Reed Beds Research Articles

Olive Mill Wastewater Treatment by a Pilot‐Scale Subsurface Horizontal Flow (SSF‐h) Constructed Wetland

Performances of a pilot-scale reed bed for the olive mill wastewater (OMW) treatment were investigated, by monitoring influent and effluent pH, total suspended solids (TSS), chemical oxygen demand (COD), total Kjeldahl nitrogen (TKN), total phosphorus and polyphenols. In order to reduce the suspended matter concentration and to avoid clogging, OMW was pre-treated by adding lime putty, calcium hydroxide and hydraulic lime. The best results were obtained with 2 g/L of hydraulic lime. Pre-treated OMW was dosed in the reed bed at dilution ratios of 1/3 and 1/10 (v/v), pointing up that the latter only did not give rise to reed suffering and allowed to obtain good and durable removal efficiencies, above all for COD (74.1+/-17.6%) and polyphenols (83.4+/-17.8%). Recycling of the effluent was quite effective for the improvement of the wastewater quality, allowing a further removal of 26-70%, depending on the parameter taken into account. A post-dosage study, carried out by feeding the reed bed with the effluent of an activated sludge plant, pointed up a rapid decreasing of the outlet concentrations of the investigated parameters to values compatible with Italian regulations concerning wastewater discharge in surface water. Polyphenols were the exception, being their outlet concentration at the end of post-dosage study around 2 mg/L.

The Wheal Jane wetlands model for bioremediation of acid mine drainage

Acid mine drainage (AMD) is a widespread environmental problem associated with both working and abandoned mining operations. As part of an overall strategy to determine a long-term treatment option for AMD, a pilot passive treatment plant was constructed in 1994 at Wheal Jane Mine in Cornwall, UK. The plant consists of three separate systems, each containing aerobic reed beds, anaerobic cell and rock filters, and represents the largest European experimental facility of its kind. The systems only differ by the type of pretreatment utilised to increase the pH of the influent minewater (pH <4): lime dosed (LD), anoxic limestone drain (ALD) and lime free (LF), which receives no form of pretreatment. Historical data (1994–1997) indicate median Fe reduction between 55% and 92%, sulphate removal in the range of 3–38% and removal of target metals (cadmium, copper and zinc) below detection limits, depending on pretreatment and flow rates through the system. A new model to simulate the processes and dynamics of the wetlands systems is described, as well as the application of the model to experimental data collected at the pilot plant. The model is process based, and utilises reaction kinetic approaches based on experimental microbial techniques rather than an equilibrium approach to metal precipitation. The model is dynamic and utilises numerical integration routines to solve a set of differential equations that describe the behaviour of 20 variables over the 17 pilot plant cells on a daily basis. The model outputs at each cell boundary are evaluated and compared with the measured data, and the model is demonstrated to provide a good representation of the complex behaviour of the wetland system for a wide range of variables.

Optimising the performance of a lab-scale tidal flow reed bed system treating agricultural wastewater

A gravel-based tidal flow reed bed system was operated with three different strategies in order to investigate its optimal performance for the treatment of a high strength agricultural wastewater. According to the three strategies, individual reed beds were saturated and unsaturated with the wastewater for different periods while reasonably stable hydraulic and organic loadings were maintained. Experimental results demonstrated that the system produced the highest pollutant removal efficiencies with a relatively short saturated period and long unsaturated period, highlighting the importance of oxygen transfer into reed bed matrices during the treatment. Significant removals of some major organic and inorganic pollutants were achieved under all three operational conditions. Nitrification was not the major route of ammoniacal-nitrogen removal when the system was under high organic loading. Due to the filtration of suspended solids and the accumulation of biomass, gradual clogging of the reed bed matrices took place, which caused concerns over the long-term efficiency of the tidal flow system.

Anti-sized reed bed system for animal wastewater treatment: a comparative study

This paper presents a comparative study on the treatment of high-strength animal wastewater in two parallel lab-scale constructed reed bed systems, namely progressively-sized system and anti-sized system, which have same configuration but different arrangement of bed media. The reed bed systems were operated in a tidal flow pattern to treat diluted pig slurry. Detailed analyses were carried out for the removal of some key pollutants including COD, BOD 5, NH 4-N, P and suspended solids. The results showed that both systems have considerable capacity for the removal of solids, organic matter and inorganic nutrients. The formation of biofilms on the surfaces of gravel media in both reed bed systems was monitored by scanning selected gravel samples using scanning electron microscopy. In general, no significant difference was detected with regard to the percentage pollutant removal in the systems. However, the anti-sized system demonstrated a clear advantage in its ability to slow down the clogging of bed media and avoid the impairment of long-term functioning and sustainability of the beds. A conceptual model was developed to predict the occurrence of the clogging. The validity of the model was tested using data from this study and from the literatures.

