1 Hydroécol. Appl. (1993) Tome 5 Vol 1, pp. 77-96

l Plant management as an integrated part of l 's aquatic resources

Gestion des herbiers en tant qu'élément des ressources aquatiques d'Irlande

Joseph M. Caff rey* Central Fisheries Board, Mobhi Boreen Glasnevin, Dublin 9. Ireland.

Abstract. - Aquatic plants are fundamental to the structure and functioning of many freshwater habitats. In undisturbed aquatic habitats plants serve a multitude of important functions including the provision of substrates for colonisation by periphyton and macro- invertebrates, a direct and indirect food source for macro-invertebrates, fish and wildfowl, spawning substrates for coarse fish species, shelter for young and adult fish oxygen as a by-product of photosynthesis. In disturbed habitats, however, the natural balance among aquatic plants is altered and an excessive, often monospecific, vegetation commonly develops. This can adversely affect the beneficial use of the watercourse by diminishing its amenity and recreational potential, exacerbating water supply and flooding problems and resulting in excessive siltation in river or canal habitats. This paper outlines the nature and extent of the more important weed problems in lrish aquatic situations and describes a range of control strategies that may be adopted to rehabilitate these habitats and to enable their exploitation by a diversity of user groups. In certain instances paucity of aquatic vegetation presents problems, including bankside destabilisation and inchannel siltation. Reed transplantation trials have been conducted in denuded lrish canals in an effort to rehabilitate these habitats and the results from these are presented. Weed control procedures generally embrace four broad categories. These are mechani- cal, chemical, environmental and biological control. Of these only biological control is not widely adopted in Ireland. The effects that these procedures have on specific weeds were quantitatively investigated at a number of aquatic situations throughout the country. Results from experiments among a diversity of aquatic plant species and involving the use of mechanical weed cuîting apparatus show that, unless the plant is cut to a depth at which light penetration is significantly diminished, rapid regrowth will occur and higher standing crops than might otherwise be expected may be recorded. Experimentation with a small number of herbicides has revealed that, when used with discrimination, these provide effective weed control, with little or no detrimental effect to the aquatic ecosystem. Results from trials with dichlobenil and glyphosate on aquatic and riparian species are presented. The effect that shading has on macrophyte biomass and the effect that rotted barley straw has on algal growths in watercourses is also described.

Key words. - Aquatic plants, management, rehabilitation, weed control, herbicides. 78 J.M. Caffrey

Résumé. - Les plantes aquatiques sont des éléments fondamentaux pour la structure et le fonctionnement de beaucoup d'habitats d'eau douce. Dans les habitats aquatiques non perturbés, les plantes ont de nombreuses et importantes fonctions, incluant la four- niture : - de substrats pour l'implantation du périphyton et macroinvertébrés, - d'une source directe et indirecte de nourriture pour les macroinvertébrés, les poissons, les oiseaux aquatiques, - de frayères pour certaines espèces de poissons, - d'abri pour les alevins et les poissons adultes, - d'oxygène en tant que sous-produit de la photosynthèse. Dans les habitats perturbés, pourtant, l'équilibre naturel entre les plantes aquatiques est altéré et il se développe communément de façon trop abondante, une végétation aqua- tique monospécifique. Ce phénomène peut affecter l'utilisation du cours d'eau pour les loisirs et la fourniture d'eau potable; il peut ainsi entraîner des problèmes d'écoulement et favoriser la sédimentation. Cette note met en évidence la nature et l'ampleur des problèmes d'herbiers aquatiques les plus importants d'Irlande et décrit une série de techniques de contrôle qui peuvent être adoptées pour réhabiliter les habitats concernés et permettre un usage à buts mul- tiples. Dans certains exemples, la pauvreté en végétation aquatique présente des problèmes, incluant la déstabilisation des rives et une sédimentation du chenal. Des essais de trans- plantation de roseaux ont été menés dans des canaux irlandais dénudés afin de réha- biliter ces habitats; les résultats sont présentés ici. Le mode de contrôle des herbiers se présente généralement de quatre façons: méca- nique, chimique, environnementale et biologique. En Irlande, seul le contrôle biologique n'est pas largement en vigueur. Les effets de ces modes de contrôles ont été quanti- tativement évalués dans divers types de situations réparties dans tout le pays. Les résultats de ces expériences sur une grande variété d'espèces de plantes aquatiques et à l'aide d'une faucheuse montrent que si la plante n'est pas coupée à une profondeur correspondant a une luminosité très faible, la repousse est rapide et atteint une hauteur plus importante que celle à laquelle on aurait pu s'attendre. Les expériences avec un petit nombre d'herbicides ont révélé que lorsque cette technique est utilisée avec dis- cernement. le contrôle de la végétation est effectif, avec peu ou pas d'effets néfastes pour l'écosystème aquatique; les résultats des essais avec le dichlobenil et le glyphosate, sur les espèces rivulaires sont présentés ici. L'effet de l'ombre sur la biomasse des macrophytes et l'impact de la paille d'orge pourrie sur la croissance algale dans les cours d'eau est également décrit.

