We investigated the effect of the presence of an experimentally generated thermocline on the vertical distribution of larval Strongylocentrotus droebachiensis, Asterias rubens and Argopecten irradians. Vertical distributions were recorded over 90 min in rectangular plexiglass thermocline chambers designed to regulate the temperature of a central observation compartment to the desired values. The temperature in the bottom water layer (B) and the temperature difference between layers (ΔT) were manipulated in an orthogonal design. We used, for S. droebachiensis: 4 levels of ΔT (0, 3, 6 and 12 °C) and 3 levels of B (3, 6 and 9 °C); for A. rubens: 3 levels ΔT (0, 6 and 12 °C) and 2 levels of B (6 and 12 °C); and for A. irradians: 3 levels of ΔT (0, 5 and 11 °C) and 2 levels of B (5 and 11 °C). The difference in temperature between water layers did not affect the vertical distribution of echinoderms consistently, while the distribution of A. irradians was limited to the bottom layer when any thermal stratification was present regardless of strength. Our results suggest that the vertical position of larvae of S. droebachiensis and A. rubens is related to the temperatures of the surface layer and that the presence alone or the steepness of the thermocline has less influence on their distribution. Consequently, in the field, echinoderm larvae would aggregate at the surface unless temperature extremes were encountered. In contrast, the position of A. irradians was limited to the bottom layer in the presence of a thermocline of at least 5 °C (the shallowest used in our study). Such thermoclines are common in a natural setting and could affect the vertical distribution and horizontal dispersal of larvae by acting as a barrier to vertical migration.
survival data
We measured larval mortality for S. droebachiensis at 3 different temperatures (3, 10 and 21 °C) reared in 4-l culture jars (n = 3). Temperatures in the jars were maintained either by placing them in a water bath (3 and 21 °C) or in a temperature-controlled room (10 °C). These temperatures encompass ambient and the extreme temperatures used in the experimental thermoclines. Six-day old larvae (each replicate was from a single parental pair, and reared at 10 °C) were used for this experiment. To quantify mortality, 30 larvae from every treatment for each replicate were transferred to a Petri dish at 0, 24 and 48 h, and categorized as live if swimming was observed using a Nikon SMZ 1500 dissecting microscope. Data set includes jar (replicate 4-l cuture jars), temperature (°C), time (h), and the number of live and dead larvae from a sample of 30 larvae
survival.xlsx
vertical distribution
We manipulated both the temperature in the bottom water layer (B) and the temperature difference between layers (ΔT) in an orthogonal design. We used, for S. droebachiensis (n = 5): 4 levels of ΔT (0, 3, 6 and 12 °C) and 3 levels of B (3, 6 and 9 °C); for A. rubens (n = 4): 3 levels ΔT (0, 6 and 12 °C) and 2 levels of B (6 and 12 °C); and for A. irradians (n = 4): 3 levels of ΔT (0, 5 and 11 °C) and 2 levels of B (5 and 11 °C). These combinations of factors were chosen to represent the conditions in the local environment. Since there were up to 12 treatment combinations for a particular species and we only had 4 thermocline chambers, not all replicates could be run simultaneously. For S. droebachiensis, different cohorts (from unique parental pairs) were blocked in time and all treatments were completed in 3 randomized groups of 4 within 26 h. For the other species, all replicates of all treatments were conducted in randomized order with a single larval cohort (from multiple parental pairs) within 48 h. Due to limitations of the seawater system at the time, the surface water was ~ 20 °C rather than 21 °C in the treatment combination of ΔT = 12 °C and B = 9 °C for S. droebachiensis.
Once the thermoclines were established, 50 ml of seawater containing 250–350 larvae were introduced to each chamber within 1 cm from the bottom of the observation compartment (filled with 0.45 μm-filtered seawater) by gently pouring into a funnel attached to a small tube (2-mm inner diameter). Before being introduced to the experimental chamber, larvae were acclimated to the respective bottom temperature for 15 min. Larval position was visually determined to the nearest cm at 5, 15, 30, 45, 60 and 90 min after introduction for S. droebachiensis and A. irradians, and at 10, 30, 60 and 90 min for A. rubens. Larvae in the entire 50 cm water columns were counted in < 2 min, making repeat counts of individual larvae highly unlikely. Data set includes experimental species, date, temperature of the top and bottom layers of the water column (°C), the temperature difference between layers (°C), time (min), and the number of larvae counted at each depth (cm)
