The efficacy and ecological impacts of herbicide and cutting regimes on the submerged plant communities of four British rivers. I. A comparison of management efficacies

Summary

The vegetation of many UK rivers is managed to reduce the damage to adjacent land which results from flooding. Field trials to compare the effects on aquatic weed communities of weed cutting and herbicide (diquat-alginate) management techniques were carried out in four U.K. rivers. These rivers varied in water quality but all supported macrophyte communities that included species of Ranunculus subgenus Batrachium (water crow-foots). The efficacies of the management techniques were quantitatively evaluated by plant biomass sampling and by the mapping of permanent transects.

Study sites: Four rivers in England and Scotland with populations of a crow-foot species known to be susceptible to diquatalginate were selected. The rivers had a range of water hardness i.e. calcium ion concentrations, to investigate possible effects of calcium on the herbicide. The efficacy of diquat-alginate in two highly calcareous southern limestone streams (Rivers Windrush and Coln) was compared with two less calcareous northern rivers (River Petteril and Mouse Water).

The River Windrush is a calcareous river of southern England (National Grid ref. SP 393058-SP 384067). Low species diversity and predominance of fennel pondweed Potamogeton pectinatus suggested organic enrichment. The study site had not been managed since the 1970s. Prior to this, aquatic vegetation had been regularly cut.

The River Coln (SP 146030-SP 147037) is a calcareous tributary of the River Thames. The study site is managed as a sport fishery by the manual removal of specific weed beds.

The River Petteril is a moderately soft-water stream (NY 498349-NY 492367) in northern England. It did not receive regular aquatic plant management.

The Mouse Water (NS 941491-NS 936496) is a tributary of the River Clyde in southern Scotland. This small river was organically enriched from adjacent farmland.

Experimental design: Herbicide, cutting and uncut treatments were compared (cutting was not applied to the Mouse Water). On the River Coln and Mouse Water two replicates of each treatment were divided into two blocks with a buffer zone of at least 250 m in between. Data from previous work indicated that the herbicide was unlikely to have any effect beyond 200 m downstream of application. Additional cut and untreated replicates were sampled in the Windrush and Petteril.

Cutting and removal: In early July 1984, weed cutting in the Petteril was carried out using sickles to cut crow-foot and flick it onto the bank. In mid-June, three 100-m sections of the Windrush were cut using a chain-scythe, with cut material removed by raking. Weed beds cut in the Coln throughout the summer by manual scything. Because of requirements of the fly-fishery, some cutting of control sections could not be avoided, and cutting was not uniformly applied to the cut sections. However, all plants in the permanent transects were cut or left unmanaged in appropriate sections.

Herbicide application: Herbicide was applied early in the summer, when vegetation was actively growing but had not yet reached the water surface (late May in Rivers Windrush and Petteril, mid- to late June in the Coln and Mouse Water). It was applied to 200-m sections in all rivers except the Windrush (where 100-m lengths were treated). Only the middle third of each section was subsequently surveyed, to minimize edge effects.

The commercial formulation of diquat-alginate was applied to the water surface (using a
modified knapsack sprayer with a nozzle of 2 mm orifice diameter, in accordance with the manufacturer's instructions to give a nominal concentration of 1 mg diquat/1). At regular intervals after spraying, 1l water samples were collected downstream of the area to analyse for diquat residues which were quantified using u.v. spectrophotometry.

Macrophyte surveys: Aquatic macrophytes were surveyed just before, and at intervals ranging between 8 and 129 days after, treatment. Two biomass samples were taken per replicate section using a Lambourn sampler (0.05 m² sampling area). Samples were sorted into species and dried. Permanent transects, not destructively sampled, were established in each replicate. The vegetation was mapped and the mapping data were used to calculate the percentage frequency of each species or substrate in a transect.

Physical characteristics of the rivers: The channel width of each section was measured at each sampling time on the downstream edge of the transect. Depth and water velocity were recorded at 3-5 regular intervals across this transect. Light extinction coefficients were calculated from readings of photosynthetically active radiation (PAR) taken with a PAR meter placed just below the water surface and from a known depth.

Water chemistry: The following were recorded: temperature, pH and dissolved oxygen at the downstream edge of each section at each sampling time. Total water hardness, concentrations of calcium and nitrate ions, and reactive phosphorus were estimated in the laboratory.

Cutting: The cutting treatments in two of the rivers removed most crow-foot from the channel centres. In the Petteril, the cut in July was sufficiently late to prevent regrowth, but in the Windrush a significant regrowth was evident 6 weeks after the June cut.

Herbicide: Diquat-alginate was very effective in crow-foot removal in the shallow, swiftly flowing and moderately calcareous (60-80 mg/l) River Petteril and Mouse Water. Crow-foot plants were removed or damaged for considerable distances downstream of the point of herbicide application. Less susceptible species, such as filamentous algae and broad-leaved pondweed Potamogeton natans, showed herbicide damage only in the treated areas. The maximum concentrations of diquat residues detected exceeded 1 mg 1l for at least 40 min.

The herbicide appeared ineffective in the Windrush, but in the Coln, of similar size and calcium content (110 and 120 mg Ca/l, respectively), a large proportion of the crow-foot was removed. Differences in the plant species composition, lengths treated and water turbidity may have led to these differences. Adsorption and inactivation of diquat ions by suspended clay particles may have been responsible for the low concentrations and short persistence of diquat in the Windrush.


Note: If using or referring to this published study, please read and quote the original paper, this can be viewed at:

http://links.jstor.org/sici?sici=0021-8901%28199008%2927%3A2%3C520%3ATEAEIO%3E2.0.CO%3B2-N