Mots-clés. - Plantes aquatiques, aménagement, contrôle de la végétation, herbicides.

1 INTRODUCTION commercial and recreational activities. This trend is likely to continue as com- In recent years there has been mercial and industrial water users increased demand for clean, well- demand water of better quality and as management waterways to service a the use of waterways for recreational wide range of domestic, agricultural, pursuit increases. Aquatic plant management in lreland 79

One of the principal problems in revegetating and stabilising facing those charged with responsi- denuded canal banks using trans- bility for the maintenance of clean and planted reed species is also de- functional watercourses is that of scribed. obstructive or unsighlty aquatic plant growth. In natural, undisturbed aquatic habitats plant species intergrade and 2 PRINCIPAL WEED-RELATED integrate, ~roducing com~lex and PROBLEMS IN IRISH AQUATIC well-balanced communities. These HABITATS l plants and communities hold a central position in the network of ecological relations between nutrients, plankton, Serious weed problems were not a macroinvertebrates, fish and avifauna feature of Irish watercourses prior to (de Nie 1987). The many functions of the 1970's (Caffrey 1991a, in press). aquatic plants in aquatic systems Since that time the effects of Man's are described in Marshall and interference with water and water- Westlake (1978) and Caffrey courses has altered the finely-tuned equilibrium that existed among plant (1991a). Where this natural balance is disturbed, however, conditions communities and created conditions favouring the establishment and that favour the spread of aggressive proliferation of invasive plant taxa species or groups. These have, in commonly results. Among the an- turn, impacted upon Man's desired thropogenous factors that are most use of the affected waterways. In often implicated in the explosive some circumstances this interference development of aquatic plants are has resulted in the loss or elimination of plants, with consequent adverse pollution, cultural eutrophication, abstraction, impoundment and impacts on species diversity, fisheries drainage. In this condition the enhancement and bankside stabilisa- aquatic vegetation no longer com- tion. plements the functionality of the In Irish rivers the most serious in- watercourse but now serves to Stream problems are presented by depress biotic diversity and seriously Scirpus lacustris (Common Clubrush), detract from the aesthetics and use- Potamogeton pectinatus (Fennel fulness of the system. As such, the Pondweed), broad-leaved Potamoge- aquatic plants have now become ton spp. (Pondweeds) and filamen- "weeds" (Navas 1991). tous algae. The enrichment of many of Our lotic systems from agricultural The aim of this paper is to de- and municipal sources has made con- scribe the functional problems that ditions for the proliferation of pollution- weed species can illicit in a variety tolerant species, such as P pectinatus of aquatic habitats and to show how and filamentous algae, more many of these problems may be favourable (Newbold and Palmer ameliorated. The success achieved 1979 ; Caffrey 1985, 1986a). The ef- 80 J.M. Caffrey fects of arterial drainage on many caused widespread concern among river catchments have served to water managers, amenity groups and create trapezoidal channels with al- conservationists. The reasons for the tered flow, depth and substrate rise to prominence of this plant in the characteristics. These physical fea- past decade, cognisant of the fact that tures are known to have a primary in- it has been in the country for many fluence in the distribution and growth years, is unknown. This plant repre- of macrophytes in rivers (Caffrey sents a serious human health hazard, 1990b). This, in combination with in- impedes access to rivers, overgrows creased sediment and water nutrient and shades-out indigenous her- availability, creates conditions suitable baceous plants and can cause serious for the spread of submerged and bankside erosion problems (Caffrey in emergent Scirpus lacustris and of press). broad-level Potamogeton spp.. In lrish lakes and ponds, planktonic The proliferation of these species and filamentous alqae-. present serious and grouPs and the ConSequent problems, by posing threats to recrea- restrictions of water flow, increased in- tien, water treatment and supply and, channel siltation, flood incidents and indirectly, to a wide range of lentic hazards, reductions in biotic diversity, biotic communities. In small and fouling of treatment plants, and other ponds (< 10 hectares) problems effects, have been a cause for serious posed by the encroachment of reed concern among water and drainage fringes or the occupation of the water engineers and others responsible for surface by free-floating Lemna spp. or the management of watercourses. floating-leaved lilies (normally Nuphar The fact that the problems are in- lutea) are commonplace. The loss of creasing in lrish watercourses tourist revenue to local communities highlights the urgent requirements for as a consequence of these weed judicious control practices. growths is often considerable and In some rivers, notably those slow- highlights the necessity for urgent re- flowing watercourses that receive habilitation. drainage water from bogs under Ireland supports an extensive net- development for milled peat produc- work of man-made canals, used for tion, siltation has the effect of at- navigation, angling, swim- tenuating light and eliminating the ming, nature study and other water- submerged flora (Caffre~986b) This based pursuitS. The facf that the effect is reversible. however, and a canais are shallow, slow-flowing, submerged flora will recolonise within possess clear and nutrient-rich water two to three years of the resolution of and a deep, muddy substrate means the siltation problem. that practically al1 the conditions re- In recent years, the unprecedented quired for optimum plant growth are spread of Heracleum mantegazzi- fulfilled (Caffrey 1991b). Weed man- anum, a riparian plant species, has agement practices operated in the Aquatic plant management in lreland 8 1 past concentrated on the use of one spraying equipment, funds available herbicide, with the result that suscep- for the project, the desirability of total tible plant species were eliminated for or partial weed removal, among most of the growing season, while others. Thus, in watercourses used tolerant species were permitted to es- more or less exclusively for the supply tablish and spread (Caffrey and of potable water, any method except Monahan 1991). This seriously im- chemical control should be accep- pacted the recreational fishery (Caf- table. On the other hand, where a frey in press) and other amenity channel is used principally for the pursuits along the canals. More in- speedy transportation of water from formed weed management practices land to sea, and is little used for other operated in recent years have purposes, the use of approved, broad- redressed some of the problems and spectrum herbicides might be most a special research focus is being appropriate. In most situations, how- directed at outstanding weed prob- ever, the use of a sensible combina- lems. tion of al1 control methods available provides the most satisfactory and cost-effective solution to invasive weed problems (Caffrey in press). 