Reed cutting affects arthropod communities, potentially reducing food for passerine birds

Winter cutting of Common Reed Phragmites australis (Poaceae) is increasing in Camargue, southern France, potentially affecting wetlands of high conservation value. We studied its impact on arthropods by comparing four cut and uncut sites with special respect to avian prey availability in the breeding season. The two most important prey groups for breeding passerines, spiders (Araneida) and beetles (Coleoptera), were identified to species in trap catches and diet samples. The arthropod assemblages differed significantly between cut and uncut sites. Phytophagous and saprophagous species showed contrasting responses. Numbers of homopterans increased in cut reed beds, where green Phragmites stem density was higher. Saprophagous woodlice decreased, presumably due to the reduced amount of litter. Densities of some vegetation-dwelling spider and beetle species were lower at cut sites, including two of the most preferred prey species for passerine birds. Consequently, large-scale mechanically harvested reed beds host altered arthropod communities, missing major food components used by vulnerable passerines. However, reed cutting on a small scale may increase habitat heterogeneity and species richness on a landscape level. To contribute to reed bed conservation, EU schemes should reward management that leaves uncut reed patches in the proximity of cut areas to permit their recolonization by arthropods.

Dynamik der Schilfröhrichte am Bodensee unter dem Einfluss von Wasserstandsvariationen

On the basis of different sets of aerial photos the dynamics of the reed bed areas of Lake Constance were investigated in relation to the dynamics of the water levels. The objectives of the study were to quantify the changes of reed areas due to different flood events in the last decades and their recovery in the time periods between these events. The results should given information of the relevance of water level variations on reed bed dynamics and the regeneration times of reed beds after extreme disturbance events. Following the extreme flood at Lake Constance in 1999 the reed belts of Lake Constance lost approximately 30 ha (24%) of the lakeside reed beds. The loss is comparable to the situation in the late 1960s, when approximately 40 ha died back due to the extreme flood in 1965 and the high spring water levels in the subsequent years. In the time period between the extreme floods of 1965 and 1999, the reed areas expanded to nearly 85% of the area before 1965. The expansion rates increased with increasing distance to the flood event of 1965. Especially in periods with series of years of low spring water level the expansion rates were high. The damage degrees of the reed areas in the years 2000 and 2002 showed a clear relation to the elevation (i.e. average water level) of the stands. The damage degree increased with decreasing elevation. Furthermore the regeneration process of severely damaged stands was related to the elevation level of the stands. Whereas stands at high elevation regenerate fast, those at low elevation died off completely in the years after the extreme flood. This supports the hypothesis that the water level flutuations play a major role in the reed dynamics of Lake Constance. As a consequence of the climate change an increase in the frequency of high spring water levels is expected. Thus, it seems unlikely that reed stands will ever expand again to the same area as before 1965.

Seasonal and diurnal hydro-chemical variations in a recreated reed bed

Abstract. To help stem this decline, wetland areas are now being preserved and where possible created. This study investigated water quality in a recreated reed bed in south-west England designed specifically to provide habitat for birds. Since June 2000, continuous measurements of temperature, pH, electrical conductivity and dissolved oxygen have been logged. These data enable the partial pressure of CO2 to be calculated. This paper looks at the data collected up to November 2001. Despite similar solar radiation and water temperature conditions in both 2000 and 2001, peak summertime values of dissolved oxygen dropped from &gt;100% to &lt;50% and pH values dropped from &gt;7.7 to &lt;7.2. Both also show a reduction in diurnal variation. Over the same period, the average partial pressure of CO2 increased from 19.7 to 56.1. Analysis of the diurnal dissolved oxygen curves indicates a decrease in photosynthesis which, in the absence of a change in nutrient concentrations in the water, may result from surface plant growth or from a reduction in submerged plant cover. The implications of these changing hydro-chemical conditions are that the site will be less able to support such a wide diversity of aquatic fauna, which will reduce the effectiveness of the site as a habitat for some bird species. Keywords: wetland, reed bed, hydrochemistry, dissolved oxygen, pH, continuous monitoring, Hamwall, Somerset Levels

