Scaling up Ecological Measurements of Coral Reefs Using Semi-Automated Field Image Collection and Analysis

Abstract

Ecological measurements in marine settings are often constrained in space and time, with spatial heterogeneity obscuring broader generalisations. While advances in remote sensing, integrative modelling and meta-analysis enable generalisations from field observations, there is an underlying need for high-resolution, standardised and geo-referenced field data. Here, we evaluate a new approach aimed at optimising data collection and analysis to assess broad-scale patterns of coral reef community composition using automatically annotated underwater imagery, captured along 2 km transects. We validate this approach by investigating its ability to detect spatial (e.g., across regions) and temporal (e.g., over years) change, and by comparing automated annotation errors to those of multiple human annotators. Our results indicate that change of coral reef benthos can be captured at high resolution both spatially and temporally, with an average error below 5%, among key benthic groups. Cover estimation errors using automated annotation varied between 2% and 12%, slightly larger than human errors (which varied between 1% and 7%), but small enough to detect significant changes among dominant groups. Overall, this approach allows a rapid collection of in-situ observations at larger spatial scales (km) than previously possible, and provides a pathway to link, calibrate, and validate broader analyses across even larger spatial scales (10-10,000 km 2 ).

An evaluation of semi‐automated methods for collecting ecosystem‐level data in temperate marine systems

“…Benthic images were collected and analysed according to González‐Rivero et al. (), at depths of 2–12 m depending on reef topography. This system provided images collected every few seconds, while traveling, ensuring that overlap was avoided (mean distance between images was 3.8 m).…”

“…High-volume data collected by diver or vehicle-mounted cameras and analysed with machine-learning techniques will soon become commonplace, but there exists little detailed examination of the efficacy and appropriateness of this new data (Beijbom et al, 2015). One of the first examples of this "scaling-up" of observations in marine systems demonstrated strong evidence to support the use of long (~2 km) photograph transects and automated image annotation software for change detection in coral reef communities (González-Rivero et al, 2016). Our study deployed a DPV-propelled camera system (González-Rivero et al, 2014) to survey temperate reefs for the first time.…”

High Resolution Orthomosaics of African Coral Reefs: A Tool for Wide-Scale Benthic Monitoring

“…As a result, coral reef protection has become a priority in many governmental agendas (e.g., [12]). This has resulted in the development and implementation of scientific surveys and monitoring programs aimed at evaluating the causes, effects and extent of coral threats (e.g., sedimentation, bleaching and climate change) [13][14][15].…”

“…In general, wide-area (>1500 m 2 ) coral reef monitoring methods are based on remote sensing approaches that range from autonomous or semi-autonomous underwater vehicles [13,[16][17][18][19] to satellite and airborne imagery [20][21][22][23]. For the particular case of wide-area (>1500 m 2 ) in situ monitoring, the key limitations are cost, access to the site and the spatio-temporal extent to be covered [15].…”

“…In situ methodologies provide higher spatial resolution than wide-area (>1500 m 2 ) remote sensing methods (e.g., satellites) and are essential for the understanding of very fine ecological processes [15,24]. New technologies and analytical solutions (e.g., autonomous and semi-autonomous underwater vehicles, boat-based systems and close range Structure-from-Motion (SfM)) have been developed and can help fulfil the need for wide-area (>1500 m 2 ) high resolution data provision [13,18,19,25,26]. This scientific advance in coral reef monitoring relies on the integration of low cost off-the-shelf cameras and state of the art photogrammetric techniques (i.e., SfM) [27,28].…”