3 WEED MANAGEMENT This integrated control procedure is STRATEGIES widely adopted internationally (Mur- phy et al., 1987; Willis 1990; De Steno 1992; Caffrey in prep.) and has met Where the beneficial use of a water- with considerable long-term success. course is interfered with, whether through the presence of excess A more detailed account of the in- vegetation or the total lack of an dividual methods and procedures that aquatic flora, some form of habitat re- constitute the four broad weed control habilitation will be required. This may categories, and the relative merits and be achieved, in cases where vegeta- demerits of each, may be found in tion overgrows the watercourse, Seagrave (1988), Barrett et al. (1990), through the judicious employment of Schmidt (1990) and Caffrey (1 991 a, in methods embraced within the follow- press). The results from trials con- ing broad categories : mechanical, ducted on problem plant species chemical, biological or environmental under Irish conditions, using selected control. weed control procedures, will be pre- sented in Section 5. Decisions relating to the appro- priateness of these methods in any In aquatic habitats where a paucity one instance depend on a variety of of inchannel or aquaticlriparian factors. These include the nature and vegetation presents difficulties for extent of the weed problem, the use faunal and floral communities, in ad- to which the water or watercourse is dition to exposing the banksides or to be put, accessibility for cutting or shorelines to destabilising and 82 J.M. Caffrey erosional effects, urgent remedial emergent or floating-leaved habit measures are required. It is rarely (Caffrey 1990b). The adoption of the sufficient to await natural plant re- latter form enables the plant to colonisation and the planting of seeds colonise, establish viable populations along the eroding margins merely rep- and alter the hydraulic regime in mod- resents a waste of time and money. erate to fast-flowing rivers. Having es- The transplantation of living in- tablished appreciable floating- leaved digenous plants, preferably species stands, the consequent reduction of with spreading rhizomes and a capac- flow velocities within and immediately ity to rapidly consolidate soil, presents downstream of the plant stand en- a satisfactory solution. Results from courages suspended solids deposi- such a transplanting experiment are tion. The reduced flow and the summarised in Section 4. formation of silt/sand shoals permits the establishment of semi-emergent forms of S. lacustris, which further ac- celerate sedimentation. Dense and 4 PRACTICAL WEED spreading stands of emergent S. la- MANAGEMENT IN IRELAND - custris may now occupy the full chan- CASE STUDIES ne1 width and seriously impede any beneficial use of the watercourse. ln In this section the problems for en- fact, the presence of these obstructive gineers, water managers, water- vegetation stands represent a signifi- based recreational enthusiasts and cant flood hasard in any affected wa- conservationists that result from the tercourse. over- or under-representation of Long sections of the , aquatic and riparian plants are de- and other Irish rivers, are totally over- scribed. The methods selected for grown with S. lacustris. Attempts by remedying the problems are set-out Office of Public Works (O.P.W.) en- and the quantitative results from trials gineers to control this vegetation are presented. using weed-cutting boats met with little success, for two reasons. Firstly, the vegetation was so dense that 4.1 Control of Emergent weed cutting boats were unable to Monocotyledonous Weeds make progress within the weeded channel and, secondly, the cut plants The principal nuisance plant species regrew rapidly and often presented in- in this category in lreland are Scirpus channel problems within the same lacustris, Phragmites australis and growing season. Glyceria maxima, although the former species is by far the most trouble- In August 1991, at the request of some. Scirpus lacustris is heterophyl- O.P.W., an experimental plot measur- lous and, depending on the flow ing circa 20 m, and including the full regime in the channel, may adopt an channel width (30 m), was treated Aquatic plant management in lreland 83 with glyphosate at the rate of 51 ha-'. Further control, when necessary, may Glyphosate is a broad-spectrum, foliar be applied using mechanical or en- herbicide which is cleared for use in vironmental methods. The adoption of or near watercourses by the Ministry these integrated control strategies of Agriculture, Fisheries and Food should minimise problems with S. la- (MAFF) in the U.K. The chemical has custris in rivers for the foreseeable fu- a very low toxicity (Bronstad and ture. Friestad 1985, Goldsborough and Beck 1989) and is rated as practically non-toxic, the lowest classification 4.2 Reed Transplantation Trials available, by the World Health Or- ganisation. Following a major breach on a section of navigable canal in 1989, the af- At the time of treatment 284 shoots fected channel was reconstructed and per square metre occupied the chan- re-opened for navigation in March, nel. The following summer the treated 1990. By 1991, through the action of plot was practically clear of S. la- wave-wash, the clay used to line the custris (table l), except for three small canal bed and banksides had begun stands that had been missed during to erode. In order to expedite the nat- the spraying operation. Work con- ural bankside stabilisation process, it ducted at the Aquatic Weed Research was decided to "seed" the exposed Unit in the U.K. (P. Barrett pers. banks with roots and rhizomes from comm.) and at canal sites in lreland established emergent monocotyle- (Caffrey and Monahan 1991) sug- donous plant colonies. The species gests that three or even four years ef- selected for this purpose were Glyc- fective reed control may be achieved eria maxima, Scirpus lacustris and with one application of glyphosate. Phragmites australis. All three species Large-scale treatment of S. la- have intricate and consolidating root custris-infested rivers with glyphosate and rhizome structures and are in- should re-open these channels for digenous to the Irish canal habitat. recreational and flood relief purposes. Transplanting operations were con- ducted in october, 1991, and a total of 5 km of channel was revegetated. Table 1. - Mean number of Scirpus lacus- Before the operation began planting tr~s (Cornrnon Club-rush) shoots per bays along the channel, roughly at square meter in sections of the River Boyne at tirne of treatrnent with glyphosate water level, were prepared. These (1991) and one year post-treatrnent measured circa 3 m in length and (1992). were separated from each other by 1 m. When site preparations had been Year completed clumps of root and rhizome were planted and firmed-in with top- 1991 soil. Each species was allocated al- 1992 - ternate 100-m long sections. 84 J.M. Caffrey

