Among the techniques available for sampling invertebrates in the hyporheic zone, the methods based on pumping (Bilge or Bou-Rouch types) are still by far the most practical and most frequently used. Thus, it is important to understand as much as possible and to visualize more clearly what happens during water uptake in the alluvium, from both physical and biological points of view. However, despite the extensive method testing, many issues and methodological uncertainties still remain. Consequently, it is crucial to discuss these issues, but inevitably with an approach supported by hydrological sciences. The comment by Roy and Danielescu [2014] is very interesting because it provides a graphical and modeled representation of hydrological processes and water flow circulation patterns occurring within the sediments of a gravel bed stream during a Bou-Rouch pumping. To our knowledge, this is the first time since 1967, the date of publication of this method [Bou and Rouch, 1967; Bou, 1974], that so formal and precise hydrologic representations are given. As pointed out in our article, obtaining hyporheic quantitative samples remains a difficult task. This problem increases when sampling living organisms, among other things, because of an additional uncontrolled dimension related to organism behavior, including intra and interspecific variability. The second comment by Roy and Danielescu [2014] relates to the spatial variation of pore water velocity during pumping. The authors provide a vectored representation of such spatial variation and indicate that it might lead to a spatial variation in the extraction capacity of invertebrates by the pump, which is sensible. They conclude that the volume of water in which organisms undergo sufficient shear stress to be drawn in is reduced to the immediate vicinity of the well screen. This phenomenon is relatively well known and is reflected, among other things, by the sharp decrease of the abundance of organisms collected during pumping (e.g., organisms collected in ten consecutive 1 L samples [Boulton et al., 2004] or six consecutive 10 L samples [Danielopol, 1976]), which mirrors a decreasing amount of sand drawn in time. The issues related to pumping efficiency are precisely those raised by our work, because we wanted to know what fraction of the actual density and diversity is provided by a Bou-Rouch sample. Roy and Danielescu [2014] suggest that our system does not take in the entire volume of voids isolated by the Hess cylinder, as assumed during the implementation of the sampling design. Nevertheless, the error is certainly lower than suggested by Roy and Danielescu's model, because a negligible amount of invertebrates were drawn in through the bottom. Their comment is an illustration of the need of interdisciplinary efforts to better understand the hydrological processes that occur during faunal sampling, and generally for a more integrated ecological research in hyporheic and groundwater ecology. Our article may suffer from a weak hydrological support, but it nevertheless illustrates the shortcomings of the available pumping techniques to evaluate hyporheic invertebrate density and diversity. Any improvement in the current methodology that reduces the error highlighted by Roy and Danielescu [2014] is highly desirable. In order to minimize the methodologically sampling error, we suggest to eliminate the small volume of surface water trapped during the installation of the cylinder and to fit the size of the cylinder with both the desired pumped volume and the zone of influence of the pump. Obviously, the exact adjustments should be proposed after close monitoring of the hydrological processes and by modeling the interstitial water flow inside the cylinder.
Read full abstract