Nitrogen removal from domestic effluent using subsurface flow constructed wetlands: influence of depth, hydraulic residence time and pre-nitrification

This paper describes two studies into the BOD and TN removal performance of horizontal subsurface flow wetlands (reed beds) in subtropical Australia. The aim of the first study was to determine the influence of HRT and vertical position on BOD and TN concentration and removal performance in a 0.5 m deep reed bed (System 1) by taking samples from three levels (or layers) in the water column at five points along the length of the bed. The aim of the second study was to investigate the TN removal performance of a treatment train consisting of a vertical flow intermittently dosed sand filter preceding a reed bed (System 2). Both systems were dosed with primary settled municipal wastewater (BOD 194 mg L(-1); TN 49 mg L(-1)). System 1 achieved a TN load removal of 58% under a HLR of 22 mm day(-1) (HRT 10.5 days), producing effluent BOD concentrations consistently less than 8 mg L(-1). There was no significant difference in BOD attenuation rate between the three layers. While there were differences in both the nitrification and denitrification rates between the three layers, the TN concentration was found to decline steadily in all layers up to an HRT of 8.7 days. System 2 reduced TN influent load by 33%, less than half of which was removed by the reed bed. The lack of substantial TN removal within this reed bed was attributed to the low concentrations of BOD and consequent lack of dissolved organic carbon to drive the denitrification process.

Seasonal variation in phosphorus removal processes _within reed beds - mass balance investigations

The phosphorus (P) removal processes in two pairs of High and Low Loaded reed beds were investigated during five periods within a 27-month study. The uptake/release of P was measured in seven mass balance compartments. With the exception of the first year of operation, the reed beds consistently removed over 96% of the influent P load, with total phosphorus (TP) concentrations being reduced from 0.5 mg/L to generally less than 0.005 mg/L across the range of loading rates and seasons studied. During the first year, uptake by Phragmites australis accounted for greater than 75% of P removed, and was equally distributed between above and below-ground biomass. During the second and third years, three seasonal stages were identified in the uptake and cycling of P by P. australis. A period of rapid above-ground growth and uptake occurred during spring fuelled partly by P reserves accumulated in rhizomes during the previous year. During summer, uptake by above-ground biomass was governed by the influent P loading rate, while the amount of P held in below-ground biomass remained relatively stable. During autumn and winter, P appeared to be translocated from senescent shoots to reserves in the rhizomes. Approximately 85% of the below-ground biomass P occurred in the top 20 cm of the substrate. Gravel fixation increased in importance from 12% in the first year to approximately 30% of P removed in the second year, with a highly significant correlation between the influent P loading rate and P fixed by the gravel. The weakly-bound P fraction from a sequential extraction was the dominant form of P fixed by the gravel. HCI extracts were inappropriate for the examination of sorption processes as they dissolved large amounts of mineral P from within the basaltic gravel. The bottom 30 cm of the substrate became the most important site for gravel fixation during the second year. Incorporation of P into the detritus/microbiota/other compartment increased after the first year to become one of the most important P removal processes, probably consisting mainly of leaf litter and slowly accreted organic sediments.

Compact vertical flow constructed wetland systems - first two years’ performance

In order to move towards optimising wastewater treatment performance by vertical down-flow reed beds, user and installer demands were applied to design, resulting in the "compact vertical flow reed bed" (CVF bed). Performance data are presented for the two years since commissioning a CVF bed and pond system serving eight residents. Particular emphasis was placed on investigating oxygen transfer rate and susceptibility to hydraulic overloading. Average percentage concentration reduction effected by single passage through the bed exceeded 97% for suspended solids and biochemical oxygen demand, and 90% for ammoniacal nitrogen and total Kjeldahl nitrogen. The occurrence of nitrification and denitrification (up to 58%) within the same bed was clearly indicated. Percentage ortho-phosphate removal averaged 92% over the first 17 months' operation and thereafter fell to approximately 45%. Faecal coliform density underwent an almost four log reduction. Oxygen transfer rate within the reed bed was between 28-36 gO2/m2 x d. Normal operation was maintained until six times dry weather hydraulic load rate was exceeded, at which point surface flooding occurred. The capacity of a single vertical flow reed bed to achieve effective treatment of domestic sewage was demonstrated, with no requirement for reduced load rates, promising substantial advantages over established approaches.

