Site length for biological assessment of boatable rivers

There is increasing international interest by water resource management agencies worldwide in developing the capacity for quantitative bioassessments of boatable rivers. This interest stems from legal mandates requiring assessments, plus growing recognition of the threats to such systems from multiple and co‐varying stressors (e.g. chemical pollutants, physical habitat alterations, altered flow regimes, channel modifications and alien species). The elevated cost and inefficiencies of jurisdictionally‐ and taxonomically‐segregated assessments is widely recognized, as is the desire to obtain comparable data that can be easily shared among political jurisdictions and ecological regions. The objectives, sampling methods, indicators, site‐scale sampling designs and geographic extent of the resources being sampled differ among programmes, thereby limiting such data exchanges. Our objective in this paper is to review major biological assessment design alternatives for boatable rivers, with special attention given to the sample site length from which data are collected. We suggest that sufficient site length determinations should be based on the survey objectives, the relative heterogeneity of the habitat template, and the quality of data necessary for meeting programmatic data quality objectives. Future sampling effort studies should be designed to allow separate samples of several short sub‐sites at many diverse sites to generate multiple data points for each site. Data from those multiple sub‐sites should be analysed using randomization‐based data evaluation methods. We hope that our recommendations will be useful to the maximum number of institutions, including those with limited funds and a purely local focus, as well as those responsible for sampling at continental geographic extents. Copyright © 2010 John Wiley & Sons, Ltd.