Concurrent Sessions A: Passage Effectiveness Monitoring in Small Streams III - Automated Monitoring Stations to Detect Passage of Individually-Tagged Warmwater Fishes in Headwater Streams of the Ouachita National Forest

The Interior Highlands of Arkansas support an extensive logging industry, which requires a vast network of unpaved roads (e.g. U.S.D.A. estimated almost 7,000 culvert road crossings in the Ouachita National Forest). Previous studies have detected adverse effects of culverts on passage of batch-marked fish. However, conclusions were hampered by few recaptures and timing of movements was not measurable which contributed to uncertainty relative to the ecological significance of reduced passage. During 2011 and 2012, we marked over 3,700 fishes >85 mm total length representing nine species with 12-mm, half-duplex, RFID tags (also known as PIT tags). These fish were from two streams with road crossings and two reference streams without road crossings. We installed solar-powered, RFID logging stations at a mid-point on each stream to continuously monitor fish movements (centered at a road crossing on the non-reference streams). Each stream’s RFID station included two, swim-through antennas encircling the stream with one antenna upstream of the road crossing or reference reach and the other located downstream. The two-part antenna array was designed to precisely record timing and therefore movement direction of each fish passage. However, this was the first attempt to employ this approach with such small fish in streams this large and we have learned many practical lessons. For example, we detected cross-talk between the paired antennas on one stream (data contamination that negated efforts to discern movement direction and passage). Approximately 10% of the fish tagged in each stream have been detected by at least one of the antennas. We are still collecting this type of data, but it appears that we average approximately 120 unique fish detections per month which allows more robust conclusions than previous efforts with mark and recapture. Preliminary results indicate the least passage at a piped-culvert on Bear Creek which is also a stream prone to seasonal drying. A nearby reference stream with a similar tendency to dry, Crystal Prong, showed more movements across a similar distance without a road crossing (preliminary projection of about one dozen versus over fifty fish per year). Annual projected passage through a box-culvert on Long Creek was over 90 for this perennial stream. We are still trying to salvage the equivocal cross-talk data on the Little Missouri, a perennial reference stream. So far, an average of only 1.3% of tagged fish have shown passage past an antenna pair though smallmouth bass, northern hogsucker, and highland stonerollers passed at higher rates and creek chubsucker, longear sunfish, yellow bullhead, striped shiner, and creek chub passed at lower rates. We are still investigating the relationship between water level fluctuations and timing of passage events.
Research Projects
Organizational Units
Journal Issue
Publisher Version
Embedded videos