Comparison of Underwater Video with Electrofishing and Dive Counts for Stream Fish Abundance Estimation

Abstract

Advances in video technology enable new strategies for stream fish research. We compared juvenile (age‐0) and adult (age‐1 and older) Brook Trout Salvelinus fontinalis abundance estimates from underwater video with those from backpack electrofishing and dive count methods across a series of stream pools in Shenandoah National Park, Virginia (n = 41). Video methods estimated greater mean abundance of adult trout than did one‐pass electrofishing, but video estimates of adult abundance were not different than estimates from three‐pass electrofishing or dive count methods. In contrast, videos underestimated the abundance of juvenile trout; we suggest that this is because predator avoidance behaviors by juvenile trout limit their use of microhabitat locations visible to cameras. Integrated abundance estimates from two cameras increased correspondence to comparison methods relative to estimates from single cameras, demonstrating the importance of an expanded field of view for video sampling in streams. Geomorphic features helped to explain methodwise differences: more adult Brook Trout were estimated with video than with three‐pass electrofishing as riffle crest depth and boulder composition increased, indicating habitat associations with trout escapement from electrofishing. Our results demonstrated that video techniques can provide a robust alternative or supplement to traditional methods for estimating adult trout abundance in stream pools.

Disentangling effects of habitat on salmonid abundance in thermal refuges while accounting for imperfect detection

“…Our estimates of detection using UACs were relatively high (0.521–0.634), likely due to the spatial scale of thermal refuges (<352.02 m 2 ) and the channel deflecting boulders limiting escapement from most refuges (e.g., Figure 2a). Previous studies indicate that turbidity or depth (Campbell et al, 2015; Ebner & Morgan, 2013) or sampling in more complex habitats (Hitt et al, 2021) can limit fish detection using UAC surveys. We found that detection was highest at lower brightness estimates, perhaps due to light backscattering during brighter conditions increasing glare.…”

“…We extracted counts of trout from each of the 161 video surveys similar to other studies in freshwater systems (Ebner & Morgan, 2013; Hitt et al, 2021). For each 15‐min video, one observer identified trout to the lowest taxonomic level possible using body coloration and morphometric characteristics (e.g., presence of adipose fin), and enumerated fish every 90 s, totaling 11 counts (i.e., surveys) per video.…”

“…We indexed counts of trout as the maximum count of individuals observed simultaneously in a single video still frame. Using the maximum count of fish for an individual video frame reduces the chances of double‐counting individual fish (Thurow et al, 2012) and provides a conservative count of the minimum number of fish observed during a survey (e.g., Campbell et al, 2015; Hitt et al, 2021). To aid species identification and enumeration, we made simple video adjustments (e.g., contrast and enlargement) as required.…”

“…Here, we investigated short‐term patterns in trout abundance across several thermal refuges during a severe heat wave in a dynamic, marginal riverscape. We employed underwater action camera (UAC) surveys, a less‐stressful, passive alternative to sampling fish in streams (Hitt et al, 2021; Struthers et al, 2015), every other day during a two‐week heat wave to count thermoregulating trout in thermal refuges. We then used count data to initiate a Bayesian N ‐mixture model to derive abundance estimates and determine influential habitat characteristics.…”

Lake-wide mapping of littoral habitat using underwater videography

“…Side-scan sonar technology provides opportunities to map underwater lake features such as substrate and coarse woody habitat (Koeller, 2014), and more recently, autonomous surface vessels have been developed to monitor invasive macrophyte growth through imagery and collection of depth information (Codd-Downey et al, 2021). Underwater videography has also shown continued utility to estimate fish richness and abundance (Ebner and Morgan, 2013;Wilson et al, 2014;Hitt et al, 2021), behavior (Coghlan et al, 2017) and habitat use Davis et al 1997(Pratt et al, 2005, largely because it is cost-effective and more widely tractable. Its use to conduct spatially-continuous assessments of littoral habitat, to our knowledge, has not been fully explored.…”