The "Tracked Roaming Transect" and distance sampling methods increase the efficiency of underwater visual censuses

Abstract

Underwater visual census (UVC) is the most common approach for estimating diversity, abundance and size of reef fishes in shallow and clear waters. Abundance estimation through UVC is particularly problematic in species occurring at low densities and/or highly aggregated because of their high variability at both spatial and temporal scales. The statistical power of experiments involving UVC techniques may be increased by augmenting the number of replicates or the area surveyed. In this work we present and test the efficiency of an UVC method based on diver towed GPS, the Tracked Roaming Transect (TRT), designed to maximize transect length (and thus the surveyed area) with respect to diving time invested in monitoring, as compared to Conventional Strip Transects (CST). Additionally, we analyze the effect of increasing transect width and length on the precision of density estimates by comparing TRT vs. CST methods using different fixed widths of 6 and 20 m (FW3 and FW10, respectively) and the Distance Sampling (DS) method, in which perpendicular distance of each fish or group of fishes to the transect line is estimated by divers up to 20 m from the transect line. The TRT was 74% more time and cost efficient than the CST (all transect widths considered together) and, for a given time, the use of TRT and/or increasing the transect width increased the precision of density estimates. In addition, since with the DS method distances of fishes to the transect line have to be estimated, and not measured directly as in terrestrial environments, errors in estimations of perpendicular distances can seriously affect DS density estimations. To assess the occurrence of distance estimation errors and their dependence on the observer’s experience, a field experiment using wooden fish models was performed. We tested the precision and accuracy of density estimators based on fixed widths and the DS method. The accuracy of the estimates was measured comparing the actual total abundance with those estimated by divers using FW3, FW10, and DS estimators. Density estimates differed by 13% (range 0.1–31%) from the actual values (average = 13.09%; median = 14.16%). Based on our results we encourage the use of the Tracked Roaming Transect with Distance Sampling (TRT+DS) method for improving density estimates of species occurring at low densities and/or highly aggregated, as well as for exploratory rapid-assessment surveys in which divers could gather spatial ecological and ecosystem information on large areas during UVC.

Quantifying the rare: Baselines for the endangered Napoleon Wrasse, Cheilinus undulatus, and implications for conservation

“…Most underwater visual census (UVC) fish transects use survey distances of 25–150 m. With the apparent rarity of Napoleon fish (there can be fewer than 1 fish km −1 of transect survey), these lengths were deemed inappropriate (e.g. Irigoyen et al, ). Instead, the sampling strategy involved long snorkelling transects (several kilometres in length) along specific features of adult habitat, such as the reef shelf edge, that are identifiable in satellite or aerial images and easy to follow (Oddone et al, ).…”

“…Global Positioning System (GPS) to quantify their abundance and sizes, and to allow future reassessments of the same areas. Most underwater visual census (UVC) fish transects use survey distances of 25-150 m. With the apparent rarity of Napoleon fish (there can be fewer than 1 fish km −1 of transect survey), these lengths were deemed inappropriate (e.g Irigoyen et al, 2018)…”

How many fish? Comparison of two underwater visual sampling methods for monitoring fish communities

“…Other recent studies have examined and/or compared the performance of various UVS methods ( Bosch et al, 2017 ; Irigoyen et al, 2018 ). Bosch et al (2017) compared the output of three different methods for studying fish assemblages, UVC (strip transects of fixed length and width), Fish Traps, and Baited Cameras.…”

“…The authors mainly focused on species diversity; they did not assess the performance of the methods to estimate species absolute or relative abundance. Irigoyen et al (2018) used conventional strip transects of various fixed width and the distance sampling technique in combination with the “Tracked Roaming Diver” technique (a sampling method that maximizes the length of the transect and thus the area studied). They proposed that the combination of the two latter methods provides a more efficient way for UVS.…”

Peer Review #1 of "How many fish? Comparison of two underwater visual sampling methods for monitoring fish communities (v0.1)"

“…Similarly, 399 Einsing et al (1995) showed that distance sampling, compared to quadrat sampling and strip 400 transects, produced density estimates that were closer to true densities, while Thresher and Gunn 401 (1986) proposed that distance sampling should be preferred for the assessment of secretive 402 species. Irigoyen et al (2018) also proposed the distance sampling as an appropriate method to 403 survey medium-and large-sized fish species, although they also discussed some disadvantages of 404 the specific method that should be taken into consideration. Best fit model, maximum width of line transect after truncation (w) and value of detectability (Pa) of the DISTANCE analysis for each species.…”