In September, 1992, the percent- same growing season if unobstructed age survival among the transplanted navigation or water movement is to be species was assessed. The survival maintained. This rapid rate of vegeta- among G. maxima and S. lacustris tion regrowth normally occurs where plants was high (table 2) and most a shallow cut is executed. During plants had flowered and set seed. these operations the canopy vegeta- More importantly, from a bankside sta- tion, to a depth of between 0.5 and bilisation viewpoint, most of the in- 1 m, is cut. This removes the imme- dividual plant clumps had extended diate obstruction but permits light their lateral range and sorne had penetration to the sections of stem, coalesced with adjacent clumps. The cornmonly between 0.5 and 0.75 m root masses were well developed and tall, that remain in the water column no evidence of soi1 erosion in the following the cut. The stimulatory vicinity of the plant stands was ob- effect of the cut, in addition to the in- served. creased availability of photosyntheti- The percentage survival with cally active radiation, results in rapid Phragmites australis was significantly vegetation regrowth (figure 1). less (table 2) and, even where the Where a deep cut is applied, close transplanted clumps survived, only to the substratum and leaving only few individual shoots were recorded. short (10 to 20 cm) lengths of stem There was no lateral expansion within in the water column, vegetation re- the planted area and none of the growth is generally less rapide and plants flowered. The reason for the obstructive stands rarely develop poor survival is unknown but may re- within the same growing season late to the position, relative to water (figure 1). This probably reflects the level, at which the transplanted roots low level of light penetration to the were buried. Further experimentation depth at which the cut was applied. with Phragmites will continue in the The fact that only one cut per season autumn of 1992 and 1993. is required reduces the level of dis- turbance to the habitat and its biotic communities while also significantly 4.3 Mechanical Cutting reducing labour costs.

Long sections of weeded canal and Table 2. - Percentage survival of emergent river in lreland are annually cut using monocotyledonous plant species trans- planted along denuded sections of the boat-mounted apparatus. Experience between September, 1991 with this control procedure has de- and 1992. monstrated that regrowth among cut plants is often rapid and may exceed Plant Species % Su~ival! IS.E. that of uncut plants growing in the in 1992 - same channel (Caffrey 1990 a,b, Glyceria maxima 89.28 z 16.1 1991a, in press). In many situations Scirpus lacustris 81.15 - 11.1 Phragmites australis : 12.82 2 4.8 it is necessary to apply two cut in the -. Aquatic plant management in lreland

Shallow Cut

Date

Figure 1. - The effect of shallow and deep mechanical cutting operations on the biomass of sub- merged vegetation in sections of the Grand Canal in 1991. 86 J.M. Caffrey

