The question is tougher than it seems, and the answer depends on the type of natural flow regime (snowmelt/glacial? seasonal rainfall? both? ), what kind of changes (smaller floods? lower base flows?) and why you're concerned. That said, fish and insects are perhaps easist to measure and most likely to show population changes rapidly. Fish tend to be better described (so we know what the river used to have or what similar rivers without diversions look like), but it may be resource intensive to sample enough to make a good assessment. Macroinvertebrates are commonly used for bioassessment, frequently of water quality rather than quantity however (and the use of measures like EPT richness depends on established regional patterns of fairly well characterized taxonomy). Insect responses to flow changes Are still very much a research area. This link may be of some use along these lines for the general US approach.

http://water.epa.gov/scitech/swguidance/standards/criteria/aqlife/biocriteria/index.cfm

Riparian vegetation can be relatively easier to survey and is well known to respond to flow changes (see Nilsson & Berggren 2000 and Rood et al 2005 for good reviews), but depending on the species in your area and the amount of diversion plants may not quickly exhibit population-level changes. Again, it depends on just what the diversion is doing - if it's running the channel dry during hottest times, then it doesn't matter what you measure. If it's a more subtle change then the answer probably depends on what's already best described in the area.

Below a water diversion point, the flow regime generally tends to decrease the water body in the channels and to become more regular. As a result, the first reaction is that of vegetation, starting with invading weeds and ending with woody types. New associations of plants are installing and the areal of riparian vegetation extends. It could be useful to investigate seedling extend and their age for some species as (ex. is for my country climatic conditions) Alnus, Populus, Salix.

Fish are good candidates for such a study because under altered flow conditions, fish reproduction have been observed to often shift from simple nesting to nest-guarding or broadcast-spawning strategies in streams with either form of flow alteration. In streams with diminished minimum or maximum flows, active swimmers replaced benthic-oriented and streamlined fish species.

Having a reference habitat for such a study will allow for comparison and obtaining valid conclusions.

Macro-invertebrate taxa are also useful for such a study because some with the ability to temporarily leave the aquatic environment or move quickly within it (eg strong swimmers, fast crawlers) replaced taxa lacking these traits; moreover, pool (ie relatively slow currents)-loving macro-invertebrate taxa that prefer fine substrates replaced riffle (i.e turbulent flowing)- loving macro-invertebrate taxa that prefer coarse substrates.

Streams with inflated minimum flows have also been found to have an apparent increase in macro-invertebrate taxa that prefer erosional (i.e relatively high current velocity) habitats.