Figure 1: Early competent sand dollar larvae do not display 'desperate' behaviors at settlement, with or without turbulence exposure. We either exposed D. excentricus larvae 10 day post fertilization (dpf) (reared at ~20°C) to 3 min of 6 W/kg turbulent shear or did not. Then, we transferred exposed and control larvae into one of two settlement conditions: 24 hrs in MFSW alone (left side of graph) or a 1 hr exposure to 40mM excess KCl in MFSW –to assess competence– followed by a 24 hr recovery in MFSW. We detected an effect both of turbulence exposure (F1,12=5.36, p<0.02) and settlement medium (F1,12=228.93, p<0.001) on proportion of larvae settled, but no clear interaction (F1,12=3.41, p=0.09).
Figure 2: Fully competent sand dollar larvae exposed to turbulence are less choosy about settlement substrate, and thus behave like 'desperate' larvae. We either exposed D. excentricus larvae 11 dpf (reared at ~20°C) to 3 min of 6 W/kg turbulent shear or did not ("no turbulence"). Then, we transferred exposed and control larvae into one of two settlement conditions: 0% extract of sand from sand dollar aquaria (MFSW; no inducer) or 40% extract of sand from sand dollar aquaria (strong natural inducer), and counted the numbers settled at 1 and 16 hrs. We also exposed a separate set of control (no turbulence) larvae for 1 hr to 40mM excess KCl in MFSW, followed by recovery in MFSW. More than 95% of these latter larvae settled, confirming that the larvae in this experiment were indeed fully competent.
Figure 3: Fully competent sand dollar larvae exposed to turbulence are less choosy about settlement substrate, and thus behave like 'desperate' larvae. We either exposed D. excentricus larvae 40 dpf (reared at ~14°C) to 3 min of 6 W/kg turbulent shear or did not ("no turbulence"). Then, we transferred exposed and control larvae into one of two settlement conditions: 0% extract of sand from sand dollar aquaria (MFSW; no inducer) or 200% extract of sand from a beach without sand dollars (poor quality natural inducer), and counted the numbers settled at 1.5, 3, 7.5 and 19 hrs. We also exposed a separate set of control (no turbulence) larvae to a strong natural inducer (30% extract of sand from sand dollar aquaria; right side of graph). This treatment not only confirms that the larvae were fully competent (100% of larvae settled by 8 hrs), but indicates the expected rate of settlement response in a strong cue as a basis of comparison to the sub-optimal cues described above.
We performed all statistical analyses using R (version 3.4.2) and the lme4 package.