4.4 Shading 4.5 Control of Potamogeton pectinatus (Fennel Pondweed) in Bankside shading is the most com- Irish Rivers monly adopted environmental tech- nique used for the control of Potamogeton pectinatus (Fennel submerged plants in river habitats Pondweed) is a submerged macro- (Dawson and Kern-Hansen 1979; phyte that grows luxuriantly in organi- Jorga et al., 1982; Madsen and cally enriched habitats (Haslam 1978; Adams 1989; Barrett et al., 1990). The Caffrey 1985, 1986) and is rapidly success of this control procedure is spreading through many of Ireland's based on the fact that shading by tall main river systems (Caff rey 1990a, b). banks or overhanging trees can re- Once established at a site, P pecti- duce the level of incident irradiance natus grows vigorously and rapidly by 35 to 95%, depending on the produces dense vegetation stands. height of the banks or the aspect and This aggressive coloniser competi- leaf development of the trees (Owens tively excludes most other plant spe- and Edwards 1961). Under dense tree cies and soon establishes a virtual cover, summer irradiance may be re- monoculture. Because the plant in duced by 95% (Westlake 1975) and Irish rivers is perennial and maintains few submerged plants are able to es- a significant vegetation presence over tablish. the winter period, little opportunity for recolonisation by indigineous plant The effect that dense bankside species is afforded. The presence of shading along sections of the Royal dense and obstructive stands of P. Canal has on the growth of Myriophyl- pectinatus impairs the beneficial use lum spp., the most prolific and trouble- of infested watercourses and can some species in this system, is result in serious ecological, hydrologi- presented in figure 2. In shaded sec- cal and amenity resource problems tions of canal practically no vegetation (Caffrey 1991a). was recorded throughout 1992. In an adjacent unshaded section, however, Figure 3 shows the normal growth dense and obstructive stands of My- pattern for P. pectinatus in the lower riophyllum spp. were present for most . Cognisant of the fact of the growing season. While this that a plant biomass exceeding 300 g result is favourable from a strictly dry weight m-2 will impair the full weed management viewpoint, the beneficial use of a watercourse (West- virtual elimination of aquatic plants lake 1981; Kern-Hansen and Holm from amenity watercourses is not rec- 1982), it is clear that serious vegeta- ommended and can have serious ad- tion problems are experienced in this verse impacts on associated faunal watercourse over most of the growing communities (Caffrey in press). For season. Treatment of designated sites this reason, partial rather than total with dichlobenil in early-May evoked shade is preferred. no response among the treated plants Aquatic plant management in lreland 87

Jun Jul Sept

Month

Figure 2. - Effect of bankside shading on the growth of Myriophyllum spp. (Milfoil) on the in 1992. 88 J.M. Caffrey

700 Dichiobenil Trial 600 -9 E 500 Dichlobenil s i? i? 400 73 300 *a'n g 200 .- 100

O Mar Jun Jui

700

600

500

400

300

200

100

O Mar Apr Jun Jul Dec Month l

Figure 3. - The normal growth pattern and the effect of dichlobenil treatment and mechanical cutting on the biomass of Potamogeton pectinatus (Fennel pondweed) in the Lower River Shannon in 1991. Aquatic plant mar lagement in lreland 89

(figure 3). This may reflect the fact The successful establishment of the that dichlobenil-containing granules plant in this habitat highlights the im- were swept downstream with the flow portance of water-mediated seed and did not reach the substratum, dispersal. The spread is a cause for from which the chemical is absorbed concern because of the serious by root uptake. However, care was human health hazard that the plant taken to ensure that flow velocity at represent and because of the ecologi- a number of the treated sites was min- cal and amenity disruption that its imal (< 10 sec-'). This result suggests widespread establishment can have. that I? pectinatus, under Irish condi- Heracleum mantegazzianum is a tions, may be resistant to the activity perennial plant that takes three or four of dichlobenil. Further work with years to mature. Seedlings first ap- dichlobenil, and other aquatic herbi- pear in February each year. These cides, will continue at infested river produce immature plants that may sites. achieve a height of 0.4 m in their first In late-June, 1991, cuts were ap- year of growth. In September or Oc- plied to a number of /? pectinatus - tober this foliage dies back. Early re- infested sections on the lower River growth in February from tap roots is Shannon. These cutting operations ofien vigorous in the plants second reduced the biomass to circa 50 g m-* and third years. If the plant is to ma- and permitted unobstructed angling in ture, flower stalks begin to elongate the channel. By early-August, how- in May. Peak flowering occurs in late- ever, a significant vegetation had re- June or early-July and seeds are dis- established and again interfered with persed by the end of September. In the beneficial use of the channel the year of flowering the plant can (figure 3). Work in the coming sea- achieve a height of 4 m (figure 4) and sons will focus on the effects that up to 100,000 seeds per plant may early season and late season cuts be produced. Thereafter, the plant might have one the subsequent dies. If the plant is not to mature in growth of FI pectinatus and efforts to that year, it may achieve a height of identify critical phases in the plant's up to 1.5 m before dying back. life cycle, when weed control pro- cedures might prove more effective, In trials where maturing H. mante- will continue. gazzianum plants were cut, to ground level, rapid vegetation regrowth fol- lowed (figure 4). In fact, cutting in May l 4.6 Control of Heracleum induced early flowering and seed pro- mantegazzianum (Giant duction among the cut plants. Where Hogweed) along River Corridors flower andlor seed heads were removed from mature plants, they Heracleum mantegazzianum is an in- either established new seeding struc- vasive weed species that is spreading tures later in the season or postponed rapidly along river corridors in Ireland. seeding until the following year J.M. Caffrey

