Field sampling and experimental design: Experiments were performed at University of Connecticut’s Avery Point Campus in the Rankin Laboratory, a seawater facility adjacent to eastern Long Island Sound. Ripe adult M. menidia were collected on 1 May 2015 from Mumford Cove (41 19.25’ N, 72 1.09’W), a shallow embayment dominated by eelgrasses (Zostera marina) and open to the Long Island Sound. Adults were sampled with a 30 m × 2 m beach seine, separated by sex, transported live to our laboratory, and held for 48 hours in large aerated tanks (17 degrees Celsius, ambient CO2, no food). On the day of fertilization (3 May 2015), greater than or equal to 20 ripe individuals from each sex were strip-spawned and eggs evenly distributed onto window screens (1 mm fiberglass mesh) submerged in plastic dishes with clear seawater. Strip-spawned adults were measured for standard length (mean SL, lower 0.5 cm; females 9.7, males 8.7). Fertilized embryos quickly attach to the screens via chorionic filaments, which facilitates precise enumeration and even allotment to treatments and replicates. Following established protocols for rearing M. menidia offspring (Murray et al., 2014), replicate containers (20 l) were filled with filtered (to 1 um) and UV sterilized seawater (31 psu) from Long Island Sound and placed in water baths (~300 l) controlled for temperature and light conditions (17 degrees Celsius, 15h light:9h dark) throughout the duration of the experiment. Within 2h of fertilization, each of four replicates per treatment received exactly 200 embryos for measure early life survival, while four other replicates per treatment each received ~400 offspring for long-term rearing. Larvae hatched ~14 days post-fertilization (dpf) and were immediately provided with standardized rations of newly hatched brine shrimp nauplii Artemia salina (San Francisco strain, Brine Shrimp Direct) and a commercial larval powder food (first four days, Otohime Marine Weaning Diet, size A, Reed Mariculture). At 2 days post-hatch (dph), living larvae from survival replicates were counted by gently scooping small groups into replacement containers. Between 1 to 14 days post-hatch (dph), all containers were cleaned daily with partial (10%) water exchange.
At 16 dph, larvae from the survival replicates were counted and a sub-sample (Ncontrol = 37, Nhigh = 33) was preserved in 10% formaldehyde/seawater solution for later total length (TL) measurements (nearest 0.01 mm) via calibrated, digital images (ImagePro Premier V9.1). All surviving larvae were transferred to larger (50 l) tubs and maintained under the previously described protocol. At 33 dph, larvae from the survival replicates were counted and then all larvae transferred to 50 l tubs fitted with screen-covered holes (1 mm mesh) to promote water exchange from a 300 l seawater bath. Due to space constraints, from 33 to 54 dph larvae from the survival replicates were pooled into a single container per CO2 treatment. Larvae were provided rations of nauplii and supplemented with commercial powder food (Otohime B1, Reed Mariculture). At 54 dph, all juveniles from survival and grow-out replicates were counted and pooled at equal numbers into 300 l circle tanks (two tanks per treatment, ~615 fish per tank). Juveniles were provided equal rations of newly hatched nauplii and B1 commercial powder food. Tanks were siphoned for waste daily and partial water changes completed twice weekly. Additional sub-samples for length measurements (TL, nearest 0.01 mm) were made at 36 dph (Ncontrol = 20, Nhigh = 20), 68 dph (Ncontrol = 20, Nhigh = 20) and 100 dph (Ncontrol = 28, Nhigh = 28).
At 122 dph, the experiment was terminated and all surviving juveniles were euthanized via an overdose of Tricaine-S (MS 222, Western Chemical) for preservation. While some juveniles from each treatment were immediately frozen at -80 degrees Celsius for fatty acid analyses; ~75% of the samples were fixed in 10% buffered formaldehyde/seawater solution for TL (Ncontrol = 1,025; Nhigh = 1,100, nearest 0.01 mm) and weight measurements (Ncontrol = 720; Nhigh = 786, nearest 0.01g).