The design, operation and performance of a municipal hybrid reed bed treatment system

The first Irish municipal application of a hybrid reed bed treatment system (RBTS) was designed and constructed at Colecott, County Dublin. The hybrid design for Colecott RBTS was based on modifications to the Max Planck Institute:Process (MPIP). The constituent parts of the design were a septic tank, a pump sump, two stage vertical flow beds, a secondary settlement tank, a horizontal reed bed, an outlet chamber and outfall pipe. The population at the time of design was 48, but with future development in mind 60 was used as the population equivalent (pe). Construction and operational characteristics are outlined. Results of a two year monitoring study are presented, showing good performance and producing a final effluent with 88% BOD removal, 89% COD removal and 99% SS removal. Percentile compliance with EU discharge standards were BOD 60%, COD 100%, SS 100% with 50% for nitrogen and 0% for PO4. Surface water input increased flows to twice the average, resulting in increased maintenance visits without affecting results. The ability to deal with the excess inflow is demonstrated by the system. The method of distribution to the primary vertical beds was satisfactory while that on the secondary vertical beds was not.

The dietary regimes of two contemporaneous populations of Hippotherium primigenium (Perissodactyla, Equidae) from the Vallesian (Upper Miocene) of Southern Germany

The first occurrence of tridactyl, hipparionine horses is presently constrained in Eurasia between 11.2 and 10.8 Myr. Hippotherium primigenium is the only species recorded up to now in the early Vallesian–early Turolian of Germany. Two Vallesian populations of H. primigenium are investigated with respect to their trophic regimes. One of these populations derives from the deposits of the Upper Miocene Rhine river exposed at the locality of Eppelsheim (Germany). The other population (Höwenegg) is represented by lacustrine deposits of a former crater lake about 180 km south of Eppelsheim. Both localities represent two almost contemporaneous Vallesian (MN9) populations of a single equid taxon (H. primigenium) from different geographic regions within western Germany. Trophic regimes are reconstructed using the mesowear method [Fortelius and Solounias, Am. Mus. Novit. 3301 (2000) 1–36] and the derived ‘extended’ mesowear method [Kaiser and Solounias, Geodiversitas (2003), in press]. These analyses lead to the identification of a recent trophic reference taxon for each of the two hipparion populations. The common waterbuck (Kobus ellipsiprymnus), a grazer inhabiting reed beds and also foraging into woodlands is identified as the trophic analog for the Eppelsheim population. The reference taxon of the Höwenegg population is the Sumatran rhinoceros (Dicerorhinus sumatrensis), a browser, inhabiting mainly hilly country and preferring saplings, fruit, leaves, twigs, and bark as major food items. The marked differences in the trophic regimes found for the two populations of H. primigenium are interpreted as reflecting the range of possible dietary regimes, here termed the trophic spectrum of this equid species. The trophic signal of H. primigenium is therefore interpreted as an ecological signal reflecting properties of the habitat. The paleohabitat of Eppelsheim would then be recognized as representing at least seasonally extended grassy areas such as reed flats in the proximity of the Miocene Rhine river. The paleohabitat of the Höwenegg, however, was most likely an immediate part of the subtropical mesophytic forests, that covered large parts of Central and Western Europe, Central Asia and southern China during the Vallesian period.

The removal of chemical oxygen demand from primary-treated domestic wastewater in subsurface-flow reed beds using different substrates.

Subsurface-flow experimental reed beds were designed and built based on a combination of two design methodologies. Four different growing media were used with a combination of topsoil, gravel, river sand, and mature wastewater biosolids compost to determine the best substrate for chemical oxygen demand removal. Eight units were constructed, two for each material. One bed for each pair was planted with Typha latifolia plants commonly known as cattails. Primary-treated domestic wastewater was continuously fed to the beds for more than 6 months. The best performance was achieved by the gravel reed beds, with an average removal rate higher than 50%. Soil-based beds containing topsoil and sand only managed to attain removals of approximately 10%. The reed beds containing compost in their substrate produced the worst treatment, mainly because of leaching of organic substances from the compost. Primarily as a result of channel flow, all beds showed significant deviation from the designed retention time. There was no significant difference in the performance of planted and unplanted reed beds.