P1owaing l seed production 350 i

" Feb Mar APr May Jun Jul Aug

350 - 300 -O Aowering 1 seed production 250 EM g 200

C) C 150 acd 2 'O0 5 50

O Feb Mar May Jun Jul

Month Figure 4. - The normal growth pattern and the effect of cutting and glyphosate treatments on the normal adult growth of Heracleum mantegazzianum (Giant Hogweed) along the River Mulkear in 1992. Aquatic plant mar jagement in lreland 9 1

(Caffrey in press). Application of gly- of filamentous green algae, primarily phosate at the recommended dose Cladophora glomerata, impeded rate (51 ha-'), on the other hand, navigation, obstructed angling, inter- achieved a 100% plant kill. At these fered with the functioning of lock gates sites no plants matured and no seeds and detracted from the aesthetics of were produced (figure 4). Because the this unique habitat. seeds of this plant are capable of lying In October, 1990, loose bales of dormant in the soi1 for up to 15 years, barley straw were submerged along and because they can be recruited the margins an algal-infested section from untreated sections of river bank of canal. Bales were spaced at ap- located upstream, maintenance proximately 50 m intervals along the programmes must be implemented trial site. The biomass of filamentous and conducted over a number of algae at this and at adjacent untreated years if the plant is to be effectively sites was determined at this time controlled. If the plant is to be elimi- (figure 5). Replacement bales were nated from a river, sites of infestation added to the trial section at roughly throughout the channel must be tar- 6-month intervals through 1991 and geted and included in any main- 1992 and biomass determinations tenance schedule. were made regularly during this period. 4.7 Algal Control using Barley Straw The results, presented in figure 5, show that the biomass of algae in the Excessive growths of filamentous untreated site broadly follow a cyclical algae in lakes, rivers and canals reg- pattern, with peak crops produced ularly present problems for water- between July and September and low based leisure activity, as well as for biomass values recorded in February industrial and municipal users. In re- and March. By contrast, in the section cent years a considerable research ef- containing the straw, the biomass of fort has been focused on the ability algae decreased from March to July of decomposing barley straw to inhibit 1991 and, thereafter, no algae were algal growth (Welsh et al., 1990; Gib- recorded (figure 5). This has greatly son et al., 1990). While much remains enhanced the amenity potential of this to be learned regarding the mode of channel and suggests that this weed action of straw on algae, research has control procedure should have much shown that the primary effect is due wider application in algal-infested to a factor generated during the watercourses throughout the country. decomposition of the straw which in- hibits the growth of algae (Gibson et al., 1990). The identity of the factor 5 CONCLUSIONS remains unknown. Work in lreland on this subject was concentrated on In recent years aquatic weeds have canal sections where dense carpets presented some of the most serious J.M. Caffrey

UO - 2- 1w r so -0e -U) 3 60 j .-a 40

m

O Mar Jul Frb Jul

Figure 5. - Biomass of filamentous green algae in sections of the Royal Canal that were treated (dark) and untreated (grey) with barley straw between 1990 and 1992. + Straw introduction/replacement.

problems in the area of water man- commended for use in or near Irish agement in Ireland. As the conditions watercourses to four. These are di- that favour the proliferation of these chlobenil, diquat, terbutryn and gly- weed species are unlikely to amel- phosate. These herbicides effectively iorate in the short-term it is, therefore, control their respective target species important that environmentally safe and, when used with discrimination, and effective strategies for controlling have little impact on non-target orga- the spread of this vegetation are nisms. While a great deal of time and urgently developed. These control money has been spent researching strategies must, however, be suffi- the efficacy and environmental im- ciently sensitive to control the target pacts that these herbicides have on vegetation without adversely im- the aquatic habitat (Pieters and de pacting non-target flora and fauna and Boer 1971; Brookers and Edwards without interfering with the beneficial 1982; Wade 1982; Caff rey 1988), less use of the watercourse. work has been conducted on the de- Extensive research on the environ- velopment of fast, precise and econo- mental impact that herbicides have on mical herbicide application devices. aquatic ecosystems has narrowed the This probably reflects the relatively li- range of active ingredients re- mited market for aquatic herbicides by Aquatic plant management in lreland 93 comparison with that available for tion in relatively uniform river chan- agricultural products. The absence of nels, are less successful in dealing suitable equipment for applying gly- with robust submerged species in phosate to reeds in relatively inaces- fast-flowing, complex river habitats. sible, wide rivers is one case in point These plants (e.g. Potamogeton pecti- and is an area which merits urgent at- natus) present serious inchannel tention. obstructions in long sections of Irish rivers and, if these channels are to be Advances in mechanical weed cut- exploited to their full extent, more ef- ting and harvesting apparatus in fective forms of control must be recent years have greatly improved developed. Engineering work in this the efficiently of control operations. area is ongoing. However, results presented in this paper demonstrate that, unless there One of the most significant ad- is a commitment among water vances in the area of aquatic weed managers to apply deep cuts, it may management in recent years has ben be necessary to apply two or three the discovery of the ability of rotted cuts during each growing season. The barley straw to effectively control the fear among operators of breaking growth of algae, both phytoplankton blades on debris lying on the river, and filamentous forms, in aquatic hab- canal or lake beds commonly deters itats. This means that one of the most them from cutting the vegetation to a troublesome groups of aquatic plants depth where light penetration and can be managed without the involve- plant regrowth will be restricted. Effi- ment of costly chemicals or mechani- cient and economical mechanical cal cutting and harvesting machinery. weed cutting operations can only be The fact that the straw provides a conducted, in most circumstances, habitat for fish-food macroinverte- where the cut is applied close to the brates represents and additional substratum and provisions must be benefit to the use of this technique. made to facilitate this type of opera- Future research should focus on in- tion. This may involve removing rub- vestigating the availability of further bish prior to the weed cutting weed control methods that are en- operation or fitting blade guards to the vironmentally safe and may reduce cutting equipment. Our dependence on more costly and less environmentally-sensitive pro- The ability to effectively cut ob- cedures. structive vegetation in rivers where riffles, glides and pools regularly al- ternate and where grave1 shoals, 7 REFERENCES rocks and boulders are distributed along the channel remains a serious Barrett P.R.F., Murphy K.J. and Wade problem. Amphibious weed cutting P.M., 1990. The management of aqua- boats are available on the market and, tic weeds. In, Hance R.J. and Holly K, while these can effectively cut vegeta- (Eds), Weed Control Handbook: Prin- 94 J.M. Caffrey

