*2.1. Study Area*

The Carlsbad Desalination Plant, built and operated by Poseidon Water, is the first and currently only large-scale SWRO desalination facility in California. Operations began in December 2015 with a capacity of 180 × 10<sup>6</sup> L day−1. The plant is located in an industrial area at the southern end of the Agua Hedionda Lagoon, adjacent to the Encina Power Station (Figure 1A). Seawater enters the lagoon through a dredged channel at the north end of the lagoon, about 1 km from the seawater intake used by the power station as cooling water (since 1954). The desalination plant diverts ~10% of this water released by the power plant post-cooling for SWRO (Figure 1B). The brine effluent is then returned to the cooling water further downstream, resulting in a 1:10 dilution of the brine before the mixture is discharged to the ocean through a 10–15 m wide channel (outfall) between two rocky walls extending ~50 m offshore at Carlsbad Beach (Figure 1A) [39]. Carlsbad Beach is a relatively high-energy beach with wave heights averaging ~2.5 m in winter and ~1.5 m in summer [40–43]. The nearshore habitat is dominated by a sandy bottom with scattered small rocky reefs and seagrass patches at the northern end of the beach. The nearshore region is shallow with depths of 5–10 m up to 800 m offshore, and with deeper water (~20 m) starting ~1 km offshore [41].

**Figure 1.** (**A**) Aerial view of the study site around Carlsbad Desalination Plant. The SWRO plant and power plant, intake and outfall channels, and sampling sites (stars) are marked; (**B**) schematic of water flow from the Pacific Ocean to Laguna Aqua Hedionda and the intake and discharge of water by the power plant and SWRO facility (not to scale). White arrows indicate seawater flow and dark-blue arrows represent discharging brine and its mixture with the power plant's seawater return flow.

### *2.2. Sample Collection and Biological Surveys*

Water and sediment samples were collected, and benthic surveys were conducted twice in the year prior to the start of SWRO operation (pre-operation), and twice in the first year after brine discharge began (post-operation). Pre-operation samples were collected in December 2014 and September 2015, and post-operation samples were collected in May (five months post-operation) and November 2016 (11 months post-operation). Samples were collected by scuba divers, starting ~100 m offshore (closer if conditions allowed). Samples were taken along two transects perpendicular to the beach at 25 m intervals out to 200 m, and then at 250, 300, 400, 600, 800 and 1000 m from shore. Sampling was conducted close to shore as the chemical footprint and possible biological impacts were expected to be greatest close to the discharge site and quickly dissipate with distance due to mixing. A transect parallel to shore was conducted as close to the beach as possible (~200 m offshore), with samples collected every 100 m from 200 m south of the outfall to 200 m north of the lagoon entrance (Figure 1A). Bottom and surface water samples were collected at each sampling site in 1 L acid-cleaned, sample-rinsed, Nalgene HDPE bottles. Water temperatures were measured on the boat, immediately after collection, with a YSI 30SCT probe (YSI, Yellow Springs, OH, USA). Water samples were kept in a cooler until they were filtered (using various filters, see Section 2.3 below) and processed, within 4 h after collection. Sediment samples (from the upper 10 cm) were collected in 250 mL plastic jars by pushing the jars into the sediment and samples were kept frozen until analysis (see Section 2.3).

Benthic surveys of epifauna were conducted along two continuous perpendicular transects each 1 m wide, from ~100 to 200 m offshore, and then extended with 1 × 1 m quadrats (1 m2) at ~100 m intervals from 200 to 1000 m offshore. Along the continuous transects, a 100 m measuring tape was laid on the sea floor and all visible benthic organisms within 0.5 m of each side of the tape were recorded for each 10 m long section (10 m2). At stations running parallel to shore, organisms were recorded in ten randomly distributed 1 m<sup>2</sup> quadrats per station. Visible benthic epifauna were enumerated separately by two scuba divers, and counts were calculated to species abundance per m<sup>2</sup> and grouped in counts close to the discharge channel (0–200 m offshore) and in the surrounding area (250–1000 m offshore). Organisms were assigned to 10 categories: Porifera (sponges), Anthozoa, Gastropoda, Bivalvia, Cephalopoda, Polychaeta, Echinodermata, Arthropoda, fish, macroalgae and seagrass.