Role of hydraulic retention time and granular medium in microbial removal in tertiary treatment reed beds

The main objective of this paper is to evaluate the role of hydraulic retention time (HRT) and granular medium in faecal coliform (FC) and somatic coliphage (SC) removal in tertiary reed beds. Experiments were carried out in a pilot plant with four parallel reed beds (horizontal subsurface flow constructed wetlands), each one containing a different type of granular medium. This pilot plant is located in a wastewater treatment plant in Montcada i Reixac, near Barcelona, in northeastern Spain. The microbial inactivation ratios obtained in the different beds are compared as a function of three selected HRTs. Secondary effluent from the wastewater treatment plant was used as the influent of the pilot system. The microbial inactivation ratio ranged between 0.1 and 2.7 log-units for FC and from 0.5 to 1.7 log-units for SC in beds with coarser granular material (5–25 mm), while it ranged between 0.7 and 3.4 log-units for FC and from 0.9 to 2.6 log-units for SC in the bed with finer material (2–13 mm). HRT and granular medium are both key factors in microbial removal in the tertiary reed beds. The microbial inactivation ratio rises as the HRT increases until it reaches a saturation value (in general at an HRT of 3 days). The value of the microbial inactivation ratio at the saturation level depends on the granular medium contained in the bed. The specific surface area necessary to reach 2–3 log-units of FC and SC is approximately 3 m 2/person-equivalent.

Baseline hydraulic performance of the Heathrow constructed wetlands subsurface flow system

A constructed wetland treatment system has been commissioned by BAA (formerly the British Airports Authority) in order to attenuate airfield runoff contaminated with de-icant and other potentially polluting materials from Heathrow Airport. Airfield runoff containing de-icants has the potential to impose significant oxygen demands on water bodies. The site consists of a number of integrated treatment systems, including a 1 ha rafted reed bed canal system and a 2 ha sub-surface flow gravel reed bed. This research project is concerned with the performance of the subsurface flow reed beds, though attention will be paid in this paper to the operation of the whole system. Prior to the planting of the subsurface flow reed beds, flow-tracing experiments were carried out on the three different types of subsurface flow beds, so that the baseline performance of the system could be quantified. In association, data regarding the soil organic matter content was also collected prior to the planting of the beds. As expected, soil organic matter content is observed to be negligible within the bed, though a small amount of build up was observed in localised areas on the surface of the beds. This was attributed to the growth of algae in depressions where standing water persisted during the construction phase. Few studies exist which provide detailed measurements into the cause and effect of variations in hydraulic conductivity within an operational reed bed system. The data presented here form the baseline results for an ongoing study into the investigation of the change in hydraulic conductivity of an operational reed bed system.

Disturbance and rehabilitation of lakeside Phragmites reeds following an extreme flood in Lake Constance (Germany)

After an extreme flood in Lake Constance in 1999 the Phragmites australis belt showed a severe decline in area and vitality. A three year monitoring project was installed in 2000 to document the die-back and rehabilitation process immediately afterwards, to identify the co-factors of the damage, and to find out significant stressors that may impede the recovery of the reeds. The monitoring is based on CIR aerial photo interpretation, quantitative GIS analyses and field investigations on shoot density, stand structure and biomass production in 50 monitoring plots, grouped in five degrees of damage. In result we found that c. 0.306 km 2 of aquatic reed bed area died back at Lake Constance-Untersee (i.e. 23% of the former area in 1998). Among the stands which had survived the severely damaged stands were mainly composed of secondary shoots, whereas primary and insect infested shoots dominated in less damaged stands. The development from 2000 to 2001 was characterized by an overall decrease in shoot density, a change in the composition of the shoot population in favour of primary shoots, and in a recovery in culm stature. All variables depended on the degree of initial damage by the extreme flood. A conceptual model is proposed to assess the future development of Lake Constance reeds.

Nitrification in reed beds - capacity and potential control methods

Vertical-flow reed beds (VFBs) are known to be very effective regarding nitrification. However, a generally accepted design formula for dimensioning reed beds for a controlled nitrification process has still not been found. Especially for the purpose of nitrification a vertical-flow reed bed (VFB) has been installed following an existing unaerated pond-system. The paper presents the results concerning the limits of the VFB's performance as well as the main factors influencing the nitrification process gained by balancing the conversion rates under different loads. Even under high loads the VFB provides an excellent nitrification performance, which is mainly influenced by the wastewater temperature. The ammonia oxidation rate is about 90% at temperatures over 10 degrees C; at temperatures below 5 degrees C the average nitrification rate is still approximately 50%. The hydraulic load and the TKN load have almost no impact on this efficiency (the maximum load has been: 180 mm/d, 7.1 g TKN/(m2 x d)). The redox potenial, which is continuously measured in the effluent of the reed bed, proved to be dependent on the current nitrification performance and the oxygen supply of the VFB, and therefore appears to be a suitable control parameter for the operation of VFBs.