ciples. Blackwell Scientific Publica- (Eds.) The Grand Canal lnchicore and tions, Oxford 473-490 p. Kilmainham. lnchicore and Kilmain- Bronstad J.V. and Friestad H.O., 1985. ham Development Project/Office of Behaviour of glyphosate in the aquatic Public Works 66-68. environment. In Grossbard E. and At- Caffrey J.M., 1993. Aquatic plant mana- kinson D. (Eds.), The Herbicide Gly- gement in relation to lrish recreational phosate. Buîterworth, London. fisheries development. J. Aquat. Plant Brooker M.P. and Edwards R.W., 1975. Manage. 31 : 162-168. Aquatic herbicides and the control of Caffrey J.M., 1993. Aquatic weed wana- aquatic weeds. Wat. Res.9, 1-15. gement practices using dichlobenil: an Caffrey J.M., 1985. A scheme for the as- lrish experience. Pol. Arch. Hydrobiol.: sessment of water quality using aqua- 00-00. tic macrophytes as indicators. J. Life Caff rey J.M. In press. The spread and ma- Sci., R.D.S., 5: 105-11. nagement of Giant Hogweed (Hera- Caffrey J.M., 1986a. Macrophytes as bio- cleum mantegazzianum Sommier and logical indicators of organic pollution in Levier) along lrish river corridors. In, lrish rivers. Proc. RIA Symp. on Bio- Wade M. (Ed.) Ecology and Manage- logical lndicator of Pollution: 77-87. ment of Invasive Riverside Plants. John Wiley and Sons, Ltd. 00-00. Caffrey J.M., 1986b. The impact of peat siltation on the marginal macrophyte Caffrey J.M. In press. Management of comunities in the River Suck: an lrish aquatic and invasive riparian vegeta- coarse fishery. Proc. EWRS/AAB 7th tion in lrish amenity watercourses. Symp. on Aquatic Weeds: 53-60. Proc. Conf. on Water of Life, R.D.S., Dublin: 00-00. Caffrey J.M., 1988. Trials to Evaluate the Environmental Impact of Chemical Caffrey J.M. and Monahan C., 1991. Weed Control, using Casoron and Aquatic Plant Management in lrish Midstream, in Selected lrish Waters. Canais. Annual Report 1990-199 1. Of- Office of Public Works commissioned fice of Public Works commisioned re- report, Central Fisheries Board, Du- port, Central Fisheries Board, Dublin blin. 56 p. Caffrey J.M., 1990a. Problems relating to Champ W.S.T. In press. - Lough Sheelin the management of Potamogeton pec- - a success story. Proc. Conf. on Wa- tinatus L. in lrish rivers. Proc. EWRS ter of Life, R.D.S., Dublin: 00-00. 8th. Symp. on Aquatic Weeds, Lough- Dawson F.H. and Kern-Hansen U., 1979. borough: 61 -68. The effect of natural and artificial Caffrey J.M., 1990b. The Classification, shade on the macrophytes of lowland Ecology and Dynamics of Aquatic streams and the use of shade as a Plant Communities in Some lrish Ri- management technique. Intern. Rev. vers. Ph.D. Thesis, University College der ges. Hydrobiol. 64: 437-455. Dublin 254 p. De Nie, H.W. 1987. The decrease in aqua- Caffrey J.M., 1991a. Aquatic plant mana- tic vegetation in Europe and its conse- gement in lrish rivers. In Steer M.W. quences for fish populations. EIFAC/ (Ed.), lrish Rivers: Biology and Mana- CECPI Occassional paper No. 19: gement, Royal lrish Academy, Dublin 52 p. 85-98 p. De Steno F.M.,1992. Controlling water Caffrey J.M., 1991b. Aquatic plants and milfoil infestations using an integrated plant management in the lnchicore approach. Aquatics 14(2): 4-8. area of the Grand Canal. In Conna- Gibson M.T., Welch I.M., Barrett P.R.F. ghan M., Gleeson O., Maddock A., and Ridge l., 1990. Barley straw as an Aquatic plant mariagement in lreland 95

