TRAIT AND FITNESS MEASUREMENTS

We measured selection on three traits: larval mass, and metabolic rate at two stages during early ontogeny: two hours postsettlement and 24 hours postsettlement (hereafter referred to as metabolic rate early (MR_E) and metabolic rate late (MR_L)). MR_E and MR_L occur during crucial stages during the life history and have previously been shown to be under differing selection in this species (Pettersen et al. 2016). Larval mass is a key life history trait and a well‐known predictor of performance; however, the offspring size‐performance relationship is often context dependent (Marshall et al. 2018). We measured the diameter of newly spawned larvae to the nearest μm and calculated larval mass (μg) using previously developed protocols (Pettersen et al. 2015). Metabolic rate was measured for individuals using the common proxy, rate of oxygen consumption or V̇O₂, as per previous methods (Pettersen et al. 2015). In summary, individuals were settled in a drop of seawater onto small sheets of acetate and placed into glass vials containing pasteurized, 0.2‐μm‐filtered seawater and a nonconsumptive O₂ sensor spot. We used 24‐channel PreSens sensor dish readers (Sensor Dish Reader SDR2, PreSens, Germany) with 24‐chamber 200 μL glass microplates (Loligo Systems Aps, Tjele, Denmark) to measure V̇O₂ (rate of change of O₂ saturation over time; %h⁻¹), where readings were taken every 2 min over 3‐h intervals. We then used the package “LoLinR” to objectively and reproducibly estimate monotonic V̇O₂ from our readings (Olito et al. 2017). All analyses were conducted in R version 3.6.1 (R Development Core Team 2016).

We used survival to reproduction (viability), ability to reproduce (fertility), and cumulative reproductive output during the first 20 weeks of the life history (fecundity) as our measures of fitness. Survival and the presence of reproductive structures (ovicells) indicating ability to reproduce were recorded once per week—individuals were considered to be alive if they were still attached to their settlement plate and >10% of the colony contained feeding zooids. The fitness measures of viability and fertility were treated as binary data—individuals that survived to the average reproductive age (viability) and those that reproduced (fertility) were assigned “1,” whereas individuals that died before reproductive age or the onset of reproduction were assigned “0.” Reproductive output (fecundity) was measured as the cumulative number of ovicells throughout the duration of the study, which were counted using a dissecting microscope (×10) once per week, from the onset of reproduction at approximately six weeks post‐outplant. In a previous study of this population, Bugula survived up to nine months. Reproductive output during the first four months of the life history reliably predicted lifetime reproductive output (cumulative reproductive output 120 days post‐outplant explained 94% of variance in lifetime reproductive output for this same population; Pettersen et al. 2016). In addition, we measured several life history traits related to fitness over the duration of the study: growth (number of colony bifurcations per week; Keough and Chernoff (1987), longevity (number of days >10% colony remained alive), and age at onset of reproduction (days) up until 140 days post‐outplant.