Social system of the Alaotran gentle lemur (Hapalemur griseus alaotrensis): genetic characterization of group composition and mating system.

Group composition and mating system were investigated in wild Alaotran gentle lemurs (Hapalemur griseus alaotrensis) using genetic markers. These small-bodied herbivorous and cathemeral primates are endemic to the reed and papyrus beds around Lac Alaotra in Madagascar. They live in small groups in small, defended territories. Data were collected during the rainy seasons in 1996, 1997, and 1999, and include 99 individuals from 22 neighboring social groups and an additional 30 animals from other areas representing most of the geographic range. Animals were located by researchers canoeing in the marshes bordering the lake. After a group's size was determined by direct observation, all individuals were captured, marked, and released. During handling each individual was weighed and sexed, and hair samples for DNA extraction were collected. A 342 bp mtDNA control region sequence and 10 nuclear microsatellite loci provided multilocus genotypes that were used to assess pedigree relations and population structure. Alaotran gentle lemurs were found in groups of two to nine individuals (mean: 4.3), comprising one or two breeding females, their offspring, and one reproducing male. Solitary individuals of both sexes were encountered rarely. Breeding females were the permanent core of the social groups, whereas intergroup transfer of reproductive males was relatively frequent. Forty percent of all reproducing groups contained two breeding females, which were related to each other as closely as mother-daughter or full sisters. Parentage assessment revealed a variable mating system ranging from serial monogamy to polygyny within social groups. At least 8% of paternities involved extragroup males. Additional data on life history and reproduction are presented, and the social system of the Alaotran gentle lemur is discussed in the light of the new genetic findings.

The Removal of Indicator Microorganisms from Primary Treated Wastewater in Subsurface Reed Beds using Different Substrates

Subsurface, horizontal flow, experimental reed beds, were designed and built based on a combination of two design methodologies, that of the WRc and Severn Trent Water Plc and that of the USA, EPA. Four different growing media were used with a combination of top soil, gravel, river sand and mature sewage sludge compost, aiming to determine the best substrate for enteric pathogen removal. Eight units were constructed, two for each material. One bed for each pair was planted with Typha latifolia plants commonly known as cattails. Primary treated domestic wastewater, was continuously fed in to the bed for more than six months. All beds achieved a high reduction of Escherichia coli and faecal coliforms with the best results recorded in the gravel reed beds with an average removal above 3.3 log for E. coli and 3.0 for faecal coliforms. There was no significance difference in the performance of planted and unplanted reed beds.

Auswirkungen des Extremhochwassers 1999 auf die Flächendynamik und Bestandsstruktur der Uferröhrichte des Bodensees

In consequence of the extreme flood at Lake Constance in 1999 the reed belts showed a severe loss in vitality so that a permanent damage could be expected. The monitoring project presented here focuses the documentation of the damage following the disturbance event, the identification of co-factors, and is designed to find out significant stressors that may impede the recovery of the reeds. The monitoring is based on CIR aerial photo interpretation, quantitative GIS-analyses and field data. The design of the study implies the definition of five degrees of damage using the interpretation of CIR aerial photo series from July 2000. The aquatic reed beds were digitised and areal sums were calculated. These data were compared with the corresponding data from 1999 and 1993. Stand structural and biomass attributes were investigated in 50 randomly distributed permanent plots covering all site types and damage degrees. The results showed a loss of about 24 hectares of aquatic reeds at Lake Constance Untersee alone. The stands which had survived showed a great patchiness in terms of culm density, shoot population composition, culm architecture, and aboveground biomass production. Severely damaged stands were mainly composed of secondary shoots, whereas primary and insect infested shoots dominated in less damaged stands. One co-factor which explained the patchiness was the vertical position of the stand, i.e. the probability of being severely damaged increased with increasing distance below mean water level. The monitoring programme will be continued till 2003.