Limnologica 29 (1999) 120-127 LIMNOLOGICA http://www.urbanfischer.de/journals/limno © by Urban & FischerVerlag Institute of Ecology, Friedrich-Schiller-University, Jena, Germany Use of Colonization Baskets for the Investigation of Disturbance Phenomena in Streams under Model Conditions PAUL ELSER With 2 Figures and 5 Tables Key words: Stream benthos, disturbance, recolonization, resilience, seminatural substrate, colonization basket Abstract Seminatural substrates were used to mimic conditions of a denuded Hence, many investigators choose the experimental ap- stream bed after a substrate-moving spate. For this purpose gravel- proach to simulate the effects of natural disturbances on an filled wire baskets were placed on the sediment surface of the Ilm experimental scale (e.g. BOULTON et al. 1988; DOES et al. river (Thuringia, Germany). Invertebrates colonized these structures 1989; LAKE & SCHREIBER 1991; MATTHAEI1996; MATHOOKO very rapidly. The results indicate that resilience of the stream ben- 1996). One possibility is the use of artificial colonization thos after physical disturbances, such as spates is high. Colonization units as models for freshly disturbed substratum patches (e.g. patterns of numerous taxa differed significantly between summer and winter exposures. Seasonally varying colonization rates indicate REICE 1985; ROBINSON & MINSHALL 1986; ROSSER & PEAR- that the disturbance response of taxa varies throughout the year and SON 1995). throughout the life cycle of lotic invertebrates. If these changes are In the present study gravel-filled wire-mesh baskets were ex- correlated with seasonal discharge dynamics, they can be interpreted posed on the sediment surface of the Ilm river (Thuringia, as an adjustment to the disturbance regime. Findings are discussed Germany). The aim was to mimic conditions of a partly de- in regard to the practicability of small-scale experiments for the in- nuded stream bed following physical disturbance and to vestigation of disturbance phenomena in streams. launch a process similar to the recolonization of disturbed substratum patches after a bedload-transporting spate. Atten- tion was focussed on taxon specific patterns of colonization. Introduction Furthermore it was examined whether these patterns differ throughout the year. Findings are discussed in regard to the Natural disturbances are increasingly regarded as important practicability of small-scale experiments for the investiga- factors which influence the structure and dynamics of many tion of disturbance phenomena in streams. ecological communities (SouSA 1984; STRONG et al. 1984; PICKETT & WHITE 1985; WIENS 1984). Disturbance by spates is considered to be a dominant organizing factor in lotic ecosystems (POFF 1992; YOUNT & NIEMI 1990). Stream com- Materials and Methods munities frequently disturbed by variable discharge hardly ever achieve ecological saturation and tight species packing Colonization substrates which should characterize a resource-limited and niche-con- trolled structure (RESH et al. 1988; TOWNSEND 1989). The colonization substrates were made of wire-mesh baskets (length Investigations of natural disturbance phenomena in x width x depth = 20 x 20 × 15 cm; mesh 1.05 cm) that were filled with a sterile gravel mixture. The filling consisted of the same min- streams are connected with large technical and methodical eral components as the natural sediment in the study reach and pro- difficulties. A major problem is that the time when a spate vided a heterogenous substrate with interstitial spaces similar to ad- will occur can not be foreseen, and therefore, a detailed work jacent substrate conditions. 60% of the filling was composed of schedule is difficult to be put up. In addition, extreme condi- gravel with a mean diameter of 12 mm (range 8-16 mm) and 40% tions during high discharge make operating in the stream gravel with a mean diameter of 24 mm (range 16-32 ram). The grav- quite risky. el was purchased close to the study area. 120 0075-9511/99/29/02-120 $12.00/0 The baskets were arranged on the stream bed in a systematic grid Also, accumulation of silt or particulate organic matter has not been in rows of 3 baskets. Nylon rope was used to nestle the baskets to recorded, following the findings of WINrERBOURN (1978) and iron rods that where driven deeply into the sediment. In order to WILLIAMS (1980) that correlations of the content of organic matter avoid interference, baskets were 2 m apart within a row and were and animal colonization are unlikely to be detected. 