Habitat structure and fish: assessing the role of habitat complexity for fish using a small, semiportable, 3‐D underwater observatory

Abstract

In contrast to terrestrial ecology, three-dimensional (3-D) imaging technology is not well established as a method for studying species-habitat interactions in aquatic ecology. In this study, we used a semi-portable, digital, 3-D underwater observatory designed for long-term exposure in shallow water habitats to assess fish-habitat interactions to artificial structures of different complexity. The observatory was mounted on a cable-car system and was moved along a 50 m transect parallel to five artificial structures of different complexity and one control site. The optical unit took high-resolution digital images, in pairs, from each structure and the control site at a frequency of 30 min over 24 h and was then moved remotely to the next structure. The system was operated for 2 mo without being recovered and took a total of 2160 stereoscopic image pairs. The system proved highly valuable in assessing small-scale temporal patterns in fish-habitat interactions, as well as changes in habitat preferences of fish over the diel cycle. Such small-scale temporal and spatial patterns are important for a detailed understanding of specieshabitat interactions, not only in fish but also in most aquatic macro-organisms. Nevertheless, such patterns can hardly be assessed using classical sampling devices due to labor and animal protection restrictions. However, the possibility of a significant increase in quantity of sampling through the use of remote imaging technology is accompanied by a significant increase in labor costs for image analysis. Therefore, further progress in semi-and fully automated systems for image analysis is needed, including 3-D imaging tools, for ecological studies.

A practical guide to the use of consumer-level digital still cameras for precise stereogrammetric <I>in situ</I> assessments in aquatic environments

“…For both stereoscopic systems (the SLR and lowcost systems), the optical axes were adjusted to be as parallel as possible prior to the assessments, following the procedure of Klimley and Brown (1983), adapted for digital cameras (Fischer et al, 2007).…”

“…Furthermore, many algorithms used to calculate an object's length based on a pair of stereoscopic images require the parallelism of the optical axis of the two cameras. Although this was hard to achieve previously with analogue photography, digital imaging enabling the possibility of transferring stereoscopic image pairs without dismounting them from the rack provides a comparatively easy way to preadjust the optical axes of two cameras to approximate parallelism, which is often sufficient for many stereo-optical assessments (Fischer et al, 2007).…”

“…Excellent summaries of the applications of stereophotography and videogrammetry to marine biology have been published (Williams, 1969;Rorslett et al, 1978;Fischer et al, 2007;Williams et al, 2010;Aguzzi et al, 2012), which demonstrate the usefulness of both stereogrammetry-based videos and still photography in assessing the metric traits of fish stocks, for example in 'untrawlable' habitats. Stereoscopic assessments, furthermore, provide a valuable tool for SCUBA-supported non-destructive rapid assessment surveys and monitoring programmes, and are especially applicable to the assessment of endangered species, which are protected against capture (Shortis et al, 2007).…”

“…A range of other marine biology applications also routinely use stereoimaging techniques (photo as well as video), including: fish monitoring (Ruff et al, 1995;Van Rooij et al, 1995;Li et al, 1996;Tillett et al, 2000;Harvey et al, 2003;Costa et al, 2006;Costa et al, 2009;Rooper et al, 2010;Bower et al, 2011;Torisawa et al, 2011); determination of the 3D structure of fish schools (Cullen et al, 1965;Pitcher, 1975;Dill et al, 1981;Sawada et al, 2009); determination of the influence of substrate complexity on the interaction between fish and the artificial substrate (Fischer et al, 2007); and surveys of the benthos and the sea floor (Emerey et al, 1965;Evans and Norris, 1997;Chong and Stratford, 2002;Jasiobedzki et al, 2008;Shortis et al, 2008;Svane et al, 2009).…”

“…Cullen et al, 1965;Rorslett et al, 1978;Van Rooij and Videler, 1996;Costa et al, 2006;Williams et al, 2010). Photographic assessments of time series of communities also allow the estimation of quantitative derivatives, including growth or productivity over time, using an empirically determined relationship between biomass and linear measurements (Williams, 1969;Rorslett et al, 1978;Fischer et al, 2007;Williams et al, 2010, Doya et al, 2014.…”