### *2.3. Water and Sediment Analyses*

Water samples were analyzed for salinity, chlorophyll *a*, nutrients (PO4, NO3 and SiO2), particulate organic matter (POM), dissolved organic carbon (DOC) and trace metals. More detail is provided below:

Salinity: 50 mL of seawater were pumped through a Guildline Portasal salinometer (Osil, Hampshire, UK) and salinity recorded with an accuracy of ±0.001.

Chlorophyll *a*: 250 mL of seawater were filtered onto GF/F filters (0.7 μm) (Whatman, Chicago, IL, USA) and kept frozen in the dark until further analysis. The filters were immersed in vials with 7 mL of 90% acetone and extracted at 2 ◦C for 12 h before measurements. Chlorophyll *a* (Chl *a*) concentrations were measured using a TD-700 fluorometer (Turner Designs, San Jose, CA, USA) calibrated with Chl *a* according to the EPA 445 method, using CS-5-60 and CS-2-64 glass filters (Turner Designs, San Jose, CA, USA).

Nutrients: 50 mL of seawater were filtered through 0.2 μM filters and kept frozen until analysis. Samples were analyzed for soluble reactive PO4 (SRP) [44] and for NO3 and SiO2. Samples were processed on a QuikChem FIA+ 8000 Series autoanalyzer (Lachat Instruments, Milwaukee, WI, USA) using methods 31-115-01-3-A (PO4), 31-114-27-1-B (NO3) and 31-107-04-1-A (SiO2). Sample reproducibility was 0.5%, with a detection limit of 0.15 μM for all analytes.

Dissolved organic carbon: 30 mL of seawater were collected in pre-ashed glass vials (450 ◦C for 4 h). The seawater was acidified to pH 2 using concentrated trace metal grade HCl, and analyzed on a Shimadzu TOC-V autosampler (Shimadzu, Columbia, MD, USA) with prepared standards of 2.125 g KHP L−1. Detection limit was 0.3 μM with 2 mL injections.

Particulate organic matter: 400 mL of seawater were filtered onto pre-ashed (450 ◦C for four hours) GF/F filters (0.7 μm) and samples kept frozen. The filters were dried for 48 h at 50 ◦C and pressed into tin capsules. The carbon to nitrogen molar ratio (C:N) and C and N isotope ratios were measured using isotope ratio mass spectrometry (CE instruments NC2500, Wigan, UK). Analyses were done with standards of pugel and acetanilide to an accuracy of ±0.11 ‰.

Select major and trace elements: 50 mL of seawater was filtered (0.2 μM) and collected in acid-cleaned sample-rinsed LDPE bottles and acidified with trace metal grade HCl to pH 2 prior to analysis. Samples were diluted 500-fold and analyzed on an Element XR ICP-MS (Thermo Fisher Waltham, MA, USA) with standard curves prepared using NIST 1640a standard solution (National Institute of Standards and Technology, Gaithersburg, MD, USA). The average of two procedural blank values was used for blank correction. Concentrations of Li, Na, Mg, K, Ca, Mn, Fe, Sr and Ba were measured.

Infaunal organisms: Sediment samples were thawed and washed with 90% ethanol. The ethanol rinse was collected and all polychaetes and amphipods were counted immediately under a dissecting microscope. For polychaetes, individuals of the two most abundant families, Paraonidae and Capitellidae, were identified and counted; individuals from other families were counted as "other". For amphipods, the families Gammaridae, Hyperiidae, Caprellidae and Corophiidae were identified and counted. A BOPA-index was calculated for each sediment sample using methods described

in reference [27] (Equation (1)), where *fpop* is the proportion of polychaetes and *fa* is the proportion of amphipods:

$$ROPA = \log\left(\frac{fp\_{op}}{f\_a + 1} + 1\right). \tag{1}$$

Sediment grain size: The sediment was dried, and 75–90 g subsamples were sorted using twelve sieves (from 0.5 Φ to 4.75 Φ, Krumbein phi scale), and each fraction was weighed. The mean grain size and sorting factor were determined using methods described in [45].