3 m from the next row. Ambient conditions, in particular the nature of the stream bed and the current, were considered to be homoge- Study area nous throughout the study reach. The experiments were designed as time series with exposure The studies were carried out in the Ilm river, a nutrient-rich hardwa- times ranging from 6 hours up to 30 days. In order to evaluate the in- ter stream (stream order 4; STRAHLER classification) in Thuringia, fluence of season, the first experiment was conducted from August Germany. Water quality can be classified as ~-mesosaprobic with 15, 1996 to September 12, 1996, the second experiment took place nutrient concentrations amounting to 0.32 mg 1-1 Total-P and from January 28, 1997 to February 5, 1997. During the second ex- 7.31 mg 1-~ Total-N (mean values for 1996 measured at Niedertrebra; periment, attention was focussed on short exposure times to investi- Thtiringer Landesanstalt for Umwelt 1998). The study site was lo- gate the initial stages of colonization. The winter experiment had to cated in the hyporhitric zone about 2 km upstream of the village of be ceased after day 8 because a spate made further working in the Buchfart (11°20 ' E, 50055 ' N, 255 m a.s.1.). The slope was about stream impossible. 0.5%; stream width varied between 8-10 m. The substrate was In summer, a total of 15 baskets was sampled on 3 occasions (5, mainly mixed gravel with sand and flat cobbles up to a maximum 14 and 30 days after exposition), in winter 30 baskets were exposed length of 20 cm embedded near the surface. Further description of to investigate 6 colonization periods (0.25, 0.5, 1, 2, 4, and 8 days). the stream is given by KREY (1995). The study reach is partly shaded On every sampling occasion 5 randomly chosen baskets were taken by a canopy of Fraxinus excelsior, Alnus glutinosa and other trees. out beginning at the downstream baskets. While retrieving the bas- During the colder months the stretch is heavily colonized by diatoms kets, a sampling net with a square opening (net mesh 0.25 ram) was and by the filamentous algae Cladophora sp. in the warmer months set up fight downstream. Then the rope was cut and the basket sur- (SCHONBORN 1996). rounded by the net was lifted out of the water. Any invertebrates dis- A hydrograph station is located in the village of Mellingen, lodged from the baskets were washed into the net and animal losses about 5 km downstream of the study reach with a 70-year mean dis- could be reduced to a minimum. Then the baskets were opened and charge amounting to 4.2 m 3 s ~; mean discharge in summer 1996 the fillings were rinsed through a series of sieves (2.0, 1.0, 0.25 mm) was 3.27 m 3 s-1, mean discharge in winter 1997 amounted to 5.73 using an engine driven water pump. These sieve fractions were used m 3 s 1 (Staatliches Umweltamt Erfurt, unpubl.). Between the study as a simple measure to estimate the size distribution of some inverte- reach and the hydrograph station there is one tributary contributing brate taxa. about 0.1-0.2 m 3 s-1 (W. N~RB, Institute of Geography, Friedrich- After the baskets had been removed, 5 benthos samples from the Schiller-University, Jena; pers. commun.). Discharge data during stream bottom were collected using a modified HEss-sampler (area the experimental periods are given in Table 1. At the beginning and 0.043 m 2, mesh size 0.25 mm). Sampling sites were chosen at ran- at the end of each exposure period current velocity was measured dom within the study reach. All samples were fixed with 70% upstream and downstream of the baskets at a point 10 cm above the ethanol and processed in the laboratory. Invertebrates were counted streambed and at 60% depth using an electromagnetic water and identified to the lowest possible taxonomic level using a dissect- flowmeter. Due to the higher discharge current velocity in winter ing microscope (ZEISS x 50). was higher than in summer (Table 1). Before the experiments and It was assumed that the retrieved baskets contained numbers and on each sampling occasion dissolved oxygen, electrolytic conduc- varieties of animals that reflected the recolonization activity of ben- tivity and acidity were measured using portable digital instruments. thic invertebrates after a physical disturbance. No estimate of the Physico-chemical parameters showed little variation during the distance travelled before reaching the substrates has been attempted. sampling periods (Table 2). Table 1. Discharge of the Ilm river at hydrograph station Mellingen (a) and flow conditions at the sampling sites near Buchfart (b). Discharge data from Staatliches Umweltamt Erfurt (unpubl.). Mean values _+ standard deviation. Summer Winter (08/15-09/12/1996) (1/28-02/05/1997) a) Hydrograph station Mellingen Discharge [m3 s-1] mean 1.58 + 0.31 2.89 + 0.23 (daily mean values) rain 1.16 2.53 max 2.24 3.39 b) Sampling sites Water depth [m] mean 0.35 _+ 0.06 0.51 + 0.07 Current velocity [m s-~] above the baskets (at 0.6 x depth) 0.25 + 0.10 0.57 + 0.07 upstream from baskets 0.10 + 0.05 0.19 + 0.08 downstream from baskets 0.06 _+ 0.04 0.04 + 0.04 Limnologica 29 (1999) 2 121 Table 2.