedion lucioperca), pike (Esox lucius) and wels (Silurus glanis).
Manipulating the water level
Below 30 �E .cm-2 .sec-1 (i.e. photosynthetically active radiation in micro-Einstein per square cm per sec) no growth of aquatic macrophytes is possible, whereas between 30 and 400 �E .cm-2 .sec-1 the production of aquatic macrophytes is mainly influenced by the light factor (van Vierssen, 1985b). If the radiation on the bottom sediment is slightly above 30 �E .cm-2 .sec-1 the light penetration of shallow waters can be improved in spring by temporary lowering the water level. If there is sufficient light, the seedlings of aquatic macrophytes can grow more quickly during springtime. Hence the final seasonal product; on increases considerably, while the zone extends in which the macrophytes can develop (van Vierssen pers. comm.)
Sediment removal
Sediment removal is sometimes practised in small lakes:
1. to prevent sediment/water exchange of nutrients;
2. to remove (or reduce) nuisance plant growth;
3. to reduce the effect of toxicants; 4. to deepen the water body for fishing, boating and other recreational activities.
Deepening a lake, without increasing the transparency in the water column after the execution of the work, and without lowering the water level, will worsen the light conditions near the bottom and hamper the growth of submersed aquatic macrophytes. Therefore it is a risky management tool to reduce nuisance growth according to the conceptual model, because one may create a phase III situation.
Peterson (1982) reviewed 60 projects of sediment removal and five case histories. He does not give explicit recommendations for sediment removal to control plant growth because conclusive information on macrophytic regrowth is lacking. Sediment removal projects fail if only a limited part of the polluted sediment is removed and/or external loadings of nutrients and pollutants can not be controlled. The cost may range between 1 to 16 US$ per cubic metre; an average unit cost can not be calculated because the determining variables are numerous. Moss et al. (1986) estimated the cost of removing the sediment from a 63 ha large lake in the Norfolk Broads at least 1.5 million UK£ (3.4 US$ per square metre of the lake surface). Peterson (1982) stressed that every project should be preceded by a thorough study on the mass balance of nutrients and/or pollutants in the sediments and the water column.
In the Dummer lake (1240 ha) 800,000 tonnes of sediment (cost 7 US$ per cubic metre) have to be removed. To prevent the discharge of nutrient rich water from the intensively cultured surrounding arable land, the lake has been isolated from a part of the watershed (Anon., 1984). The total cost of sediment removal and other restoration measures is 10 million DM (0.4 US$ per square metre of the lake) and annual management cost is 400,000 DM (0.16 US$ per square metre).
A restoration project in Cockshoot Broad (3.3 ha; Norfolk Broads) started in 1982 by removing the surface sediments by pumps and isolating the lake from the River Bure (cost 80,000 UK£, 3.4 US$ per square metre). As a result the total phosphorus concentration decreased to 50-60 �g/l and the phytoplankton concentration to less than 15 �g chlorophyll-a/l within 2 years, while a diverse collection of aquatic macrophytes is developing. Another Broad (Alderfen) has also been isolated, but no sediment has been removed. A macrophytic community characteristic for phase II was re-established, and this mobilized much phosphorus from the sediments. After four years a large phytoplankton population appeared.
Dephosphatization of effluents
In Barton Broad (Norfolk Broads) the phytoplankton concentration was decreased by 50% after dephosphatization of effluents to the River Ant, discharging this lake. The total phosphorus concentration decreased from 350 �g/l to the designated 100 �g/l, but no recolonization of aquatic macrophytes occurred (Moss et al. 1985).
In the Dutch Loosdrecht Lakes the total phosphorus concentration ranged until 1984 between 60- 520 �g/l. A large restoration project then started. The discharge of water from the polluted River Vecht has been diverted. The lakes are flushed with dephosphatized water from the less polluted Amsterdam-Rhine Canal since 1984. The phosphorus absorption and release from the sediments and the inherent effects on the biota are to be studied (van Liere 1984, 1985; Best et al., 1984)
Landscaping
Many man-made lakes and ponds only serve for recreational purposes like windsurfing. swimming and catching sport fish. To meet these demands, usually uniformly shaped waters with uninteresting, deep water zones and rather steep profiles are made. Larger, well protected, bank zones of near-natural forms are necessary to create hatching areas for fish and amphibians and refugia for a greater variety of emergent plant species, macro-invertebrates, fish and amphibians. Akkermann (1985) presented a number of design possibilities and examples to improve existing objects or to construct new ones after digging sand and gravel.
Limiting recreational activities
Since 1969 a conservation law protects the aquatic ecosystem of the Havel Lakes area. This law prohibits treading the shore and the penetration of the reed stands by boats and sets a speed limit for motor boats (Sukopp & Kunick. 1969). In some peri-alpine lakes only a fixed number of motorboats are allowed (Klotzki 1973). Special spawning areas, where boats and other activities are strictly forbidden, have been created in the Havel lakes for protection of the fish population (Grosch. 1978).
In the Cruising and Remainder canals (England) more than 4000 mhy (boat movements per ha per year) damaged the submersed vegetation to an unacceptable extent for angling purposes. It was proposed to divert the boats from heavily used canals to small, hardly used canals. It was expected that boat traffic between 2000 and 4000 mhy will regulate the amount of vegetation within limits acceptable for sport fishing activities (Murphy & Eaton. 1981). The protection of the reed and other emergent vegetation against the negative effects of pleasure boat traffic is also important for the conservation of the submersed vegetation. Klotzli (1973) fixed 20 m as the minimum distance for passing motorboats with a maximum speed of 5 km/h. These findings may be useful elsewhere to assess maximum limits to human recreational activities where protection of the aquatic vegetation is necessary.