inhibitor of algal growth II: laboratory Further crop studies and estimates of studies. J. Appl. Phycol. 2: 241-248. net productivity of macrophytes in a Goldborough L.G. and Beck A.E., 1989. chalk stream. J. Ecol. 49: 11 9-126. Rapid dissipation of glyphosate in Phillips G.L., Eminson D. and Moss B., small forest ponds. Arch. Environ. 1978. A mechanism to account for ma- Contam. Toxicol. 18: 537-544. crophyte decline in progressively eu- Hannon D., 1992. Vegemats - home of trophicated freshwaters. Aq. Bot. 4: big bass. The In-Fisherman 108, 42- 103-126. 46. Pieters A.J. and de Boer F.G.,1971. Di- Haslam S.M., 1978. River Plants. Cam- chlobenil in the aquatic environment. bridge University Press, Cambridge. Proc. EWRS 3rd Internat Symp. on 396 p. Aquatic Weeds, 183-193. Jorga W., Heym W.D. and Weise G., Schmidt J.C., 1990. How to ldentify and 1982. Shading as a measure to pre- vent mass development of submerged Control Water Weeds and Algae. Ap- plied Biochemists, Inc., Wisconsin macrophytes. Int. revue ges. Hydro- biol. 67, 271 -281. 108 p. Kern-Hansen U. and Holm T.F., 1982. Seagrave C., 1988. Aquatic Weed Control. Aquatic plant management in Danish Fishing News Books, Surrey, England streams. Proc EWRS 6th Symp. on 154 p. Aquatic Weeds, 122-131. Simpson P.S. and Eaton J.W., 1986. Madsen J.D. and Adams M.S., 1989. The Comparative studies of the photosyn- distribution of submerged aquatic ma- thesis of the submerged macrophyte crophyte biomass in a eutrophic Elodea canadensis and the filamen- stream, Badfish Creek: the effect of tous algae Cladophora glomerata and environment. Hydrobiol. 171: 111-119. Spirogyra sp. Aq. Bot. 24: 1-12. Marshall E.J.P. and Westlake D.F., 1978. Tiley G.E.D. and Philp B. In press. Giant Recent studies on the role of aquatic hogweed and its control in Scotland. macrophytes in their ecosystem. Proc. In Wade M. (Ed.), Ecology and Mana- EWRS 5th Symp. on Aquatic Weeds: gement of Invasive Riverside Plants. 43-51 . John Wiley and Sons Ltd. 00-00. Murphy K.J., Fox A.M. and Hanbury R.G., Wade P.M., 1982. The long term effects 1987. A multivariate assessment of of herbicide treatment on aquatic weed plant management impacts on macro- communities. Proc. EWRS 6th. Symp. phyte communities in a Scottish canal. J. Appl. ECO/.24: 1063-1079. on Aquatic Weeds, 278-286. National Rivers Authority, 1990. Toxic Welch I.M., Barrett P.R.F., Gibson M.T. Blue-green Algae. National Rivers Au- and Ridge l., 1990. Barley straw as an thority, London 128 p. inhibitor of algal growth 1: Studies in Navas M.L., 1991. Using plant population the Chesterfield Canal. J. Appl. Phy- biology in weed research: a strategy COL 2: 231-239. to improve weed management. Weed Westlake D.F., 1975. Macrophytes. In Res. 31: 171-179. Whitton B.A. (Ed.), Studies in Ecology, Newbold C. and Palmer M., 1979. Trophic Vo. 2 : River Ecology. Blackwell Scien- Adaptions of Aquatic Plants. CST Note tific Publications, Oxford, 106-128. No. 18. Interna1 Report, NCC, London Westlake D.F., 1981. The development 12 p. and structure of aquatic weed popula- Owens M. and Edwards R.W., 1961. The tions. Proc. Aq. Weeds and their effects of plants on river conditions. II. Control, 33-47. 96 J.M. Caffrey

Williams V., 1992. Lake design for recrea- tems. In Hance R.J. and Holly K. tional use consideration. Land and (Eds.), Weed Control Handbook: Prin- Water 36: 44-45. ciples. Blackwell Scientific Publica- Willis A.J., 1990. Ecological conse- tions, Oxford 501-520 p. quences of modern weed control sys-

Adapté d'une communication présentée au colloque international de Besançon

(16-19 novembre 1992) (( Les acquis de la limnologie et la gestion des systèmes aquatiques continentaux ,,.