He. Pehofer, A new quantitative air-lift sampler for collecting invertebrates designed for operation in deep, fast-flowing gravelbed rivers, ARCH HYDROB, (2), 1998, pp. 213-232
Conventional methods for collecting benthos from gravel substratum usually
fail when quantitative samples have to be taken in deep and large fast-flow
ing rivers. As a purchased air-lift sampler did not perform satisfactorily
under the special conditions encountered in alpine gravelbed rivers, a new
sampler was designed. This sampler was used successfully on the armed grave
l substratum of rivers with current velocities of >2 m/s and depths of 1.5
to 6 m. To operate the sampler a stable floating platform or a boat large e
nough to accommodate sampler, generator, compressor and the crew needed for
handling (3-5 people, depending on the conditions in the river) is needed.
Sampling is possible without anchoring, provided the sampling craft is abl
e to hold its position in the current for the time needed to retrieve a sam
ple (normally 10 to 20 seconds). Quantitative samples can be collected from
sediments containing predominantly sand (>100 mu m, the mesh size of the c
ollecting net) to cobble nearly the diameter of the suction pipe (10 cm). I
n fine muddy sediment the collecting cylinder penetrates very fast and the
collecting nets tend to clog rapidly and to rupture by the abruptly growing
pressure. Therefore, in fine mud an adequate core sampler normally is pref
erable. Penetration depth into the substratum is controlled by the time the
sampler is left in operation. In our studies mean penetration depths usual
ly lay around 20-25 cm. Sampling efficiency of the new air-lift is excellen
t: When compared to other deep water samplers (Petersen grab, slurp gun, de
ep water freeze corer) the air-lift samples regularly yielded by far the hi
ghest benthos numbers. Regarding sampler precision, when stratified samplin
g techniques were employed, the sample variances with the air-lift samples
usually lay in the range of those achieved with stratified core samples fro
m lakes or with stratified MESS-samples from gravel bars. A relative measur
e of precision (95% confidence intervals expressed as a proportion of the m
ean values), calculated for the air-lift samples from two impoundments, sho
wed values between 42 and 99% for the abundance of total macrobenthos. Comp
arison of a deep mid-channel site and an adjacent shallow gravel bar, while
yielding similar overall benthos densities, showed significant differences
in the faunal composition of the two sites, emphasizing the importance of
the deep main-channel areas for quantitative assessment of the benthic biot
a of a large river.