Methodology:

Panel deployment:

The panels were deployed in the Menai Strait, Wales, UK (53.2307 N, 4.1537 W). This is a special area of conservation (SAC) where the diversity of habitats and the strong tides have promoted high benthic diversity. The area is typified by muddy sand substrates with some boulders and rocky outcrops, creating islands for settlement and development of hard substratum communities. The dynamics of the water flow in the strait are dominated by strong tides, causing the water column to remain vertically mixed throughout the year, with little difference between surface and bottom water temperatures. Sea surface water temperature (SST) in the Menai Strait can vary annually between 5°C - 18°C, with SST temperatures recorded during the period of this study of 6°C - 16.3°C (https://seatemperature.info/menai-bridge-water-temperature.html) provided by the daily SST satellite readings from NOAA.

The panels used in this experiment were of the same design and specification as described in Ashton et al., (2017) Current Biology 27: 2698-2705. In total, 12 panels were mounted on to 4 steel frames (3 panels per frame, 4 replicates per treatment (control, +1°C, +2°C)). The thickness of the layer of water warmed to the same temperature as the panel surface varies with the flow rate of water across the panel. Trials in a flume system showed it is always more than 1.5 mm and can be as much as 3-5 mm in low flow conditions (Peck, pers. comm.). No animal grew beyond the 2mm layer for the duration of each deployment. The panels were deployed on the seabed at a depth of 10m below the lowest tide mark and 50-100m from the shore. Power to, and control of the panels was conducted from a nearby building. The experiment was conducted for 13 months from June 2015 through June 2016. Panels were first deployed on the 05/06/15 and photographed regularly during each 3 month deployment period. Each deployment lasted 3 months, at which point the panels were retrieved, photographed and sampled before being scrapped clean and re-deployed. This interval was dictated by the growth rate of colonisers in summer, which reached high levels of panel surface coverage within this time but were still growing inside the heated boundary layer of the plates.

Panel photography and sampling:

At regular intervals, the panels were lifted onto the side of a boat and brought to the shore near the sampling sites but held under water at all times. Panels were photographed using a Nikon D7000 with a Nauticam Underwater housing and a 60 mm macro lens (following Ashton et al., 2017)). Panels that were deployed during the winter period (December-March) were not photographed in between the start and end of the deployment, due to low recruitment and slow growth during the winter months. At the end of each deployment all panels were brought back to inland aquaria at the School of Ocean Sciences at Bangor University for sampling. After sampling, panels were scraped clean and re-deployed.

Analyses of community composition and species:

For each deployment, species composition was assessed under low power microscopy (10x) with the use of local taxonomic guides. Where species with unconfirmed taxonomic IDs were present, these were sampled and stored in 95% ethanol for subsequent DNA barcoding using18S rRNA primers (NSF4 and NSR581). The community growth on panels was analysed for the percentage area and species composition across the different treatments (control, +1, +2) seasonally (summer, autumn and spring).

Species richness and diversity:

Presence and percentage cover data for sessile taxa on panels at the end of each sampling period were extracted from photographs using the colour threshold plugin in Fiji. A two-way analysis of variance (ANOVA) was performed in the R environment for statistical computing to examine differences in percentage cover across sea...(13)

Data collection:

Image processing: Photoshop CS5

Image data processing: Fiji (Image J)

Statistical analysis: R studio

Data quality:

Panel instrumentation was regularly checked.