NUMBER TWENTY NINE - OCTOBER 1998

The triggers that convert plant-dominance to algal-dominance fall into two groups (Table 2) - those that directly destroy the plant structure, and hence the buffer mechanisms dependent on it; and those that prevent the buffer mechanisms involving other organisms, like zooplankton, from operating

(i) Direct destruction of plants

The first group includes hand or mechanical clearance of the plants, sometimes carried out to create more open water for boating, swimming or angling and which can be overdone, and deliberate or accidental application of herbicides. At Zwemlust, a small lake used for organised swimming in the Netherlands, plant growth was so great in 1961 that herbicide was used to kill the plants, resulting in a rapid switch to dominance by blue-green algal blooms which persisted until deliberate attempts were made to restore the lake in the 1980s [148,150]

(a) Dependent on direct destruction of plants

Mechanical cutting

Boat damage

Herbicide use or accidental runoff

Heavy grazing by artificially high aggregations of native animals or of introduced species, including large invertebrates (crayfish), fish, birds and mammals

Raising of the water level to place plants at lower light intensities

(b) Dependent on interference with buffer mechanisms

Destruction of zooplankton activity by ingress of pesticides, or other organic or inorganic toxins

Increased salinity to more than about 5 parts per thousand

Reduction of piscivorous fish to zooplanktivorous fish ratio by deoxygenation in summer or under ice

Overfishing of large fish so that small size classes are favoured

Table 2 Forward switches which can convert plant-dominance to algal-dominance

Boat propellors can be equally damaging; in the canals of many Dutch cities, which are effectively small lakes because water replacement rates are low, parts of the system frequented by tourist boats are usually turbid, whilst backwaters, connected with the rest of the system, but away from the main boat routes, have attractive lily beds and other plants. Boats may have had similar effects in the slow-flowing rivers of the Norfolk Broadland, but it is unlikely that the boat activity has been sufficient to cause the switch in the greater expanses of the Broadland lakes. Lake Apopka, in Florida, provides a possible case where natural mechanical damage has caused a forward switch, following a hurricane whose violence severely damaged the plant beds which were unable to recover.

Vertebrates additional to man can also be destructive. The shallow lakes of many urban parks are frequented by large populations of ducks and geese, often including large Canada geese (Branta canadensis) which have been introduced and are thriving in many areas. Together these birds may graze the vegetation or trample it and result in switches to algal dominance. The characteristic state of most urban lakes is certainly turbid! A mammal, the coypu (Myocaster coypus), introduced from Argentina to Europe and Africa for fur-farming, may also act as a forward switch, though it mainly eats the underground parts of emergent reedswamp plants. The disturbance it causes in doing this, however, may have consequences for the submerged plants, though properly documented evidence is lacking.

There is a strong relationship between common carp (Cyprinus carpio) stocking and switches to algal dominance [24,32,38,54,93]. Common carp grow to large size and in areas where they are not native, including the UK and Australia, have, as adults, no predators. They forage in sediment, greatly disturbing the habitat by stirring up the mud, and feed directly on plants, often uproooting those they do not eat. Large bream (Abramis brama) have similar effects. In a survey of lakes in Sites of Special Scientific Interest in England, some 25% were found likely to have been damaged by carp introduction [35]. On a wider basis the proportion is likely to be higher (Box 4).

Fortunately common carp do not breed readily in NW Europe and need artificial hormone treatment in nurseries for their propagation. They are essentially warm-water fish, so their spread is controllable. There is some danger, however, that, with global warming they may breed more readily. Chinese grass carp (Ctenopharyngodon idella), were they to be freely released would be even more damaging [138], but their use (in weed clearance from canals and drainage ditches) is closely controlled.

A final example where an introduced species has caused loss of vegetation and switch to algal-dominated conditions may be Lake Naivasha in Kenya [65,66,67,68,136], where introduction, in the 1970s of the Louisiana red-spotted crayfish (Procambrus clarkii) (Fig 17) as a fishery resource has coincided with loss of submerged plants, which this omnivorous animal favours as its food. Data are few but the switch in the lake, against a background of increasing nutrient loading from agriculture and clearance of nutrient-retarding swamps from around the lake, has many of the characteristics of better-documented temperate examples.

Fig 17 In Lake Naivasha, Kenya, a switch from plant-dominance to algal-dominance may have been caused by the introduced Louisiana red spotted crayfish. Other mechanisms have probably also been involved, including severe overfishing and poaching which has removed most of the piscivorous fish.

(ii) Switches that destroy buffer mechanisms

The second set of switch mechanisms includes those that reduce the zooplankton grazing potential of the system by killing Daphnia. and probably other grazers. These include the leakage of pesticides [60,122,141], or other organic or inorganic toxins [83,84] to the system. Daphnids are particularly susceptible to organochlorines and some switches in lake state in the 1950s and 1960s may have been associated with these [141]. Recent use of pyrethroids in sheep dips has proved lethal to stream organisms and may have had ultimate effects in lakes.

There are instances of accidental contamination of lakes resulting in severe reduction of Daphnia populations and immediate rises in algal populations [141]. There is also evidence correlating changes in the zooplankton community detected in sediment cores and pesticide residues [141] and direct experimental evidence for some heavy metals and organic compounds [83,84]. Daphnids may indeed be lost from garden ponds following the application of waterproofing paint to concrete surfaces, and it is highly likely that heavy metal contamination of waters could lead to daphnid loss. There is enormous potential for this mechanism, as substances are spilled from accidents on highways or at factories, and may reach waterways but such consequences are always difficult to reconstruct in retrospect.

Increased salinity may also be a forward switch [4,71,81,100,110]. It may result from contamination of fresh or mildly brackish waters by seawater, due to pumped drainage of catchments near the sea or contamination of floodplain lakes by surge tides coming up river from estuaries. Daphnids will not tolerate more than about 3-5 parts per thousand salinity. At Hickling Broad in Norfolk, this appears to have been the main mechanism leading to loss of the daphnids and of algal dominance in the 1970s. This was a very interesting example, involving also the eutrophication of the Broad by roosting gulls, attracted by a nearby municipal waste tip, and fish kills caused by a brackish-water alga, Prymnesium parvum, favoured by the increased salinity.

Changes in the native fish community brought about by eutrophication may also be effective forward switches [29]. Large growths of aquatic plants may result in overnight deoxygenation in summer, or to production of organic rich sediments that readily deoxygenate the water under the winter ice of continental lakes. Piscivorous fish tend to be eliminated more readily than zooplanktivorous fish in such conditions, leading to greater zooplanktivore pressure the following year as young-of-the-year (those born that year) zooplanktivores are not removed by predators to the extent that they would normally be. This increased pressure on the zooplankton may be sufficient to cause a forward switch, but again evidence for such mechanisms remains largely circumstantial.